Btw, the reason for this post is that I'm on one of my regular boat trips from Stavanger to Bergen, and I need to kill some time!
So what influences headers won? Several forum posts go a long way indicating that jumping is the only attribute needed to win a header.
Well, let the numbers answer the question.
Myth:
Jumping is the only attribute needed to win headers. Forum post link to one of the threads claiming this.
Data set:
1932 random players in Voller (sorted by apps, excluding GKs)
Attributes looked into:
acc agg agi ant cnt bal cre det dri hea jum off pac pas pos sta wor str fla mar tac
Result:
Jumping is the most important attribute for winning headers by a landslide, over 5 times as important as the next attribute. The importance of attributes for headers won in ranked order: Jumping - Heading - Strength - Anticipation - Marking
Conclusion:
FML Myth confirmed!
Jumping is the only attribute needed to win headers!
What about height?
ReplyDeleteDid. Not. Check. :)
ReplyDeleteBut I'm willing to bet a day's salary that the result would have been the same, because height and jumping are strongly correlated.
Why would the heading attribute have ANYTHING to do with winning headers? That seems particularly dense...
ReplyDeleteIt can partly be explained because Heading has a strong correlation with Jumping. Strength, however, is even more correlated, so there has to be other factors influencing the result too.
ReplyDeleteInteresting.
ReplyDeleteI thought that jumping was ability to reach height and connect with the header.
I'd have thought that strength would be needed to determine who would get to the header once on target to reach the ball.
That would leave heading as the accuracy of the header if the player actually makes nice contact?
That was expected, but it's really sad that SI aren't doing anything quicker to fix the ME...
ReplyDeleteI wonder who is testing this ME, and if SI haven't lost the plot...
Voller - BG UNITED
You could be right Rik. If jumping is equal, perhaps strength (or heading) decides who wins the header.
ReplyDeleteIt sure looks like most times a jumping 16 beats a jumping 15 to the ball irregardless of other stats though.
Heading, as far as I have understood it, is exactly as you describe it, the ability to control the ball once the header is won.
Remember though Hristo, we are a lot more competitive compared to the casual FM2010 player. There are a lot of things under the hood of the ME that we don't know about, and it must be immensely difficult to balance it all.
ReplyDeleteBy highlighting stuff like this, I'm confident that SI will take notice and rectify the flaws in due time.
I think what Hristo (and I) is most frustrated about is what we perceive to be the slow pace with which these changes get into the Live client. I understand you need a period of testing in Beta before you can actually push the changes to Live (I think the testing process is severely flawed but that's another conversation) but the fact that SI really only have 1 coder who is full-time committed to the ME (PaulC) is a joke in my opinion when it is the single most important feature of the game!
ReplyDeleteThe amount of time from the identification of these issues by users to the day that they are finally changed in the Live client is currently unacceptable.
my 19jmp 16str 16hea 12ant 12mark DC loses to a dc with 17jmp 19str 19hea 18ant 16mark without exception, so I guess theres a limit to what only jumping can achieve :(
ReplyDeleteThis is probably where a statistician would kill me, but the outcome from my regression gave the following T-values:
ReplyDeletejum 29.1
hea 5.5
str 4.7
ant 3.8
mar 2.9
If you do the maths on the two players mentioned, I would actually expect the second one to win the header:
29.1*19+4.7*16+5.5*16+3.8*12+2.9*12 = 796.5
29.1*17+4.7*19+5.5*19+3.8*18+2.9*18 = 809.1
So you could be right, it's not only jumping by itself, but it is by far the most important factor.
Just ran this very equation on the player db i have for Vol and Vald. The top guy that came out of the equation in Vol had 92% headers won, the top guy in Vald had 91% won, so I don't think I'm too far off..
ReplyDeleteOf course, they both had jumping 20.. :D
ReplyDeleteJumping isn't the only attribute needed to win headers but as far as the ME is concerned it may as well be :)
ReplyDeleteI guess the pass also has a big effect on the outcome, at least on corners and crosses. I just had the DC I previously mentioned beaten twice by someone inferior in every stat except strenght, which was 1 higher. I think its quite usual in corners that inferior jumpers get to head the ball, and yeah I mark the players personally so no issues there.
ReplyDeleteI'd love to see how big an effect different attributes have on the succesfulness of long shots, though I'm not sure that can be measured.
ReplyDelete