next »

[trewej]

May be lower the necessary xp limit a little to have some more High xpgains games :).
Because at the moment its very rare to see one of these.
I vote for 1100 average xp . That avgxp is reached quite often but not to often .

[JourneysEnd]

I dont like that idea at all, its way way too low for it to be high gains, I'd don't like to play every xp game with +30-30~ gains.

Maybe it could be possible to !highcm/cd challenge eachothers for high gains.

[zzark]

Ye I miss the high-xp gains games!

But I think 1100 is to low, maybe 1150?

[HaSu]

Ye I miss the high-xp gains games!

But I think 1100 is to low, maybe 1150?

+1

[noobwar]

1150 or back to 1200, cuz there are really no high gains games even with avg xp of like 1260

[Kenjin`-]

never seen a high xp gain game so far

[Slund]

I think it should be 1200xp.

[Leo]

There are 1 very wrong and 1 wrong things here:
1) The wrong one - static high-gain threshold
At the moment, the difference between 1st and 100th players is 1687-1030. If you remember before the reset is was like >2000 - >1200. This means, that the system of ratings is dynamic, not static, and putting a static and constant threshold is wrong(maybe now 1100 is high, but after a month it can be low...).

2) Te very wrong thing - when is it high and normal gain?
http://en.wikipedia.org/wiki/Elo_rating_system#Most_accurate_K-factor
The K-factor is actually the thing we call it here "gain": normal gain - K=40; high gain - K=60.
In the "Most accurate K-factor" part, it is discussed how and what K-factors to be used in different cases.
The rating system's goal is to rate players regarding their "true skill" - to find their place among the others. To achieve that more efficient, when a player is too far from his "true skill", he should "move" faster to it and when he is close to it, he should "move" slower.
Nevertheless, the only only I want to point from the article above is: the more rating/xp a player has, the lower his K-factor is, which is exactly the opposite here.

3) Disregarding 2) and looking for a solution about 1)
Without any analysis, a good way is to use 50th player's rating...

[NeeeeK]

should be around 1.2k imo.
and it should be around +25%more xp.

[Cip]

Well curently you didn't see high gains because it is off. As Leo said, turning it on too early after a reset will screw up the system. Because it's actually hard to make a static point where the high gains start, when everyone is at 1000 gaining 20 or 30 is a huge difference and u can't say "ok 1050 is high gains" then after 1 week is "1100" etc..

Last season it was set late at 1180. And it raised the K factor as Leo said (in other words, you gain/lose 50% more). It was mainly used to make games with high level players more decisive and more demanding (ppl tend to focus more when they risk 30 xp).

It is a good suggestion to make the limit dynamic. I currently have a dynamic limit implemented: the !add msg on euro for top 50 players. I can use same value for high xp gains average. Atm it would be 1144.

[BaBaUTZ]

• vote for dynamic values

[HaSu]

1150 or back to 1200, cuz there are really no high gains games even with avg xp of like 1260

just look at this

(20:10:26) (@Sara[1337]) CM Team Setup - xp.cm93: Sent: IamMorden + BrixmarK fnatic\iZnoGouD yeskoN JourneysEnd ---VS--- Scourge: Grandiose + PeymaTheGreat Seepferdchen HaSu ToT|KnowMe - AvgXp 1269 -> Normal Gains - Sent...|.'.....Scou - IamMorden type -cm 1 - GID2805

NORMAL GAINS, thats a joke for such a game :(

[Kalb]

you were in it. so i guess it was just right :P

[ArT-mohonk]

• vote for dynamic values

#2

[coyote]

Imo. If you are an admin you gain 50 extra for each game and only lose 5.
All the time.