A quick reversal, then, of the Insights post.
The Insights post discussed which away formations obtained the best results against the aggregated home formations, when I had FM21 playing against itself. The data told me that when the away team set up with some sort of 451, it was enough to get 113 results (wins or draws) in 168 matches – that is, all 28 formations, repeated six times.
What about the other way around? What home formations were most vulnerable to the aggregated set of away formations?
This is useful to know, because when your scout tells you that your next opponent is likely to play one of the vulnerable formations, you might consider easing off tactical training that week. If any old formation, any at all, is getting 110 results against the way the AI is setting up, then whatever tactical options you choose are likely to have a good chance of exploiting its vulnerabilities, and so you can tailor your training towards other things. The reverse is also true, obviously: more tactical training when the computer is likely to serve up a more difficult formation, as well as more looking at the AI’s match replays and chalkboard stats, in the hope of finding individual weaknesses in the opponent, to even things up.
Here, then, is the story of the data:
It looks to me as though all seven of those most vulnerable formations – the ones where the away team got most results – are light on bodies in central midfield, and central defensive midfield, areas. The screenshots below don’t take any interest in either roles, or duties, but just the distribution of bodies in particular areas of the pitch. Things like roles and duties, what to do out of possession, how hard to press, and where to set the line of engagement, can come later.
This is how the formations look in terms of the distribution of players, followed by a brief (non-data-driven) comment.
Only two in central midfield here. You won’t get much defensive help from the two higher up the pitch, although the wing-backs can make a decent contribution.
Central midfield totally vacated here, relying on just two defensive midfielders, though again, the wing-backs can be expected to take some of the defensive responsibility.
This one is interesting. The xG for the home team is massive here (+81.52 in only 168 matches), with a goal difference which is also the best of the seven formations I have highlighted as being the most vulnerable. I will look into this in more detail, and see if I can find a reason why something which appears to be comparatively strong, is actually weak. Perhaps it has something to do with two strikers. Even though I am all about killing my preconceptions, they die hard, and I don’t like more than one up front. If Football Manager reflects modern football, it will likely punish you for daring to commit so much resource to attack.
Again, the defensive midfield area has been left completely unoccupied, and, as before, I am suspicious of the two forwards.
I think that the two strikers are probably at risk of being bystanders here – there’s a huge gap between the most advanced midfielders, and the strike pairing.
This formation has more-or-less given up the box which represents the central midfield, and central attacking midfield bands. Based on what I concluded from the Insights post, that’s probably not a good idea.
Whether you’re prioritising defensive depth, or attacking depth, a basic 442 is unlikely to satisfy either requirement. All your central midfield roles and duties are carried out by just two players – they have to hold back the tide, as well as getting forward to support the attack. There isn’t the security of three central defenders behind them, and it doesn’t look to have the different bands you need to help build or suppress attacks.
I am now even more content that I need to choke up the defensive midfield, and central midfield, areas, as much as I possibly can. When the away team has lots of players in those areas, it tends to do better, and when the home formation does not, the team tends to struggle.