EPTPower Rank 05/03/2010 - Page 10
| Forum Index > Polls & Liquibet |
|
o[twist]
United States4903 Posts
| ||
|
okum
France5778 Posts
 | ||
|
Holgerius
Sweden16951 Posts
On May 07 2010 20:15 o[twist] wrote: gosh guys i'm soooooooo worried about flash's tough schedule His schedule got a lot less tough when he didn't even have to go to the Proleague game, let alone play a game/ace game. But ya, based on how he raped both MVP and Pure so easily he probably could've done that too. | ||
|
SuperArc
Austria7781 Posts
 But I think if Flash had to prepare for PL too the bo5s could have been closer. So it was a really good decision to skip them. As long as Kal doesn't get EotS twice he has no chance to beat Flash. And Effort will be just raped. | ||
|
Monkeyshark
United States406 Posts
| ||
|
TwoToneTerran
United States8841 Posts
| ||
|
Cpadolf
Sweden1199 Posts
On May 07 2010 21:01 TwoToneTerran wrote: He'd get 500 a piece purely for winning the titles, so yes. If he beats Jaedong and Kal in both finals, both fairly high kespa dudes, he'll probably be close to 5000. That's actually not how it works. You don't get 500 points for winning a league, and the kespa rank of the players you beat does not matter. He'll still be well over 4000 if he wins both finals though. http://www.teamliquid.net/forum/viewmessage.php?topic_id=118167 | ||
|
bentnormal
112 Posts
| ||
|
Holgerius
Sweden16951 Posts
| ||
|
blazinggpassion
United States27 Posts
| ||
|
Musoeun
United States4324 Posts
On May 07 2010 14:25 o[twist] wrote: that's circular reasoning. i think this month's was needlessly disruptive. No, not really. Circular reasoning means you are assuming your conclusion somewhere as a premise. I was just pointing out - or trying to point out - what I now apparently have to spell out distinctly: The Power Rank is a (monthly) rank based on how well players are playing (during the last month). Currently, by which I mean for the last couple months, no one except Flash and *maybe* Jaedong is consistently playing solid S-class or near S-class Starcraft. But, there are a large number of players who are playing "pretty well": Kal, free, Stork, Midas, BaBy, HiyA, RorO, great, EffOrt, Calm, Violet, ZerO, Kwanro, Stats, Anytime, and possibly a few others. There are also those, like Pure, MVP, sKyHigh, Jangbi, fOrGG, etc. who are either known strong players apparently slowly working out of slumps or weak players having great runs. But still, no one has really stood out from that crowd for more than a couple weeks. In short, what we should expect right now is that the PR gets fairly well shaken up every month as one or the other of these players really makes an impression for the month - because that's how people are playing. What we're seeing right now in the proscene is a lot of streaks and good play for a few weeks before fading back down to averagely-good - and the PR's reflected that. You say this month is "needlessly" disruptive? Well, here's last month's: Flash, Light, Sea, Jaedong, BaBy, Stork, Kal, Kwanro, free, Leta. Flash, Jaedong, Kal, Sea, BaBy, and free (that's 60%) are still on (even though Sea probably shouldn't be and maybe not BaBy). Light has been playing at 50% (no leagues); Stork I believe went 5-4 (no leagues), Kwanro is out of the leagues and playing average, Leta went 4-4 this month (no leagues) - they don't belong on the PR any more. Of the new players (Midas, by.great, MVP, HiyA) every single one did better than the guys who got knocked off. You can argue that EffOrt, Calm, or Pure should have also been in the discussion, but that would still be "new" to this month. So sure, the PR's been a little unsteady lately - but that just reflects reality. | ||
|
Lebesgue
4542 Posts
On May 07 2010 23:54 blazinggpassion wrote: plexa critisizing the most perfect starcraft player in the history of starcraft. What guts. He is as ballsy as Stork and Flash just shows that he is just that damn good by winning both bo5. Take that Stork! | ||
|
Nightmarjoo
United States3360 Posts
| ||
|
o[twist]
United States4903 Posts
| ||
|
Musoeun
United States4324 Posts
On May 08 2010 04:08 Nightmarjoo wrote: ... but Hiya and Great don't belong on top 10 imo. I'd rather put Flying and ForGG there instead. What? I'm tempted to call troll, not so much on your wanting to take HiyA and great off but on your trying to put Flying an fOrGG (especially fOrGG) on. I'm not really sure who's done better than great and HiyA anyway, but Flying and fOrGG are poor replacement candidates. I would be looking more at Calm, or EffOrt, or even Pure or Movie or something first (for the April rank). In descending order of belonging on the PR for April: great: 7-3 in April, made the MSL Ro8, wins over Kal, HiyA, and Movie, no bad losses. So far in May is 3-4, losing to Jaedong and Stats (in a very poor game) and taking Calm to 5 games in their MSL match. HiyA: 7-4 in April, favored to advance to the MSL semifinals (over free, a solid player), wins over Kal, Calm, Shuttle, and of course free. Only one bad loss, to Horang2. So far in May is 1-1, beating firebathero and losing to fantasy. --- Flying: 4-1 in April, no individual leagues, no standout wins, loss to the only good player (RorO) he faced. So far in May is 1-0, beating YellOw[arnC]. fOrGG: 4-6 in April, advanced from OSL groups only to get mutilated by Kal. Only good win was against Jaedong in Proleague. So far in May is 1-1, losing to Hyuk (in a terrible game) and beating Jangbi (solid game from fOrGG, poor decision-making from Jangbi). | ||
|
Musoeun
United States4324 Posts
On May 08 2010 04:27 o[twist] wrote: lol museon. i quoted two lines of yours; you edited out the circular reasoning and then said there was no circular reasoning. i'm sure your post here is fine, but the circular reasoning was saying "it's fine that there's so much switching up in this month's PR. look at other PRs; that'll show you the difference in consistency between then and now." but a perfectly valid alternate explanation for the difference in consistency between PRs is that those were done more accurately. using past PRs as evidence for a difference in consistency when what's at issue is how consistent the PR is and should be is classic circular reasoning, which is, i assume, why you edited it out. Oh, I see. I was guessing at what you were referring to, so I edited to only include what I thought you were talking about - and voila! - we find the misunderstanding. I still think you're looking at this sideways though. First, about your post: it's true that I only mentioned the PR rankings being fairly stable. This is because I didn't think I had to mention that "two years ago" (2008) was a period when we had several players who held on to top spots in reality for a significant amount of time. Apparently I should have. Now, let's go through the argument as I see it. If I expressed myself poorly before, apologies; my argument hasn't changed. You tell me where you see a problem: We agree - I think, and at least for the sake of argument - that (at least some) past PRs have been more "consistent" from month to month. Where we differ is on what we think this indicates. (By the way, define consistent. You seem to be saying that the PR should not change that much from month to month, but you could also be referring to changing criteria used to determine the rankings.) -> I say the PR seemed "more consistent" because the players were more consistent at those points. As a corollary, I deny that this PR is less accurate because of any issues of "consistency" with the last PR. -> You say the PR was more consistent because it was "done more accurately". On me, the burden of proof is to show that players were more consistent. While negative proofs are dangerous, I challenge you to find anybody right now - apart from Flash (and possibly Jaedong) playing with the consistency displayed by, say, Taek-Bang-Lee-SSang over Summer-Fall 2008. It's just not happening. We don't even have an Unholy Trinity like we did in Winter 2008-Spring 2009, where Bisu, Jaedong, and Flash destroyed everyone, only challenged by Stork, Leta, and possibly EffOrt and fantasy. We don't even have Summer/Fall 2009 where Jaedong, Flash, fantasy, Stork, Light, Sea, Calm (and possibly a couple others) all were playing very well, but no one quite seized the "best player in the world" title (and it remained JD's by default). Apart from Flash, there simply is no consistency at the top right now, and we should expect the PR to reflect that. For your part, if you contend that the earlier PR was more accurate, you need to show how this PR is inaccurate - and for extra credit, how making it more consistent would help that accuracy. (For myself, I feel that a really accurate PR for April would probably drop Sea completely and drop BaBy further, but that's hardly "consistency" and so doesn't help you much.) | ||
|
o[twist]
United States4903 Posts
under this view, this PR was less accurate simply because it predicted top-10 play from people who could not deliver it. seriously, if MVP actually plays better this month than Leta, Light, Fantasy, Really, etc., i will retract this. and if Great actually plays better than Calm, Effort, Kwanro, etc., same. this reasoning at least begins to suffice the burden of proof you've placed upon me - it suggests that, at least in a "predictive" power ranking, an increase in consistency would be rewarded with greater accuracy (and thus greater consistency next month!). anyway, there you go - like i said, we understand each other, so there's no real problem. | ||
|
Shikyo
Finland33997 Posts
| ||
|
Lebesgue
4542 Posts
On May 08 2010 05:10 Shikyo wrote: A Semi-finalist and 8-2 in his last 10 games, yet Effort still isn't anywhere in the PR, but some people like Hiya and MVP and Great are, who all are way worse. If he beats Kal in a convincing fashion then you can expect Effort to come back to PR. Otherwise, well, he kind of managed to sneak into semis. It is for the same reason that Pure is not on the PR. So far he has only proved that his ZvZ has returned. His ZvT is still untested (and he was posting terrible terrible results in that match-up) and he hasn't played a ZvP in like a month... | ||
|
SuperArc
Austria7781 Posts
| ||
| ||
