Warrior General Games

Originally posted by Corndog
Every team does have an equal chance. Some might take slightly longer than others because they had some bad luck with matchups.

This is what baffles me, you keep saying "over time it will even out" but probability theory exists to say "no, that's not how it works". Did you completely miss the example about flipping a coin 100 times and expecting 50/50?

What makes you so confident that in general, over time, opportunity will balance out as evenly as you describe?

Originally posted by Galithor
So you want more volatility? More ranking based on the very recent results, and less ranking based on more distant results.

More like the more you win a ramping difficulty kicks in and it starts matching you with noticeably higher ranks and therefore more elo for the win.

Originally posted by Laggo
Did you completely miss the example about flipping a coin 100 times and expecting 50/50?

No, I read it, and completely disregarded it because it was a really bad analogy.

Originally posted by Laggo
What makes you so confident that in general, over time, opportunity will balance out as evenly as you describe?

Because teams will eventually be properly ranked and won't win every single game ever.

If a team continues to win every single game ever, then it's hard to dispute that they should be ranked number 1.

Originally posted by USC_Trojans
More like the more you win a ramping difficulty kicks in and it starts matching you with noticeably higher ranks and therefore more elo for the win.

That's not how ELO works.

I like the ladder system.

Corn, are the last ladder games of the season within your tier and 1v2 3v4 etc?

Originally posted by o The Boss x
Corn, are the last ladder games of the season within your tier and 1v2 3v4 etc?

Nope.

Originally posted by Corndog
That's not how ELO works.

This is a funny response considering one of the basic tenets of Elo is that people are playing a fair game (as in equal access to resources on both sides). SP Value between tiers pretty much breaks Elo as it was designed, as a 1600 Rookie VS 1700 Rookie is not nearly the same game as a 1600 Rookie VS 1700 Sophmore.

Originally posted by Corndog
Because teams will eventually be properly ranked and won't win every single game ever.

If a team continues to win every single game ever, then it's hard to dispute that they should be ranked number 1.

I'm guessing you have never heard of the Gambler's fallacy.

Originally posted by Corndog
Every team does have an equal chance. Some might take slightly longer than others because they had some bad luck with matchups. How far you make it is still based on you, not the system.

I think seasoned is a great example. The continentals have managed to hold the two spot because in their wins this season their oppents record was something close to 55-109 before the Sacramento game. While Tampa has had to face opponents with a significantly better record.

Originally posted by Laggo
I'm guessing you have never heard of the Gambler's fallacy.

I'm guessing you've never heard of ELO.

ELO is a marathon, not a sprint. If someone is ahead of you 5 minutes in doesn't matter if they can only run half as far as you.

Originally posted by Corndog
That's not how ELO works.

Thats exactly how elo works in league of legends, the more you win the faster you climb.

Originally posted by Corndog
I'm guessing you've never heard of ELO.

ELO is a marathon, not a sprint. If someone is ahead of you 5 minutes in doesn't matter if they can only run half as far as you.

This has nothing to do with the opportunity given to certain teams over others, I dunno why you keep placing your hands over your ears on this point. There is nothing in the system to encourage those opportunities to balance out over time. Some teams get them and some teams don't. The Gambler's fallacy here is real. This is exactly what is happening.

Originally posted by USC_Trojans
Thats exactly how elo works in league of legends, the more you win the faster you climb.

It also was added to some chess ranking variants to better rate weaker players, as it was determined they usually have a slightly better chance to win than the traditional Elo formula predicted.

It's kind of odd we keep talking about the "Elo system" as some set thing, when there are probably a dozen different variants and weightings you can use to maintain the basic formula. Saying something shouldn't be added because it's not the "Elo system" is dumb.

Originally posted by Corndog
Nope.

That's the only thing I'd like to see happen. Hard to find out the best without playing one another at least once.

Originally posted by Laggo
This is a funny response considering one of the basic tenets of Elo is that people are playing a fair game (as in equal access to resources on both sides). SP Value between tiers pretty much breaks Elo as it was designed, as a 1600 Rookie VS 1700 Rookie is not nearly the same game as a 1600 Rookie VS 1700 Sophmore.

Except you play within your own tier in Rookie and Sophomore.

Originally posted by Laggo
I'm guessing you have never heard of the Gambler's fallacy.

Do you believe that chess is a game of chance too?

Originally posted by Laggo
This has nothing to do with the opportunity given to certain teams over others, I dunno why you keep placing your hands over your ears on this point. There is nothing in the system to encourage those opportunities to balance out over time. Some teams get them and some teams don't.

Their ability to win games is what evens out opportunity.

If the entire game was just a flip of a coin whether a team won or not, sure, your whole probability argument is amazing. Flawless, even. Except winning games isn't really just a coin flip. A better team will continue winning games where an inferior team loses games. The inferior team getting to the point of losing games before the better team passes them has almost no relevance in the long run.