The Puritans

[Malicious Wraith]

Chi stacks to inflate his Helo so he looks like hes a vet when he is really a low-skilled voob.

/tongue.gif" style="vertical-align:middle" emoid=":P" border="0" alt="tongue.gif" />

[takingarms1]

Ironically, HELO said that game was a massive stack but it was Kofi who said we had a lot of underrated players and we'd have a really good shot at beating the other team (who were largely SysX).

I'm going to go out on a limb though and guess that IB wasn't talking about you.

[takingarms1]

QUOTE (ImmortalZ @ Dec 29 2007, 01:59 PM) Was fun. Didn't know you were AngryAngel, TA /tongue.gif" style="vertical-align:middle" emoid=":P" border="0" alt="tongue.gif" />

Heh, befoer we launched I almost said, "Don't worry Immz, I won't boot you this time" but I thought you might not fly for me /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />

[MrChaos]

QUOTE (Ksero @ Dec 29 2007, 11:23 AM) So what you're saying is (my replies start with •):

* HELO ranks are inaccurate because some players have inflated their ranks by stacking. This means that judging stacks by team helo is inappropriateJudging stack by "players" is hard, weighing twenty pilots on one team vs. twenty pilots on the other. Especially if you don't play much.Yes. If we start forcing helo to be even, players who have inflated their ranks by stacking will lose more initially (since Awesome(30) isn't equal to AngryO + Nooblet), until their ranks are deflated. After some time of auto-balancing, the ranks will settle down and become more accurate, and Awesome(22) might be equal to AngryO + Newbie(7).* Everyone has stacked. Therefore, we can't blame specific players.Some players do it more often than others. See also my posts in this old thread. There you'll see how easy it is to find players who stack consistently, since they win games more often than they lose (ie. they win more than 50 % of their games).* Play the game and have fun instead of bitchingSure... but I don't mind discussing how we can have more even games and more fun

HELO being inaccurate is wider then just people who blatantly and actively attempting to be on the winning side no matter what they have to do to achieve it. HELO, as implemented and based on the community's behavior , can have 1,000,000 games and still not achieve balance.
AllegSkills is a different story all together [ assuming we get the coding bits right ] and what I believe you were referring to in your post. I'm just making sure we clarify that HELO and AllegSkills are different and not interchangable in discussions of this nature.

[Ksero]

MrChaos, you say that HELO ranks can never become accurate, yet you don't offer any arguments in support of that statement. Can you say more exactly why you think so?

Still, Trueskill is a better rating system, that can achieve accurate ratings more quickly.

[sgt_baker]

Helo provides a 15 point reward (to the stacker) for winning even when a rank 30 player goes head to head with a rank 0 player. Thus it is entirely feasable to inflate one's rank purely through joining, and winning with, a stacked team. We'll leave out the plethora of other stability issues and effects of Helo's averaging (the stack-bonus is shared between everyone on the team) and ask this: What type of 'accurate' system provides a significant reward for an activity it is designed to prevent?

Edit: The same situation in trueskill actually results in *a reduction* being applied to the stackers rank. Try it out with MS's online Trueskill calculator /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
Edit 2: If you do, the 'Draw Probability' for Alleg is 1.01%.
Edit 3: http://www.fe23.co.uk/alleg/AS/StabilityExample.png

[MrChaos]

QUOTE (Ksero @ Dec 29 2007, 02:28 PM) MrChaos, you say that HELO ranks can never become accurate, yet you don't offer any arguments in support of that statement. Can you say more exactly why you think so?

Still, Trueskill is a better rating system, that can achieve accurate ratings more quickly.

KSero

Patience my man patience. Im well aware of your credentials and respect them quite properly. Please read Baker's wonderful explaination in the development forum.

[ImmortalZ]

QUOTE (takingarms1 @ Dec 30 2007, 12:32 AM) Heh, befoer we launched I almost said, "Don't worry Immz, I won't boot you this time" but I thought you might not fly for me /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />

Well, if you were on as takingarms, I would've had reservations flying for you. But it would've been a toss-up between my contempt at the SysX stack and (my respect of ACE - my pissed-offedness at you).

Still, you didn't make any game-breaking mistakes that game, so I'll fly for you whenever you're on as AngryAngel /laugh.gif" style="vertical-align:middle" emoid=":lol:" border="0" alt="laugh.gif" />

[IB_]

QUOTE (takingarms1 @ Dec 29 2007, 11:01 AM) I'm going to go out on a limb though and guess that IB wasn't talking about you.

[Ksero]

Baker,

First you say that helo ratings can be inflated by stacking. We've covered that already, and MrChaos said there were more reasons for why helo ratings are inaccurate. That's what I wanted him to elaborate on.

Secondly, you ask why helo rewards stacking when it was designed to prevent that behaviour. Helo is not designed to prevent stacking. Helo is just designed for measuring skill. Helo + autobalance can prevent stacking. I don't think AllegSkill can prevent stacking by itself, without using autobalance.

Then you say that when a high-ranking player goes head-to-head with a lower-ranked player and the higher-ranked player wins, his rank is reduced. That sounds preposterous to me. I tried the calculator with the following figures:
CODEplayer . Before . After
| mu sigma | mu sigma
Alice | 30 8.333 | 30.362 8.106
Bob | 5 8.333 | 4.638 8.106

Draw probability: 1 %
As expected, winning increases the winner's rating.