Subject: New match engine discussion

" excessive weight of the random part of random in the creation of actions. " I was replying to this. So maybe we should focus on the shot count versus resulted which seems to bring concern to most people.

First we must agree that failing to score in 10 attempt out of 10 is NATURAL. For example, let's assume a striker score once in an average of 4 shots (Actually no player in real life can be that efficient), the chance of that strikers failing all 10 shots is 5.6%, which means naturally you would expect to see a match which have shot count 10-0 ending up in 0-0 in every 15-20 matches.

Whether it is FUN is another question and is entirely personal. I personally find it weird to elinminate the situation you described as 'not acceptable' since it's so unnatural.

If we really were to change this, maybe first we would build a database of expected goals based on shot types and locations for different skills combination of keepers and strikers. Then in a match if the expected goal deviates from the actual goal by too much, the next shot would be generated until a goal is resulted (maybe through a do-until loop). Parameters would be tweaked so game balance is not destroyed.

The procedures for all these are likely to be very tedious that I cannot imagine anyone would do it unless he/she has a lot of time and gets paid really handsomely. Anyway, I strongly believe that developers have a lot of better things to do instead of messing with this.

This is based on my knowledge in scientific computing, real game coders may find a lot of inaccurate descriptions above.

That's indeed the problem.

And this 'discussion' keeps on going, doesn't matter how many times all is explained to these people, they will keep posting the same nonsense and even lies over and over again.

The biggest problem is, others will believe these people who post this nonsense about the ME :/ And these people are most of the times angry managers who post these angry feeling because they lost a match they expected to win.
(edited)

First we must agree that failing to score in 10 attempt out of 10 is NATURAL.
I agree

For example, let's assume a striker score once in an average of 4 shots

you better remember that sokker matches last 20 minutes.. not 95'... once on 4 shots can't be a good tuning..

(Actually no player in real life can be that efficient)

it is a useless to compare sokker and real football.

the chance of that strikers failing all 10 shots is 5.6%, which means naturally you would expect to see a match which have shot count 10-0 ending up in 0-0 in every 15-20 matches.

see above. That is a wrong tuning leading to a excessive weight of random on results.


Whether it is FUN is another question and is entirely personal. I personally find it weird to elinminate the situation you described as 'not acceptable' since it's so unnatural.

That's personal for sure, I agree.
Still I'm not the only one who feels it.

If we really were to change this, maybe first we would build a database of expected goals based on shot types and locations for different skills combination of keepers and strikers.

actual ME does it, but it can be done better maybe.

Then in a match if the expected goal deviates from the actual goal by too much, the next shot would be generated until a goal is resulted (maybe through a do-until loop). Parameters would be tweaked so game balance is not destroyed.

nonono.
I don't want correction over results during match.
I want better tuned probabilities in game, from the start.
A double superdivine passing/playmaking player (superdivine form) how many times can make a wrong pass? 1 on 5? no it's too much. 1 on 20


The procedures for all these are likely to be very tedious that I cannot imagine anyone would do it unless he/she has a lot of time and gets paid really handsomely. Anyway, I strongly believe that developers have a lot of better things to do instead of messing with this.

Oh yes, they should produce useless and old smartophone games.. and see users of sokker fall at 10k every year. no?

This is based on my knowledge in scientific computing, real game coders may find a lot of inaccurate descriptions above.

I find very funny the fact that someone believes that you can't use logic if you are able to code.
I think they don't even knows who are they talking to sometimes..
(edited)

The biggest problem is, others will believe these people who post this nonsense about the ME :/ And these people are most of the times angry managers who post these angry feeling because they lost a match they expected to win.

as always happens when someone criticize sokker,
someone appears to claim that who criticize is crying for his matches.
That's not my personal case:

I feel very lucky in sokker. I almost win without deserving it.

you better remember that sokker matches last 20 minutes.. not 95'... once on 4 shots can't be a good tuning

I can remember sokker matches last 20 minutes. Measures are already done to ramp up the number of shots to value close to real life (around 12 shots/team) (which means shots are created 4 times as frequently as in irl). Or are you suggesting we should lower this to 3 shots/team while giving each shot a high % of going in? This will give rise to more randomness.

it is a useless to compare sokker and real football.

In the last statement you are already comparing real life to sokker matches (which end up in the conclusion of each shots should have a high % of going in). Anyway, I think in a football manager we cannot totally ignore real life football. Sure we can take away undesirable elements like drug abuse/fraud/career-ending injuries etc., but some core elements should remain. Ultimately whether to keep an element can be determined by two things: whether the elements bring fun and whether they are something functional. See below and my previous post for discussion on the function of variability.

A double superdivine passing/playmaking player (superdivine form) how many times can make a wrong pass? 1 on 5? no it's too much. 1 on 20


That will lead to the problem stated in my previous post. Skills of players become more dominant over tactical choice in determining the results. A certain degree of variability is required to keep the tactic making aspect of the game functional.

nonono.
I don't want correction over results during match.


The other user JBlaze seems to want that. (Or I should say this is the only way I can think of to achieve what he wants)

The procedures for all these are likely to be very tedious that I cannot imagine anyone would do it unless he/she has a lot of time and gets paid really handsomely. Anyway, I strongly believe that developers have a lot of better things to do instead of messing with this.


I mean the developer, if they really have the time and motivation, should spend efforts on fixing some other issues, which is easier to solve and less constroversial. For example, can we make the players recognise the situation of open goals and take less risky shots against open goals? Can we fine-tune the behavior of ball launching especially near the side-line? etc. Not to mention issues others than ME.

I find very funny the fact that someone believes that you can't use logic if you are able to code. I think they don't even knows who are they talking to sometimes


umm... Scientific computing does (heavily) involve modeling and coding.

it is a useless to compare sokker and real football.

Yep, but it is a football simulation, not some other sport. So for example the number of goals should somehow be the same as normal football and not like basketball :/

A double superdivine passing/playmaking player (superdivine form) how many times can make a wrong pass? 1 on 5? no it's too much. 1 on 20

I know, you will say 'it's not like real football', but football players often pass incorrect in reality so nothing wrong if the same happens in Sokker. Super divine doesn't mean Superman!

as always happens when someone criticize sokker,

Double correction: As often happens when someone post nonsense based on feeling which has nothing to do with data and facts.


EDIT: and maybe you should check the posts from JBlaze again. The things he posted is not called criticizing Sokker, I would call that pissing on the forum.
(edited)

I find very funny the fact that someone believes that you can't use logic if you are able to code. I think they don't even knows who are they talking to sometimes


umm... Scientific computing does (heavily) involve modeling and coding.

I expressed it wrongly!
Exscuse me.
I want to say this:
"I find very funny the fact that someone believes that you can't use logic if you are UNable to code."

There are users that pretend that changes are impossible or that we already have the best possible sokker for mistical (and undisputed) "coding reasons" .
I think that is a quite stupid thing to say.

I think that is a quite stupid thing to say.

What is? It seems you are answering your own comment :/ strange

EDIT: and indeed, it is a stupid thing to say, but it is you who posted the comment, no one else said these things! So I'm glad you agree with yourself.

Now please stop putting words in people's mouth they didn't say and stop the insinuations; it's unfair and lying behavior.

(edited)

Not much time ago I posted some matches i dislike. 7 weeks separate them, if i remeber well, the second i lost 1x7, the first i won 5x2. And both matches were played with same tactics, same players, same orders.

People said it's because i don't know the ME. It seems that it's "you don't know the ME works" is the easy sentence to defend ME. For me is subjective. Some might like that same players, same tactics, same orders might result 5x2 or 1x7. I'm not in this side.

The discussion if you like the ME or not is something completely different, that's a matter of taste.

But when you post for example a comment something is wrong with the ME because you don't like surprising match results, then you are mixing things that are completely different subjects.

I believe there are more factors than your tactics.
It´s also the actual form, the actual team discipline, the actual teamworkrate of each player, perhaps also the weather conditions, the pitch condition, the stadium size etc.
You just speak about 2 games with same tactics and think that all parameters where the same. They where not.

there were already 2 identical arcade matches because of a bug and results were quite like this one....but we don't know if the engine is crap or that was 0.01% possibility of results being completely different like 1-5 and 7-2.

Indeed, there are so many parameters, it is impossible to play 2 exactly the same matches. As Faitas also posted, 2 arcade matches at the same time are already different.

The random factor is much more than 0.01%.

I'm not going to have another discussion about random :) But the number of surprise results, if that's what you mean, is most probably a lot higher as 0.01%. I don't know the exact numbers, but 1 in every 10.000 won't come close.

I think team discipline, teamwork, weather etc. can be treated as noise and have small effects. (Form is more tricky though)

I do get an impression that when two teams having similar strength encounter, their relative performance seems to be oscillatory. For example, team A is better at min 0-10, then team B at min 10-25 then team A at min 25-38 etc.

I hypothesize that the ME will lead to the dynamics being self-organized into such behavior (Just an educated guess with no proof whatsoever).

Then maybe some choices of tactic will amplify this effect such that once a team got the upper hand due to randomness, it is difficult for that team to lose the upper hand due to the self-organized dynamics. That would lead to results with bimodal distribution explaning the 2-5, 7-1 stuff.

I may be completely wrong. Actually it's a something very complicated. Interested plus users may arrange many arcade games with same set-up to let us get a better idea.
(edited)