AC Le Havre – 2022-23: Sailing out to a European waters?

The 2022-33 season saw us fall into an initial slump, with poor results against our opponents in the opening fixtures, seeing four defeats in the first seven games and only two wins. This form saw us sat in 15th. Only after around nine games did we find our rhythm. Reducing the number of goals conceded by minimising the quality of chances presented to the opposition helped significantly, and as confidence and morale improved, so too did our goal scoring. Working on our defensive set up in training, as well as dropping Václav Jemelka after doing some metric analysis into our performances in favour of Dalle Mura, on loan from Fiorentina, helped to solidify our back line, dramatically bringing down our xGA.

This is reflected in our eventual finish of 6th, one place outside of European qualification (due to the winner of the Coupe de France), and in the amount of points we gained above that of those that my self-calculated model said we should have attained come the end of the season. The dip in form after the crazy run of fixtures following the Winter World Cup is mostly due to the need to rotate our small squad (more on this later). A strong finish to the last third of games saw our league position soar.

If we look at the cumulative net points (points minus expected points using xG differentials), then we were largely outperforming the number of points we should have had, with a big dip after the 4-1 battering at the hands of Lille. Earning three more points by the end of the campaign than we were predicted to using xG differentials helped us to overachieve by way of our expected league position finish.

I alluded to the heavy rotation that we had to adopt due to the fixture congestion thanks to the Winter World Cup. This can be plainly seen in our squad profile – taking minutes played and player ages. In trying to protect our plethora of young players against burn out, only goalkeeper and captain Juan Soriano played more than 80% of potential minutes in Ligue 1 (in fact he played every minute). With the minutes shared out, you’d think that the development of player attributes would have been evenly shared, but because of the number of midweek games and a near month long holiday in December, player development largely stagnated because they weren’t on the training field as often as they would have been during a normal season.

This leads me onto a player I want to hone in on and almost pay homage to as a result of his performances.

Player Focus – Hannibal

Hannibal Mejbri, known simply as Hannibal, has been on loan with us at Le Havre for the last two seasons. Over that time he’s made 58 appearances and 27 goal contributions, one every 160.22 minutes, or roughly 0.56 per 90. He has therefore understandably come to make the 8 position his own on the right-hand side of midfield.

Despite his young age, he already has an eye for a pass, with fantastic vision and technical ability. His underlying technique, flair and first touch enable him to make defence splitting passes for players running through on goal.

This is reflected in his third place finish in the Ligue 1 Meilleurs Passeurs for 2022-23, after racking up ten assists, behind Neymar (20) and Marseille’s Maxime Lopez (12).

His attributes naturally lend themselves to playing as an advanced playmaker, and to an extent, the team has somewhat been moulded around him in this role.

With a false nine ahead of him (Ali Akman), a winger on the left to flank (typically either Nordin or Pité), an inside forward cutting in from his right flank (either Dia or Šaranić) and a mezalla acting as another 8 to his left (normally Nascimento), he’s not short of passing options. Combine this with the right sided full back going forward on a supporting duty, and he has a gamut of choices.

With the adopted tactic set to play at a slow tempo with shorter passing, this does act as something of a constraint on his creative abilities. However, it does mean that the team can build possession together to look for an opening rather than have to continually press in exhaustive fashion when the ball has been lost in an effort to recycle possession. Equally, if the ball is turned over, because the team have progressed up the field together, it is then easier to enact a meaningful press by blocking multiple passing lanes rather than a solo press.

This helps to hide Hannibal’s main weakness – he’s a poor defender. He can shirk his defensive responsibilities, with 2.96 defensive actions per 90 (tackles and interceptions adjusted for possession) over the last season, a metric which puts him in the bottom 24th percentile for Ligue 1 central midfielders. This is part of the reasoning for opting to play a defensive midfielder in behind him and the other 8, as Nascimento is similarly uninspired by having to defend (with 1.82 defensive actions per 90 – in the bottom 2%).

With this cover, Hannibal is able to play the creative role, and was amongst the league leaders in all manner of passes (attempted, completed, key and assists per 90), dribbles and in being fouled (leading to the potential for attacking free kicks, which he himself takes). His pass completion is poor, but I think in part this is because he takes set pieces, but also because he’s looking for the through ball which will often be cut out by opposition defenders.

The below shot you can see Hannibal’s ability to deliver a cross to pin point perfection. Following an interchange with right back, Godswill Ekpolo, which created separation for Hannibal away from the opposition defender. Hannibal then swung in a cross, picking out Pité who has eluded his marker at the far post for an easy finish. This is only made easy because of the quality of the delivery itself from Hannibal.

It’s at this point that I wanted to delve into Hannibal’s fantastic passing ability and vision further, but sadly I’ve come across (yet) another bug in the game with a divergence between recorded assists outside the ME to what is shown inside the ME when going back to old fixtures, so instances in some matches are wrong, including passing maps, times of goals, goal scorers, assist makers… I’m hoping it was the update that threw this all out, but who knows.

Club update

After receiving the prize money for our league finish, I immediately request that the board reinvest the proceeds into the recruitment of youth prospects, and improve our training facilities for both the senior and youth sides. This should aid player development amongst the senior side who have stagnated as previously mentioned thanks to the lack of time on the training ground. Fingers crossed it will also help develop our own youth players as this is a key part of the club focus and as you can see by the squad profile, I’ve not done a great job of bringing any through, instead favouring external recruits.

The problem of this long term strategy was that it left us without much immediate cash with which to extend contracts of key staff, and a tiny budget to initially spend on transfers. As a result, early incomings on the playing staff were nil –  we simply didn’t have the funds to bring anyone in. This was fine, as I felt that bar left back, with Carole departing following the expiry of his contact, I had two players of adequate or good quality in every position.

This though, didn’t solve the issue with staff. I was unable to negotiate contact extensions with a number of staff because their weekly wage expectations had risen following the club’s success on the field. Yet this wasn’t matched with the wages I was in a position to offer them. As such, a number left and were replaced with unemployed coaches, analysts, physios and scouts who were willing to join for lower compensation for their gainful employment.

My thoughts regarding squad depth held until I had an offer for Guilherme Montóia from Arsenal. I knew as soon as Arsenal bid for him that I was not going to be in a position to reasonably stand in his way. Whilst the Club is just about washing its face when it comes to finances, player sales were going to be required to pay for the growth and development of the Club as a whole. The initial £7m bid was negotiated up to £12.25m, including £5m in £1m payments staggered over semi-annual payments for the next three years, which should help with our cash flow. The big bonus though was the fact that Arsenal were happy to loan Montóia straight back to us to continue his development. Montóia could remain first choice whilst his eventual successor bedded in.

Looking into the finances of this deal in more depth, because we can book the proceeds of the sale of Montóia straight away, i.e. record the receipt of funds immediately and not when actual payment is received. This results in us making an instant paper profit of the £12.25m that Arsenal agreed to pay because Montóia was signed on a free transfer – his book value to us was nil, so the transfer fee represents pure profit.

The other transfer that I couldn’t possibly refuse was that of Brahima Ouattara. Juventus were negotiated up to an offer of £11.5m, including three guaranteed payments of £1.33m over the next three years to further bolster future cash inflows. The Ivory Coast international made twenty three appearances but never really shone in the mezalla role. Given the depth we have in central midfield, I was more than happy to let Ouattara go to Turin and try to break through into i Bianconeri’s first team.

The fee we had paid RC Abidjan for Ouattara was a meagre £275k and he had signed on a four-year deal. Given that there were still two-years left on his contract, this meant that the booked profit on this transfer amounted to a whopping £11,362,000 (whopping relative to our club size at least).

In booking this £23,612,500 profit, this provided the Club with funds to reinvest back into the transfer market where we could find more ‘wrinkles’.

With Carole leaving, I went searching for a left back. After much deliberation, largely because no one was as good at the soon-to-be outgoing Montóia, we brought in Jonathan Augustinsson from Djurgården in Sweden for £1.4m, rising to £1.8m after 50 league appearances. His appearances for Sweden and his relatively older age compared to those around him in the squad I’ve assembled at Le Havre (note the number youngsters in the earlier squad graphic above), should add experience to the side. His personality of a model professional certainly made my mind up when deliberating over his transfer – hopefully he can climb up the player hierarchy to become a team leader so that he’ll make a perfect mentor for any defender in the side.

We also looked to invest into a back up for Ali Akman and brought in Leon Bosnjak, a young Croat from NK Varaždin, for his £2.6m release fee. Looking at his attributes, he looks well developed for an 18-year old. Whilst his finishing is below what I’d normally look for, I realise I can’t have everything at this level, especially not for an initial back up player. His composure and off the ball movement, along with his excellent with rate and flair should mean that the can bamboozle defenders to increase the quality of chances he creates for himself/find himself free in pockets of space to be find by the likes of Hannibal.

Chema Núñez signed on a free transfer from Albacete to improve the quality of our attacking output down the left hand flank. The pacey Spaniard likes to dribble down the left flank, yet also moves into channels, so should confuse the opponents he finds himself up against when coupled with his flair, technique and dribbling ability. His vision and passing is also excellent, with his trait of playing one-twos and killer balls, he should offer a fantastic threat going forward.

Dual national, Giulian Biancone came into to offer a challenge to Godswill Ekpolo. Ekpolo will remain the first-choice, but at £750k following his transfer-listing at Monaco, this looked a good deal for the former French U21. He’s attacking in his nature as a full back, looking to bomb on down the right flank, so will help to add width given the inside forward that is played ahead of him. His decisions, technique and composure could be better, but there won’t be many, if any, better players at this price in his position.

Just prior to the start of the season, Juan Soriano picked up an injury which meant that he was set to miss the opening three games of the forthcoming season. With only the under-developed Yahia Fofana as a back-up option, I brought in Alfred Gomis from Rennes as a reserve goalkeeper. This is quite some come down for the player chosen by his previous club Rennes to replace Edouardo Mendy after the latter signed for Chelsea. The 6’5″ Senegalese international joined for just £1m, after his transfer listing. Having played just four games in 2021-22 and zero in the 2022-23 season, Gomis was more than happy to cut the cord from Rennes and be the second choice behind Soriano when he returned to fitness.

Hannibal and Della Mura both had their loans renewed, so will be with us for the 2023-24 season. All deals saw us spend £5,750,000 in total on first team players, with a further £1,855,550 spent on U19 potential prospects. This was in part trying to becalm the directors who were ‘devastated’ at my inability to purchase players from lower leagues in France, develop them to the first team before selling them on for a profit. In total, five such players were brought in but they will not be expected to feature in the first-team and anyone older than 18 will be loan listed to encourage their development.


Will this addition of depth to the squad help our onward march to European places? Find out in the next blog post.

AC Le Havre – 2021-22 Season Review

HAC Foot – Safe Haven in Ligue 1

The winter break saw HAC Foot in 8th place and it is there that we finished in our first campaign in Ligue 1. Quite some effort for a team expecting to be relegated back from whence we came.

The second half of the season actually went better than our first. Picking up nine wins and four draws saw us achieve four more points we did before the Winter break. Notable performances included a 3-0 victory at home to Lyon, 1-0 win against Nice and a 3-3 draw versus Monaco where we raced into a 3-0 lead only to see Monaco rage back against us.

The results enabled us to finish just outside the European places – perhaps as well given the lack of depth we have in our squad. This is evidenced by the fact that eight players played in excess of 75% of the possible minutes in the league – six of which were new additions to the team. It’s great that the players we signed had the natural fitness and stamina to last the season. However, this will be taken into consideration when looking to utilise the £6.78m we’ve been allocated in transfer spending and nearly £100k p/w in wages.

Gibaud’s, Gorgelin’s and Club captain Fontaine’s contracts are all expiring and they will not be signed to new deals. This may be a little harsh on Fontaine who has been with HAC Foot throughout his career. When delving into his data for the 2021-22 season, his first ever in Ligue 1, it looks even more ruthless. Fontaine, despite playing as a deep-lying playmaker in the defensive midfield strata, produced some elite level metrics, delivering key passes, being efficient with his use of the ball and managed a goal involvement 0.41 per 90, almost a goal or an assist every other game. It is worth noting that this sample size isn’t huge with only 14.62 90s played, but he did play in a good number of the games in the second half of the season alongside our improvement in results.

However, at 33, his physical attributes are starting to wane. Whilst Ligue 1 isn’t the most physical of leagues, a two-year deal did not seem like a good use of the clubs finances, would block the player pathway of Briand, and could open wage budget for other recruits too to fill the gap between Fontaine and Briand, besides just Lekhal.

For Gorgelin, whose performances saw us keep eleven clean sheets – the decision could have also have been said to be heartless. Again, the contract that he was demanding was not acceptable to the Club in terms of how he was valued. That might seem strange given the number of clean sheets and the fact that we conceded just 42 goals as a club. However, as Sir Arthur Conan Doyle wrote in ‘A Scandal in Bohemia‘, “It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts”. So let’s take a much closer look at the data.

Further investigation into our metrics demonstrates a clear down turn in form in keeping out goals over the second half of the season compared to those the data tells us we should have conceded (GA-xGA). We conceded nearly four goals more than we ought to have done according to xG – not awful, but perhaps there’s room for an upgrade in this area. This would so easily be overlooked given the improvement in results from the pre-Winter break.

If we take a deeper dive still into the data, you’ll notice that over the course of the season the 10-game rolling average goals, xG, goals against and xGA drifted together. Once we are beyond the “noise” of the first ten games, we see that bar one small spell, there never was much gap between all statistical measures, both actual and theoretical. In other words, we were performing roughly as expected over a rolling ten game period and we were perhaps a little lucky to be ultimately as high as 8th. Therefore, if we can upgrade in the goalkeeping position, this could see us improve further still, should we bring the goals conceded back towards xGA and xGA remains low next season.

To an extent, this is backed up on the cumulative xPts data too. We broadly achieved the number of points we should have done, with us dropping only two points using the cumulative differences between G-xG and GA-xGA. Accordingly to the data, we statistically could have picked up two additional points to those we achieved.

Therefore, it could be said that although 8th place was something beyond our wildest dreams before first fixture kicked off, this is an indication that we actually underperformed in our first season in Ligue 1. This comes despite the bookmakers having us odds on to be in the bottom two and relegated. Definitely some to ponder on when deciding the balance of the squad over the Summer transfer window.

Stand out player

Before we progress onto the Summer dealings, it is worth highlighting the speed at which Ali Akman became an integral part of the squad. Eighteen goals on an xG tally of fifteen, and an xG per shot of 0.25, meaning one in every four of his shots would likely result in a goal. The latter statistic was the best in Ligue 1, and for a 20-year old that is seriously impressive. Especially so given he was playing as a withdrawn striker in the false nine role. His actual xG tally was fifth best of all strikers, behind Ligue 1 top goal scorer Harry Kane, and Akman ranked second by shot to goal conversion. So despite playing in a far deeper positions than any of his peers would likely been playing in, Akman looks to have made a fantastic start to his time on the French coast. If he can improve his interlinking to set up his teammates to up his assists, his time be at HAC Foot might be limited as bigger clubs start to take notice.

Transfer planning

With the aforementioned contracts expiring, deals needed to be done to replace these players and to deepen the player pool in terms of quality.

Before the transfer window opened, a deal was done to make Arnaud Nordin’s stay a permanent one, with him happy to stay on at HAC Foot after his deal with Saint-Étienne expired. Saint-Étienne coincidently were relegated. He signed a 3-year deal on a basic salary of £9,250/week with various bonuses and clauses attached. His hard work as part of the forward line to press and win the ball back with his tackles and interceptions (T&I/90) and the willingness shown to take on his opposite number with a dribble saw him create seven assists. These metrics are easily good enough for a Ligue 1 player, so to acquire him for no transfer fee is a big bonus.

Also prior to the transfer window re-opening, Manchester United were approached to extend Hannibal’s loan. Whilst he didn’t see much over 50% of the possible minutes, his importance to the system was growing, despite his inconsistencies in performances. Both the club and player accepted the offer, so Hannibal stays in his home country for another year and on a lower wage contribution to boot. It’s not hard to see why I was so keen either given his metrics. In the first year of his loan spell, he was elite at goal involvement and key passes per game when on form. Whilst his passing success rate is low, as an attacking playmaker, this is not that much of an issue if he is creating chances and assists for his team mates. He was second in the list of key passes played by either defensive or central midfielders with more than 1,000 minutes played in Ligue 1 – he will be welcomed with open arms upon his return from holiday.

It was adjudged that Yahia Fofana did not have the required skills and attributes to step up to the No. 1 jersey, and as such, scouts had been focussed on finding a new goalkeeper. Scouts had for a long time been keeping tabs on Spaniard Juan Soriano. He’d spent the 2020/21 season on loan at Malaga from his parent club, Sevilla. Over that season he kept seven clean sheets from 41 starts, conceding 50 goals, with four player of the match awards. However, over the 2021/22 season, he had returned to Sevilla and found himself a third wheel behind Tomas Vaclik and Yassine Bounou, who shared the goalkeeping duties. With the picture clear that he wasn’t going to receive first team action at Sevilla, when we approached the once U21-capped 24-year old with a contract and promise of being the first-choice goalkeeper, he didn’t hesitate in signing a three-year deal on £6,500/week. Although a tad eccentric for my liking (15), he is excellent at one-on-ones (16), has a high level of concentration, reflexes and agility (all 15),

To replace Fontaine, we looked for an experienced player to offer a mentality of leadership and positive attitude, alongside a considerable number of minutes in Ligue 1. Defensive midfielder Abdoulaye Touré became available on a free signing after being released by Nantes. He took some convincing to join, with negotiations being protracted over a number of weeks after the player initially turned us down on two occasions, but we eventually managed to come to an agreement with the Frenchman signing a 3-year deal on £17,000/week, making him our highest paid player. At 28-years of age, he has made 175 appearances in Ligue 1 and has a fairly determined mentality with strong attributes for determination, teamwork and work rate.

Mathieu Cafaro had also been impressing domestic scouts in the Stade de Reims team. Following his transfer listing, we saw an opportunity to sign a quality young French player who could rotate with Hannibal and had the versatility to play in a number of other roles if required. A fee of an initial £2.6m, climbing to £3.2m, was agreed with Stade de Reims, with Cafaro signing a deal worth £13,000/week. If we can flip the player for a profit, this could prove a shrewd deal.

With Umut Meraş not receiving many minutes because of his consistently poor performances, another back up to Carole was required. Carole’s age of 31 meant that the Club were in a position to sign a young player with Carole passing on his experience to him on the training pitch. Guilherme Montóia was available to be able to discuss a pre-contract as S.L. Benfica had neglected to renew his contract. A considerable number of top clubs were after his signing – yet because we were in a position to offer a clear pathway for him to be a key first team player within the next two years Montóia looked favourably upon our contact offer. He did seek a reassurance that we were going to be an established Ligue 1 side over the next few years, but having achieved 8th in our first year, we felt confident we could match that promise.

Scott Fraser’s first season at the club was less than successful. It became apparent that he did not play well alongside Hannibal. Both players weren’t significant contributors to defensive actions (tackles and interceptions), and Hannibal outperformed Fraser when it came to goals and assists over the season. Whether this was due to a language barrier, the jump from League One’s MK Dons being too great or finding himself playing in a slower, a less direct and slower playing style than he was used to, it’s hard to determine, but it was clear that a replacement option was required.

Diogo Nascimento was identified as the player to come in and take Fraser’s place in the starting eleven. Available on a pre-contract agreement after running down his contract with S.L. Benfica, much like Montóia, the young Portuguese starlet was similarly wanted by a host of top European sides from most of the Big Five leagues. We moved quickly to secure his signature on a five year deal. The left footed midfielder will hopefully adapt to life quickly in Le Havre so he does not have the same fate as Fraser. The Scot will be offered out on loan to try to drum up some transfer interest with the aim of recouping a transfer profit from the sale.

If Fraser was a failure, then Boulaye Dia couldn’t have been anything more different. Dia had a fantastic campaign – his superb metrics show how important he made himself to our HAC Foot side. His twenty four goal involvements saw him in the top 90% of all wide attacking players with more than 1,000 minutes. His attacking nature saw him drive at defences, and have a better shot to goal conversion rate than players like Mbappe and Neymar. Yet when Dia wasn’t available, the drop off to players like Pité was notable. Therefore, a second option to add depth to the playing squad was required.

Scouts had spotted Ivan Šaranić playing in the UEFA-21 qualifiers, alongside fellow Croatian Tomislav Duvnjak and came back raving about both. Šaranić possesses the attributes that mark him out as an ideal initial back-up to Dia, but he was keen to impress upon us that if he was to move to France at such a young age, then he would want to do so with a fellow Croat to help him settle. Duvnjak, who was playing alongside Šaranić at GNK Dinamo Zagreb also looked like someone that we could sign to develop and then sell on, hitting the Club vision of signing players under-23 who we could flip for a profit. A bid was then tabled for him to, and whilst Duvnjak had his own reservations, his wages demands were reasonable. Both players signed for a combined £5.3m split across four years payments, rising to £7.2m after various criteria are met, largely player performance-related.

In analysing the performances of our three main centre backs, whilst Mesonero is showing the possible ability to make the step up, with the Swiss defender making the most interceptions per nine across all central defenders, Jemelka and Mayembo look to have struggled. Their tackles per ninety and tackle success rates are very poor, with Mayembo’s seeing him in the bottom five percent. I haven’t included heading data within the below player data profiles, largely because it’s not possible to discern between an attacking heading attempt and a defensive effort, but all of the HAC Foot defenders struggled to win the ball in the air. With only three defenders as an option, more depth was required, but if we could improve the quality of defenders then this would be beneficial too, especially with the metrics making it clear that Mayembo has been ground to be lacking at this level.

Scouting reports had highlighted the availability of David Costas who had not accepted contract offers with Celta Vigo. Costas likely won’t necessarily improve us in the air at only 6’0″ tall, but he will add to our ball-playing ability as he is a natural passer. Given our desire to play passing, controlled football, this should allow us to build up from the back. To add greater depth in the left centre back position, Christian Dalle Mura was signed on loan from Fiorentina as a squad option, with the intention that he goes in behind Jemelka to rotate in and out of the squad when the fixture list piles up, with nine games inside January alone after the Winter World Cup.

With these new additions to the squad, we’ve hopefully achieved our objectives in the transfer market to both strengthen and deepen the team to help prepare for the fixture madness of January and February to catch up from the games that have been moved from November and December. This was despite having only been issued a transfer budget of £6.78m – spreading payments out over four years with the use of installments helped us meet our objectives. With an overall spend in the region of £8m with regards to transfer fees, with further in add-ons if certain targets are met, were should find ourselves better resourced to avoid the so-called difficult second season.

As to how we’ve progressed against the Club objectives of signing under-23 players to sell on and from domestic lower leagues, well five out of eleven players were under the imposed age bracket. Where we missed the prescribed goals was not signing anyone from Ligue 2 or below bar technically Nordin from Saint Étienne. It’s not for the lack of player search, there just wasn’t the talent there, or if there was in the relegated teams, they wanted wages well above what we could reasonably offer with our wage structure. We’d already stretched that with some of the above signings. Hopefully this won’t be too much of an issue for the Board. To address the problem head on, two scouts were immediately assigned to Ligue 2, National leagues and the junior age group leagues for a better overview.


The next post in the series will go onto look at the 2022-23 campaign, as well as highlight some youth prospects starting to emerge from the academy, skipping the usual half season analysis. I hope you’ll return for that post. Until then, au revoir et bonne santé.

AC Le Havre 2021-22 Mid-Season Update

HAC Foot Steps

Having been promoted to Ligue 1 ahead of schedule, and having spent only around £4m in improving the squad, the bookmakers had us in the bottom five of the league after we made our additions to the squad. The Board and the players weren’t holding out too much hope when it came to avoiding the drop either, with the expectations largely to fight bravely against relegation. Understandable when our wage expenditure was a Ligue 1 low of £9.63m (see later graphic for more on this).

Being overpowered by our fellow competitors in the Ligue 1 labour market is all well and good, but with an encouraging pre-season behind us, spirits are high. If we can remain tactically solid, then we stand half-a-chance against some of our opponents.

Tactical analysis

The defensive set-up is the most obvious place to start in analysing how HAC Foot are going to go about trying to survive in Ligue 1.

A traditional back five, sees a sweeper ‘keeper with clear guidance to distributor the ball out to those directly in front of him in order to maintain possession, with our aim for purposeful and considered build up. With a more aggressive left wing back and a right back playing as a full back on support provides width in attack but also tactical solidity out of possession because of the instinct of the right back to return back to his defensive position as possession is lost.

Since none of the options at centre back are especially gifted with the ball at their feet, they’re told to play it simple when passing. If they can reduce the risk of turning the ball over to the opposition, they cut the likelihood of the opponent being able to score should they win the possession with only a solitary defender or two between them and the goal. José Mourinho riled against possession statistics after defeat to Liverpool, “It is a little bit like the efficiency of players and sometimes you say: ‘The stats say Player B had 92% of efficiency in his passing.’ But the stats don’t say that player only made passes of two metres, they don’t say that the player was a centre-back who only passed to the other centre-back”. Whilst he has a point, I am still more comfortable with a centre back who can repeatedly make that pass, potentially away from the press, than one who is like a panicked chicken with a fox bearing down on him. This approach has seen us average 57% possession in our games up to the half way point – second only to PSG.

To provide some defensive rigidity, a defensive midfielder operates as a single pivot. However, the role varies depending upon those entrusted with their name in the playing eleven. If either of the HAC Foot stalwarts, Lekhal or Fontaine play, then they act as a deep lying playmaker, looking to recycle possession and keep the ball moving as we methodically look to pick our way through an opponent’s defence. Should we pick youth team graduate, Sébastien Briand, then he plays in his more natural role of a ball-winning midfielder, acting as a combatant, utilising his bravery, teamwork and aggression.

The remaining central midfielders are more set in their roles, a mezalla on the left and an advanced playmaker on the right. The latter has been rotated between another youth graduate, Abdelli, and loanee Hannibal, whilst Scott Fraser has been assigned the mezalla role he was signed for. When on song, Hannibal can take a game by the scruff of the neck and dictate play, looking head and shoulders over anyone else on the pitch. He is only young, but if he can improve his consistency, Manchester United could have a world beater on their books.

Down the left flank, on loan Nordin plays as a winger, utilising his pace and dribbling ability to stretch opposition defences. Dia has taken well to the inside forward role on the right, cutting inside with regularity, either to meet a cross or with the ball at his feet driving at the defence.

This leaves lone front man, Ali Akman, playing almost as a hybrid false nine. His desire to push the boundaries with offside sees him more advanced than a typical false nine, but he will still often drop back into the hole between the defence and midfield of the other side to link up play with, Hannibal, Nordin or Dia. Constantly busy, his off the ball movement and anticipation has aided our style of play.

So far, by and large, it’s a system that it seems is working.

Half-season break results

As you can see from the results and our current League position, things have been going much better than anyone outside and even inside the Club would have initially expected. With the defeats, all but three (Angers, Dijon & Lens) were anticipated and of those, all of those losses were away. Further still, bar the Monaco loss, in none of the matches were we trounced. Not even away against PSG – though admittedly we were playing for the 0-0 and hardly registered a shot, never mind one on target. We were undefeated at home until Lille turned up in November, something I don’t think anyone could have dreamed about.

It’s worth revisiting the start of this review post at this point. Our wage expenditure is just £9.63m for the entire squad, yet we find ourselves 8th.

Having a strong start to the opening fixtures certainly helped bring a feeling of belonging in the top tier of French football. The new signings were clicking well. Hannibal had a goal contribution of five – or one every 198.8 minutes if you prefer; Ali Akman was performing well in the false nine role, with seven goals from fifteen appearances, and Boulaye Dia had also made fast start to life a HAC Foot, with a goal contribution of eleven (six goals and five assists) – registering a goal or an assist every 110 minutes.

This gave the recruitment team the confidence in their talent spotting abilities, with so many of the Summer transfers having seen substantial first team action and doing well. Six of those signed in the Summer had been involved in at least 80% of potential minutes thus far.

Window shopping

Their talents were to be put into action again in the January window. Prior to the closing of the Summer transfer window, Ertuğrul Ersoy demanded to leave because he felt that there was too much competition for places at centre back, presumably threatened by Vaclav Jemelka’s arrival and starting berth. He left for Kasımpaşa for £925k. Without the necessary time to replace him, the decision was made to wait until January and spend the first few months of the season identifying a series of potential candidates within the allotted wage and transfer budget. The situation was exacerbated further still when Yanga-Mbiwa was unsettled by his lack of first team action after his physical attributes taking a downturn after his lengthy last off in the previous season. He asked to be sold – with his contract expiring at the end of the season, an offer of £205k from Auxerre was gladly accepted.

Since the Board were keen to sign and develop youngsters, a number of scouts were given the remit of finding central defenders who were at most 23 years old, not paid in excess of £12k/week and valued at no more than £2m. Below are the profiles that made the final shortlist:

Elias Mesonero: Pros – Top-ranked by averaged percentiles across defensive metrics for those of whom the club had knowledge and who were interested in signing for the club, considered a leader, driven in pursuit of goals, high level of determination, anticipates situations well, likely to be a good fit with the squad, fluent French speaker, 4 U21 caps. Cons – none.


Emin Bayram: Pros – Impressive jumping reach, balanced/normal personality, fairly consistent performer, good in the air, 9 U21 caps. Cons – needs to work on first touch, won’t fit in easily to any social group, would need to learn French and Galatasaray aren’t willing to listen to any offers, whilst he wins a lot of tackles and interceptions, his tackle success rate is <80% and a good deal lower than other potential signings.


Dimitris Nikolaou: Pros – No problems adapting to another country, very brave, balanced/normal personality, enjoys big matches, fairly consistent performer, good in the air, 1 senior cap (20 U21 caps). Cons – high agent fee, poor first touch, won’t fit in easily to social group, would need to learn French, appears to give the ball away too frequently judging by passing efficiency.


Kamil Piątkowski: Pros – Strong player, fairly determined attitude, good stamina, 17 U21 caps. Cons – peripheral figure in social group, would need to learn French, looks pretty solid statistically but falls outside our transfer budget with Raków demanding his minimum release fee be paid up front for permission to speak with him.


Jan Sobociński: Pros – Model citizen, adaptable to living in new country, enjoys big matches, good in the air, good at marking, 12 U21 caps for Poland. Cons – demonstrates a lack of composure, has a competitive streak, would need to learn French and likely a peripheral figure in social group, makes almost as many fouls as he does tackles – likely to be booked frequently and be suspended/cost the team goals from free kicks when matched with his competitive streak.


As you can see, detailed player profiled were put together for each potential signing. Everything from their personality, recurring injuries, likely wage demands to details on their existing contract were assessed, beyond simply their player data – both attributes and metrics.

The stand out from the above and primary target was Elias Mesonero. Mesonero’s metrics shone – with the highest average percentile across all the key defensive statistics considered of not just the shortlisted players but of all players aged 23 or less with 250+ minutes. Other things that were looked upon favourably when evaluating Mesonero was that he had the second best average rating in the Swiss Challenge League and led the league for blocks, indicating his eye for reading the development of play. A bid was tabled, structuring the deal over three seasons to alleviate any cash flow concerns that were brewing, with Grasshoppers Zurich for £1.1m. rising to £1.4m. He joined after agreeing a contract over four and a half year for £5,750/week.

The second part of the process was to agree never to face a situation where a first team squad member was to leave and not have an identified target – either a youth team player to promote or a new player to come in from outside the club. This led to the establishment of a shortlist of players who were added to the ‘back up squad’ list. Players added to this list were deemed of sufficient quality to at least match the existing first eleven, or could have the potential to do so in the very near future. Scouts who were not assigned the duty of finding a replacement centre back were asked to monitor Central and Eastern Europe, as well as Scandinavia and Africa for information gathering with a broad net. Anyone who was highly thought of would then be monitored by the Director of Football, Jean-Michel Vandamme. His overview would help to determine whether or not they were the kind of player HAC Foot would possibly look to potentially sign.

Whilst the Club laid out their desire for young signings, older players weren’t ruled out of being shortlisted, but they had to come in under budget and be a clear and obvious improvement in the current player in possession of the shirt. It was also agreed that the shortlist was to be reviews on a six monthly basis, prior to the opening of both transfer windows so that those deemed not to be of a good fit for HAC Foot would be delisted and new additions brought on following recommendations from the scouting reports.


The next blog post will be the end of 2021-22 season. Will HAC Foot’s season continue on the right path and will Mesonero settle in to life on the French coast? Time will tell.

AC Le Havre 2020-21 – La joie de vivre

Ligue 2 – Fais toujours de ton mieux même si personne ne regarde

After the change in formation to the 433DM Wide, rolling goals against, i.e. average goals conceded over a rolling period of 10 games, plummeted to well below 0.5 goals per game. In part, it was this up-tick in the number of clean sheets which went a long way towards HAC Foot’s rise in the table from a low of 12th to the summit of Ligue 2. It is there that we stayed for a total of eight game weeks, winning Ligue 2.

It’s true that the increased frequency of goals scored meant that we were winning games relatively comfortably, but it’s clean sheets that helped deliver the wins too. In truth, as you can see from the comparison between the rolling G v xG, we considerably overachieved against the number of goals that we should have scored. This overperformance isn’t particularly concerning now. After all, promotion to Ligue 1 when we were only meant to finish 5th is fantastic. HAC Foot’s Board duly offering me a new contract, which I gratefully accept, even putting in a relegation clause that lowers my salary should we be relegated, to ensure the financial security of the club.

The most pleasing aspect of the below radar is oddly probably the pass completion. Setting the team up to retain control of the ball so that we could facilitate chance creation has yielded winning results, off the back of a high shot frequency and positive gap between xG and xGA.

On a game-by-game basis, the difference between cumulative G-xG rose to the extent that the goals we were scored at the back end of the season overcame the deficit that had been created at the half-way stage. This big upswing over and above goals we were expected to score evidentially helped our rise in fortunes. When you combine this with keeping GA under control, then this highlights the secret to success for the side.

The self-calculated xPts rose to over four points above what we should otherwise expected to have achieved. The in-game analysts agreed that we overperformed but their xPts still had us top (72.1 xPts – +4). We were truly the best team in Ligue 2 and deserved winners of the title.

Squad analysis

Taking a look at the squad profile using the number of minutes played gives the chance to review progress in utilising the club’s academy prospects.

The opportunity to blood more home grown youngsters increased as the season went on and rotation options were needed. The change in formation suited one in particular – Himad Abdelli. He became the first choice in the advanced playmaker role on the right hand side of the central midfield pairing. He was on the pitch for a total of 1,621 minutes, and contributed six goals and five assists. Given that HAC Foot have a history of naming youth academy graduates in their captaincy roles, with Fontaine currently the captain and Lekhal the vice-captain, perhaps Abdelli is future captain material?

Planning for Ligue 1

Promotion secured, and news of an £11.53m share of TV revenue from Ligue 1 to come our way over the course of the next season, the Board made a £3m transfer budget available.

With numerous first team contracts expiring on players who weren’t going to make the grade at the higher level (shaded light blue in the above graphic), this was an opportunity to free up the wage budget. This included players like Bonnet who had had his testimonial at the start of the season after 12 years of service. The alleviated expenditure on wages could then be reallocated towards new recruits to come in who would improve the quality and possibly depth of the squad. With Ba and Meddah seeing some minutes as indicated above, but far from ready for Ligue 1 football, they too were placed on the loan list. After assessing the situation with the squad and discussing about the future and the quality of the squad, it was clear recruitment was needed.

Under Chief Scout, Bernard Pascual, the recruitment team have been out looking for possible player acquisition targets and these missions yielded some recommendations that were favourably looked upon.

Four deals had already been agreed prior to the transfer window opening and before it became clear that we were to be promoted – those shaded in purple. Two of the three signings were from Africa, in particular Côte d’Ivoire – Jean N’Guessan and Ibrahima Ouattara – both prospects signed for their release fees of £275k each from RC Abidjan. Both will be added to the development list with the newly hired Loan Manager tasked to find them appropriate playing time at another club to aid their development. The third deal was one that was probably actioned too soon. Matthieu Saunier was signed to provide depth at centre back but does not look good enough to be back up at a Ligue 1 side. Hindsight is a wonderful thing. The remaining deals were going to have to be both shrewd and more considered.

The other pre-agreed deals that comes under that category was for Lionel Carole, who joined from Strasbourg. He came in to provide competition in the left back role for Meraş, who was largely unimpressive following an injury hit year. The 30-year old French former youth international signed a two-year deal after his contract expired. A well-rounded player with no real weakness to his technical, mental or physical attributes, Carole adds experience to the youth that’s already been added to the side.

A deal was struck with Manchester United to bring in 18-year old French midfielder, Hannibal. The French U20 player will slot straight into the advanced playmaker role as an upgrade on Abdelli, with ratings of 15 for a lot important attributes. Paying only £5k of his wages seems like a bargain for a player of his talents. His technical attributes make him stand out from the rest of the squad – his ambitious mentality and consistency should hopefully provide him with the drive to succeed at HAC Foot.

With Romain Basque the main option in the mezzala role, Scott Fraser was brought in on another free transfer to step into the starting eleven. Being post-Brexit, Fraser counts as a non-EU player (something that’s quite that’s difficult to achieve when you look at the list of countries that are treated as EU citizens by the French). The left-footed Scot offers flexibility in playing positions given his versatility, but it’s hoped he will form a good relationship with Hannibal and either Lekhal or Fontaine in behind. Whilst not blessed with acceleration, his willingness to try high risk/reward passes and penchant for arriving late into the opponents box mark him out as a good fit for a mezzala.

Ali Akman is another teenager brought in to strengthen the first team. As previously stated, the Club is seeking to exploit transfer markets where there are ‘wrinkles’, and Turkish scout, Burul, highlighted Ali. Goals from the front man during the previous season were hard earned, and even when they did come, they came from on-loan striker Simon Banza (whose loan was renegotiated for another year as a precaution, after some initial reflectance from his parent club). Therefore, a player with not only a higher ceiling but also an improved ability in front of goal was sought. Signing on a free transfer, after being not agreeing a contract renewal with Buraspor, 19-year old Ali has the ability to play as a false nine, which could be a pivotal change in HAC’s approach play, enabling Ali to drop into space in front of the opposition back line whilst the two attacking wide players push on beyond him. Despite his young age, he looks a talent destined for bigger things than here at HAC Foot.

Speaking of wide players, new recruits were added here too. On the left, Arnaud Nordin was signed on season-long loan from Saint Étienne to provide competition for Pité who may be moved across to cater for the other attacking acquisition – more on him soon. Nordin is relatively quick and a good dribbler and although he will start out in the winger role, he will look to cut inside towards goal. Given the switch to a false nine, this should increase the threat upon goal beyond the traditional wide man. His fear of big matches is a concern, but overall consistency and relatively low cost of only £6k/week mean that he will be a valuable acquisition for the team.

The other attacking player signed is none other than Boulaye Dia. Signed from Reims for a bargain fee of just £900k, Dia hits the club’s ambition to sign players under the age of 23. Heavily backed to do well in real life, Dia looks like a good player in the making. Rather than playing up front through the middle, he will be trained to play as a right-sided inside forward. His pace, strength and finishing ability should enable him to score goals at this level.

At right back, improving the quality of options was more important. With transfer fees limited as previously stated, the search turned to more unfancied markets, in this case, Scandinavia. BK Häcken’s Nigerian defender, Godswill Ekpolo had caught the eye of scouts with his solid defending ability, work rate and physicality. The most expensive deal done over the summer window, Ekpolo arrived on a 5-year deal for £1.5m, rising to £1.7m after appearances, broken up in a series of payments over the next four years.

A left-footed centre back was missing from the first team squad and a player search using attributes for a central defender. One of the players who cropped up within budget was 26-year old Vaclav Jemelka. A naturally fit and physically strong player, with good positional sense and tackling, the Czechian was signed for £975k, rising to £1.2m, from SK Sigma Olomouc. He is somewhat limited with regards to his mental attributes, fingers crossed he won’t be exposed too often due to his poor decision-making and anticipation of what is going on around him.

After a busy transfer window with no fewer than eight additions to the first team, and more to the wider squad/HAC 2s, the squad looked like this as the season began.

Using some relatively basic accounting principles, using amortisation to divide the cost of the transfer over the duration of the contract, the total basic expenditure, before player agent fees, loyalty bonuses and player performance/appearance bonuses, all of the players brought in cost in the region of £4.6m for the first (or only) year of their contract. Less than £5m spent on, hopefully, improving the team for Ligue 1. Thanks to spreading the payments over several years, this still left some funds in the transfer budget of the original £3m for the January window, should we need it. Think on this when PSG have signed Harry Kane.


The next blog will review the half-way point of the Ligue 1 2021-22 season – time will only tell if HAC Foot have managed to get themselves in through the door of the top league in France.

AC Le Havre 2020-21 Mid-Season Update

This mid-season update sees a review of the transfers that were made in the opening transfer window and how the season has progressed thus far at HAC Foot. There’s also a sneak peak at the youth intake.

Transfer Methodology & Practice

With a limited budget and no room in the wage budget, the available transfer funds were reallocated towards wage spending. A clear focus on transfer dealings was needed – no spare cash means next to no mistakes can afford to be made. As such, free transfers and loans would be the chosen methods of bringing players in and bids would be considered for players who were considered surplus to requirements.

As shown on the previous blog, HAC’s squad is a relatively young one, with players who have come up through their youth academy being given the chance to play first team football. The identified youth prospects that are in the wider squad don’t yet look ready for first team minutes. Only Meddah and Fofana have played any minutes this far, with Fofana seeing action in an opening day 3-1 defeat to Châteauroux. The youth prospects will be kept in the reserves to play together or given the chance out on loan to learn from first team action where possible.

With some hope that these players will be ready to represent HAC in the future, the best option is to bring in players on relatively short-term contracts, which have the added bonus of protecting the club if the recruitment were to fail. Accordingly, where possible, incoming players will be given a two year deal maximum. With gaps in terms of the level of experience in the side, older players are not automatically ruled out given the increased likelihood of these players being without a club.

Some additions have been made to the squad in line with the above restrictions that were self-imposed. Mapou Yanga-Mbiwa and Dorian Dervite were both added to the squad to provide a challenge to Mayembo and Ersoy. As the only two centre backs in the squad and with no youth prospects to promote in this position, depth was required. Neither player wanted eye watering contracts and appeared better than other options available after Simunovic turned the club down. Fans of Newcastle United, Roma or Lyon should recognise Yanga-Mbiwa’s name – sadly, he broke a leg during one of his first games for the club and is yet to regain his fitness. Dorian Dervite’s name may be familiar to Tottenham supporters as he came up through their youth team and earned caps for France’s youth teams. However, it’s likely that Bolton Wanderers fans will be more familiar the Frenchman. Dervite’s time thus far at HAC has proved to be less than successful as the centre half has made a number of mistakes and has found himself on the bench more often than in the starting eleven as a result. As stated earlier, mistakes cannot afford to be made with transfers so time will only tell how costly these ‘misses’ will be.

The best ‘hit’ signing in terms of output would have to go to Pité. Picked up on a free, the Portuguese player wanted £11k/week, which is more than any other HAC player. However, his qualities have shone through and are seemingly more than justifying this outlay. With six goals and four assists, he has been a big upgrade on any wide player already at the club. As his minutes played demonstrate, he’s already become one of the first names on the team sheet. A versatile player, he has been playing as an inside forward on the right wing, cutting in on his favoured left foot and appearing at the back post to finish crosses from the left. His conversion rate of 12% is pretty good, and bested only by two of the other forward players in the squad.

Yannis N’Gakoutou is the other permanent addition to the side, again being picked up on a free transfer. Whilst initially unfancied by the coaching staff, his attributes appear well suited to the wing back role on attack, at least if you ignore his lack of natural fitness. He’s so far contributed a number of useful crosses, taking advantage of the space in front of him created by Pité cutting inside and dragging the opposing defender with him.

Goals were seemingly impossible to come by for HAC in the early part of the season for striker Thiaré. With an xG of only 2.68 in 901 minutes and no goals scored, Simon Banza was brought in on loan from RC Lens after impressing off the bench with three goals in three substitute appearances. Banza has added more dynamism to the front line with more movement off the ball and has been instrumental to a change in fortunes at HAC since his arrival (more on this later). Within 487 Ligue 2 minutes, he’s already scored five goals, with an xG of 3.31 – 0.78 goals/90. His goal conversion rate of 30% is exceptional – scoring his goals from just fourteen shots. Fingers crossed Banza’s performance can be maintained over the rest of the season.

£215k was raised from player sales, the bulk of which came from Khalid Boutaïb (£95k – KV Oostende) Ayman Ben Mohamed (£76k – AC Ajaccio) and Abdelwahed Wahib (30k – Châteauroux). All of these players were adjudged to be surplus to requirements due to better players being available in their respective positions.

League Table – on the right track?

As you can see from the graphic below, it hasn’t always been plain sailing for HAC over the first half of the season. With intermittent form, a loss against Guingamp saw HAC slip to 12th in Ligue 2 – well off the target of play-offs.

The gap between the rolling number of goals for and against, both in terms of actual and xG/xGA, was narrowing quickly and so a tactical switch was implemented. Shifting away from the 4-2-3-1 to a slightly more considered 4-1-2-2-1/4-3-3 DM Wide saw a rise in the number of chances being created. With the ball being retained more easily, with wide and advanced players having the safe go to option to play back to the defensive midfielder to re-establish an attack, fewer balls were being forced and lost, leading to fewer turnovers and counter-attacks from opposition sides. If you take a look at the below graphic, the 4-2-3-1 was actually hurting the side, with opponents creating more clear cut chances per 90 than we were creating for ourselves. The bottom right part of the graphic shows the last five games under the 4-3-3 DM Wide – only one goal conceded in the last five games with an xG of just 2.55 – a significant improvement in both the quality and quantity of chances being created against us has improved our standing in Ligue 2 no end.

This adjustment might look defensive, taking a player out of the attacking central strata, but it actually added to the attacking fluidity of the team because there was now more space ahead of them to run into and meant that the opposition defenders now had to mark players that were running at them rather than those trying to find pockets of space. Given that many teams were sitting catenaccio-type approach against HAC, space within this area was at a premium. This also helps to explain why Thiaré has such a low xG, the players around him were struggling to progress the ball through to him to take advantage of his elite speed and put him through on goal. The under performance against xG is also explained in the below radar graphic demonstrating our performance against the Ligue 2 average. Clearly we’re above average (in a good way), which of course is explained through our league position. Our xG/Game to Goals/Goals drop off is a concern though.

Indeed, using the in-game graphics, our analysts have us down as being shot happy but wasteful, which is curious given our tactical instruction to work the ball into the box. Perhaps our players grow frustrated by the defensive low block in front of them and then shoot front distance. Yet when we do score, it’s from within the box, with just two exceptions (see the bottom right of the graphic). You can also see how important the right hand side and central parts of the pitch are to our build up with the frequency of assists from these areas (see top left of the graphic).

As part of the move to 4-1-2-2-3/4-3-3 DM Wide, some player roles were also changed, with the deep-lying playmaker who previous sat on the right side of central midfield switching to an attacking playmaker on attack, and the advanced forward changing to a pressing forward to reduce the gap between attack and midfield. This helped to avoid the forward becoming isolated in the build-up play, linking up a little more with those around him, as well as pressing opposition defenders to stop them lumping long balls for their attackers to run onto over our high defensive line.

Whether playing 4-2-3-1 or 4-3-3 DM Wide, we haven’t faced many shots per game, but have seen opposition forwards score at a rate of just shy of 14 shots per 100 taken. This may not sound bad, but that’s the second worst in Ligue 2. The fact that we don’t face many shots is therefore a blessing. Our mantra of keeping hold of the ball in possession is clearly a necessity to reducing the frequency of shots faced. Through balls and set pieces account for 63% of the goals conceded – fingers crossed the changes made to the tactic help minimise the regularity of both, with an extra player in the defensive midfield pivot.

Despite this, the gap between the xG and the actual number of goals scored is somewhat of a concern. Hopefully, as players adjust to this tweak to the system and adapt to their new roles, the frequency of goals will rise and the under performance against the xG will disappear.

You would also think that having seen fortunes seemingly improve following the change in tactic that the expected points (xPts) would be broadly in line with actual points, and to a degree, you’d be right. Mikel Arteta would be relatively pleased. Yet according to some calculations taking from here, using both xG and xGA then looking at the difference between the two across each game, there is still some improvement to be had from this HAC side. The change in tactic does appear to have steadied things from the precipice that they were headed down following the defeat to Guingamp, but the side are still light around three points from where xPts has them at the half way point in the season. Only an over-performance against both xG and xGA will see that gap disappear. With both on the ‘wrong side’ thus far on a cumulative basis, this does not look likely to be eradicated any time soon.

Indeed if we take things a step further and look at cumulative G-xG and GA-xGA then we see a chronic under-performance in terms of goals scored against what the metrics tell us that HAC should have scored. Around six goals more should have been scored. Additionally, the team has conceded two extra goals than they ought to have done given the probabilities. A swing of eight goals over a nineteen game period could have had a dramatic impact upon HAC’s fortunes over the season so far. If they are to reach their goals of teaching the play-offs, it’s likely that they will need to rectify their relative waywardness in front of goal and look to be more clinical.

Youth intake – first glance

Looks promising given that this is rather the point of the save. Let’s hope it’s as good when it comes to March.


The next blog post will focus on the above youth intake.

FM21 Is Coming

For those that followed my AC Milan save in Football Manager 2020, you may have been wondering who I’ve chosen for Football Manager 2021. Keep reading to find out what I have decided.

Data Analysis within Football Manager 2021

Given new data available in FM21 compared to previous editions, I am keen to further integrate this into my writing and review of my save game. Those that read through my FM20 save with AC Milan will know that I took a deep dive into the data analytics around player recruitment, squad performance, but also within a tactic and across the league in general.

The new staff analysis roles, especially the Head of Performance Analysis, pretty much mirrored the approach I adopted with my writing and this year promises to be different only by the level of data available to me. We know that xG is in FM21 and of course I will be including that into my analysis of strikers/wide forwards when recruiting players and analysing my own players. It’s great that the conversion rate (the goals/shots as a percentage) is in there too as this is something that I had to calculate within Excel myself. I may also look to use this to try computing the difference between Expected Goals from Shots on Target and Goals conceded to assess goalkeepers. The below radar graphic is certainly appealing and a fantastic addition to the game, but as yet, doesn’t fully cover all the data that I would like to be available.

The shot map is also an encouraging edition to the game and I’m also keen to see whether these will be available just on a game by game basis or whether we will be able to map multiple games onto the same shot map to assess a team’s shot quality over a multitude of games.

I do have a mental list of different metrics that I’d like to be in the game, and it’s good that Sports Interactive have disclosed that there will be additional data included, such as clearances for defenders on their FMFC site. Yet there’s no mention of xA (Expected Assists), which is somewhat curious and disappointing at the same time. Hopefully, if it’s not included in this edition of the game, it will be in the future editions (during the live stream on Twitch, Miles did say that they were looking to fold in more metrics going forwards).

As such, I do expect to be producing my own data analysis from within the game to create a method of analysing player performance, team performance and tactical analysis once more. This has helped to improve my use of Excel and PowerPoint no end, so it’s been pleasing to develop my own skill level as well as becoming a FM content creator. So you can expect more graphics like the ones I created below:

The last graphic in particular is one that I’m keen adapt to utilise the xG in some way, perhaps to assess the xG differential between the two teams as a crude method of xP.

With all this in mind, I presumably have to be at a club that has a good history with regards to data analytics to ensure that the analytics facilities and staff are up to scratch. With AC Milan I did manage a fallen giant within the European game, and I’m not against that idea again, but I’m looking for metrics to play an even greater role in my save than it did in FM20.

With the beta, I’m tempted by a save with Brentford for fairly obvious reasons – their analytics team have unearthed a number of ‘low hanging fruit’ players via metric analysis and highly detailed scouting, offering the players a clear pathway and acknowledging that they will be willing to sell players if the offers for them are to the benefit of the club. Equally, I’m tempted by a quick save with AS Roma or Fiorentina as a number of their players were regular transfers I picked up in FM20, so it would be good to manage them ‘at source’, whilst learning about the new metric analysis and how to integrate that into my playing and writing style.

With regards to my ‘full save’, I’m going to dig around a bit more and see what smaller teams in non-English (or indeed most likely non-UK) have good data analytics, or at least enough cash in the bank to improve these facilities. Cash may of course be hard to come by for many clubs, with the impact that Covid-19 will have had upon the cash flows of most teams that aren’t in the very elite leagues. Therefore, I intend to be very selective and take my time upon making my decision. I don’t mind saying I’ve considered teams in Italy, Spain, Austria and France – time will only tell if I choose one of those nations to manage in.

I would not expect to be able to progress too far into a save game – being a teacher and my ability to play the game is very much dependent upon how heavy my workload is. Only teaching exam groups and having to manage balancing teaching pupils both in and out of school at the same time has doubled my workload, so I can’t expect to be progressing twenty to thirty years into the future. The SI guys on the first look stream did say that the game processes much faster now, and Miles has said something similar on Twitter, but I don’t expect that will make much difference to me.

That said, I would like to be able to have some good youth prospects come through my academy this year that I can actually develop and then use. In FM20, I only had one newgen player who was nearly good enough for the first team, not with AC Milan but with a Liverpool save I started after my AC Milan save ended. With around 10 years played in game across two saves, this was a big disappointment. However, ultimately he never displaced Alisson from the starting eleven. So a club with good youth facilities and at a level where it’s easy/easier to bring them through would be high on my list too. Therefore, I expect to spend some time researching around different leagues and seeing what I can find that fits with what I want to do.

Equally, I have a moral compass. So, for instance, I will never manage Lazio – their far right fans do not sit with my political view points and so I can rule out quite a number of clubs or fans who follow similar political standpoints.

In other words, yes – I’ll be blogging Football Manager 2021, but as of yet, I don’t know who I’ll be managing or when my blog will launch with a true save reveal. Keep your eyes out around December/January is my best guess – sorry it won’t be before then!

AC Milan – 2022-23 Season Review: La Risoluzione

You read this after the 2022-23 Champions League Final. AC Milan have just lost to Manchester City at Wembley after a 119th minute extra time goal by Bernardo Silva secured the trophy for the Citizens, after Piątek had missed a penalty in normal time. Within two days AC’s double scudetto-winning manager was fired. Why, you ask? Let’s find out.

Numerous slips saw their Serie A points tally fall from 90 to 76, costing them the scudetto, which returned to the trophy cabinet at Juve’s Allianz Arena. A sixth place finish was not enough for a place in the Champions League, with the Club having to settle for the Europa League in the 2023-24 season.

La Difensori

Before the opener against Perugia, Mattia Caldara damaged his cruciate ligament. Caldara had been a stalwart for the team over previous seasons but was going into the final year of his contract, with AC officials said to be considering negotiating a new deal for the 29-year-old Italian international. News from the medical department that he was expected to miss between seven to nine months, and, as it turned out, he missed the entire season. This left the club with no option but to effectively write off the final year of his deal and look to quickly moved on in the transfer market, following the sale of Mussachio just days prior. This led directly to the acquisition of Corbo from Bologna in the Summer transfer windown.

Consequently, the Club were left with captain Romagnoli, plus new boys Upamecano and Corbo as the main central defenders, with Wöber able to deputise at left centre back should he be required.

In analysing at least one of the contributing factors of what went wrong, focus can immediately be placed here – la difensore centrale – the central defence.

Their apparent inability to engage in tackles, saw the trio of main centre backs complete an average of just 1.09 tackles/90 across their minutes played – Upamecano – 0.84, Corbo – 1.10 and a Romagnoli – 1.33 (see below for the images or click on the hyper-linked player names above for click-able graphics). With the level of interceptions per 90 just as unimpressive, opponents were able to build attacks and run at them, seemingly secure against being dispossessed by either of the centre back pairing no matter who played. Whilst their ball playing capabilities were almost without parallel across the top five leagues, their defensive metrics ranked close to, if not at, the very bottom.

Whilst Technical Director, Paolo Maldini would argue that to make a tackle is to already have made a mistake, it’s hard to accept that this is the case here. This cataclysmic failure to defend against attacks by making a tackle, or intercept the ball by skilful anticipation, reading the game playing out in front of/around them, saw AC Milan’s goals conceded tally rise by nine compared to the previous campaign.

Widening the focus out to the rest of the back line, the frequency of mistakes makes for some shocking reading. Fullbacks Cucarella, Calabria and Conti averaged an alarming 2.39 mistakes per 90, or 37.93 minutes per mistake. When a team can’t rely upon the flanks of the defence not to make a mistake during a half of football, then the team is liable to be at risk of seeing their play break down. As a consequence, the opposition are presented with an opportunity to act upon those failures, in this case in wide areas, and look to strike. The turnover of possession and the inability to complete an interception or win a tackle/header saw teams play against I Rossoneri without the accustomed fear that had been there over previous seasons.

As a team, AC Milan completed merely twenty five fewer tackles compared to the previous season, but crucially, their tackle win percentage also dropped off by 2%. Donnarumma conceded 0.63 goals per 90, compared to 0.36 the season prior, whilst his minutes on pitch per goal conceded went from 248 minutes/goal to 143 minutes/goal. Shots faced/90 was roughly comparable to the 2021/22 campaign, but his save rate/90 fell, as did the number of shots he held, indicative of his drop-off in handling as his shots parried/90 rose. This means opposition attackers may have had opportunities to score from easy tap ins with the goalkeeper out of position to respond to the shot.

This, at least partially, was behind the nine additional goals conceded. It culminated in eight losses, six more than the previous year. Below is a graphical representation of the season, indicating the minute goals were scored by AC Milan players and against them. Whilst AC did record twenty-two clean sheets, six out of the eight defeats were by a single goal. Three of the eight were also from a winning position, perhaps signifying an over-confidence amongst the squad and the inability to close out a game.

With games having to be scheduled around the Winter World Cup of 2022, domestic and European fixtures were crammed into the early season and January to allow for the fact that no club football would be played in November and only one game in December after Christmas. With a number of players away on international duty at the World Cup and lacking fitness on return, there is perhaps the two points collected out of nine following the return of Serie A should be looked back upon as the downfall of AC’s campaign.

Using a ten-game rolling average of goals scored and conceded highlights the rough patch that Il Diablo went through just before and after the World Cup. The rolling average goals scored drops off, falling to just over a goal a game midway through the season, with only a marginal difference to the rolling average goals conceded. It is this stage that the damage was done, as shown by the league position. By March, AC were 7th and adrift of the Champions League qualification places and had lost any hope of retaining their title.

So what else went wrong? Why were they more open? Misfortune or poor tactics?

Stile di gioco

As hinted at in the last blog post, following the new board’s recruitment of Declan Rice, the coaching team felt compelled to play the England international in his preferred position in the defensive midfield strata. As a result, the team moved away from the nominally flat midfield three by shifting the central deep lying playmaker back into the pivot. The issue with doing this was that this subsequently allowed the opponents more time when in transition to stabilise possession to launch attacks as they had more space within the central midfield zone, providing them with more opportunities to make defence splitting passes, either through or over the top of the back line.

Rice’s arrival meant fewer minutes for Tonali, who over previous seasons had been the regista of the team, if not in actual player role then certainly in reality. The problem with Rice is that he was neither a regista nor a mediano, in the style of say former AC Milan player and manager, Gennaro Gattuso. He lacked the guile and vision to be a top level playmaker and the ‘palle’ to be an aggressive stopper.

Looking at the below graphic, the lack of creativity from midfield becomes apparent. The paucity of chances created by the central midfielders underlines the lack of playmaking by these players. The low frequency of passes per ninety amongst the more attacking/advanced central midfield players could be indicative of the fast verticality of AC Milan’s play, but equally it could be a sign that when the play slowed down in attacking phases that these players weren’t able to find space to be involved in play. Bruno Guimarães’s metrics in this area are notable – to have averaged 44.39 passes/90 but to have only created 0.18 chances/90 and made five assists speaks volumes for someone who is meant to be offering himself in central positions when in the attacking phase.

This is where the move towards the single pivot with Rice could have been a critical flaw with AC’s tactical set-up – allowing the two more forward central midfielders to sit narrower and increasing the gaps between their wide forwards, resulting in opposition midfielders have more chance to place themselves in the passing lanes so that play was broken up. Consequently, this would help to contribute to the 25% fall in chances created from the season prior. This theory looks to be supported by Dani Olmo, Lucas Paquetá, Hamed Junior Traorè and Calvin Stengs (metrics are available for each by clicking on their names, bar Stengs’s whose is below).

Stengs was the sole central player who created above 0.4 chances/90, and this was only enough to find him in the bottom 40% of all central midfield players across the Big Five leagues. With Stengs completing a meagre 25.03 passes/90, his assists/90 metric is somewhat remarkable but you have to wonder if he could have been supplied the ball more by his teammates to unlock his creativity. Yet his passing completion statistic of 80% again indicates that the creative players in AC Milan often do lose the ball when trying to deliver key passes.

Indeed average possession and passing accuracy fell compared to previous seasons as players were more wasteful with the ball, seeing loose passes being cut out by the opposition, with Piątek once more choosing to isolate himself despite clear instruction to be more engaged in the build up beyond just throw-ins and shooting on goal. This is visible in his pass completion per 90 metric, completing only 14.56 passes per 90. His NPG/90 was lower than in last campaign, seeing his average drop below 0.50 NPG/90. His goal conversion (goals to shots ratio) fell too to 12.99%.

It’s obviously impossible to know the road not travelled, but it’s hard not to think about what might have been had Suso not been sold to PSG. Have the team missed his creativity and set piece delivery?

Perhaps Buendiá wasn’t given sufficient minutes given his apparent output, but it was Chiesa that impressed more in the advanced winger role down the right wing. Chiesa looks to have been a shinning light in what was otherwise a dour Serie A. His goal conversion of 16.83% placed him just outside the top 10% across the top five leagues, with seventeen goals and his seven assists, giving him a scoring contribution of 0.86 NPG&A/90. Not quite Suso levels of the past season, but still exceptional, so it’s hard to blame the Italian international for his part in proceedings.

Czech youngster, Hložek, had a good session too given his relatively young age. At just 20, his dribbling and tackles and interceptions combined demonstrated his high level of workers for the team both in the attacking and defensive phases of the game. His NPG&A/90 was in the top 20% of all players in attacking wide positions across the Big 5 leagues in Europe at 0.57 non-penalty goal involvements per ninety. If he was to be criticised, his chances created for his team mates could have improved, especially given his attacking intent from his dribbling exploits. By being selfish and shooting himself, he missed providing opportunities for his team mates who were in potentially better spaces to score. His understudy, fellow youngster, Gabriel Veron, played fewer than 1,000 minutes so is not analysed here as a result of the small sample size.

The curious thing is that AC Milan actually scored the most amount of goals in Serie A during the 2022-23 campaign then across the rest of the three year tenure – 78 – despite creating the fewest chances – 89, or 2.34/90. Shot efficiency also peaked at 10.36%. Therefore, it looks to be the case that the jump in number of shots not being held and defensive errors played a bigger part in AC Milan’s failings.

Reclutamento e fidelizzazione – nessun professionista senior, nessuna prospettiva per giovani

Something worth highlight is the lack of AC Milan youth coming through from the academy but also how unsuccessful the scouting set up appeared to be. With regards to the scouting, Chief Scout Geoffrey Moncada had been tasked with being in charge of organising the scouting team on the search for potential targets. Yet any recruitments that were brought in were identified by the manager spending time looking for players. The lack of appropriate targets being suggested by Moncada, with the same names put forward time and time again frustrated the coaching set-up. Whilst the scouting team were sent out to assess the managers targets, they weren’t producing their own suggestions who were likely to be signed by I Rossoneri.

The problems with recruitment and retention could have been another contributing factor towards the sacking. Whilst AC Milan are stacked with players in their prime years, this is also perhaps a weakness. An imbalance across age ranges meant that the squad lacked the guidance, guile and leadership from older, wiser heads who are experienced in seeing games out and dealing with pressure situations. With Andrea Conti being the oldest player in the squad at 29 speaks volumes. Whilst the Club board had instilled a vision of not signing players over 30, this should not have prohibited the manager from retaining older players who were in the inherited squad. Bonnaventura was jettisoned immediately, despite being the one of the most creative players and his replacement, Sandro Tonali, played fewer than 50% of the minutes over the course of the season. Whilst the manager’s hand was forced by the Board’s signing of Declan Rice, it’s clear an oversight not to have retained older players to act as mentors, ensure that the mentality of winning games ‘ugly’ when the opposition are breaking down the play or sitting in a low block.

Equally, the Club’s academy produced no players of potential that were ever close to being in with a chance of being first-team ready. Milan Academy’s initial hope, Daniel Maldini, lacked the progression in attributes to ever make the squad, never mind the first eleven. Not since Gianluigi Donnarumma’s emergence some six years previous, have the Academy created any great prospect. Whilst recruitment was targeted towards remedying this issue, with the likes of Hložek, Tonali, Corbo, Gabriel Veron and Junior Traorè being brought in to add youth to the squad, to not have had any home grown players to supplement these additions must have been a source of great disappointment to those in and around the Club’s hierarchy.

In truth, given the expectations of new board – having spent £196m in the previous Summer transfer window, which yielded no trophies nor a Champions League place for the 2023-24 season – it’s can easily be argued that the sack was warranted.

Ultimately, the Club Directors decision to fire their manager left him looking for a new job. It would be a year before he found a position he thought worthwhile taking, and whilst he dabbled with the idea of another rejuvenation project, trying to help Sampdoria return to Serie A. Instead, it was Juventus that came calling for his services, recognising a serial winner of Serie A when they see one. The road between Milano and Turin is one that has been travelled before by managers as they look to further their career at I Bianconeri, with the likes of Trapattoni, Allegri, and Capello (via Roma) having previously made the relatively short trip across the North of Italy. Tasked with restoring Juve’s status as the top team in Europe, this was a grand job and provided the opportunity to demonstrate to the AC Milan board that they had been too swift in their dismissal.

This marks the end of the Football Manager 2020 blog series, I hope you enjoyed reading it.

AC Milan – Season Preview 2022-23

Before the new transfer window opened, AC Milan owner, Paul Singer announced his retirement, triggering a sale of the Club. This news was somewhat of a surprise in terms of his retirement but it was not a total shock as takeover rumours had been doing the rounds for some time. Indeed if you have read the preamble on this series, a takeover was very much expected early on in this save, as it was anticipated that Singer would likely want to divest AC Milan from his investment portfolio.

Two rival consortia vied for ownership of the Milanese Club, with an Italian-led group winning out backed by the existing chairman, Paolo Scaroni. Upon news of ownership switching to Scaroni being released, he vowed to invest into the Club, putting forward an additional £42m into funds and investing into the youth training facilities. Disconcertingly, in taking over control the Club, Scaroni also stripped out all of the directors who were AC Milan through and through, including Zvonomir Boban and Franco Baresi, leaving only Paolo Maldini behind as Technical Director. This is particularly curious given Scaroni has been the chairman of the Club since 2018.

Additional to the successful takeover, bids were also made for Declan Rice and Dayot Upamecano. These were done without any involvement of the AC Milan manager or the transfer committee that had previously signed off on deals, with the new board seemingly keen to please I Rossoneri’s tifosi with some headline signatures.

The announcement of these transfer bids and subsequent negotiations were met with a degree of incredulity by AC Milan’s manager. Declan Rice was not a player he felt the side needed, with Sandro Tonali the go-to-guy for the central pivot in midfield and throughout the last season. Additionally, AC Milan had not played with a player in the defensive midfield strata, the only area where Rice is a natural. As such, this signing could risk squad imbalance and upset squad harmony as Rice could very well not play the amount of minutes he might expect according to what is promised in his contract negotiations. Combine this with a British/Irish player coming into a league where he does not speak the language nor understand the culture, this could represent a very poor investment, particularly given the size of the fees involved.

His metrics, set against players aged 27 or younger, with 1,000+ minutes across the Premier League, La Liga, Serie A, Bundesliga and Ligue 1, are relatively underwhelming, bar his pressure adjusted tackles and interceptions – his raw figures (pre-possession adjusted) are actually very impressive, but West Ham’s 46% average possession impacted them heavily. This perhaps goes some way to explaining his passing and key passes numbers, which were well below Tonali’s, although Rice’s chances created were vastly higher. So if AC Milan are to look for someone to break play up, perhaps against stronger opposition, he could be the pick, albeit a reluctant one.

To an extent, there was more understanding toward the bid for Upamecano. The metric analysis of the squad’s performance had highlighted a potential for improvement within the central defensive partnership, with Romagnoli having to do the bulk of the work. However, at £54.75m for the overall transfer, plus signing on fee and wages, there was some doubt as to whether this is a signing that represents good value for money. When comparing Upamecano’s defensive metrics against other central defenders in the top five leagues who played more than 1,000 minutes, he comes out well in his ball playing abilities and in winning the ball in the air, at least in terms of aerial duals as a percentage. Yet his heading data also highlights that he was somewhat below average for central defenders in frequency of aerial duals. After further inspection, to rule out styles of play in the Bundesliga, Upamecano’s aerial challenges are sitting at edge of the bottom 20% – evidence he could look to compete more in the air, utilising his not insignificant strength. Nonetheless, he does look something of an upgrade on both Caldara and Mussachio.

Both deals were confirmed by the new board without any oversight from the transfer team, complete with a 5% transfer fee sell on clause to each player. Only time will tell if these new signings can mesh into the AC Milan system or whether tactical adaptations will need to be made to fit them into the starting eleven.

No high potential youth prospects had come through the AC Milan Academy over the last three years, with no youth recruits deemed anything like good enough for first team minutes, as indicated by the below graphic of player ages and minutes played across the 2021-22 season. As a consequence, the transfer committee recognised the need to focus on young players with room for growth when bringing in players to provide depth to the squad. This will help support those players who are in their peak years, giving them a rest, whilst developing new talent for the future.

Further strengthening of the central midfield was identified as a priority to provide greater squad depth in the box-to-box role, behind Bruno Guimarães. Sassuolo’s Ivorian international, Hamed Junior Traoré was the main focus. The 22-year old played 36.98 90s for relegated for I Neroverdi, and showed some reasonable promise. His total potential transfer fee of £54.75m was probably too much, especially given Sassuolo’s relegation, but for such a well-rounded young player with potential to develop his game yet further. This could be a good deal for the Club if his game was to kick on when surrounded by better quality players in a more aggressive side/system than Sassuolo – a team which managed only 29 goals (2nd lowest) from 474 shots (by comparison, AC Milan managed 69 goals from 811) and conceded 55.

The next transfer business was to sell both Rodrigo De Paul and Mateo Mussachio. With Upamecano coming in, Mussachio was now surplus to requirements, especially with twelve months left on his contract. He went to Valencia for £7.5m. Rodrigo De Paul had proved over the last two seasons that he was out of his depth at AC Milan, with a poor statistical performance. He was moved onto Napoli for £15.5m.

With just one year left on his contract, and staunchly unwilling to sign fresh deal whilst wanted by PSG, Suso, AC Milan’s Player of the Year in 2021-22, would either have to be sold to raise funds for reinvestment into the squad or leave on a free at the end of the season. Given his transfer value, it was decided it was best to cash in on his value, despite his importance to the squad. A deal was struck with PSG which netted the Club £55m upfront with a further £15m in possible bonuses.

In order to replace Suso, rather than opt for an out-and-out replacement, it was decided that Chiesa should be given the chance to switch flanks and play as a winger out on the right. As such, another left-sided player should be brought into play back-up to last year’s emerging break through talent, Adam Hložek. After some extensive scouting, with impressive reports, and with a free non-EU player slot at start of the new season, Gabriel Veron was signed on a four-year deal for his release fee of £18.5m. The 19-year old had been impressing in the Campeonato Brasileiro Série A for Pamerias, with 0.5 non-penalty goals per game in the 23.59 90s played so far in the 2022 season. These metrics stack up well against all players in the same position from the aforementioned top five leagues and also the Brazilian Série A. With an outstanding 0.29 assists per 90, it is hoped that he can push Hložek for his place and make the step up to a tougher European league. If he can quickly adapt to the Italian style of living like so many Brazilian’s before him – with many having played for AC Milan including Cafu, Kaká, Ronaldinho, Dida, Robinho and the Ronaldo – he could be a star there for the next decade. Fellow countrymen, Paquetá and Guimarães should help him to adjust to life playing at the San Siro.

With a dearth of potential back-up left back options available which would allow Wöber to slot in behind Romagnoli as back up left-sided centre back, it was thought best to bring in a young central defender who could develop at the Club under the guidance and mentorship of club captain, Romagnoli. 22-year old Gabriele Corbo had long been impressing the Club scouts and analysts with his performances in the Serie A, despite his relatively young age for a centre back. His ability to read the game is reflected in the PAdj interceptions and dominance in the air is shown by his headers won per 90. His key tackle metric was similarly impressive, leading the Serie A table for this statistic, and will likely have gone some way to Bologna keeping twelve clean sheets over the previous season. If he can improve his ball retention and look to be more comfortable on the ball, there’s no reason why the Italian youth international cannot be a long-term success at the Club. He joined for a £33.5m transfer fee on a five-year deal and concluded the transfer dealings.

This final transfer tipped AC Milan’s spending, including the deals done by the directors, to £196m – the second most in Europe’s top five leagues, behind only PSG. With Serie A spending totalling £630m, AC Milan’s was 31.11% of all spending – the new directors were putting their money on the line. Time will only tell to see if these new signings deliver after the team has lost its main creative talisman in Suso.


The next blog post will reflect upon the 2022-23 season. Will AC Milan achieve a third scudetto in a row? And will a deeper squad enable the team to go further in the Champions League? More on that soon…

AC Milan Tactical Set-up 2021-22

This post sets out to look at the tactical set-up of the double scudetto winners, AC Milan. Below you’ll see the players from the squad with hyperlinks to their metric radars – click on those to open a new window if you are interested in seeing them.

Goalkeeper – Sweeper Keeper (Support)

AC Milan academy graduate, Gianluigi Donnarumma, was the first-choice goalkeeper. Over the course of the season, he kept twenty two clean sheets out of a total of thirty three Serie A appearances – 0.67 CS/90. An 88% save percentage indicates his fantastic abilities protecting the goal, using his large frame and agility to prevent the opposing side from scoring. Of the shots he faced in Serie A, he held onto 63.28%, thus securing the ball for his side. 14.85% of his saves were tipped around/over the post/bar and 22.77% were parried (including rounding).

When it came to distribution, the sweeper keeper on support had the team instruction to play out to his back four to relatively safely establish possession of the ball, in part helping explain his 92% pass completion. However, using his passing abilities and the manager instruction to his players to be more expressive, he was willing to override this instruction so that he could be the instigator of a swift and direct counter attack by spraying the ball out to the wide attackers. When this happened, it helped overcome issues faced with deep lying defensive set-ups, not uncommon in Serie A, catching a team out when they have overcommitted players to their attack, leaving space for the wide forwards to exploit. An example of such a counter is shown below.

Defence

Right full backs, Conti, Palencia and Calabria all played in the attacking full-back role to provide width, with the added playing instruction to stay wide to stretch the defence of the opposition. With a narrow central fulcrum in midfield and an inverted winger in front who will look to cut into the box, the full back was a key asset within the tactical set-up.

The player often found himself overlapping his teammate, running towards the byline to put in a cross, as seen by the passes received by Calabria in the match analysis above against Perugia. In offering the forward option, they typically were afforded space whilst the left-sided defender was drawn in to cover off the infield run of the inverted winger. This gave them time to deliver a cross to the centre forward or the in drifting left-sided forward and provided goal scoring chances. Players playing in this role ranged between 1.24 to 1.75 completed crosses/90 over the course of the season.

With Donnarumma instructed to pick out any of the four defenders in front of him, if he opted for the right-sided full back this occasionally led to swift attacks, when they were in space to drive forward with the ball against narrow formations. This meant that counter attacks could be somewhat swift, if not necessarily direct to the attacking trident. Should the opposition have regrouped and adopted a narrow defensive formation, this then led to the full back having room in front of them to dribble down the right flank. Both Palencia and Calabria were above League average from successfully completed dribbles per 90. In part this signifies their attacking contribution to their side, by their willingness to take on defenders and commit their man.

When the opportunity for the cross wasn’t on, either because of a blocked crossing lane or a lack of credible options in the box, then the player needed to be proficient with the ball at their feet, so their passing abilities were also important to ensure that possession wasn’t squandered with the player being potentially out of position deep in the opponents half. Whilst the counter press was set for this AC Milan side, opposition sides could evade the press by going long down their left flank with the space vacated in this area, with the full back well forwards. AC Milan’s full backs in this role had a pass completion percentage upward of 85% – above the 82% League average for fullbacks. Equally, these players needed to be prepared to commit tactical fouls in order to stop the fast flow of opposition counters and allow the team to re-structure into their defensive arrangement. If the ball was lost away from their flank, they need to be prepared to run back and cover their position, so it’s no surprise that all three right backs averaged more than 12.7km/90 and Conti and Palencia committed above League average fouls/90 at 2.05 and 2.23 fouls/90, respectively.

On the left-hand side, the preferred role was the wingback on support – a role that Cucurella and Wöber fulfilled. This provided a little more defensive stability, only looking to go forward when the opportunity presented itself, otherwise hanging back if the attacking full back down the right had gone forward. Again, the provision of width to the team is a key fundamental principle of this role. The player must be available to deliver crosses into the box and help give defenders dilemmas about closing down their attackers. With the mezzala and inside forward on his flank, he can also often be afforded time to deliver his cross because of the overload created ahead of him.

Evidence of the overload in the left half-space and the importance of both full backs to provide width is seen in the below graphic. When in possession, this tactic can turn into a 2-3-3-2 or a 2-3-2-3 depending upon the speed of the transition and development of the play.

Yet whilst their attacking contributions were important, both full backs still had to be more than capable in their defensive skill sets. Whilst AC Milan have succeeded in winning the last two scudetti, there are still five other big clubs in the League and Champions League football to contend with. Only Wöber had fewer than 4 PAdj tackles/90, though he was in the top 10% for PAdj interceptions/90, so perhaps he preferred a more front foot approach to winning the ball back, using his anticipation to read the play. It’s also worth noting that only Palencia had a higher than League average 5.11 tackles/90 for full backs but is worst in the league for PAdj interceptions/90 – perhaps his style of defending was more reactive due to his poor positional play?

Club captain, Alessio Romagnoli, played on the left-hand side of the AC Milan defence, utilising his left footedness. Mattia Caldara and Mateo Mussachio rotated in the other centre back slot. Due to his inferior passing, vision and technique, Caldara played the less technical role of centre back whereas Mussachio and Romagnoli are both suited to playing the ball-playing defender role. Yet this did not seem to mean that Caldara’s passing network/distribution was any less risky than that of his central defensive peers, as demonstrated in the above passing network graphic against Fiorentina. Caldara played penetrative passes into the wide right-sided attacker, seemingly more so than Romagnoli did.

Caldara’s defensive output, using PAdj tackles and interceptions looks poor, and it possibly is but that is a harsh criticism for a player that played 1,915 minutes (21.3 90s) and was a part of a team that only conceded 15 goals in Serie A. Nonetheless, there may be room for an upgrade given Mussachio’s weaknesses too. It is interesting to note that Romagnoli led the League for headers won per 90 but neither Caldara nor Mussachio are anywhere near league average – Mussachio being at the very bottom of the League for heading percentage and in the bottom 2% for headers won/90. Perhaps Romagnoli took it upon himself to cover for his partners weaknesses – nonetheless this looks to be an area to address in the transfer market going forward. As to why none of these players stand out with regards to pressure adjusted (PAdj) tackles per 90 its hard to say, expect perhaps players further forward implementing a press as soon as the ball is lost meant they had fewer tackles to make and found themselves intercepting more long clearances.

When they were forced onto the back foot, I Rossoneri opted for a relatively standard defensive width. By not sitting to narrow to allow crosses to pepper their box, nor too wide to provide space for playmakers to enjoy putting their centre forward(s) through clean into goal, this helped to restrict the number of shots Donnarumma had to face. Additionally, given Mussachio’s seemingly perpetual aerial weakness, this seems to have been a sensible ploy – try to minimise the frequency of crosses arriving in the box without being carved open centrally.

Midfield

In the central midfield triumvirate, the heartbeat of the team was the single pivot – the deep-lying playmaker, set to defend. This is the solid and reliable base of the trio, sitting back and dictating terms and the play. Offering himself to his team mates when the ball was won back either in defensive or attacking positions to find a pass. Italian playmaker, Tonali, was the primary pick for this position, playing just over 2,000 minutes. Completing 48.71 passes/90 with an 89% pass completion demonstrates his abilities with the ball. His key passes/90 statistic underline this further – the deep-lying playmaker is there to play the defence cutting pass if it’s on, otherwise, they’ll look to recycle possession.

This is very evident in the below passing map from AC Milan’s 4-1 victory over Cagliari. Here, Tonali offered himself in a central position to his team mates, either advancing the ball to his central midfield team mates, or, more frequently spraying the ball out to the left as indicated in the graphic. This ability to recycle the ball, maintain possession and look to pass the ball into space for his team mates to run onto allowed attacks to continue and placed pressure upon the opposing team. Here you can also see the long progressive passes that he made to quickly advance the ball forward, cutting out the Cagliari defenders.

The other squad member who performed this duty was Nicolás Domínguez. Whilst his passing percentage is lower, his key passing, and PAdj interceptions show that this role needs to be prepared to put in a defensive effort to shield the back four. Neither player registered significant goals/90 or assists/90, but this simply isn’t their role – they’re often too deep to provide direct goal scoring opportunities from their passes and similarly too far away from goal to have high percentage goal scoring chances. This is indicated by the fact that they registered less than one shot on goal/90 (Tonali had 0.55 S/90 and 0.96 S/90 for Domínguez) over the course of the season.

On the left, the mezzala on support offered a more attacking role to hit the half space and overload this area in combination with the left-sided attacker. Throughout the season, Calvin Stengs, Lucas Paquetá and Dani Olmo were the rotational choices for this position, using their superior dribbling and creative abilities and their eye for goal. Given that the attacking play can bypass them through the direct playing style with swift counter attacks, their goal involvements are not to be overlooked. Stengs’s first season in the Serie A was impressive, with 0.33 goal involvements per 90 (four goals and six assists) and 0.60 chances created per 90.

The right-hand side of the midfield trio saw a box-to-box role adopted. This runner, often fulfilled by Bruno Guimarães, Olmo and Nicolás Domínguez, provided support during both the offensive and defensive play, covering substantial distance. As Guimarães’s metrics show, this role requires a real all-rounder – someone who is capable of being creative but is also prepared to role his sleeves up and put in the hard yards, putting his efforts towards defending, breaking down the oppositions attacking phases. Guimarães is truly an exemplified box-to-box midfielder – his 0.36 chances created/90, 2.04 key passes/90, 12.4km covered/90, 3.67 PAdj tackles/90 and 1.80 PAdj interceptions/90 are stand-out, though his passing accuracy could be improved.

The speed of play that AC Milan exhibit meant that the central midfielders often looked to advance the ball quickly, either dribbling to put pressure on the defenders, or often opting for vertical forward passes. Further evidence of this is indicated in their passes completed over the season – AC Milan ranked only 12th in terms of the number of passes completed, despite having 52% possession (4th best). This is further backed up by the two graphics below – the central midfielders were expected to make high risk passes and as such, their metrics were lower than average for pass success. Perhaps this is further evidence too of their inability to pick holes in a defence once it has become entrenched.

Transition – defence to attack

The below set of graphics help to explain the link between the defence and the midfield, as well as help to show how their roles work within the tactical framework as the team transitions from defence to attack.

In the first instance, the AC Milan side have just won back the ball and Cucarella looks to advance down his flank. Here you see how advanced the right full back (Calabria) was, transitioning into an attacking forward position close to Suso. It’s also notable that Tonali dropped back into the gap between the two centre backs, acting as the single pivot, despite nominally playing in the central midfield strata. Roma elected to engage in a very light, almost meagre, press as most players looked retreat into their defensive set-up.

Cucarella choose to play a relatively safe pass inside to the supporting Paquetá who then dribbled beyond the half way line. As Milan’s defensive line moved up, it is evident how wide the two full backs were, providing the width as the two wide forwards were already positioned vertically along the edge of the 18-yard box. The central midfielders were also very narrow, but crucially there was verticality between them. As such, they had options to pass across the plains between themselves and also beyond them to the forwards.

As it is, Paquetá passed inside to his right to the box-to-box, Domínguez, as his dribble was cut shot due to the narrow, compact defence that Roma established. Since Domínguez was in space, he had the time to use his vision identifying that he could pick out Suso with a first time pass. Here you can really see the verticality of the tactic – within seconds the ball has quickly advanced from the defensive phase into the Roma penalty area.

Suso’s dribble was cut short due to a tackle from the Roma left back, but notice how far advanced the full back for I Rossonerri was, as the box-to-box player, Domínguez, holds himself back and Tonali has maintained his withdrawn position too. Should Roma win the ball back, as they did, AC Milan team had four central players covering the middle area of the pitch as the press began from the advanced players. This would allow others to potentially retreat, or catch the Roma players out as they look to advance believing they have a chance to counter if AC Milan are able to win the ball back.

As it turned out, Calabria’s advanced position actually saw him win the loose ball back. The two wide forwards maintained their advanced positions, but notice how the left-sided inside forward retreated a little to create spacing between himself and Roma’s right back. This allowed him time and separation to attack this full back at speed to out jump him should Calabria have looked to deliver an immediate cross.

However, instead, Calabria played a quick one-two with Domínguez, putting him into space where the left-sided full back for Roma had to re-engage. The cross was blocked and the play resulted in a corner for AC Milan but this gives you an idea of the transition and positional play that is typical within this tactical set-up for AC. The left-sided inside forward still had the critical separation to run at the opposing defender and note too that Piątek had not opted to be on the deepest defensive line so that he had space and just enough time to choose his area to attack the ball should it have come into the area.

Attack

The two right-sided inverted wingers are also two of the set-piece takers for I Rossonerri. As such, Buendía and Suso have somewhat inflated metrics for assists (Suso – 12, Buendía – 9), with eight goals coming from corners for AC Milan. Despite this, both players have outstanding metrics – with Suso totalling 0.48 non-penalty goals per 90 from 3.29 shots per game. Buendía saw something of a come down from his 2020-21 metrics, but perhaps he reverted more towards his mean metric output rather than dipped below it. His NPG/90 dropped by 0.21, despite his shots on target actually rising, and his passing accuracy worsened – in fact he was the worst in the lead amongst his peers. This is slightly concerning, but only slightly. He is in a position to make risky passes, as such it is expected that some of these will be cut out by the opposition, and clearly with 0.48 assists/90, behind only Suso’s 0.52 A/90 in the League, this is indicative of the creativity that came from the attacking right flank for AC. They must be prepared to take on defenders, dribbling at them at pace, and either lay on passes/crosses for the striker or the left-sided forward to score or create chances for themselves to score.

Here you can see how frequently the respective attacking wide forwards find the space to shoot, and also Suso’s out-performance with regards to his NPG/90. This clearly demonstrates the importance of the wide forwards in terms of their shot frequency – Hložek’s metric here is also stand-out but more on this below.

The same can be said of the left-sided inside forward too, yet these players, Federico Chiesa, Rodrigo De Paul and Adam Hložek did not reach the echelon that Suso and Buendía established. Originally brought in as a youth prospect, Hložek’s rapid development saw him rack up far more first-team minutes than originally intended. And his output shows promise – with more shots per 90 than both Suso and Buendía, demonstrating that he found the space to attempt a shot on goal, but perhaps needs to be more clinical/selective given his 8.5% conversion rate – far inferior to Suso’s 15%.

As demonstrated above, Chiesa’s first season has been somewhat disappointing, as it was hoped that he would be the player that readdressed the imbalance between the two outer prongs of the AC Milan trident. Perhaps, after more time at the San Siro, his talents will shine through and the Italian attacking midfielder will show the curva why the AC Milan manager sanctioned his purchase following his transfer listing at the end of last season. He will certainly need to be more creative, with only 0.11 assists per ninety and only 1.33 shots on target per 90.

Rodridgo De Paul’s poor form continued throughout the 2021-22 season, following on from his below par performance in the 2020-21 season after his transfer from relegated Udinese. As a result, of both this and Hložek’s faster than expected development, he will be moved on during the off season.

The centre forward, playing in the complete forward on attack role, led the line and looked to link up play where possible. Krzysztof Piątek is a somewhat selfish striker, not engaging with the build up play as much as desired, but his movement off the ball and finishing ability saw him once more land the capocannoniere. His numbers were marginally below the fantastic metrics of 2020-21, but he still found himself in position to register 4.23 shots/90, with only a 1% drop in his goal conversion. André Silva played Piątek’s deputy when the Pole was in need of a rest, and played some games in the Coppa Italia. The graphic below (including goals scored in all competitions) illustrates Piątek’s importance to the overall tactical set-up for AC Milan with his finishing skills. He may not have many touches on the ball in open play, but when he does find space in and around the area, he is deadly.

Piątek’s off the ball movement and the verticality of the play from the tactic can be seen in the clip below. If you watch his movement from the moment he comes into view, he makes a run across to the right-sided centre back dragging the left-sided centre back with him, presumably because he is instructed to man mark him. Piątek then catches him out of position by creating separation initially and then he ghosts in behind him. Suso lays on the perfect weighted slide rule pass for Piątek to run onto and the Polish striker does the rest, slotting it past the on-rushing keeper for the only goal of the game.

Hopefully this deep dive into the tactic enables you to have an idea as to the tactical set-up AC Milan utilised throughout their scudetto winning season. Any questions, please use the comments section or contact me on Twitter @afmoldtimer.


The next post will focus on the player sales and recruitment that took place over the off season in preparation for 2022-23 season.

AC Milan 2021-22 Season Review

The 2021-22 Serie A season saw AC Milan defend their title, with a six point gap to second placed Napoli. I Rossoneri improved upon their points total, picking up an additional two, and maintained their frequency of clean sheets when compared to the 2020-21 season. With 69 goals scored (one fewer) and 15 goals conceded (two fewer), AC Milan’s dominance over their Serie A rivals looks to have become rooted once more.

Last season’s runners-up, Atalanta didn’t even make the Europa League places, finishing 27 points worse off compared to the previous campaign, despite creating more chances and having more possession compared to 2020-21. 18 fewer goals was their biggest issue, with forwards Gabriel Barbosa (6), Duván Zapata (12), Musa Barrow (5) and Josip Ilicic (2) scoring just 25 goals between them. In 2020-21, Zapata (16) and Gabrigol (19) managed 35 between them – a remarkable drop-off from Gabrigol.

Napoli improved their points total by seven, which was enough to see them rise up the table by two positions, thereby solidifying their Champions League status. Lazio and Juventus joined them both ending up on 80 points.

At the bottom of the table, two of the promoted sides, Udinese and Lecce, returned back to Serie B. Whilst Sassuolo’s defence was around average for the bottom half of the table, their inability to put the ball into the back of the net led to their relegation – their failings catching up with them from last season, with just 29 goals per season. The other promoted team, Frosinone, faired much better, finishing 13th and a full twelve points above the relegation zone.

The haves and have-nots

Taking into account expenditure upon salaries and looking at the number of points gained, Juventus once again are out on their own with their salary spending, but this was not still enough to land them the Serie A title. I Bianconeri actually reduced their salary expenditure by £20.71m compared to last year and yet gained 3 extra points. Comparatively, to pick up the extra two points that AC Milan earned, they spent an extra £13.63m on wages – the second largest salary hike behind city rivals Inter Milan (£14.53m) in their forlorn efforts to attain Champions League qualification.

Lecce’s salary outlay (£22.10m) was somewhat reflected in their overall points total and relegation, but credit should go once more to Perugia (£16.42m) and in particular Frosinone (11.75m) given their lower middle-table finishes on such small salary budgets. Their backroom staff clearly overachieved given the financial resources available to them, eking out more from their respective squads than could otherwise be expected.

It’s worth pointing out that the correlation between salary expenditure and total points has actually strengthened between 2020-21 and 2021-22, with the R2 figure rising from 0.4368 to 0.6123, suggesting that wage expenditure has become increasingly important to be successful – broadly in line with Szymanski’s findings.

Success – with or without the ball?

The 2020-21 season demonstrated a reasonably strong correlation between possession and PPG with an R2 of 0.427. Yet the season just past saw that relationship entirely disappear, with the amount of possession having essentially no relationship to the PPG that teams picked up. Given Serie A’s relatively low average goals per game (AGPG) of 2.45 compared to those of the EPL (2.76 AGPG), Bundesliga (2.65 AGPG), Ligue 1 (2.63 AGPG) or La Liga (2.58 AGPG), it’s clear most managers favoured a defence first approach, and had their teams adopt a low-block to form a solid defensive system in response to much larger, more financially powerful teams and also perhaps down to the physical, mental and technical capabilities that their players possess.

Lazio and Inter Milan appear to be the main exponents of playing successful football without the ball, with Lazio having an average of 47% possession over the season, yet picking up 2.11 PPG. This is an impressive metric, clearly proving their defensive solidity – underpinned by their 13 clean sheets, 7th best in the League. Both teams eschewed the ball, with Inter (11,834 completed passes or 311.42 per 90) ranking 16th in terms of passes completed and Lazio lower still at 18th – behind only Benevento and relegated Udinese – with 10,990 completed passes (289.21 per 90). For context, Napoli (ranked 1st) completed 16,984 passes (446.95 per 90).

Inter Milan averaged marginally more possession, at 48%, and enjoyed 2.08 PPG. Their defensive capabilities were far superior to that of I Biancocelesti – registering just 22 goals conceded and 21 clean sheets. However, I Nerazzurri were not as strong in front of the opponents goal – scoring 19 fewer goals than the sky blue team from the capital.

Given the respective managers in charge of the two teams, this approach to their teams’ football makes sense. Both Simone Inzaghi (Lazio old-boy) and Diego Simeone (former Inter Milan great) favour counter-attacking football, and from the metrics, their very styles have yielded very similar results. Inzaghi’s preference for vertical tiki-taka and a 5-1-2-2 DM WB formation may mean that possession is turned over more frequently than is typical average for the League, but it created goal scoring opportunities of far higher quality – with their players scoring from more than 10% of their overall shots. In contrast, Simeone’s tactical rigidity actually yielded more shots per ninety, yet the same number on target per ninety as Lazio. With only 7.23 goals/100 shots, their strikers Lautaro Martínez and Andrea Belotti appear on the face of it to have been somewhat wasteful. Yet Martínez scored 0.49 goals per 90s, with 22 goals in total, and a not awful shot conversion rate of 12.2% on the back of a 47% shot accuracy. This might go some way to alleviate any pressure upon him and put it back more onto Belotti (4 goals at 0.18 NPG/90, shot conversion rate 5.8% and 42% shot accuracy) and/or Simeone’s chosen tactical system.

This is backed up by the below graphic looking at the metrics for the number of chances created per ninety against shot efficiency (goals/shots) that were created for them – more than three chances per ninety – joint fourth in the League. This graphic also seems to underpin the failure of Atalanta’s forwards – the chances were there for them to put away, they simply seem to have failed to take advantage of them.

Equally, there appears to be no significant correlation between the amount of time that teams have the ball and the number of goals that they concede. Frosinone are a good example of this – joint fourth in the possession metrics, but the fourth worst defensive record, albeit a minnow in a sharks tank.

Clean Sheets = Points

The importance of keeping clean sheets to picking up points is clearly highlight in the below graphic. A correlation of 0.86 between keeping clean sheets and the number of points a side were able to collect illustrates this further still.

It’s notable that only Roma out of the ‘Big 6’ failed to keep a significant number clean sheets and pick up more than 1.6 PPG. A sign that they have perhaps still failed to redress the loss of Alisson to Liverpool in 2018. Clearly investing into a top level defence and goalkeeper yields results and helps sides to compete for the European places.

Looking at Roma’s number of goals scored, it could be argued that I Giallorossi recruitment team have other areas to address first. Their number of goals scored to frequency of shots highlights that they could do with an upgrade on Odsonne Edouard and Patrik Schick, with Pierre-Emerick Aubameyang likely to see even fewer minutes given his age. Both of their younger forwards had a goal conversion rate of less than 10% – 8.3% and 6.8%, respectively. Comparatively, rival Lazio forwards Ciro Immobile (24 goals) and Carlos Vinicius (17 goals) had 15.8% and 12.9% accordingly.

Ciro Immobile was crowned the 2021-22 Capocannoniere – his 24 goals outranking the 22 scored by both Piątek and Lautaro Martínez. Whether or not the 32-year old Lazio forward can continue to out-compete his rival strikers to achieve this feat will remain to be seen in 2022-23. Given his previous proficiency in front of goal and Serie A’s history of more senior strikers having great longevity, there’s every possibility.

The new blog post will take a look at AC Milan’s tactical set up to analyse their continued success.