The Numbers Game is a significant step on in football analytics writing from Simon Kuper and Dr Stefan Szymanski’s Why England Lose. Both books follow on from Moneyball and all three are highly recommended for anybody who believes that the beautiful game can also be a thinking man’s game.
Football Manager, of course, has been a stat-driven game from its very inception, so when Sports Interactive kindly sent me a Beta version of its new FM14 iteration, I wondered if some of the evidence-based advice offered to real-life managers in Chris Anderson and David Sally’s excellent book would make sense if used in the computer game and help me bring success to a virtual Derby side.
The O-Ring Theory
TNG discusses a recruitment theory derived from the tragic example of the Challenger space shuttle, a vastly expensive project which came to a terrible end because a humble O-Ring seal failed to do its job properly. All of that expenditure, planning and hard work was, very sadly, completely undone by something which must have seemed comparatively minor to the designers and possessed deficiencies which were overlooked, with disastrous consequences.
Based on analysis of team performances across various top leagues around the world, the O-Ring theory asserts that a team is really only as good as its worst player. As good as Robin van Persie is, he can only score you so many goals – if you’ve got Claude Davis at centre back, you’re simply not going to win every week.
Therefore, rather than blowing your transfer budget on a star striker, you should establish where your team is at its weakest and reinforce that position as a priority.
Applying this advice to my FM14 Rams squad, I find that the lowest rated player in my team is Craig Forsyth, who, according to the game at least, is not suited to playing at left back, as he doesn’t have the necessary defensive abilities. Having identified my O-Ring, I place him on the transfer list and start searching for his replacement.
Looking at my team, it becomes clear that the best system involves a midfield three of Hughes, Bryson and Eustace or Hendrick; but there aren’t too many options to play further forward in wide areas. Johnny Russell is my best bet to play on the right, but isn’t particularly well suited to that position, so I start looking for a new right winger. Once that deal is done, I will look again to see who is now the weakest link – and so on, until I’ve spent my meagre budget…
Defenders and goalkeepers are undervalued
TNG makes it clear that the popular fan obsession with glamorous, megabucks centre forwards is misplaced. When you look at the data, as Anderson and Sally have done, it turns out that you are more likely to win the game by keeping a clean sheet than you are if you score two goals.
Not conceding, an analysis of ten seasons of Premier League games showed, resulted in an average of around 2.5 points per game, whereas scoring two goals yielded an average of just over 2 points per game. Even conceding one goal per game dropped your average points score to about 1.5.
So, although it might not be anywhere near as ‘sexy’, the numbers imply that you are better off fortifying your defence before you start dreaming of pushing the boat out to sign a Jordan Rhodes.
The real-life Rams have found this out to their cost this season. Despite being the top scorers in the division at the time of writing, they are only eighth in the league and it could easily be argued that this season’s defensive frailties are what ultimately cost Nigel Clough his job.
Looking at the FM14 Derby squad, I find that Sammon, Martin, Ward and Russell are all pretty good – certainly good enough to ensure that a better player would be expensive. With limited funds to spend, it is more likely that I can strengthen at the back, so my first transfer deal sees Paul Coutts sent to Blackpool for about £500,000 plus left back Bob Harris, who is considerably better rated than Forsyth. Although not all of the fee is up-front, it still solves two problems for me in one deal – in the game, Coutts can’t play central midfield and doesn’t fit into my plans, while Harris improves my back four.
The theory that forwards are overpaid in comparison to defenders also guides my decision to sell Jamie Ward. I offer him a contract extension, but his bolshy agent demands £22,000 a week plus bonuses for the little Brummie. While he is good, he is not that good and the agent is not prepared to negotiate the wage down by enough to make a deal possible.
Plenty of clubs express an interest when I offer Ward out, but none are prepared to pay the amount I want. I therefore start looking for player swaps and manage to arrange a straight exchange with Birmingham for Chris Burke, who is what Harry Redknapp might call a ‘top, top’ Championship right winger. Free agent Albert Crusat, released by Wigan, then comes in to play on the left. I’m pleased with this, as I’ve essentially got two good wingers for about the same level of wages that Ward’s agent had demanded.
58 / 73 / 79
One of the most remarkable pieces of research in TNG focuses on the timing of substitutions. A theory has been developed which suggests an optimal time to make changes during a game, especially when you are behind. To give the best chance of turning the result around, the numbers suggest, the first sub should be introduced before the hour mark and the other two brought on before the 80th minute. In other words, you need to think of it as a 14-man game and give all three subs a chance to make a difference.
Anderson and Sally contend that managers are too often biased against taking off one of their starting XI, due to their natural faith in the players they chose – to change things too early would be tantamount to admitting that they got it wrong in the first place. Moreover, performance data shows that players pace themselves through games, in order to give themselves a better chance of lasting the 90+ minutes. The last thing they want to do is burn themselves out, so they generally operate at about 90%. By the time they are visibly flagging, it is much too late – they haven’t been able to push themselves to their physical limits for some time.
TNG states that the manager should be prepared to overcome his in-built selection bias and bring a player off before it’s evident that he needs to be replaced, to give a fresh player a run and maximise the chances of a positive result.
I decide to strictly obey this theory during my FM14 tenure at Derby. During pre-season, it helps to make sure that none of my first-choice players pick up an injury and it also gives me a chance to look at plenty of fringe players. Into the season, it also seems to make a real difference in terms of keeping players fresh.
Contractual wrangling has become ever more byzantine in the new FM and Financial Fair Play has been introduced, with your FFP cash limits displayed while you are making offers for players.
I am annoyed but not really surprised to find that new signings are now demanding unused sub fees as well as substantial appearance fees. When I promote Eric Steele to assistant manager, he requests a clause waiving any compensation should he be offered a manager’s role elsewhere.
For matches, the tactics interface has been simplified considerably – and is much better as a result. The team tactics tab has been stripped down – no more timewasting or defensive line sliders, no more tick boxes to use a playmaker or target man. You just set the overall mentality and then use buttons to add any extra, simple team instructions you might want to use – push up, drop deeper, play wider, etc. This means that setting each player’s individual role and mentality properly is now of paramount importance. The changes have made things much less fiddly and I think are good on the whole, although I still have a bit of a beef about the way set piece instructions are handled.
The set piece wizard should be brought up automatically whenever you make a substitution – if it’s a straight like-for-like, you could just skip through, but if you’ve just introduced Ward for Sammon, say, you’ll probably need to make sure Ward doesn’t try to man mark at corners – if you forget to go to the relevant screen to change it, disaster could ensue.
I also find that for some unknown reason, formation changes and some tactical tweaks seem to erase all of my carefully set opposition instructions, which drives me bonkers.
Moaning aside, a wealth of new player roles have been introduced, including enganche (the Argentinian term for an attacking central midfield playmaker, as exemplified by Juan Román Riquelme – the game says that Hughes is suitable to play this position) and regista, an Italian term for a ‘quarterback’ playmaker operating from deep, like Andrea Pirlo (Eustace can do this). During a pre-season friendly at Burton, I amuse myself by asking Conor Sammon to play as a ‘false nine’.
Football Manager is a game which has always had the potential to completely take over my life, although these days, I only seem to play it in sporadic binges. It’s very different to a traditional football video game like Pro Evo or FIFA and a mate of mine once very rationally explained to me that he thought FM was for losers – “What’s the point of a game if you can’t taste your friends’ tears?”.
I understand that an online version is now available, meaning that you can pit your team against other people, but this doesn’t really appeal to me. It’s not as if you control the players, as in FIFA or Pro Evo.
There’s a time and a place for both styles of game – you can’t beat a bit of Pro Evo and a few beers with your mates, but sometimes, an evening in on your own with Football Manager is a proper treat. Unless, of course, you can’t buy a win, in which case, it becomes a mind-warping, obsessive experience.
Enjoy Football Manager responsibly…