Jump to content

Eae rating


faction02
 Share

Recommended Posts

Eae,

I suggest here an unformal eae rating for op player. You can compute your own aea rating using the simple following formula:

     eae rating = (Number of times you won againt the top player)/(Number of times you played against that top player)

If we call this top player borg- for example, then borg- would have a constant score of 0,5 and he would be the benchmark to beat. Other players would have a score between 0 and 0,5. For those with a 0 score, keep trying beating the top player or just keep on having fun! 

Eaaaaaaaaaaaaaaaaaaaaaaaaeeeeeeeeeeeeeeeeeeeeeeeeeeeeee !!!!

f02

  • Thanks 1
Link to comment
Share on other sites

On 12/3/2018 at 11:21 PM, faction02 said:

that's not worse than most.

Congratulations !!

1 indeed, statistically speaking, that score would be poorly distributed (massed around 0 and some few values above). 

2 what would happen if a player wins more than half of his games vs borg ?  Feldfeld, Val, Aristol ..  ? 1 game is enough is to be 100%  He would be the new benchmark ?

3 this score is calibrated vs 1 player and not all of them. So the gaming style matters and borg can decide to improve your score or even refuse to play vs you

 

To stick to your idea, better take a well rounded player like ffm as benchmark  and make this

           ln (Number Wins /  Number Fails)      

So it will gravitate around 0 for those with same level than ffm ;  positive for those who won most games ; negative for those who lost most and no DIV/0 or infinite problems. The score distribution will look better

Edited by JC (naval supremacist)
  • Like 1
  • Confused 1
  • Sad 1
Link to comment
Share on other sites

12 hours ago, JC (naval supremacist) said:

To stick to your idea, better take a well rounded player like ffm as benchmark  and make this

           ln (Number Wins /  Number Fails)      

So it will gravitate around 0 for those with same level than ffm ;  positive for those who won most games ; negative for those who lost most and no DIV/0 or infinite problems. The score distribution will look better

This formula isn't better in any way, and it makes the infinity problem worse.

Link to comment
Share on other sites

12 hours ago, JC (naval supremacist) said:

well, i wrote  " most " to say i want to avoid the extreme case in which 1 of the player has 0 wins (thus avoid 0 and DIV/0) . I should have been more clear.

The problem is not you being unclear, but you being wrong. The original formula has only one scenario when it is undefined, that is when no game has taken place. Your formula, on the other hand, has an infinite number of cases when the result is undefined (all cases when the player has never won or never lost).

12 hours ago, JC (naval supremacist) said:

Yes; its better formula as its centred on 0 and has symmetrical shape. But seriously,  idc if its better or not.

It's not really clear what you mean by symmetrical shape of the formula or why your formula should be rated higher in respect to this property. This is where you should be clear.

Except from the definability problem detailed above, I don't think your formula is an improvement due to these additional reasons:

  • Most players' rating is gonna be a negative number, generally unbounded, but typically in an interval too narrow to consider rounding to an integer. The original produces ratings between 0 and 1 and can be naturally thought of as a percentage.
  • With the exception of zero, your formula always yields a transcendental number. I'd rather say that my eae rating (i.e. win ratio against the reference player) is 0.1 (i.e. 10%) instead of having to use -2.197 as an approximation.
  • Your formula doesn't give an intuitive idea of how good a player is. You can mostly only tell if they beat the reference player more often than they didn't. When you get a simple 5% instead, you just see they have won one in every 20 games so far. Your formula doesn't provide any more information than that anyway.
  • The original formula is simple enough for anyone to compute in their head; yours isn't.
Link to comment
Share on other sites

I like the Pavlovian reaction of those 3 .. auto-congratulates each other with smileys even if they don't really get the point. Lazy minds and poor personalities ; facebook generation.

Boudica, sorry to say that but it's better u just keep "liking" the post of 2 others to avoid wasting your time and mine..  i said i really don't care if this is better or not as the idea in the first place was a joke : making score based on 1 player... (even the name eae is a joke)   But i feel i have to answer simply to avoid people to get me wrong.

5 hours ago, Boudica said:

The problem is not you being unclear, but you being wrong. The original formula has only one scenario when it is undefined, that is when no game has taken place. Your formula, on the other hand, has an infinite number of cases when the result is undefined (all cases when the player has never won or never lost). 

nope .. that makes only 2 cases my friend .. not an infinite number of cases. ( exemples : 10-0  and 0-10  or  7-0 and  0-7  or .. etc)   And those 2 cases can be labelled as +Inf and -Inf .. just like there are with the kd ratio. Am I wrong or unclear ?

5 hours ago, Boudica said:

It's not really clear what you mean by symmetrical shape of the formula or why your formula should be rated higher in respect to this property. This is where you should be clear.

- it's symmetrical and centred on 0.. .. i give you a simple example : ln(0/10) = -Inf ,   ln(3/7) =  -0,8473  ,     ln(5/5)  = 0  ,  ln(7/3) = 0,8473  ;  ln(10/0)  = +Inf   ..        I hope its clear now.

- I was just saying in first place that taking ffm as the benchmark could be a better idea to avoid having like 99%  similar score whatever the formula taken;  then i precise that taking the "ln" scale softens the distribution of values (to avoid the massing effect).  But im agree with you to say that it will be hard to interpret.

 

@PhyZic the fnny fact is : I teach math and i create a video game ; while you mostly spec games of others and write crap on this forum. I have probably more respect for a "Retard" than something like you

@ the forum admins :  the least, it's not to delete the answer of someone being insulted. Personally,  I won't answer on this topic anymore as it's going nowhere

  • Like 1
  • Sad 1
Link to comment
Share on other sites

  • Stan` locked this topic
Guest
This topic is now closed to further replies.
 Share

×
×
  • Create New...