The Massively Gigantic Thread On Elo Comments

Post by **Raveen** » Fri Jul 14, 2006 6:46 pm

That's not the point Terr, the point is to promote a balance of skills on both teams and avoid having all the int ho's on one team that get's an instant bonus if it goes Exp vs a balanced team of equal ELO. Whether it would make much difference I'm not sure, maybe selecting tech based on the skills is one of the vital com skills.

Freeza · Post by **Freeza** » Fri Jul 14, 2006 7:09 pm

QUOTE Yes... if we could hire a team of judges that could watch over all the games played and assign grades to all players, that would probably be more accurate. But we can't. And though our community is still small by online standards, it's still sufficiently large to make remembering the individual skills of all or even most of the active players unfeasible. And then you have the hider nicks... even more names to remember.
Computers, on the other hand, are good at number crunching. All I want is some help in estimating how good other players are. ELO might be far from perfect, but it's still better than nothing.[/quote]
You don't need a group of judges. I personally know who's good, who's average, and who isn't very good. I don't see the big deal about knowing a few hundred names and who they corespond to.

I'd prefer 'nothing' over ELO. Age rank is probably as accurate or more so than ELO, why not just use that if you want some pretty numbers to look at?

Team balance isn't the problem anyway, it's commander balance. Good commanders can draw good teams. One bad commander and one good commander means...unbalanced teams. Why aren't there enough good commanders commanding? People being idiots and not listening.

If you want balanced games, don't use ELO. Get people to listen to the good commanders, so that they command more. I bet we could easily make a list of the top 25 commanders (If there are that many), who people should always listen to.

MrChaos · Post by **MrChaos** » Sat Jul 15, 2006 12:39 am

"probably, might, could be, nothing is better, I personally know".

My bad. Your offhand thoughts and feelings trump the cold, hard, mathematical based papers with examples.

Could they be full of crap? Ab-so-lute-ly. How the HECK do you know. If the math goes ClayPigeon (math teacher) you keep going and come on back here. I got questions too a lot of them. Will pool our resources, and work as a * gasp * team.

Im no genius as my posts prove so for crying out loud read MSR's website like Bad. He saw the inherent weakness of TrueSkills, 90 plus games. It got me thinking, comparing it to Elo and thus my reply.

This is the way it's been done for six looooong years. It certainly passes the insanity test doing the same thing over and over expecting a different answer.

Stacking stinks therefore:

Autobalancing /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
Elo ranking current

Elo for TEAM games /huh.gif" style="vertical-align:middle" emoid=":huh:" border="0" alt="huh.gif" />
TrueSkill > Elo /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
Players being honorable /mrgreen.gif" style="vertical-align:middle" emoid=":D" border="0" alt="mrgreen.gif" />

As one of the great Alleg philosophers said:

QUOTE #3 you jerks will find a way to ruin games no matter what poor pook does so send him pictures of Lima and jars of honey to help him make it through another day.[/quote]

So while not nirvana, TrueSkills ranking works better then Elo for ranking/autobalancing.
MrChaos

jgbaxter · Post by **jgbaxter** » Sat Jul 15, 2006 8:13 am

MrChaos, I have no accusations, I was even tongue in cheek. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />

I think you're biased, and you are, we all are, humans are biased. /tongue.gif" style="vertical-align:middle" emoid=":P" border="0" alt="tongue.gif" />

No hard feelings mate. /mrgreen.gif" style="vertical-align:middle" emoid=":D" border="0" alt="mrgreen.gif" />

MrChaos · Post by **MrChaos** » Sat Jul 15, 2006 2:14 pm

baxter

now why would you say that? Just because I say please discuss and every post swoop down from perch squacking "TrueSkills TrueSkills" ?
[Deng is busy so I'll translate myself. Baxter's right]

MrChaos

Why is it whenever someone from the commonwealth calls me mate I want to buy them a beer?

I can't use it sounds wrong
* spoken in MrChaos' Midwest American Accent * mate

see

jgbaxter · Post by **jgbaxter** » Sun Jul 16, 2006 2:11 am

Mrchaos, drop it, you can't win, our postings aren't a contest.

Thanks for the beer. /mrgreen.gif" style="vertical-align:middle" emoid=":D" border="0" alt="mrgreen.gif" />

MrChaos · Post by **MrChaos** » Sun Jul 16, 2006 5:53 pm

Baxter 1
MrChaos 0

< couldn't resist >

MrChaos

Ksero · Post by **Ksero** » Fri Jul 28, 2006 11:58 am

Freeza wrote:QUOTE (Freeza @ Jul 14 2006, 08:09 PM) You don't need a group of judges. I personally know who's good, who's average, and who isn't very good.

We have a volunteer Judge! Can you always be online when I play, then, so I can ask you? An ELO-based balancing feature will always be available, while the Freeza-based one won't be. ELO will never be 100 % accurate. But it'll be a fair assessment.
QUOTE I don't see the big deal about knowing a few hundred names and who they corespond to.[/quote]I think it is. In the heat of battle, I'm usually too busy to evaluate how good the situational awareness of my teammates is. When I scramble to defend our tech base, I generally don't remember who placed the strategic probe that spotted the bomber, giving us a few precious seconds to react. ELO does, in a roundabout way.
To put it another way... How large can the community be before you think it's unfeasible to keep track of everyone's individual skill? Twice as large? How much would you need to play per week to keep these estimates accurate and up-to-date?
QUOTE Age rank is probably as accurate or more so than ELO, why not just use that if you want some pretty numbers to look at?[/quote]I think that ELO, while far from perfect, is way better than age ranks. Allegiance is not like World of Warcraft. You don't automatically get better by just spending time in the game.
QUOTE Team balance isn't the problem anyway, it's commander balance. Good commanders can draw good teams. One bad commander and one good commander means...unbalanced teams. Why aren't there enough good commanders commanding? People being idiots and not listening.[/quote]Yes, I agree that commander balance is vital to having an even game. But I disagree about why there aren't more commanders. Essentially, you're saying that the good commanders don't comm because the players suck. If all the Allegiance players got better, then the good commanders would command again. Why hasn't this been an issue before? If anything, with the academy and the cadet program, the new players are better than they used to be. In pick-up games there will always be people of different skills and ambitions, and I hardly think you'll improve anything by screaming "Get Better!" or "Listen to the commander!".

I think that the new auto-balancing system will encourage fledgling commanders to try their wings, ensuring that the match isn't over before it has even started. We should also encourage commanders of superior skill to step down if they can't find an appropriate opponent. With better sportsmanship / lighter attitudes in the lobby (One can always hope /wink.gif" style="vertical-align:middle" emoid=";)" border="0" alt="wink.gif" /> ), I think everyone will enjoy Allegiance more.

In the end, since the development effort is voluntary, the developers are quite free to pursue the issues they feel are most important. This seems to have made their list. In the end, it'll just be another tool that we can use to have more fun, even games.