dode74 wrote:I know. I was talking specifically about this post, which was itself a counterexample for the probability-based comparison Veggente was talking about. I know the current formula doesn't use probability distributions, obviously.
But are you aware of why it's not relevant to seasonal ranking, and why things like TrueSkill are also not relevant to it? The way you're discussing it with Veggente, who doesn't even understand the concepts you're discussing with him, I'm not convinced you are.
dode74 wrote:That's not what I was thinking, actually.
Then why talk about it at all? Seasonal performance is not about using a sample of games to determine the overall ability of a team.. if we cared about that we wouldn't restrict ourselves to the performance during a given season. We'd end up with teams playing in the final tournament based on how they generally play, not how they performed during that season... so we'd have to wonder why we were bothering with seasons at all.
Scram Lyche wrote:Your level of condescension is starting to irritate me mike, if you can't ditch your rude arrogance and insults, don't bother speaking to or about me.
Or what, you'll cry? In both instances where you and I have interacted in this thread it has been you who initiated it, and you who specifically made it a conflict. Maybe you need to work on stopping yourself from engaging in conflict with people you can't handle being in conflict with?
Scram Lyche wrote:And if anyone thinks they can dissuade me from voicing my opinion by using snide comments and cheap put downs.. well think again, I ain't going to be bullied away from the discussion.
Awesome, keep up the good work. The same goes for me pointing out that your opinions are uneducated and ill-considered... especially when you voice those opinions toward me by name ;)
Veggente wrote:Calm down.
I'm perfectly calm - I know what I'm talking about and you don't... also, you have no direct influence on the future of CCL. I'm not sure why you think I'd be upset over you making erroneous statements. I'm mostly amused when someone goes on a rant and says a bunch of stuff that is totally out to lunch... and I like to point it out when they do, unless there's more fun to be had by not doing so.
Veggente wrote:But, if you want to compare the performance (measured simply as a win rate) of someone with 112 played games with the performance of someone with 30 played games, you are comparing theoretical performances.
Nope. You're comparing actual performances - that's what seasonal play is. You can certainly theorize that the performance during a season is not indicative of one coach's true skill, but that's irrelevant when ranking them for that season - we rank them on their time-limited performance.
To reduce the randomness associated with the outcome of games we need to set a minimum number of games being played before we take their limited performance seriously... or we end up with someone winning one game and being top ranked because of it. Other than that taking games played into account is just a preference - one that Focus/Cyanide opted for. I personally think the amount that games played contributes should be explicit and tightly controlled, but that's not how the current formula works.. which is why I don't like the current formula... but I don't inherently disagree with the idea of having games played contribute to ranking.
Veggente wrote:What I tried to say in my previous posts, and probably I wasn't able to (given your reaction), is: if you compare the theoretical performance of players, you should account for the fact that the central tendency of the player with 112 games is more likely to be close to his theoretical performance (or his "theoretical central tendency", surely not a good name) than the central tendency of the player with 30 games.
You can't "account for it" - that's just how probability distributions work. You set ONE confidence range and that's the range you'll have for everyone... if you try to change the range for different people you're being a crazy person who doesn't understand statistics.
Likewise, who the hell is trying to compare theoretical performance? This is a seasonal system! If team A that is normally a really good team, plays crappy during a season.. should they go to the playoffs anyway because we feel they're still good enough to be one of the playoff teams? No, we base it on that season's performance.
Veggente wrote:Because of that, a given win rate with a lot of games should be valued more than a slightly higher win rate with much less games.
Well, that's one opinion I suppose. Lucky for you the way you'd accomplish that is by making games played contribute to the ranking value... and that, for different reasons, is what already happens. Ta da, you already got what you wanted... so why are you arguing with dode?
Fashbinder wrote:*(not one that gives a new player against a 20 year vet an equal chance of winning the game)
As a 20 year vet you just want to win every game you play? That makes you sound like a douchebag single-player gamer, not a veteran multiplayer competitive gamer. What being a veteran player should do is get you ranked higher than a neophyte, and that can be done even under a system that lets you play that neophyte and both have to play your best to have a chance to win.