HTs on Global/Chances distribution

From Hattrick

Keywords: (Chances distribution), (Engine)

By: flameron 12547512.11 as reply to 12547512.1
To: Dj-DuDu 14.2.2009 at 16:36
Dj-DuDu wrote:

Hi , I played a match today which i got all 10 regular chances , Played no tactic and yet i'v got strange distribution between all sides attacks .

4 chances left
4 chances right
2 chances middle

in percentage its goes like
40% left
40% right
20% middle

the common percentage (befor indirect free kicks) was something like
25% left
25 % right
40% center
10% set pieces
obvious the Indirect free kicks change it a little , And set pieces (as much as i know) are something like 14% (if i recall correctly).

I never had anything like that , and i was wondering is this rare , Or not.

Edit - match link : 207345087


Using your %.
The chances not to get any free kick is only 35%.
The chances to get 2 attacks in the middle and 8 on the sides assuming the above happened is only 2%.
The chance to get out of 8 attacks on the wing 4 on each side is 27%.
If I use all the above (without the exact numbers) I get ~0.2% chance for such distribution.

Not common, only 1:500. The engine is probably broken :)

Keywords: (Chances distribution), (Engine)

By: flameron 12547512.14 as reply to 12547512.12
To: Catalyst2950 14.2.2009 at 16:49
Catalyst2950:

Well, there are only 30 matches a team can play per season, so most people should never see a match with a 1:500 occurance of events, especially not since they won't get all 10 chances - much less all of them reported - in all of those played matches.

:)


Actually very few games have a chance distribution with much higher probability.
Taking the guy numbers the most common distribution is:
1 free kick - 38%
5 attacks in the middle - 26%
4 attacks on the side 2 in each side - 37.5%

This is the most common and will give you only 3.75%

I think people have no understanding of probability and this is why quite frequently they find strange things in quite reasonable scenarios.

Keywords: (Chances distribution), (Engine)

By: flameron 12547512.21 as reply to 12547512.19
To: busawreck 14.2.2009 at 18:30
busawreck wrote:

The normal distribution gives (close to) 50% chances in the middle?


I just used the first post numbers.

Keywords: (Chances distribution), (Engine)

By: flameron 12547512.23 as reply to 12547512.20
To: Oligarf 14.2.2009 at 18:37
Oligarf wrote:

The most common distribution only hits 3.75% in this case.


Well, There are 5 ways to get 4 wing attacks (4r 0l, 3r 1l, 2r 2 l etc.).
When using only one option you get 3.75% but if you use all of those you get ~10%

Add cases when you have 0 free kicks and 5-6 middle attacks and 4-5 wing attack no matter how they spread and you around 20% or above that (did not bother calculating it).

I think that if we add the chance distribution to the equation we get even more options.

Overall I would say that the chance distribution sides wise is not problematic in hattrick but the chance distribution is too sensitive to random which cause most of the noise on the conf.

I know some see it as a flaw and I might agree to certain degree but my gut feeling is that this is part of Hattrick magic as well. That can become an interesting topic by its own.

Keywords: (Chances distribution), (Engine)

By: flameron 12547512.24 as reply to 12547512.22
To: busawreck 14.2.2009 at 18:40
busawreck wrote:

Well I'm talking about probabilities here. If the most common distribution is 5 chances in the middle with 10 chances, then it seems the probability for central attacks should be at least close to 50% (or am I wrong here?) which is quite interesting since the general view is that it is somewhere around 40% (rw25/c40/lw25/sp10).


I feel we misunderstand each other in some way so I will say again that I used the initial post numbers which are exactly like yours in my calculations. I am not sure what you refer here exactly. Do you feel I used 50% in certain place in my calculations by mistake? Please try to explain again.

Keywords: (Chances distribution), (Engine)

By: flameron 12547512.28 as reply to 12547512.26
To: Oligarf 14.2.2009 at 19:09
Oligarf wrote:
flameron wrote:
Overall I would say that the chance distribution sides wise is not problematic in hattrick but the chance distribution is too sensitive to random which cause most of the noise on the conf.

I guess part of that is just in the fact that we have more influence on how chances are distributed (through midfield) and hence it is more frustrating that something we perceive to be controlling doesn't go our way.

I wonder if you have any opinion on the situation that it is inherently harder to make 'the difference' at the higher rating levels.


I think the most interesting is to hear what you have to say about it. I hold the opinion that even though I know more about the game real formulas I have no superior understanding of such issues compared to the community as a whole and in some cases not even superior to some individuals. By saying what I think on such complex game design issues I think I can only make negative contribution to the discussion.

Not in a strict way I think it has some negative effects and some good effects. It is harder to be sure to win what you deserve team strength wise in the top. The best team might not finish first more often due to this and this might feel wrong. On the other hand it seems to create a much better competition on the top with lower power to short term strategies compared to a situation where it does not exist.