|
On May 16 2009 12:42 JWD wrote:Show nested quote +On May 16 2009 12:42 frogmelter wrote:
Hi JWD
I have read the entire thread [every single one of the posts] and just have one question.
If Jaedong and Zero played ZvZ at the time that this PR was written [Best out of 5] who do you think would win? Jaedong, 3>2. But I should add I think this matchup is almost impossible to call right now. If it happens in OSL or MSL, I'll be on the very edge of my seat. A major advantage for JD over ZerO in series play is his mental toughness.
Huh? Zero got 3-0ed by Lux..... Jaedong would have a much easier time. Even if that was before, I don't see a zerg that can be called a favorite over Jaedong in a bo5 ZvZ. I don't think 2-3 against him is possible either.
|
Zero is too inconsistent to stand a chance in a Bo5. But I wouldn't automatically call a JD victory if they met in the proleague.
|
On May 16 2009 14:19 Avidkeystamper wrote:
Zero is too inconsistent to stand a chance in a Bo5. But I wouldn't automatically call a JD victory if they met in the proleague.
yes, you first have to check if the map is imba in ZvZ or not... 
|
Hurray on a spectacular writeup, JWD - amazing. Laughed, nodded, and the KeSPA dog just topped it all off.
Cheers to a very well-done summary of the tip-top sc players of the past month.
|
Nice write up, I don't agree with Jaedong's position, but the rest of it looks ok. Manifesto had the best reasoning/definition behind the power rank, and I believe Jaedong is still one scary fucker.
|
On May 16 2009 08:28 JWD wrote:Show nested quote +On May 16 2009 08:07 tfeign wrote:On May 16 2009 03:46 JWD wrote:On May 16 2009 03:33 tfeign wrote:On May 16 2009 03:24 JWD wrote:On May 16 2009 03:18 tfeign wrote:
The power rank is NOT about who has the best last month in Starcraft. If that's what you solely use to determine players in the power rank, then you are wrong again. It's almost comical that you're willing to speak in such absolute terms about something which is clearly subjective and you have absolutely no control over. Here's a proposition: I wrote this Power Rank, and that gives me the right do determine what it is based on (within reason). You've come into this thread, claimed the Power Rank is based on criteria which differ from the criteria which I had in mind when I wrote it, and then criticized my ranking based on your claimed criteria. Do you realize how ridiculous that is? Here's an exaggerated example to help: I decide to rank gamers according to the sexiness of their haircuts. You come along and tell me that my PR is garbage because gamers should be ranked according to the sexiness of their wardrobes, and then proceed to criticize my ranking based on the fact that the 4th-ranked player has a sexier wardrobe than the 3rd-ranked player. The criteria of the PR has never been in question. I'm telling you that solely and exclusively using the last month of Starcraft results alone to determine players in the power rank is not how it should be. It has never been what the PR is, and it will never be accepted by readers. People want to know the who are the best players, that's why the PR was created, and the only reason why the PR was created. The PR was not made to list who has the best stats in the last month of Starcraft. Watch your absolutes. I am not "solely and exclusively using the last month of SC results alone", and I've already covered that several times. Just saying "the PR was created because people want to know who the best players are" is too simplistic. You need a timeframe - otherwise, every PR would include Boxer, Savior, and iloveoov ("the best players" ever). Because the PR is a monthly affair, the natural thing to do is to set that timeframe to one month. Once again: that's not to say, though, that I think results from previous months should be totally ignored - just substantially downplayed vs. more recent results. I'm obviously aware that I don't have the reputation of, say, Mani or FS here at TL. This only furthered my desire to slightly change the PR's basis - while Mani or FS might be able to get away with posting the PR as what they think the results of a Bo99999 ladder between every pro SC player would look like (not that they have, this is a hypothetical for argument's sake), I surely wouldn't. Nobody wants to read "some guy"'s opinion of who the scariest progamers are right now, so to make the PR informative and relevant I figured I would need to make its focus a bit more concrete. By stressing recent results I feel like I made the PR worth reading, if only as a refresher on the major trends in individual skill over the past month of SC. Whether you disagree with my rank, I think you have to concede that it would have been a hell of a lot less useful if I wrote it with less focus on recent results and more on my abstract sense of "who is better" at SC right now. Yes, the PR was created solely for the purpose of listing who are the best players. It has never been just about who's got the best stats in the last month. It was created to list out the best players, based subjectively on many different factors -- the last month's result, though is important, is only one of such factors. There is indeed a timeframe as we all know it. Of course no one in their right mind would put Boxer or oov in the top PR, but at the same time no one in their right mind should put Zero over Jaedong either. A couple of unimportant games against lower caliber players does not in any way, shape, or form justify him as the best current zerg. It doesn't matter that Jaedong has a couple more losses than Zero, or what kind of leagues Zero is alive in. Zero can never be labeled as the best current zerg until he can prove himself far in a Starleague. The only way you can somehow make an argument of Zero>Jaedong is if you somehow only solely and exclusively just take into account the last month of results, completely ignoring what kind of games they are, regardless whether they're merely qualifiers or the final of a Starleague, and completely ignore everything beyond the last month. Hell, one can even make a point if in the last month Zero won a Starleague title, or defeated top-ranked Kespa or ELO players in a bo5 of an important league. Zero did none of that. He's accomplished nothing. Nada. Zero. No pun intended. You need to understand that to solidify yourself as the best, it takes a lot more than just a couple of extra wins in games that aren't even important to begin with. Not only is it that you overconsider the value of recent results, you also overconsider the value of the importance of the games played. A PR that places a big emphasis on just the last month's worth results usually only happens when the last month's of results included the conclusion of a Starleague or two. None of that happened, nothing even close to that happened. You need to understand that a player who loses say, 10 games, but at the same wins a Starleague title, would be considered better than a player who won 10 unimportant games but hasn't gotten anywhere far in the league. Zero has not proven anything in the leagues and until he can start putting up solid results in a Starleague against its stiff competition, labeling him as the best current zerg, on top of an S-class player who's proven himself for years, just won an OSL title, and is still putting up rock-solid results (80% win in his last 10 games), is incomprehensible and inconceivable. tfeign, I'm not really sure what else I can say to you. I've already tried several times to get across the basis for my PR but you keep insisting it's "entirely stats-based" and an attempt at ranking "the best players". If you're just going to refuse to accept the basis for my ranking, there's no possible way for me to debate it with you. If you think the PR should be written as something other than how I wrote it (a ranking of "the last month's most powerful SC players"), that's fine. Point taken. But please stop trying to insist that there is only one valid basis for a PR, or that nobody has ever challenged that basis, or that that basis means my rank is incorrect. None of those points is valid.
Power Ranks have been around in just about every sport. Your definition of a Power Rank being "who has the best stats in the last month" is not what a PR really is. It's not what a PR means, and there is no disputable definition as to what a PR is: a ranking of the best current players, not a ranking of who performed best in just the last month and only the last month.
Of course the last month's of performance are often the most important, but they're never the only factors to determining the players on the PR. Yes, there are times when the last month's worth of results go toward to determine a major if not all of the rankings on the PR, but those times are really only reserved to when the last month contained the conclusion of a Starleague or two.
|
look at how KeSPA is so happy! 
|
On May 16 2009 17:36 tfeign wrote:
Power Ranks have been around in just about every sport. Your definition of a Power Rank being "who has the best stats in the last month" is not what a PR really is. It's not what a PR means, and there is no disputable definition as to what a PR is: a ranking of the best current players, not a ranking of who performed best in just the last month and only the last month.
Yes, there are times when the last month's worth of results go toward to determine a major if not all of the rankings on the PR, but those times are really only reserved to when the last month contained the conclusion of a Starleague or two.
You're absolutely right when you say that a Power Rank is a "ranking of the best current players". But what does that mean? Clearly, you have to combine some combination of recent results, reputation, guesstimates of continuing results, and all-time performance. All-time performance isn't enough, or any top ten would change really slowly. Recent results isn't enough, or you automatically have to drop anybody who hits a slump, even if you're 99% sure it's temporary. Reputation is double-edged at best: it can put an underperformer in or keep a relatively quiet player out (Calm, great). And guessing at future performance? Luxury, we're looking at you.
Even statistically, there are tons of ways to rank players: most wins, most games, best winning percentage, etc. When you try to include anything subjective in a power rank (and the power rank isn't defined by a set of statistics, or we wouldn't need people to do it every month and it would always be on time), how things are viewed is - surprise! - subjective. Of course the PR guy tries to be objective, but we all value different things slightly more or less and that influences how the rank ends up. It's objective according to his standards, but the standards aren't set in stone.
In short, any time you have a combination of things, how different people combine them is going to vary - wildly. But how do you judge a Power Rank's accuracy? You seem to be looking for a list of predictions (that is, emphasizing the expected future performance aspect): and yes, we expect Jaedong to do a bit better than Zero in the months to come. But that's not how JWD wrote it. This is more of a way-things-are-now rank (that is, focusing more on the recent results). It may not be how you would do it, but why is it invalid?
|
Valhalla18444 Posts
On May 16 2009 09:35 Avidkeystamper wrote:
I'll always remember FS showing all the nonbelievers with his ranking of Pure over Lucifer.
that was something i was so adamant about because i knew i was right. the 10th spot has never been so fiercely defended 
|
Monthly power rank should be about the month's best players imo.
Oh well...
|
On May 16 2009 20:29 DoX.) wrote:
Monthly power rank should be about the month's best players imo.
Oh well...
I wish the power rank was less of a knee-jerk reaction and more about actually who is the best.
Luxury was laughable at no.2 last month, now he's not even in the top10.
|
Luxury at #2 last time was perfectly reasonable. He won a fricking Starleague.
|
On May 16 2009 20:42 Jaksiel wrote:
Luxury at #2 last time was perfectly reasonable. He won a fricking Starleague.
... but played much worse overall than Bisu or Fantasy.
It's not "list the players in order of single accolades this month RANK".
|
On May 16 2009 20:42 Jaksiel wrote:
Luxury at #2 last time was perfectly reasonable. He won a fricking Starleague.
the reasoning behind the 2nd place was bad, Luxury had nothing to do with "best ZvP" title, otherwise yeah, MSL gold is MSL gold + he did well overall
|
United States12607 Posts
On May 16 2009 20:29 DoX.) wrote:
Monthly power rank should be about the month's best players imo.
Oh well...
This is really what my PR is about...The month's best players.
|
JD could lose one game against Zero, but that's all. 3-0 or 3-1 in favor of JD is the most likely result in a bo5.
also:
On May 16 2009 12:15 JWD wrote:
Thanks Goragoth, interesting information. Also a pretty obvious hint as to why ZerO is where he is in the rank.
then why isn`t calm on the PR?
ur arguments fail JWD, sorry to say this, but u give examples that apply just for some specific players, not all of them. So what's the criteria?
-most fearsome player...no
-best PL performance...no
-best SLs performances...no
-top 10 overall in the last month...no
-top 10 in a boX...no
...so what is this PR based on?
|
United States12607 Posts
Jaeden, have you considered that the PR might be based on a combination of some of the factors that you've listed, not necessarily on any one single factor (and sidenote, would you even want to read a PR based on one single such narrow factor?!).
UR argmnts fail.
And as for your last question...read my comments in this thread, I promise the answer is not hard to find. Hint: you might want to start with the post directly above yours, but I've given a bit more precise definitions elsewhere.
|
On May 16 2009 23:42 JWD wrote:Show nested quote +On May 16 2009 20:29 DoX.) wrote:
Monthly power rank should be about the month's best players imo.
Oh well... This is really what my PR is about...The month's best players.
So we might as well just check TLPD for the recent best performing players per month and toss them on a top 10, with obvious exclusions to amusing things like Kespa? It seems almost arbitrary depending on a player's workload, like if in Offseason a really good player is seeded in both tourneys and doesn't have to play any qualifiers.
It's like, say Flash goes on a 14 win streak and tops the chart of a Power Rank for obviously performing the best for a month, but between that PR and the next he only ends up playing maybe 5 games (say 2 proleague games and a Bo3) and loses 1 of the proleague games and one of the Bo3 games and ends up with a 3-2 record. He still has a positive record in PL, still advanced in the tourney, but should he drop significantly for that 3-2 when he JUST came off a huge streak?
I really think it's misleading to make something as definitive sound as a Power Rank off purely "the" month's performance.
|
United States12607 Posts
On May 17 2009 00:39 TwoToneTerran wrote:Show nested quote +On May 16 2009 23:42 JWD wrote:On May 16 2009 20:29 DoX.) wrote:
Monthly power rank should be about the month's best players imo.
Oh well... This is really what my PR is about...The month's best players. So we might as well just check TLPD for the recent best performing players per month and toss them on a top 10, with obvious exclusions to amusing things like Kespa?
Do you really think "checking TLPD" can tell you who the "best performing players" are over the past month?
|
Yep. Bisu's only dropped one game and has the best record, #1. Zero and Fantasy have a better record this month than Jaedong, so he obviously drops below both, etc etc. Leta's record is very similar to Jaedong (someone posted a list earlier where they were on the same level), so he goes right beneath...Going by your version of the PR though. I'm not saying you're wrong, I just think that sounds simplistic.
Sorry for editing in the hypothetical after you responded.
|
|
|
|
|
|
EPT
