5 replies to this topic

[Pompey_Dan]

I recently thought it would be interesting to take a look at how a player from Football Manager measured up to their real life counterpart, and here is what I found when I compared Robin Van Persie's performance against Wigan at the weekend with the Football Manager version of the same player in the same fixture.

The article was posted on my blog (TheBoyDoneGood) last night but i thought it would be nice to share it with my friends at FM-View too.

All of the data for the real Robin Van Persie has been gathered from FourFourTwo magazine's excellent iPod application Statszone, the Football Manager data was gathered from the corresponding fixture played out in the match engine.

Arsene Wenger used a 4-2-3-1 formation at the DW Stadium with Van Persie the one up front, they won the game 4-0 with Robin Van Persie scoring once. Football Manager's virtual Wenger employed a 4-3-3 formation but again Van Persie took up the role of lone striker, the game was won 2-0 with RVP netting both goals.

The passing map from each game demonstrates how similar the movement and contribution of the real and virtual Van Persies were. In each he operates quite high up the pitch, combines long and short passes to equal effect and occasionally drops deep - it's also clear how both the real and virtual RVP utilised the full width of the pitch.

In the real life fixture against Wigan Van Persie completed 40 out of 48 attempted passes (83.3%), although the pass completion rate of our Football Manager version is similar (88.5%) he doesn't see as much of the ball attempting only 26 passes. This is largely down to differences in formation and strategy, Arsenal achieved over 65% possession against Wigan on Saturday with a midfield that got forward and supported the striker, by comparison Virtual Arsenal managed a lower 58% and RVP was more isolated.

When it came to the chances each player created (4 for the real RVP, 2 for the Football Manager version) the maps do look different, with 3 of 4 chances created by the real Van Persie coming from corners. The virtual player made use of space in the wide positions to play teasing balls for the on rushing midfielders.

In each version of the match Van Persie managed four shots and the maps below illustrate again how the real and virtual players got into similar positions, albeit the virtual RVP was more likely to have a pop from distance.

So there were some striking similarities between the way Robin Van Persie performed against Wigan for real and the way in which his virtual counterpart set about the same fixture in the Football Manager match engine. I like to think (in a small way) this helps demonstrate the thought, effort and research the game's developers have put into the production of Football Manager although it will be interesting to see if future experiments yield similar results - they'll be more like this coming soon!

[Murray]

Great article there, I look forward to seeing more along the same lines, as I think this is something not prevalent enough in the scene.

[Pompey_Dan]

Murray, on 07 December 2011 - 06:08 PM, said:

Great article there, I look forward to seeing more along the same lines, as I think this is something not prevalent enough in the scene.

Thanks Murray, I'm glad you enjoyed it mate... I'll be sure to pop back to FM-V with some more stuff later

[gkennedy]

Heh that's really quite interesting. Only major difference I see is 'running past opponent' which he did not do at all irl, but did 4 times in FM. I don't know van Persie's style, would he normally do that or not in real life?

Very interesting stuff. I love it. I'm a statistic geek so love that.

++edit++ actually just had a thought; do you think the difference in formation accounts for the lack of running past players? I'd say he'd have more chance to run with the ball in 4-4-3 than as playing as a lone striker (even though they are quite similar).

[killie97]

gkennedy, on 08 December 2011 - 10:11 AM, said:

Heh that's really quite interesting. Only major difference I see is 'running past opponent' which he did not do at all irl, but did 4 times in FM. I don't know van Persie's style, would he normally do that or not in real life?

Very interesting stuff. I love it. I'm a statistic geek so love that.

++edit++ actually just had a thought; do you think the difference in formation accounts for the lack of running past players? I'd say he'd have more chance to run with the ball in 4-4-3 than as playing as a lone striker (even though they are quite similar).

4-4-3? Good tactic that. A lot of managers say it's very hard to play against.

I'd think the same thing if the other team had 12 men!

[Pompey_Dan]

lol

gkennedy, on 08 December 2011 - 10:11 AM, said:

Heh that's really quite interesting. Only major difference I see is 'running past opponent' which he did not do at all irl, but did 4 times in FM. I don't know van Persie's style, would he normally do that or not in real life?

Very interesting stuff. I love it. I'm a statistic geek so love that.

++edit++ actually just had a thought; do you think the difference in formation accounts for the lack of running past players? I'd say he'd have more chance to run with the ball in 4-4-3 than as playing as a lone striker (even though they are quite similar).

Have to agree - this is likely down to the virtual RVP being that bit more isolated than the actual RVP was, which I can only contribute to the difference in formations.

Glad you enjoyed the article, I'm a stat geek myself and you can see that this is reflected in my writing style.