[FM 19] Moneyball, statistics and Vitesse

[Unigo]

Introduction
Welcome to another attempt at actually sticking with a career and topic, something i'm terrible at, apparantly. But maybe this is the one that will keep my interest!

The setup of this career (and topic) will be different from the usual ones. I'll start the game as Vitesse (not that new for me), but will play mostly based on data and performance instead of player attributes. Of course attributes still do matter, but this is new for me. Usually, i make a big spreadsheet with average values for crucial and recommended attributes and build my squad around theses values. This time we take a different approach, based on performance (goals, passing percentage, etc.). This is based more on Moneyball (the book and movie), Soccernomics and The numbers game (both books). 

The purpose of this topic is (of course) to entertain, but also share ideas on this kind of approach to Football Manager. I'm not a data-analyst so my analysis-skills are severly limited, nor am i a brilliant tactictian. Eitherway: let's work together and make this an interesting read, hopefully with some learning effects as well.

Principles
Of course we need a few principles to work with. I'm not going to re-read all the books to find em out, i'll base mine mostly on Alex Stewart his work (https://thesetpieces.com/features/football-manager-meets-moneyball-pt-1/).

• What stuck out in Moneyball to me was the way BIlly Beane approached the game. Not focusing on all kind of non-important qualities but on the true value a player had to his team, hit percentages for example. Also, at the start of the season an estimation was made on how many wins were needed to get to the play-offs. It is my intention to use this in football manager. So, we'll look at what's needed to achieve a certain result in the competition and try to get the players that contribute to these things. 
• In Soccernomics and the Numbers Game there are a few useful remarks that we'll (loosely) use here as well. We're going to try to refrain from buying very young players. Also, we're not going to spent money on new players older than 25 years old. Players that are 30 years or older are going to be sold, since their value rapidly drops off after that age. Now, this rule might conflict a bit with the data side of the game. For instance: what will we do when our 30 year old winger was responsible for 50% of the needed assists? In general this means we won't be super strickt regarding the rules, but we'll see what happens during the game.
• Other than transfer-guidelines it is mentioned in The numbers game that a goal conceded is more costly than the profit from a goal scored. This also goes for players: a bad player is more damaging than a top player can fix. So the advice would be to focus on fixing the weak links first.

Now, with these principles we face a few (or more than a few) limitations. Baseball is a vastly different sport than Football. I'm not a baseball expert, but to me baseball is a bit like the penalty shootouts in football. What i'm sure we're going to struggle with is the effect that the tactic has on our results. We can determine we need 60 points to get to the Europa League and that scoring 65 goals in a season should be sufficient to reach these points and that our attacking players should account for these 65 goals. But, if i set up a tactic that fails to enable these players to score we're in trouble. Also, the tactic employed by a players current club (if he's not playing for me) is ofcourse hugely influential to his performance. We'll have to see how this will play out. 
Another limitation are my poor data-analysis skills. No complicated regression-analysis in this topic (atleast not by me), but i hope i'll inspire folks to more sophisticated calculations. Another limitation is the lack of data. By my understanding i can see quite some stats in the current season, but i can't look at performances in previous seasons regarding minutes played, passing percentages and stuff like that. This will limit our scouting i think. The last thing i want to mention is the effect different leagues have on the stats. A midfielder can have a 95% pass completion in his current competition, but if he's playing on an amateur level and is transfered to Real Madrid he might not get that same result. We have to keep this in mind with our scouting process. 

Now to the good stuff...
We're gonna play with Vitesse, playing in the Dutch Eredivisie. In preparation for this season i've done some data-gathering at Whoscored.com. Unfortunately not all data has been gathered for the Eredivisie, but i think we got some useful stuff anyhow. Now, what have i done...

First off: i copied the end-result table of the last 5 seasons before 2018/2019 (the season we're playing). This looks like this:

I hope that these results (also did this for the other 4 seasons) can tell us something about points and goals needed to reach our goal. This isn't too complicated, so lets get a bit deeper. What is needed to get these points and goals? I've tried to get a few indicators that can help us to determine a club's performance, that we can distill from the game and that are available on statistics websites. The result looks like this:

now what are we looking at:

• Goalkeeping: saves are (duh...) the amount of saves made during the season. The saves/goals against (should be conceded...) percentage assumes that the goals conceded + the saves equals the total amount of shots a goalkeeper had to deal with. The percentage represents how many of those shots got saved, telling is something about our goalkeeper performance. (Note: i'm not sure if these saves also include the blocks. There is a 'shots blocked' category in Whoscored.com, but since this could be applying to defenders i did not include it). 
• Defense: my first idea was to use both succesful tackles and interceptions as indicators. But, it turns out the top teams have less interceptions and less tackles. I think this is due to them usually being the attacking side, thus not having to intercept/tackle as much. The ratio of succesful tackles is interesting though, so i've displayed the attempted tackles, succesful tackles and the ratio. 
• Passing: passing is (in my opinion) a stat relevant for all field players, but most specifically for the midfield. I was looking for a passing completion ratio, but this wasn't available for the Eredivisie. So i looked how it was calculated for the Premier League and did the same for the Eredivisie. So i took the total amount of passes, minus the inaccurate passes and calculated the ratio of these numbers resulting in the passing completion. 
• Attacking: Goals are hugely important, but we allready have these stats from the league table. I did add the number of assists. A shots on goal percentage wasn't available, so i went with the total amount of shots and divided this by the number of goals. This gives us a percentage on how many shots (either on or off target) lead to a goal. This should be a nice idicator for our more attacking players.

Now that we have this stuff for 5 seasons i calculated the average over 5 seasons. This looks like this:


In this table we can see what the average amount of points, goals, and all other things was over the previous 5 seasons. Now, we're assuming that a better competition performance equals better values in the different categories. To make this more visible i've created a few graphs based on the table:

What can we learn from this: well, the good news: each graph shows a (rough) declining line. Or, in case of the goals conceded an upwards line. This means that we see a rough trend that tells us that the better performing teams (in terms of competition result) are also performing better on the different indicators mentioned. 

So what?
Well, for VItesse we're going to aim to qualify for the Europa League play-offs. This is means we need to finish at least as 7th. We're gonna aim a bit higher than that, also based on the season prediction (season preview) which is 6th. So, let's get 6th as our goal to aim for. When we look at my image-spam from above, this tells us that:

• we need at least 51,4 (52) points at the end of the season;
• on average the team finishing at 6th position had 58 goals scored and 48 goals conceded, so a positive goal-difference of 10;
• the save-percentage is the least linear, so it is difficult to draw a conclusion on it. But, i'd say 69% should be a good goal to aim for;
• the same goes for the tackle-percentage, but 69% should be sufficient here as well;
• pass-completion should be above 78%, so let's go for 79%;
• The assists (table) should be above 40 a season, i'm going for 42;
• and last: the percentage of shots leading to a goal should be around 11.5%. 

Ofcourse this is all assuming that if we achieve these numbers on the indicators we indeed will get our treasured Europa League play-off place! 

Enough for now, untill next time!

Edited June 21, 2019 by Unigo

[withnail316]

Interesting concept. I watched Moneyball for the first time a couple of weeks ago, and it did make me think of FM. 

Best of luck.

[Ronaldo Beckham]

Looking forward to seeing who you bring in and how this progresses, I like your approach.

[Unigo]

Thanks for the replies guys!

After the previous opener post it’s time for a tactical break. All the statistics are nice, but we need a formation. Now, as i said i’m not a tactical wizard, so i’ll explain my thoughts, maybe someone can learn me something ;). I’ve put together a formation that i think isn’t too complicated and suits the current squad. Of course everything can be changed based on experiences in the game. My thoughts are:

• The average heading skill of our defenders is under-average. I’m going for a wider defensive play to prevent crosses. The speed isn’t too good either, so playing too high up the field could result in balls over the defense that our central defenders can’t get back. So, we’re going for the standard line of engagement and line of defense. I don’t want to engage too high, i want to avoid creating too much space between our lines.
• Our midfield has the highest average acceleration and pace. This is good for the wingers and inside forwards, we also have quite a few players capable of playing on those positions. So, we’re going two inside forwards.
• The best midfielders have good workrate. This suits Box-to-box midfielders and defensive midfielders, but also a half-back. To get an extra guy in the box if someone manages to get a cross in i’m going for the half back. I want to add a bit of creativity to the half-back and the box-to-box midfielder, so i’m also adding a deep lying playmaker to the squad.
• With one player left (4 in defense, 3 in midfield, 2 wide players leaves 1 to spearhead the attack) this strengthens my decision to go with inside forwards to get a few more players in scoring opportunity. I want one of the IF’s on a support duty to accommodate the striker, but i put a fullback on attack on that side as well to get another option in attack, namely crossing the ball in. The lone striker will be a complete striker. This might be tricky on the Eredivisie-level, since it requires a well-developed player that can play in different ways, but i want to give the striker the opportunity to both shoot at goal and facilitate our inside forwards.
• Other than that our goalkeepers are quite good at throwing the ball, and i’m not a fan of kicking the ball up field only to lose it. We don’t have an attacking midfielder and only one striker, i’m not sure anyone is going to pick that ball up anyhow. So i’m telling the goalkeeper distribute to the defenders. 
• When it comes to passing, our midfielders (and defenders and strikers, for that matter) are just below average. I don’t want to lose possession due to long balls that are intercepted, so i’m opting for shorter passing. I’m combining this with a bit higher tempo. This might increase the errors, but could be necessary to surprise and beat our opponents. 
• Finally i’m going for a counter-pressing style of play. This means instructing to counter-press, and make use of the recovery by countering. I think the countering suits our fast players, who can hopefully catch the opponents off guard. I’m not going for a more urgent pressing style, i think that’s over doing it and that the counter-press instruction on its own should be enough.

So, what does it look like? Well, like this:

Next up (either in this post by an edit or in a new reply) we're going to analyse our current players. Since we don't have any in-game data yet, i might look for some external date as well on past performance of our players.

Update: player analysis
Right, as promised here is the analysis of our players (this will be a LONG read). I’ve only looked at the ones in the main squad, i’ve also excluded the players on a loan. Since we’re only just starting we have no in-game data that could help us judging our players. But, there was a lot of information on Whoscored.com on most field players. So…. i went a bit overboard… Not all things are filled, because of a lack of data. During the season we’ll use the football manager statistics to fill a new form for the current season. The result looks like this:

So, what are we judging our players on? I make a distinction (of course) between the goalkeepers and the fieldplayers. For the fieldplayers i use the same indicators regardless of position, though the importance of every indicator does differ. So, i’ll get a successful-tackle ratio for both the central defender and the striker, but obviously for the central defender this is a far more critical indicator. We also have some ‘base’ information: the games played and the minutes played. The ‘first year’ is referring to the earliest year that had data available on whoscored.

For the goalkeepers we’re looking at the following:

• Clean sheets: this is not only down to the goalkeeper, but is widely regarded as an achievement by the goalkeeper, so we’re keeping track of it.
• Mistakes made: football manager keeps track of mistakes made for all players. I don’t know when something is a mistake or not, but i’m going to use it. Whether a goal was scored is not relevant to me, in my opinion this is also down to luck and the ability of other players to fix your mistake.
• Conceded goals: Pretty self explanatory, we want to know how many goals are scored against the goalkeepers.
• Min / conceded: this calculates how often a goal is conceded (on average). a higher number is better, meaning there is more time between each conceded goal.
• Saves tipped, saves blocked and saves held are all coming from football manager. on whoscored they are only monitoring saves. I’m not that fussed about the difference between these 3, but we’ll keep track of them. Obviously i do prefer saves held, blocked saves give a chance on a rebound and saves tipped cause a corner kick. 
• Shots / conceded: this one is calculated by adding the conceded goals and the saved shots (total shots to deal with for the goalkeeper) and calculating the ratio of the saves on the total amount of shots. Come to think of it… the title might not be totally accurate. Either Way: this is one of the most important indicators to me, it tells me how many incoming shots are saved (and thus how many are scored).

For the fieldplayers we use these indicators:

• Tackles attempted, tackles won and the tackle ratio: the attempted and won tackles are needed to calculate the ratio. In football manager we have the ratio available, so we won’t fill attempted/won ones. The ratio is most important for the defenders, but is quite interesting for the whole squad. To me this is the key indicator of a well-performing defender.
• Headers attempted, headers won and the header ratio: this one is kind of the same as the tackle one. In the future we’ll only use the header ratio. This one is also important to us for the whole squad. It’s a must for defenders, but also the midfield has to be able to deal with incoming high passes, for instance from goal kicks. Bad heading could lead to loss of possession. For the strikers the headers won ratio is also relevant in offense.
• Passes attempted, passes completed, successful passing ratio: same story as the two previous indicators, only the ratio matters to me. good passes equal preventing loss of possession and are crucial to the team. I think this indicator is most relevant to the defenders and the midfield, but it also tells us something about the strikers.
• Mistakes: Mistakes are relevant for the same reason as explained by the goalkeepers.
• Assists: Assists are a key part of the game. Over the past 5 years almost 70% of the goals in the Eredivisie was preceded by an assist. For midfielders, strikers and especially players on the wings getting their assists in is crucial.
• Min / assists: we also want to know how long it takes a player to get an assist in. This allows us to compare player performance. Only looking at the number of assists isn’t good enough, a player could have 10 assists but it might have taken him 1800 minutes (20 games). If a player manages to get 5 assists in 600 minutes he might actually be a better pick.
• Goals and min / goals: no goals, no wins. So, very crucial stat. For the min / goal number we follow the same reasoning as with the assists.
• Shots and shot/goal ratio: the last one! this tells us how effective a players shots are. the number of shots isn’t too relevant, it is used to calculate the ratio. In the future i might add if a shot was on or off target, but for now i won’t. Personally i might prefer a player who misses the target more because he always aims for the corner instead of just shooting straight at the goalkeeper. this might lead to more off-target shots. Basically it all comes down to a player having a shooting opportunity which can either be a goal, or not. The ratio will tell us how effective a player is dealing with these opportunities.

Now, with that out of the way: let’s look at the players we currently have!

Eduardo - goalkeeper

Eduardo is quite old and on a loan from Chelsea. He’s a little bit better than Pasveer, but only just. All in all he’s a decent goalkeeper and because of the loan arrangement he’s with us without getting a single euro in wage. So either way, i’m happy to have him around. I might play him and Pasveer both for the start of the season to see who gets the best numbers. The data on his previous seasons is not complete enough to tell us anything.

Remko Pasveer - goalkeeper

Pasveer is a bit of the same story as Eduardo. Old, quite decent attributes and he’s not too expensive either when it comes to wage. So, he might actually get some games this season! he still has 2 years on his contract, i would consider selling him if i can get a decent offer, but i’d prefer to use him at least for the coming season. Again, the data on previous seasons on the goalkeepers is not telling us anything.

Jeroen Houwen - goalkeeper

Houwen is definitely quite a step behind Eduardo and Pasveer. But, he’s only 22 so still has 12/13 years to get to their level. Then again, i don’t think he’ll ever exceed just being ‘good’. I might keep him around, but we have a few very talented young goalkeepers coming through the ranks. Houwen won’t get any first team football as a substitute either due to Pasveer and Eduardo being so closely matched. So, i’m going to list him for a transfer (or loan, to cut the wage bill) and see what happens. The data on previous seasons… you know the story by now.

Vyacheslav Karavaev - right back

Karavaev is rated as a good right back for the Eredivisie level. Also, he’s only 23 years old so still has some room to grow. He isn’t the top earner, but isn’t cheap either. But, given his attributes his wage is more than justified. Based on his attributes he should be one of our starting 11 and would actually be one of our team’s stars.

When we look at the statistics over the previous seasons we should be scared. The poor heading isn’t that much of an issue since he isn’t playing a centre back. But his tackling ratio is low as well (56%). Passing is a bit better, but with 72% he’s still below our average goal. These numbers are based on only 1795 minutes, so the sample size is quite low. I expect to see a huge improvement the coming season, but it is something to keep an eye on!

Rasmus Thelander - Centre back

Thelander is a fine central defender and would probably be in our starting 11 if it weren’t for the fractured ankle that keeps him sidelined for at least half a year. He’s good in the air, strong and has good attributes in every thing you’d want from a defender. He’s not one to use as a ball-playing defender though, but we weren’t intending to do that anyhow. He’s 26 years old and ‘only’ earns €19.000 a month. He still has 3 years on his contract to, so when he’s healthy he most surely will get gametime. 

In our previous seasons-statistics we only have 8 games / 720 minutes of data. This means we’re not gonna come to any firm conclusions. The first signs are hopeful though: a tackle ratio above 90% is impressive! His heading ratio is only at 64% though, let’s see what he can achieve when he’s playing this season.

Jake Clarke-Salter - Centre back

Clarke-Salter is on a loan from Chelsea meaning he’s costing us nothing! He’s quite inexperienced though, being only 20 years old. He has decent stats, but is still a step behind Thelander. One thing that stands out is his tackling rating, he surely needs to improve in that area. He has two traits that stimulate long passes, i’m curious to see how that will work.

His previous season statistics aren’t numerous enough to draw any conclusions. That’s a good thing, because both his tackling and his heading ratio are below what we need. I doubt Clarke-Salter will start a lot of games, but as a substitute he might get his share of playing time

Arnold Kruiswijk - centre back

Kruiswijk is one of our oldtimers. He’s 33 years old and that’s reflected in his wage and worth. WIth still 1 year left on his contract i could be tempted to sell him. But, that would be shortsighted: his attributes are still quite good and his mental attributes really stand out. In pure defending he is a tad behind Thelander, but compared to Van der Werff, Clarke-Salter, Doekhi and Klein-Holte he still stands his ground. Kruiswijk is with Vitesse since 2014, i think he still has some games in him so we’re going to make use of him for this season. Hell, i might even try to extend his career with one more year!

If we look at the statistics over the previous seasons we actually have quite a lot of data! With 98 games and 7739 minutes we can surely get some useful insights. WIth a heading ratio of 61% he isn’t too much of a killer in the air. He makes up for it with 80% tackling ratio and 88% successful passes though. Let’s see what he brings to the table this season!

Maikel van der Werff - centre back / defensive midfielder

Van der Werff is basicly a bit younger, bit bigger and bit faster Kruiswijk. Kruiswijk is a bit stronger in pure defending, but van der Werff surely is a good Eredivisie-level player. Even better is that van der Werff can be used as a half back. It’s not his favourite position, but he might grow into it after some games. The issue with van der Werff is that he only has 1 year left on his contract, is 29 years old and currently is worth €1.800.000 with a €23.000 a month wage. Financially it might be a good idea to sell him after this season, but: he’ll be a free-agent by then. For now i’ll see if i can extend his contract by 1 year with some reasonable terms and will try to sell him after this season. If that does not work, i might get rid of him during the winterbreak to still get some profit out of him. 

On van der Werff we have a lot of data (117 games, 8954 minutes). he has a 81% tackling ratio (great!) and a 58% heading ratio (not so great…). His passing is good though, 84% of his passes are finding their intended destination. A nice extra is that he managed to get 5 assists and 7 goals in over that period. Poor figures for a striker, but a nice bonus from a defender!

Joris Klein-Holte - Centre back / defensive midfielder

Klein-Holte is a tough one to judge. He’s 21 years old, so still has some growth left in him but also is quite one-sided. His attributes are okay-ish for a stand-in defender, but Clarke-Salter is a lot more well rounded for instance. Klein-Holte is quite big and has good jumping, sadly enough his heading isn’t that impressive. He still has 1 year on his contract and isn’t worth much (€195.000). The issue is his wage: he earns €11.000 a month currently. I’m quite confident i can find decent back-up players who will do just as good for less. I’ll probably try to sell Klein-Holte for a small fee, if that doesn’t work i’ll go for a loan and probably wave him off as a free agent after this season. We don’t have any statistics on his previous seasons available.

Danilho Doekhi - Centre back / defensive midfielder

Doekhi is 20 years old is a better than Klein-Holte and is a bit like Clarke-Salter. He has decent stats but still needs to grow some more. He’s currently just shy of Eredivisie-level, but has the potential to get there. He still has 4 years on his contract and his wage is quite reasonable as well, so we’ll keep him around. Especially during Thelander’s injury he will compete with Clarke-Salter for the stand-in spot. Regarding previous season data we only have 9 minutes of info, so we’re not going to bother analysing that.

Alexander Buttner - left back

Buttner is a special guy. Big history at Vitesse, after few foreign adventures at Dinamo Moscow, Anderlecht and of course Manchester United he’s back home. When it comes to speed and technique he’s our best left back. But, he lacks defending quality. Also, he’s a bit injury-prone and an inconsistent performer according to my coach. His situation is complicated: we’ve blown our entire wage budget, so we can’t get any new players. Buttner plays a role in this, because of his €30.000 a month wage. He’s 29 years old and is still worth €1.500.000, with a 1 year contract. It’s a bit like van der Werff but with a even steeper wage. What makes his situation a tad worse is that van der Werff definitely has a place in the starting 11, Buttner has to deal with Max Clark, who is is very decent left back as well. His saving might be the lack of other backs, so if we sell Buttner we only have Clark. Then again: for his value/wage i might find a left AND a right back as back-ups. We’ll see what we’ll do in the near future.

Regarding his previous season numbers: we have 84 games / 5838 minutes played, so plenty of information. 72% tackling and 49% heading are a bit below what i’d expect from the starter left-back. His passing is above 80% though and he has 11 assists on his name. WIth an average of 530 minutes per assist he’d get around 5 assists this season if i’d play him full time, every game. Our season-goal is 42 assists, so that could be a useful contribution.

Max Clark - left back / defensive midfielder

Clark can be used in multiple spots, which is great. He has decent defending attributes and is also a capable freekick taker (then again: so is Buttner). He’s only 22 years old so there is still some potential there too! Clark will surely be useful, but he kind of has to be: he has a wage of €34.500 a month, so he is one of our top earners. This is: if Buttner will play left back and van der Werff plays as the Half back he might not get any game time. Maybe Thelander’s injury might be his luck. As a Half back he might be just as good as van der Werff, yet his height of only 1.80 (meters) is a bit low. I guess his future is depending on what will happen with van der Werff and Buttner, he is too expensive to keep on the bench. 

Based on the previous seasons he’s falling behind on Buttner. we’re only looking at 2270 minutes played, but his tackling ratio (69%) and heading ratio (54%) are kind of the same, yet Clark has 65% passing ratio and currently no assists.

Thulani Serero - Defensive midfielder / central midfielder

Serero is one of our top players. Having an extensive history at Ajax shows in his attributes. Good teamwork, first touch and technique stand out, as does his workrate. He’s 28 years old and still has 2 years on his contract. And: with only €23.000 a month he is quite cheap as well for his level. I like his as a deep lying playmaker, but not as a half back. Serero is only 1.64 (meters) in length and also lacks strength. Yet due to his qualities i’m sure he’ll be one of our key players on midfield, either as a deep lying playmaker or as a box to box midfielder.

The data over previous seasons confirm our conclusions. WIth 8886 minutes we got a big sample size, so let’s see: tackling ratio is slightly below what we want with 70%. Heading is (as expected) well below adequate with 33%. Now the good stuff: 89% passing is great! He doesn’t have a lot of assists or goals, but when he shoots he actually scores in 12% of the attempts. This isn’t all that bad, only 2 players (Linssen and Matavz) show a better ratio.

Martin Odegaard - winger / attacking midfielder /  central midfielder

Odegaard is one of Real Madrid’s wonderboys. He’s most at home on the right wing and has great attributes for that role, except finishing. That could be an issue, since we expect our inside forwards to score as well. Either way, we’ll see what the season will bring, and if that’s a truckload of assists i’m just as happy. Unlike the loans from Chelsea we do have to pay a chunk of Odegaard’s wage, and it’s a big chunk: €41.000 a month. For that money you can be sure we’ll make use of him, but if he doesn’t give results we might have to send him back: he’s way too expensive to have as a reserve. Vitesse has some decent wingers, but i’m more than willing to give Odegaard his chance to impress!

We only have 2975 minutes of play regarding previous seasons, but they do tell us something. His defending (both tackling and heading) are very poor, his passing is at 83% though. He hasn’t racked up a lot of goals and assists, but with his talent he should improve a lot and contribute to our season.

Matus Bero - winger /  attacking midfielder / central midfielder

Bero is a bit strange. His natural position is the right wing, as a winger. Yet his crossing isn’t too impressive and neither is his speed or technique. In other words: he’s not going to play on the right wing i think, and as an inside forward on the left we have better options as well. That doesn’t mean he’s useless though: high workrate, good of the ball skill, bravery and anticipation all help, as does his decent strength. Maybe retraining him to a box to box midfielder might be worth it? He’s only 22 years old and is with us for the coming 4 years, so we have plenty of time for that. It will be costly though: his wage is €34.500 a month. Let’s see how much we use him, if he’s not playing often or contributing he’s too expensive to have around.

The statistics are a bit disappointing as well. The tackling is below average with 67%, but that’s to be expected of a winger. His heading is at 28%, which is absolute ****. Passing isn’t impressive either with 72%. His productivity is mediocre, with 2 assists (961 minutes per assist) and 3 goals (640 minutes per goal). If we’d play him every game for the whole 90 minutes this season he’d get 3 assists and 4-5 goals. Not impressive as a midfielder, but surely unacceptable as a winger. Let’s end on a positive note: Bero did score those goals in 38 shots, getting a 8% shot/goal ratio. Not good enough for a striker or inside forward, but not that disappointing for a midfielder. Let’s see what he does this season!

Hilary Gong - winger 

Gong is young, and it shows in his attributes. He has decent attributes for a winger, but would be a bit less suited as an inside forward due to a lack of composure and flair. Then again: he’s 19, plenty of time to grow! The coaches think he has great potential, so we definitely want him to stay with Vitesse. He could be a decent stand-in as a inside-forward on the left wing, but will surely face some competition. I think i’ll give him some minutes in the friendlies and start of the season to decide if he stays with us or will go on a loan (if we find him a spot). We have no data on previous season, sadly enough.

Roy Beerens - winger

Beerens is a typical wing player. right footed and good as either an inside forward or a winger. He’s an oldy with his 30 years, but still has a contract for the next 3 years. His wage is decent with €21.500 a month, especially given his attributes. great speed, crossing and technique, just lacking a bit in the mental area. Currently Beerens is worth €2.100.000, but that will rapidly decline given his age. He’s also facing a lot of competition on the left wing as an inside forward. Linssen, Gong, Musonda and possibly Buitink are all more than capable of playing on the left. On the right we only have Odegaard as a left-footed inside forward, so there is a lot less competition there. With this knowledge: Beerens might be on the way out…

Looking at his past performance his tackling and heading is nothing special. His heading ratio is even below 10%. Passing is above 82% though. In the 11001 minutes played Beerens produced 18 goals and 15 assists. If we’d play him all games for 90 minutes he’d end up with 4 assists and 5 goals, with a shot/goal percentage of 9%. I doubt this will be enough to secure his spot in the starting 11.

Navarone Foor - attacking midfielder / central midfielder / winger

Foor can play at multiple positions, that might be useful. He’s only 26 years old, has 2 year remaining on his contract and his wage isn’t too high (€21.500 a month). His favourite position (attacking midfielder) is not currently in our gameplan, so that’s a bit disappointing. I think he could play as both a box to box midfielder and a deep lying playmaker, due to his workrate, technique, passing and decent mental skills. Serero is the favourite for the deep lying playmaker spot, but Foor might come out on top when he’s competing with Bero for the box to box midfielder role. Bero is slightly better at most attributes, but his superb vision might seal the deal for Foor.

In previous seasons his tackling and heading is nothing to speak highly about. His passing is at 79% though, and he has a 10% shot/goal ratio which is good for a midfielder. All in all i could live with these figures for the current season, though a bit higher tackling would be good. Then again, that might be too much to ask given his attributes.

Charly Musonda - winger

Musonda is a young winger from Chelsea, meaning no wage bill! He has good technique, is fast and agile, yet lacks crossing or finishing skill. His long shots are good though, so that might bring something to the table. Compared to Bryan Linssen you can see the potential. At only 21 years old he’s almost at Linssen’s level. It is my expectation that Linssen and Musonda will battle for the spot in the starting 11. This also means that Gong has his work cut out for him.

The previous seasons show incredibly poor heading, surprisingly good tackling (85%, which is not reflected by his attributes) and good enough passing above 78%. His goal-scoring ability is not that impressive (5% shot/goal ratio, only 1 goal in 1577 minutes) but he needs just shy of 400 minutes per assist. Over a whole (fulltime) season this would lead to 7-8 assists.

Bryan Linssen - winger

Linssen is quite possibly the best winger we have. Better attributes than Beerens and Musonda, can cross, finish, is fast, very determined and has technique. He has 2 years left on his contract for (only!) €25.000 a month and is worth €3.200.000. He’s 27 years old so he should serve us well for the next few years. Musonda might be waiting for him to mess up, but i’m sure Linssen will play his fair share of games this season!

Over the past years Linssen played 13258 minutes. His tackling and heading are below average but good enough for a winger. His passing could use some improvement though, being at just above 67%. The true value of Linssen is his productivity though. he needs around 600 minutes per assist, but only 245 minutes for each goal. With 13% shot/goal ratio he’s a quite reliable shooter as well! if he’d play 34 matches for 90 minutes he would likely score more than 12 goals, but i’m more than satisfied if he gets in the double figures this season. 

Dauda Mohammed - striker / attacking midfielder

Mohammed is new, and on a loan from Anderlecht. Luckely he’s on a €0 wage, because i’m not sure if i’m going to use him. He is quite technical, has okay-ish finishing, but his main asset is his speed. He’s not that well rounded yet, but that’s due to his age (20 years old) i guess. I think he has no role on midfield due to his poor vision, passing and teamwork, so if he plays it will be as a striker. There is no data from previous seasons available on Mohammed.

Thomas Buitink - striker / winger 

Buitink is quite adept at all striking roles, except for the complete forward. Which is strange, given his very well rounded attributes. He can finish, has pace, is technical, can pass and has decent mentals for a 18 year old. He does need to work on his composure though if he really wants to make it as a striker. Currently i don’t think he will be good enough to start as our main striker, but i’m secretly hoping he will fight his way into our squad during the season. He’s amazingly talented and could potentially bring a big sack of money to Vitesse if he ever leaves. WIth his versatility he could be a good stand-in on the wings. Due to his lack of playing time there is very little information on his previous seasons.

Oussama Darfalou - striker

Darfalou is an interesting player for Vitesse. Good workrate, stamina and mental attributes are all in his favour. He’s not as good at finishing as i’d like to see though, his pace is also nothing special. He’s quite big (1.87) and can jump and head, so he might give us some goals with his head. He’s better than Buitink, but should really watch his back! Compared to Matavz he’s losing out, Matavz is the best striker Vitesse currently has. I’m curious to see what Darfalou can do in the early weeks of my management, with Mohammed and Buitink staying at Vitesse Darfalou is too expensive to keep on the bench (€34.500 a month, worth €1.600.000). We have records on his previous seasons.

Tim Matavz - striker

Matavz is the best striker Vitesse currently has. He’s already 29 years old, but still has 2 years on his contract. I intend to keep him around for this season and see what happens next summer. Matavz is not that well rounded, but brings great finishing, composure, anticipation and off the ball attributes to the table. The rest of the attributes are sufficient or mediocre, but i’m confident these skills will bring a lot of goals. We do play with a complete forward, which does not really suit Matavz all that well. We’ll see how that progresses during the season. In that sense, Buitink might be one to look out for given his well roundedness. 

Looking at past performances, it confirms what we already know. Tackling is poor, so is passing. Heading is at 40% which isn’t too impressive either. Yet Matavz manages to score every 244 minutes and scores 14% of the shots he takes. This season he should up that minute/goal ratio, or he’ll be stuck at 12-13 goals, but i’m hopeful of him managing that!
 

Edited June 22, 2019 by Unigo
added player analysis

[Unigo]

Part 3: time to start the season!
We're three days away from our first competition match against FC Groningen, so it's time to see what happened and where we are at now. So, what 'minor' things have i done:

• Recruited a whole bunch of staff members and replaced one or two bad ones. Especially the scouting department and the coaching teams of both youth teams needed some more staff members. The importance of the scouting department is obvious, but the youth coaching teams are also very important. If we can develop a young player well enough to get a place in the first team that saves us spending money on transfers. So, the youth teams are key to my success.
• I made a slight change in the formation after the first 4 games. I liked the complete forward, but it doesn't seem to match with my current strikers. So, i've went for an advanced forward now.
• In the player analysis i mentioned possibly wanting to sell Houwen and Klein-Holte. Houwen is still at the club and no one seems interested in taking him off our hands. Klein-Holte did get some attention though. After some negotiating we sold him for €140.000, but are still paying a small part of his wage bill. We did manage to get a clause that gives us 30% of the profit of the next sale, so maybe we’ll see some money coming our way in the future.
• Also some sad news: Karavaev played in the first friendly of the season but got injured afterwards. He’s expected to be out for 8 weeks with a broken upper arm. When i’m posting this (august the 9th) he’s still 10 to 21 days out of the running and probably won’t be entirely match fit after that. We don’t have a second right back (we’re looking for one…), so one of our youth products (Grotenbreg) is getting some gametime. We’ve also played van der Werff on the right wing with Clark taking his spot at half back, so we do have some options to deal with the absence of Karavaev. Mohammed Dauda is also out of play for a few weeks with sprained knee ligaments, then again he wasn’t really a part of our starting 11/subs anyway.
  

How did we do?
Before we dive into the statistics I’ll give you a brief look into our performance sofar. We’re still contending for the Europa League group stage, so not all our games were friendlies. We’ve played 6 games sofar:
Friendly: 2 - 1 win against Helmond
Friendly: 0 - 2 loss against Karpaty
Friendly: 3 - 1 win against FC Lienden
Europa League: 0 - 0 draw against Lillestrom (home)
Europa League: 1 - 1 draw against Lillestrom (away)
Europa League: 2 - 0 win against Partizan (home)

I was a bit disappointed with the performance against Karpaty, we were also struggling with Lillestrom, while we should be the more stronger side. The 2 - 0 win against Partizan (Beograd) was good though. We’re not there yet, i think there is another qualification round after this with much stronger teams, so we’ll see where we end up. 

The good stuff… numbers!
Interesting enough Football Manager doesn’t really gather data on the friendly games. The friendlies and games in august are quite useful to me though, because they will give me a look in player performance while i’m still able to transfer/loan players who won’t play or are not performing well enough. So, i’ve made a table for the friendly games, this is added below. For the ‘official’ matches i’m making use of exported football manager data, this is also added. Below the images we’ll look at what stands out.
Friendlies

Official games

Goalkeepers
A bit as expected, in the friendly and European matches Pasveer and Eduardo are both still in contention for the spot in the starting 11. Eduardo gets to play in the next game against Partizan, Pasveer will start in the first game in the competition. Maybe those two games will help me decide who will be the first keeper for the next few games.

Defenders
In the friendlies our Tackling ratio was quite good, our heading ratio does need some improvement though. Then again: in our player analysis we already concluded that heading could be a problem area of this team (and thus the defenders). In the Europa League these conclusions still stand: good tackling, poor heading and great passing. 

In our defense we have a few players in contention for starting-11 spots. Clarke-Salter got the spot over Doekhi in the European games and has taken full advantage of that chance. He plays well, so will stay in the starting 11 for now. Doekhi will get chances as a substitute though.

On the left back position Buttner and Clark are battling over the starting spot as well. In the friendlies Buttner displayed more accurate passing and managed to get two assists in. In the European games Buttner proves his worth by (just as Clark) showing good performance on tackling and passing, and again getting in an assist. Based on this performance i’m feeling i need to keep Buttner around, i might try to give him a new contract in the next few months as well to try to get some revenue out of selling him after this season (unless i keep him..). Clark is on a hefty wage, but i’m not ready to give him up yet. He’ll get some gametime in the next few games both as a left back and as a half back to see how he does. I don’t want to sell him, i think he’s the perfect replacement for Buttner if he leaves or retires. Maybe a loan could be an idea?

On the right back position we are lacking players. Karavaev is injured, so we pulled Mats Grotenbreg out of our youth teams to fill up Karavaev’s spot. His attributes are nothing special, nor is his potential that good. He usually plays on the left as an inverted back, but given his right-footedness he’s good on the right back spot as well. In the friendlies and the European games he played rather well! Good heading, passing and tackling ratios, and only three mistakes in the friendlies and none in the European games. Maybe i’ll stop looking for a stand-in for Karavaev…

Midfielders
As a half back van der Werff is doing a solid job, so he’s certain of his spot for the near future. He did struggle a bit against strong opposition in Europe, but should be good enough to deal with Eredivisie-level players. Serero, Foor and Bero are displaying good passing, decent tackling and (as expected) could do better in heading. Bero is the exception, he did quite well on all three departments in Europe. Serero managed to get get an assist in Europe and played quite well. He’ll be the first choice as a deep lying playmaker for now. Bero really outperformed Foor by scoring 2 goals in the friendlies and another one in Europe. He’ll get the box to box spot for now.

Wingers
On the right wing Odegaard scored once in the friendlies and another goal in the Europa League games. He’s the only left-footed player capable of playing on the wings, so he’ll be our main pick for now on the right wing. 

On the left Beerens did play two games in the friendlies and showed some decent passing. He didn’t manage to get an assist or goal though, so that’s disappointing. His contenders did a bit better: Linssen shows equal passing, but managed to get 2 assists in the friendlies. Buitink performed about as well als Beerens and so did Musonda. For now Linssen will be the main man on the left wing. I don’t want to keep Beerens, Musonda and Buitink all three to keep the bench warm. Buitink is a huge talent and isn’t that expensive, Musonda is free due to the loan from Chelsea so i intend to sell Beerens.

Strikers
And finally, the strikers. I’m mostly looking at Matavz and Darfalou for this spot, Buitink occasionally can play as a striker but is versatile enough to also play on the wings. So far neither of the two are impressing me much: Darfalou only played a little more than one game, but has no assists or goals to show for it. Matavz did get an assist in the friendlies and a goal in the European games, but that’s still below what we need. Their performance might be due to the complete forward role not suiting them though, so let’s see how they do when they can play as an advanced forward. 

 

Now, time to start the competition!
 

[Deisler26]

*ears perk up*

Welcome to the Netherlands!

Edited June 23, 2019 by Deisler26

[ElCuriosoJr]

Something new here? I’m a supporter of statistics and an excel spreadsheet-fanatic. 

hopefully you find some time to update this.

[Jimbokav1971]

I like the idea of this but of course it's very hard to do. Isn't the idea behind "Moneyball" not just the use of statistics, but finding the value in under-appreciated statistics? (or something like that). I'm sure that was the definition rather than just stats. Otherwise everyone would be after the same players. The baseball theory was very much looking at particular statistics that other were over-looking for some reason. No? 

Anyway, whatever the answer, good idea and will be interesting to follow. good luck. 

[Edit]

If I was going to do something like this, (and it's not likely at all so feel free to completely ignore me), then I wouldn't be looking at such stats as tackles, headers, passes and goals, because they are "obvious" stats. I would be looking at things like....

Interceptions/tackles/headers for central defenders. I know you are playing 3 at the back, (I think you said that and I saw mention of a HB somewhere), but even if you are playing 3 at the back and a possession based system, who cares how a defender passes really, (given the right set-up). 

I'm not a key fan of passes as a stat because it can be too easily skewed by tactics/formation/opposition. Key passes is better, but as SI are unable to provide an accurate definition, then I find it hard to trust that either. What I do like are assists, especially for central players. 

Cross completion and dribbles is something I like the look of for wide players and strikers need you look further than shot conversion? If a player scores 10 goals in 10 games but had 50 chances, does that make him better than a player with only 7 goals from the same 10 games but from only 20 chances? 

Like I said anyway, good luck. 

[Edit]

This is what I was talking about.

So basically concentrating on certain statistics that might have been undervalued in the transfer market in comparison to others. 

It's common sense that goals and clean sheets are "over-valued" statistics, (more sought after is probably a better term), but part of the reason that Spurs did so well last season was Hugo Lloris having a ridiculously low goals conceded stat in comparison to expected goals conceded, and at the other end Spurs also converting a very high number of chances at the other end of the pitch. Some might suggest that they might not be performing hugely differently this season to last, but the extremes of the stats last season make it appear otherwise. 

Anyway, apologies for hi-jacking your thread. Really interesting and will keep an eye on it. Thanks. 

Edited October 12, 2019 by Jimbokav1971

[Unigo]

On 07/09/2019 at 14:56, coolestrock said:

Ive been trying to set up a custom screen for this , where is shots to goal ratio and minutes/assists in the views section ?

Don't think it's in the game, i created it myself in google spreadsheets (dividing goals/shots and dividing minutes/assists).

On 10/10/2019 at 09:41, ElCuriosoJr said:

Something new here? I’m a supporter of statistics and an excel spreadsheet-fanatic. 

hopefully you find some time to update this.

Sadly enough i've ditched the save after my laptop went away for a repair. I want to try again when FM2020 comes out though!

On 11/10/2019 at 19:58, Jimbokav1971 said:

I like the idea of this but of course it's very hard to do. Isn't the idea behind "Moneyball" not just the use of statistics, but finding the value in under-appreciated statistics? (or something like that). I'm sure that was the definition rather than just stats. Otherwise everyone would be after the same players. The baseball theory was very much looking at particular statistics that other were over-looking for some reason. No? 

I think you're right. Though i think it's also a bit about ruling out the non-relevant stuff (he throws funny, for instance). It's kind of hard to find that undervalued statistic for football without proper data-analysis i think. BUT! i'm all open for discussion on what i could use!

Strikers
The most obvious stats for attacking players are assists and goals. But we want to look at less-obvious stats. For strikers i'm quite happy with the shots/goal percentage. Of course tactics and teammates have influence on striker performance, but this stat does tell us something about how efficient the stikers are with their chances. One could argue that the moment of strikers is important as well, because if they move in a bad spot and don't get the ball, they might not even get a shot in. So, maybe a striker with a slightly worse shot/goal percentage but with a higher number of shots is actually the better player. Then again: maybe the player taking more shots is preventing teammates in a better position from getting the shot on goal. Maybe we can also look at the shots per 90 minutes to give an indication of how triggerhappy the strikers are. But, to conclude this rambling: shots per 90 minutes can also tell us our striker has great movement off the ball and is always at the right spot...

Midfield
Midfielders are a tricky one too... (they are all tricky, to be honest). The obvious statistic is goals, assists, and maybe pass%. Pass% to me is quite useful! I also like the chances/90 minute stat that football manager provides us with. Finally, also the mistakes (calculated per 90 minutes) are a good indicator i think, because that might lead to chances for the opponents. But: i'm curious if someone has a better idea...

Defenders
Bit of the same as the midfielders i think, but the tackle% and mistakes are more important. For the backs i think pass% and crosses completed are also good indicators.

Goalkeepers
Think we have a good one here with number of shots dealth with and how many goals came from those shots. The more obvious statistics would be saves and clean sheets. But, those tell us more about the functioning of the entire team. How many shots on goal a goalkeeper actually saves is way more relevant. Maybe the goalkeeper of the top club in the league has a lot more clean sheets, but the goalkeeper of the relegated actually has more shots saved.

And finally...
I found the spreadsheet of the statistics at the end of the season. I'm not gonna share the entire table, because a lot of players are new to the roster. I do wanna share one interesting comparison though: Grotenbreg versus De Col. With the injury of Karavaev i purchased De Col to replace Karavaev, who i also succesfully sold in the winter transferperiod. De Col played almost three times the games that Grotenbreg played, but the results are still interesting:

So, what stands out:

• Grotenbreg's tackle ratio is better than De Col's. But, this might als come from a small sample size. Generally my backs didn't tackle that much, so that would make sense. Then again: Grotenbreg did manage to make a key tackle, De Col did not.
• The headers-won ratio is for both backs at 63%. Passes completed ratio is slightly better for Grotenbreg with 81% versus 75%. This is also shown by the key passes per 90 minutes, Grotenbreg is beating De Col here too with 0.48 key passes per 90 minutes versus 0.29 key passes per 90 minutes.
• But it does not stop there: both backs gave one assist, but with the playingtime factored in Grotenbreg needs less time to get to that one assist. Also in chances created De Col has no rating, because he did not create any chances. Grotenbreg creates 0.32 chances per 90 minutes. When we look at crosses completed Grotenbreg also was better, with 23% versus 14%. 
• Lets give one to De Col then: he scored, Grotenbreg did not. WIth 1 goal (out of 4 shots) he was more of a threat in front of the goal, Grotenbreg did not place any shots on goal, unfortunately.
• But now the final one, for Grotenbreg: The number of assists is uncomparable due to time played (grotenbreg 6 versus 50 of De Col). But, with time factored in De Col made 2,38 mistakes per 90 minutes, Grotenbreg made 0,96 mistakes per 90 minutes.

So, based on these numbers i think Grotenbreg has beaten De Col. If i still had this savegame, i would have played him more next season to try to get playing time more evenly divided among the two players to get a better comparison. Then again, even this season's performance is remarkable: Grotenbreg was better than De Col, but Grotenbreg is worth 7K (and not a wonderkid...), De Col is worth 3.3 million. That's 471 times more than Grotenbreg...