On Football Manager Slack (footballmanagerslack.slack.com) there’s a channel (i.e., forum) called the Transfer Committee. FM managers post player profiles and ask for views on whether to sign the player.
It’s fascinating because each commentator works from the same data yet commentators see things differently. That is, they discern different patterns.
Personally, I find the sheer amount of data overwhelming. All those numbers! Where to start?
So I’ve been trying to work out a protocol for informing decisions, derived from methods I use in my day job (in engineering communications). Here’s my protocol for strikers:
- Check the personality. I do this even before I look at the profile: I’ve set personality to show up on my search page. Along with the obvious personalities to avoid (‘Low self-belief striker, anyone?’), I’m also very wary of balanced. I want something more than that.
- Check the injury record. Just because the scouts haven’t identified him as injury-prone doesn’t mean he’s not an injury risk. If the record’s dodgy, abort.
- Convert the numbers into colours. Not everyone will like this: changing to colours loses precision. On the other hand, it can make it easier to see the bigger picture. I can then ask myself, (a) what’s he excellent at? (b) what’s he good at? and (c) what he’s weak at? Where you set the thresholds between colours can be varied according to the level you’re playing at.
- I ask myself, ‘If he gets into a scoring position, how likely is he to convert the chance?’. Finishing is the obvious attribute to look at: but Football Manager is a heavily contextual game, so looking at any datum in isolation is rarely helpful. It’s entirely possible for a striker to be have a high finishing attribute, yet not convert chances. There’re several other attributes that are relevant here. They include first touch (to get the ball under control) and composure (so as not to lose their heads). Technique (so they can execute what they attempt), heading, and balance are amongst the others.
- Then I ask, ‘How likely is he to get into a scoring position?’ Relevant attributes include acceleration, off-the-ball, jumping, bravery, work rate, and stamina. Also anticipation, which I welcome, not least because it can compensate for lack of acceleration. Dribbling too, since it enables the striker to create the chance himself.
- ‘What’s he like in the air?’ In steps 4 and 5 above I’ll have already considered heading and jumping, but it’s good too to bring them together.
- ‘Will he create assists?’ I look at distribution (passing, crossing, heading) and also team work. If a striker has no end distribution product, they’re probably only playable as poachers. If a striker is good at these things but weak at #4 and 5 above, they might still make decent false-9s or defensive forwards.
- ‘How likely is he to score in other ways?’ Free kicks, penalties, and long shots (see postscript below).
- ‘What else do they provide?’ I look for reds and yellows in the hope of finding something like strong leadership.
- In the light of steps 4-8, I ask myself, ‘What roles could he play?’ I take some notice of the role assessments on the profile, but not that much. Those assessments are heavily influenced by what roles the guy has played in the past. It may be, for example, that the role information suggests he won’t make a good trequartista, because he has little experience in that role: yet he might have the necessary attributes and so be convertible.
- I check the career record. How often do they actually score? I’ve seen countless strikers who, judging from their profile, look world-beaters – but then when I look at their career records, I find they don’t actually put the ball in the net very often. A warning sign flashes.
- On the basis of all the above, I can then make up a series of sentences to create a portrait. Take, for example, Nimba:
My portrait would be: ‘He’s fairly determined — that’s OK. He doesn’t get injured much. So he’s a candidate. He looks pretty good at finishing, except his first touch will let him down somewhat. Though he doesn’t anticipate that well, he’ll be pretty good at getting into scoring positions. He’s good in the air. Though he’ll be good at getting into scoring positions and converting chances, he won’t create much for others, except with his head. He won’t be taking penalties or free-kicks and he won’t score from long shots. He’s got good agility and fitness, which will help him steer clear of injury despite his advancing years. Target man, no question: probably TM-attack, though his heading and teamwork may mean he could play TM-support’.
The above protocol doesn’t actually make the decision for me. But, by helping me to make sense of the data, it helps me arrive at a judgement. It’s rare to find that, after working through the profile, I don’t have a pretty strong view on whether he’d be right for the team.
Essentially this is a method for converting quantitative data into narratives — which is close to being a definition of what it is we do when we play FM
PS! On long shots: If they are poor at long shots, it may be possible to train them to desist. The problem comes when they’re quite, but only quite, good. They may then be tempted to have a go, yet rarely score. If the shots are on target, that might produce corners or poaching chances from goalie fumbles. But the long shots might also waste possession. The problem is that if they’re quite good, they may be resistant to training to desist. In which case, you probably have to train them to improve. Though it’s counter-intuitive where strikers are concerned, you can also instruct them to shoot less often — though of course that can damage their effectiveness in the box. If they’re not that great at long shots, the last thing you want is for them to score a dramatic, goal-of-the-month, long shot early in the season, because once they’ve done that it’s difficult to rein them back in.