|
On July 19 2013 20:35 opterown wrote:
even in BW not many people reached 70% right? how many people would be able to do so consistently?
Flash was the only player to have a 70% win rate for basically his entire career (other than dipping below briefly near the beginning of his career). Other players like Boxer, Nada, Iloveoov, Savior, Jaedong and Bisu maintained a 70% win rate for periods of time. Possibly Best also had a 70% win rate for a bit.
|
On July 19 2013 20:35 PerSe wrote:Show nested quote +On July 19 2013 20:18 Big J wrote:On July 19 2013 19:45 PerSe wrote:On July 19 2013 18:49 HolyArrow wrote:On July 19 2013 18:04 PerSe wrote:
Flash is just a better player than Symbol. Getting a high win-ratio and the most wins in SPL is more impressive than Ro8 finishes given the amount of games you have to play in SPL and the consistency you have to show. People say Hero/Taeja are EG-TL's aces, yet they can barely get over a 50% win ratio.
Fact of the matter is, Bo3s in tournaments like the GSL have high variance due to the low number of games (especially when you happen to be in the Group of Death), whereas in an entire season of SPL, the amount of games reduce the variance to the extent that you can really tell who the best players are accurately (in SPL).
Innovation, Flash, Rain, Soulkey, sOs
Granted WCS KR has better players, but weighting WCS KR significantly over say, an entire season of SPL, doesn't make sense with regards to statistics, only with regards to hype ('cause it's an individual tournament and the GSL). Lmao this is exactly why I dislike weighing PL so heavily for Power Rank. You make that sample size argument and suddenly you can justify having a bunch of PR slots filled by players who are doing well in Proleague no matter what the results are outside of Proleague. Basically you're advocating for a Power Rank where a bunch of players get spots for doing well in their isolated little arena while disregarding any outside results because there just aren't enough games. What you're suggesting means that Kespa players will always be hugely favored to take a bunch of spots on the Power Rank solely due to the fact that they get to play in Proleague which has so many games in the course of its season. The sample size 'argument' is legit. if you want to know who the best players are, you have to look at statistics. Considering how volatile a game like SC2 is, where even the best players can't get above 70% win ratio, you need a larger sample of games to get an accurate picture. WCS KR has the highest calibre of play, but the low sample size is really volatile. It took DRG some time to get into Code S, even when he was clearly world class from his GSTL performances. At the same time, you get guys like Sniper or Seed, who have a good run and win Code S and then fall out very quickly. people moaned about OSL's bo1 format in R32. 'cause it's volatile, 'cause the better player might lose 1 game but could've won in a bo3. The same principle applies here and quite frankly in a game like SC2, even bo3s and bo5s are very volatile. If a good player goes on a good-run of say twelve games, he'll get into the ro8, yet over a larger sample of games, that result may simply be due to variance and his avg win rate may be much lower. Conversely, a very good player (i.e Life) may go on a slump for a bit, yet over a large sample, he's still world-class. The top players in SPL do tend also to be the top players in WCS KR. People have been predicting sOs to break through in WCS KR after his win ratio in SPL. Could've predicted the same with SK and Innovation as well. Rain and Flash, they haven't done as well, yet I'd expect them to eventually break through in individual-leagues in the long-run. The problem is when guys like Fantasy and maybe Jangbi, who go on a good run in the SPL and are put on the power-rank. I think that's got to do with the fact that the Power-rank weights recent games far more heavily. If one looks at their long-run win ratios, there's definitely a distinction between the cream of the crop guys: Inno, SK, sOs, Flash, Rain and the good players who can go on streaks: Jangbi, Fantasy, Jaehoon/Argo, Baby. Also eSF players need to get a team league where they actually regularly play games. sorry but that's nothing SC2 specific. German football Bundesliga last season: Even the very best teams in other sports like football will only sometimes reach like 80% winrate. E.g. german Bundesliga champion Bayern Munich had 82% winrate last season. the runner up Borrusia Dormund had already only 55% winrate. No it's not SC2 specific and football is another example of a 'volatile' game. Except bear in mind that in SC2 there's generally only two possible results: win or loss. In football, draws erode the win rate significantly. Glad you brought up the football example though. Again, you determine the best team over the course of a season, not over 1 game. Let us suppose BM = Innovation, in other words, performs very well in the national league (SPL) AND in the Champions League (GSL) and contrast to a team that performed very well in the national league but not well at all in the champions league: Manchester City (I dunno, Flash?). Actually good example because they were in the Group of Death with Borussia Dortmund and Real Madrid. I think most people would still regard Manchester City as one of the top 10 teams in the world. On the other hand you have a team like Chelsea which WON the Champions League but did poorly in their national league and in that year, not many would have considered them a top 10 team.
Yeah, the draws of course play a role. Though I'd really love to know what you consider a non-volatile game/sports? I'd say football is really, really solid - as well as Bo5 or Bo7 SC2 series. Bo3 and especially Bo1 less so, which is one of the reasons why I think teamleagues in Starcraft don't express skill as much, as they are all Bo1's.
And to your examples: I'd say that Leverkusen or Schalke could become german champions currently. Do I think that they can win the champions league? Not really. Same thing goes for every national league in soccer. I can see all of those teams do well, for as long as they play 90% of the time against lowerrated opponents and only 10% of the time against equals/better ones. But in the champions league, it's rather that 90% of the teams they face are equal/better and only 10% lower rated.
For starcraft players that means that I think every Premiere League player could achieve a top5 spot in teamleauge statistics, because the amount of wins against lowerrated opponents is mostly a result of how much they get used overall and should be very even (and high) for all of them if they get used equally often. But what really determines how good they are can only be seen by watching the highest level.
Of course there can be upsets (like Chelsea/Porto winning the CL and ManCity being defeated early), but in the special case of Flash I don't see this. He barely didn't get into Code S 2times in a row against "lowlevel Code S players". Now he is struggling to get past "highlevel Code S players" in the Ro16. His performances at the highest level are very steadily lower than top8 for quite some time now.
Yeah, ManCity is a really good comparison to Flash. Because up to now they haven't achieved anything in the CL that would rate them top10, yet many experts would do so. Still, they'd probably rate Bayern, Dortmund, Real, Barcelona and ManUtd higher, because they are all pretty beastly in their respective leagues as well AND have had better results in the CL. And I'd expect a similar treatment for Flash (and Fantasy and Jangbi) in SC2 Powerranks.
|
On July 19 2013 21:10 Big J wrote:Show nested quote +On July 19 2013 20:35 PerSe wrote:On July 19 2013 20:18 Big J wrote:On July 19 2013 19:45 PerSe wrote:On July 19 2013 18:49 HolyArrow wrote:On July 19 2013 18:04 PerSe wrote:
Flash is just a better player than Symbol. Getting a high win-ratio and the most wins in SPL is more impressive than Ro8 finishes given the amount of games you have to play in SPL and the consistency you have to show. People say Hero/Taeja are EG-TL's aces, yet they can barely get over a 50% win ratio.
Fact of the matter is, Bo3s in tournaments like the GSL have high variance due to the low number of games (especially when you happen to be in the Group of Death), whereas in an entire season of SPL, the amount of games reduce the variance to the extent that you can really tell who the best players are accurately (in SPL).
Innovation, Flash, Rain, Soulkey, sOs
Granted WCS KR has better players, but weighting WCS KR significantly over say, an entire season of SPL, doesn't make sense with regards to statistics, only with regards to hype ('cause it's an individual tournament and the GSL). Lmao this is exactly why I dislike weighing PL so heavily for Power Rank. You make that sample size argument and suddenly you can justify having a bunch of PR slots filled by players who are doing well in Proleague no matter what the results are outside of Proleague. Basically you're advocating for a Power Rank where a bunch of players get spots for doing well in their isolated little arena while disregarding any outside results because there just aren't enough games. What you're suggesting means that Kespa players will always be hugely favored to take a bunch of spots on the Power Rank solely due to the fact that they get to play in Proleague which has so many games in the course of its season. The sample size 'argument' is legit. if you want to know who the best players are, you have to look at statistics. Considering how volatile a game like SC2 is, where even the best players can't get above 70% win ratio, you need a larger sample of games to get an accurate picture. WCS KR has the highest calibre of play, but the low sample size is really volatile. It took DRG some time to get into Code S, even when he was clearly world class from his GSTL performances. At the same time, you get guys like Sniper or Seed, who have a good run and win Code S and then fall out very quickly. people moaned about OSL's bo1 format in R32. 'cause it's volatile, 'cause the better player might lose 1 game but could've won in a bo3. The same principle applies here and quite frankly in a game like SC2, even bo3s and bo5s are very volatile. If a good player goes on a good-run of say twelve games, he'll get into the ro8, yet over a larger sample of games, that result may simply be due to variance and his avg win rate may be much lower. Conversely, a very good player (i.e Life) may go on a slump for a bit, yet over a large sample, he's still world-class. The top players in SPL do tend also to be the top players in WCS KR. People have been predicting sOs to break through in WCS KR after his win ratio in SPL. Could've predicted the same with SK and Innovation as well. Rain and Flash, they haven't done as well, yet I'd expect them to eventually break through in individual-leagues in the long-run. The problem is when guys like Fantasy and maybe Jangbi, who go on a good run in the SPL and are put on the power-rank. I think that's got to do with the fact that the Power-rank weights recent games far more heavily. If one looks at their long-run win ratios, there's definitely a distinction between the cream of the crop guys: Inno, SK, sOs, Flash, Rain and the good players who can go on streaks: Jangbi, Fantasy, Jaehoon/Argo, Baby. Also eSF players need to get a team league where they actually regularly play games. sorry but that's nothing SC2 specific. German football Bundesliga last season: Even the very best teams in other sports like football will only sometimes reach like 80% winrate. E.g. german Bundesliga champion Bayern Munich had 82% winrate last season. the runner up Borrusia Dormund had already only 55% winrate. No it's not SC2 specific and football is another example of a 'volatile' game. Except bear in mind that in SC2 there's generally only two possible results: win or loss. In football, draws erode the win rate significantly. Glad you brought up the football example though. Again, you determine the best team over the course of a season, not over 1 game. Let us suppose BM = Innovation, in other words, performs very well in the national league (SPL) AND in the Champions League (GSL) and contrast to a team that performed very well in the national league but not well at all in the champions league: Manchester City (I dunno, Flash?). Actually good example because they were in the Group of Death with Borussia Dortmund and Real Madrid. I think most people would still regard Manchester City as one of the top 10 teams in the world. On the other hand you have a team like Chelsea which WON the Champions League but did poorly in their national league and in that year, not many would have considered them a top 10 team. Yeah, the draws of course play a role. Though I'd really love to know what you consider a non-volatile game/sports? I'd say football is really, really solid - as well as Bo5 or Bo7 SC2 series. Bo3 and especially Bo1 less so, which is one of the reasons why I think teamleagues in Starcraft don't express skill as much, as they are all Bo1's. And to your examples: I'd say that Leverkusen or Schalke could become german champions currently. Do I think that they can win the champions league? Not really. Same thing goes for every national league in soccer. I can see all of those teams do well, for as long as they play 90% of the time against lowerrated opponents and only 10% of the time against equals/better ones. But in the champions league, it's rather that 90% of the teams they face are equal/better and only 10% lower rated. For starcraft players that means that I think every Premiere League player could achieve a top5 spot in teamleauge statistics, because the amount of wins against lowerrated opponents is mostly a result of how much they get used overall and should be very even (and high) for all of them if they get used equally often. But what really determines how good they are can only be seen by watching the highest level. Of course there can be upsets (like Chelsea/Porto winning the CL and ManCity being defeated early), but in the special case of Flash I don't see this. He barely didn't get into Code S 2times in a row against "lowlevel Code S players". Now he is struggling to get past "highlevel Code S players" in the Ro16. His performances at the highest level are very steadily lower than top8 for quite some time now. Yeah, ManCity is a really good comparison to Flash. Because up to now they haven't achieved anything in the CL that would rate them top10, yet many experts would do so. Still, they'd probably rate Bayern, Dortmund, Real, Barcelona and ManUtd higher, because they are all pretty beastly in their respective leagues as well AND have had better results in the CL. And I'd expect a similar treatment for Flash (and Fantasy and Jangbi) in SC2 Powerranks.
I'd say a non-volatile game would be Chess, or Go. Where the better player will rarely lose to a lesser-skilled opponent.There's no existence of build-order loss like you get in SC2.
In the case of Flash. It's safe to say he's no way near his prime in BW, but I do think that he will break out in individual leagues eventually. He's been put in two groups of death, having the unfortunate luck of facing Innovation each time. Ro16 is especially bad since it's still Bo3.
The whole Team league performance doesn't express skill because it's 'bo1' is not really true over a large sample of games. Individual bo1s are volatile. 100 bo1s really aren't. In fact they're a whole lot less volatile than a bo3, bo5 or bo7. maybe one can point to differences in preparation. I'm not sure what the differences would be. In SPL, players prepare extensively, just as they would in an individual tournament. Perhaps lesser skilled players are more inclined to cheese in a bo1. I don't know.
|
On July 19 2013 20:36 Assirra wrote:Show nested quote +On July 19 2013 19:26 Emzeeshady wrote:On July 19 2013 18:49 HolyArrow wrote:On July 19 2013 18:04 PerSe wrote:
Flash is just a better player than Symbol. Getting a high win-ratio and the most wins in SPL is more impressive than Ro8 finishes given the amount of games you have to play in SPL and the consistency you have to show. People say Hero/Taeja are EG-TL's aces, yet they can barely get over a 50% win ratio.
Fact of the matter is, Bo3s in tournaments like the GSL have high variance due to the low number of games (especially when you happen to be in the Group of Death), whereas in an entire season of SPL, the amount of games reduce the variance to the extent that you can really tell who the best players are accurately (in SPL).
Innovation, Flash, Rain, Soulkey, sOs
Granted WCS KR has better players, but weighting WCS KR significantly over say, an entire season of SPL, doesn't make sense with regards to statistics, only with regards to hype ('cause it's an individual tournament and the GSL). Lmao this is exactly why I dislike weighing PL so heavily for Power Rank. You make that sample size argument and suddenly you can justify having a bunch of PR slots filled by players who are doing well in Proleague no matter what the results are outside of Proleague. Basically you're advocating for a Power Rank where a bunch of players get spots for doing well in their isolated little arena while disregarding any outside results because there just aren't enough games. What you're suggesting means that Kespa players will always be hugely favored to take a bunch of spots on the Power Rank solely due to the fact that they get to play in Proleague which has so many games in the course of its season. I wouldn't be against a separate rank for PL players as the creators of these rankings lean so heavily towards proleague results that it makes having the rest of the players pointless. Next month Fantasy will just be replaced by whatever player does well in proleague... Honestly, a seperate ranking is required. Having a top10 where half the players get basically a second chance is laughable.
Did you ever think about why power rank was reintroduced only after SPL started, even though it was one of the most required features since the start of SC2? Yes, because there was not enough games before...
btw do you also want to ignore GSTL? And why don't we also ignore MLG since it is in America and not all the Kespa players can participate? And what about Dreamhack? Or ASUS RoG? Or WCS EU/AM? Or any arbitralily chosen tournament?
For starcraft players that means that I think every Premiere League player could achieve a top5 spot in teamleauge statistics, because the amount of wins against lowerrated opponents is mostly a result of how much they get used overall and should be very even (and high) for all of them if they get used equally often.
And yet they are not. Flash is the best performer in SPL and (I guess) Ryung (or Keen) is in GSTL..
|
|
|
On July 19 2013 22:57 Ammanas wrote:Show nested quote +For starcraft players that means that I think every Premiere League player could achieve a top5 spot in teamleauge statistics, because the amount of wins against lowerrated opponents is mostly a result of how much they get used overall and should be very even (and high) for all of them if they get used equally often. And yet they are not. Flash is the best performer in SPL and (I guess) Ryung (or Keen) is in GSTL..
Flash, Rain, sOs, Innovation are all having the same stats. Flash was used a little more, so he has a little more wins (and more losses). Zest, Roro, Soulkey, free also have similar success when being used, but are again used a little less.
It's really just a matter of how many games a player gets to play, that determines the exact proleague ranking. It as little to do with "who's better".
|
On July 20 2013 00:59 Big J wrote:Show nested quote +On July 19 2013 22:57 Ammanas wrote:For starcraft players that means that I think every Premiere League player could achieve a top5 spot in teamleauge statistics, because the amount of wins against lowerrated opponents is mostly a result of how much they get used overall and should be very even (and high) for all of them if they get used equally often. And yet they are not. Flash is the best performer in SPL and (I guess) Ryung (or Keen) is in GSTL.. Flash, Rain, sOs, Innovation are all having the same stats. Flash was used a little more, so he has a little more wins (and more losses). Zest, Roro, Soulkey, free also have similar success when being used, but are again used a little less. It's really just a matter of how many games a player gets to play, that determines the exact proleague ranking. It as little to do with "who's better".
I'm on your side.
Rain had a higher Win ratio and decent amount of game played. I would argue he is the better player in Proeague than Flash.
But if you think about it, innovation is far more reliable than both of the players in critical matches. So the proleague ranking is misleading a little here.
|
On July 19 2013 22:28 PerSe wrote:Show nested quote +On July 19 2013 21:10 Big J wrote:On July 19 2013 20:35 PerSe wrote:On July 19 2013 20:18 Big J wrote:On July 19 2013 19:45 PerSe wrote:On July 19 2013 18:49 HolyArrow wrote:On July 19 2013 18:04 PerSe wrote:
Flash is just a better player than Symbol. Getting a high win-ratio and the most wins in SPL is more impressive than Ro8 finishes given the amount of games you have to play in SPL and the consistency you have to show. People say Hero/Taeja are EG-TL's aces, yet they can barely get over a 50% win ratio.
Fact of the matter is, Bo3s in tournaments like the GSL have high variance due to the low number of games (especially when you happen to be in the Group of Death), whereas in an entire season of SPL, the amount of games reduce the variance to the extent that you can really tell who the best players are accurately (in SPL).
Innovation, Flash, Rain, Soulkey, sOs
Granted WCS KR has better players, but weighting WCS KR significantly over say, an entire season of SPL, doesn't make sense with regards to statistics, only with regards to hype ('cause it's an individual tournament and the GSL). Lmao this is exactly why I dislike weighing PL so heavily for Power Rank. You make that sample size argument and suddenly you can justify having a bunch of PR slots filled by players who are doing well in Proleague no matter what the results are outside of Proleague. Basically you're advocating for a Power Rank where a bunch of players get spots for doing well in their isolated little arena while disregarding any outside results because there just aren't enough games. What you're suggesting means that Kespa players will always be hugely favored to take a bunch of spots on the Power Rank solely due to the fact that they get to play in Proleague which has so many games in the course of its season. The sample size 'argument' is legit. if you want to know who the best players are, you have to look at statistics. Considering how volatile a game like SC2 is, where even the best players can't get above 70% win ratio, you need a larger sample of games to get an accurate picture. WCS KR has the highest calibre of play, but the low sample size is really volatile. It took DRG some time to get into Code S, even when he was clearly world class from his GSTL performances. At the same time, you get guys like Sniper or Seed, who have a good run and win Code S and then fall out very quickly. people moaned about OSL's bo1 format in R32. 'cause it's volatile, 'cause the better player might lose 1 game but could've won in a bo3. The same principle applies here and quite frankly in a game like SC2, even bo3s and bo5s are very volatile. If a good player goes on a good-run of say twelve games, he'll get into the ro8, yet over a larger sample of games, that result may simply be due to variance and his avg win rate may be much lower. Conversely, a very good player (i.e Life) may go on a slump for a bit, yet over a large sample, he's still world-class. The top players in SPL do tend also to be the top players in WCS KR. People have been predicting sOs to break through in WCS KR after his win ratio in SPL. Could've predicted the same with SK and Innovation as well. Rain and Flash, they haven't done as well, yet I'd expect them to eventually break through in individual-leagues in the long-run. The problem is when guys like Fantasy and maybe Jangbi, who go on a good run in the SPL and are put on the power-rank. I think that's got to do with the fact that the Power-rank weights recent games far more heavily. If one looks at their long-run win ratios, there's definitely a distinction between the cream of the crop guys: Inno, SK, sOs, Flash, Rain and the good players who can go on streaks: Jangbi, Fantasy, Jaehoon/Argo, Baby. Also eSF players need to get a team league where they actually regularly play games. sorry but that's nothing SC2 specific. German football Bundesliga last season: Even the very best teams in other sports like football will only sometimes reach like 80% winrate. E.g. german Bundesliga champion Bayern Munich had 82% winrate last season. the runner up Borrusia Dormund had already only 55% winrate. No it's not SC2 specific and football is another example of a 'volatile' game. Except bear in mind that in SC2 there's generally only two possible results: win or loss. In football, draws erode the win rate significantly. Glad you brought up the football example though. Again, you determine the best team over the course of a season, not over 1 game. Let us suppose BM = Innovation, in other words, performs very well in the national league (SPL) AND in the Champions League (GSL) and contrast to a team that performed very well in the national league but not well at all in the champions league: Manchester City (I dunno, Flash?). Actually good example because they were in the Group of Death with Borussia Dortmund and Real Madrid. I think most people would still regard Manchester City as one of the top 10 teams in the world. On the other hand you have a team like Chelsea which WON the Champions League but did poorly in their national league and in that year, not many would have considered them a top 10 team. Yeah, the draws of course play a role. Though I'd really love to know what you consider a non-volatile game/sports? I'd say football is really, really solid - as well as Bo5 or Bo7 SC2 series. Bo3 and especially Bo1 less so, which is one of the reasons why I think teamleagues in Starcraft don't express skill as much, as they are all Bo1's. And to your examples: I'd say that Leverkusen or Schalke could become german champions currently. Do I think that they can win the champions league? Not really. Same thing goes for every national league in soccer. I can see all of those teams do well, for as long as they play 90% of the time against lowerrated opponents and only 10% of the time against equals/better ones. But in the champions league, it's rather that 90% of the teams they face are equal/better and only 10% lower rated. For starcraft players that means that I think every Premiere League player could achieve a top5 spot in teamleauge statistics, because the amount of wins against lowerrated opponents is mostly a result of how much they get used overall and should be very even (and high) for all of them if they get used equally often. But what really determines how good they are can only be seen by watching the highest level. Of course there can be upsets (like Chelsea/Porto winning the CL and ManCity being defeated early), but in the special case of Flash I don't see this. He barely didn't get into Code S 2times in a row against "lowlevel Code S players". Now he is struggling to get past "highlevel Code S players" in the Ro16. His performances at the highest level are very steadily lower than top8 for quite some time now. Yeah, ManCity is a really good comparison to Flash. Because up to now they haven't achieved anything in the CL that would rate them top10, yet many experts would do so. Still, they'd probably rate Bayern, Dortmund, Real, Barcelona and ManUtd higher, because they are all pretty beastly in their respective leagues as well AND have had better results in the CL. And I'd expect a similar treatment for Flash (and Fantasy and Jangbi) in SC2 Powerranks. I'd say a non-volatile game would be Chess, or Go. Where the better player will rarely lose to a lesser-skilled opponent.There's no existence of build-order loss like you get in SC2. In the case of Flash. It's safe to say he's no way near his prime in BW, but I do think that he will break out in individual leagues eventually. He's been put in two groups of death, having the unfortunate luck of facing Innovation each time. Ro16 is especially bad since it's still Bo3. The whole Team league performance doesn't express skill because it's 'bo1' is not really true over a large sample of games. Individual bo1s are volatile. 100 bo1s really aren't. In fact they're a whole lot less volatile than a bo3, bo5 or bo7. maybe one can point to differences in preparation. I'm not sure what the differences would be. In SPL, players prepare extensively, just as they would in an individual tournament. Perhaps lesser skilled players are more inclined to cheese in a bo1. I don't know.
And i want to add one more argument to the illegitimacy of proleague performance, at least partial illegitimacy.
1. The critical thing about proleague is how the coach arrange the players' sequence before the match. The best player to the mid-class opponents. The worst to their best...etc. Therefore, a better coach will definitely increase the win-rate of a player, while the worse opposite. On the other hand, on individual leagues, you mostly face the best of best instead of the low-tier players. So, in any giving time,i will rank Symbol higher than Flash.
2. The map plays such a important role in proleague. You got to prepare one map instead of a sequences really helped those with clear weakness in skills. A good example will be Zest or Stats, both performed excellent in Proleague, but failed in individual.
The top proleague players may not do well in individual leagues while those who do well in individual leagues did well in proleagues.
Therefore, its much accurate to judge player's level from individual leagues than from PL.
|
On July 19 2013 22:57 Ammanas wrote:Show nested quote +On July 19 2013 20:36 Assirra wrote:On July 19 2013 19:26 Emzeeshady wrote:On July 19 2013 18:49 HolyArrow wrote:On July 19 2013 18:04 PerSe wrote:
Flash is just a better player than Symbol. Getting a high win-ratio and the most wins in SPL is more impressive than Ro8 finishes given the amount of games you have to play in SPL and the consistency you have to show. People say Hero/Taeja are EG-TL's aces, yet they can barely get over a 50% win ratio.
Fact of the matter is, Bo3s in tournaments like the GSL have high variance due to the low number of games (especially when you happen to be in the Group of Death), whereas in an entire season of SPL, the amount of games reduce the variance to the extent that you can really tell who the best players are accurately (in SPL).
Innovation, Flash, Rain, Soulkey, sOs
Granted WCS KR has better players, but weighting WCS KR significantly over say, an entire season of SPL, doesn't make sense with regards to statistics, only with regards to hype ('cause it's an individual tournament and the GSL). Lmao this is exactly why I dislike weighing PL so heavily for Power Rank. You make that sample size argument and suddenly you can justify having a bunch of PR slots filled by players who are doing well in Proleague no matter what the results are outside of Proleague. Basically you're advocating for a Power Rank where a bunch of players get spots for doing well in their isolated little arena while disregarding any outside results because there just aren't enough games. What you're suggesting means that Kespa players will always be hugely favored to take a bunch of spots on the Power Rank solely due to the fact that they get to play in Proleague which has so many games in the course of its season. I wouldn't be against a separate rank for PL players as the creators of these rankings lean so heavily towards proleague results that it makes having the rest of the players pointless. Next month Fantasy will just be replaced by whatever player does well in proleague... Honestly, a seperate ranking is required. Having a top10 where half the players get basically a second chance is laughable. Did you ever think about why power rank was reintroduced only after SPL started, even though it was one of the most required features since the start of SC2? Yes, because there was not enough games before... btw do you also want to ignore GSTL? And why don't we also ignore MLG since it is in America and not all the Kespa players can participate? And what about Dreamhack? Or ASUS RoG? Or WCS EU/AM? Or any arbitralily chosen tournament?
I understand there were not enough games, however the way it is done now it is incredible biassed to proleague players and that is a fact you cannot disagree with. Want to remove the GSTL from this as well? fine by me. Foreigner tournaments? Like they ever count or in rare cases like with Polt. If they want to do it like now they should just call it Proleague rankings or KeSPA rankings since its more accurate considering the way stats are taken.
|
On July 19 2013 20:41 Daswollvieh wrote:
With proleague being a closed competition, I´d even be inclined to ignore it altogether. Sure you see a lot and it´s just an article, but in order to have a proper power rank, you have to have proper competition, and that is not given within the pond (though it´s a shark pond) of PL. So, for me, the argument of a greater number of games can very well be disregarded, because it is a closed competition, while WCS is open and thus more competitive.
Yeah that was a point that I kind of alluded to by mentioning "isolated little arena" in my post, thanks for expanding on it (although I do think ignoring it altogether is a bit drastic).
It rubs me the wrong way that they put so much weight on PL in the Power Rank when there isn't even an Up/Downs for PL, which implies that possibly not the most deserving teams are playing in it. For example, who knows, maybe MVP, IM, or Startale would do better than EGTL/T8/CJ. It simply looks a little silly when you give players so much credit for doing well in a closed competition when players in the closed competition are dropping significant amounts of games to non-Kespa players in open competition tournaments like MLG, WCS KR, and WCS AM/EU Qualifiers.
|
On July 20 2013 02:52 HolyArrow wrote:Show nested quote +On July 19 2013 20:41 Daswollvieh wrote:
With proleague being a closed competition, I´d even be inclined to ignore it altogether. Sure you see a lot and it´s just an article, but in order to have a proper power rank, you have to have proper competition, and that is not given within the pond (though it´s a shark pond) of PL. So, for me, the argument of a greater number of games can very well be disregarded, because it is a closed competition, while WCS is open and thus more competitive. Yeah that was a point that I kind of alluded to by mentioning "isolated little arena" in my post, thanks for expanding on it (although I do think ignoring it altogether is a bit drastic). It rubs me the wrong way that they put so much weight on PL in the Power Rank when there isn't even an Up/Downs for PL, which implies that possibly not the most deserving teams are playing in it. For example, who knows, maybe MVP, IM, or Startale would do better than EGTL/T8/CJ. It simply looks a little silly when you give players so much credit for doing well in a closed competition when players in the closed competition are dropping significant amounts of games to non-Kespa players in open competition tournaments like MLG, WCS KR, and WCS AM/EU Qualifiers.
Okay, but are these non-PL players doing consistently well?
Symbol is alright (with his ZvT full of all-ins not exactly inspiring), but he's definitely a few notches below SK.
Bomber...no elaboration needed.
Who else? I would gladly give that parting and life performed well for a period of time but these players have likewise fallen off the radar. If parting went 3-0 I would have no hesitation placing him at #4 or even #3. Unfortunately he didn't. Don't say we're being unfair. PL games are still games.
|
On July 20 2013 03:37 shadymmj wrote:Show nested quote +On July 20 2013 02:52 HolyArrow wrote:On July 19 2013 20:41 Daswollvieh wrote:
With proleague being a closed competition, I´d even be inclined to ignore it altogether. Sure you see a lot and it´s just an article, but in order to have a proper power rank, you have to have proper competition, and that is not given within the pond (though it´s a shark pond) of PL. So, for me, the argument of a greater number of games can very well be disregarded, because it is a closed competition, while WCS is open and thus more competitive. Yeah that was a point that I kind of alluded to by mentioning "isolated little arena" in my post, thanks for expanding on it (although I do think ignoring it altogether is a bit drastic). It rubs me the wrong way that they put so much weight on PL in the Power Rank when there isn't even an Up/Downs for PL, which implies that possibly not the most deserving teams are playing in it. For example, who knows, maybe MVP, IM, or Startale would do better than EGTL/T8/CJ. It simply looks a little silly when you give players so much credit for doing well in a closed competition when players in the closed competition are dropping significant amounts of games to non-Kespa players in open competition tournaments like MLG, WCS KR, and WCS AM/EU Qualifiers. Okay, but are these non-PL players doing consistently well? Symbol is alright (with his ZvT full of all-ins not exactly inspiring), but he's definitely a few notches below SK. Bomber...no elaboration needed. Who else? I would gladly give that parting and life performed well for a period of time but these players have likewise fallen off the radar. If parting went 3-0 I would have no hesitation placing him at #4 or even #3. Unfortunately he didn't. Don't say we're being unfair. PL games are still games.
Well, first of all, you didn't address the problem with PL being a closed tournament at all.
My point is that by the logic of people talking about sample sizes and variance, in a non-PL setting you can often claim that no one is really doing consistently well because they don't get to play enough games to show that they're doing consistently well. And even if someone is doing "consistently well" outside of Proleague (as in, winning in the few games we do see them in), Proleague will always give overwhelmingly ample opportunity for a player to show his consistency even more compellingly due to the sheer amount of games played.
Note that my issue was never with Innovation and SK being high on the Power Rank, because they've proved themselves not only in PL but also hugely in WCS. I was more talking about the placements of players like Jangbi and Fantasy - essentially, players who are on the Power Rank solely due to a strong month in Proleague. I don't dispute that Symbol is below Soulkey - I don't think any sane person would, but is Symbol, who consistently places very high in individual league tournaments (Ro8/Semis/finals) below Fantasy and Jangbi?
The main point that you simply can't tiptoe around is that again, by the logic of the "sample size" argument, you're always gonna have players doing well in Proleague significantly populating the Power Rank while players with significant individual league achievements (long-term or short-term) get marginalized due to how less games are played for them. What's the point of a Power Rank that's supposed to include everyone if you're just going to do that?
The other key point someone else made is that a lot of the records for stronger PL players are significantly bolstered by them essentially farming wins off of weaker PL players (we know that the gap between top Kespa players like Innovation, Flash, or Soulkey and mid/low-tier Kespa players is fairly significant) over the course of the season. Meanwhile, although there are far fewer games in individual leagues, the way brackets work means that as you advance further in an individual tournament, you face increasingly steeper competition, so of course it'll be harder to appear consistent. Furthermore, if you lose a couple Bo3s in an individual tournament, you're done. You're eliminated. You don't get a chance to show that it may have just been a hiccup, and that you're still a good player. Meanwhile, in Proleague, you can go on a 3 or 4-game losing streak and then win 12 games in a row after that, and then you're sitting pretty at 12-4 and still look like a beast. Individual league players don't have that luxury.
There are simply too many problems with putting that much emphasis on PL results.
|
So then, the questions stands, how would you do the power rank then? Simply ignoring the games played in the most competitive league in Starcraft (this is a fact stated by several people, including TL and EG CEOs and their players, read the article on Polygon on that topic if you don't believe me)? Or what exactly would you do?
Also, "grinding games against weak opponents" is pretty unfair - 7 players sent on regular Proleague games are mostly all very good - of course there are regulars who suck (like Zenio, Stork  , Bang...) but most of them are on very high level.
At the end, yes, I do think JangBi (and around 10 other SPL players - Fantasy not being one of them right now even though I am a fan, also none of the Kespa players that were eliminated in Ro16 OSL besides Flash) are better then Symbol, Maru or Supernova.
And one more thing - they are not putting SUCH an emphasis on PL results, they are putting emphasis on all the results. Proleague just tends to have more of them. It will probably change for coming months, since there will be almost no Proleague.
|
We need to cut the crap about sample size and that's why PL results are weighted heavily. Those are excuses, not the real reason. Had random Kespa players like Pigbaby, Zero, Turn done as well as Jangbi, Fantasy, Flash in PL they would still not have been in those PR spots that the famous players occupied. TL writers are just BW fans and are eager to bring back the BW fan favorites to PR for good old times sake, and PL results that month gave the perfect excuse. Rain had terrible PL results that month but those were explained away because guess what, he is the OSL champion and another fan favorite (although I do think he deserved his spot). The last 3 spots are just political, given to please the esf fans, the underprpivileged group in TL.
In short, the PR is rigged
|
On July 20 2013 06:48 Ammanas wrote:So then, the questions stands, how would you do the power rank then? Simply ignoring the games played in the most competitive league in Starcraft (this is a fact stated by several people, including TL and EG CEOs and their players, read the article on Polygon on that topic if you don't believe me)? Or what exactly would you do? Also, "grinding games against weak opponents" is pretty unfair - 7 players sent on regular Proleague games are mostly all very good - of course there are regulars who suck (like Zenio, Stork , Bang...) but most of them are on very high level. At the end, yes, I do think JangBi (and around 10 other SPL players - Fantasy not being one of them right now even though I am a fan, also none of the Kespa players that were eliminated in Ro16 OSL besides Flash) are better then Symbol, Maru or Supernova. And one more thing - they are not putting SUCH an emphasis on PL results, they are putting emphasis on all the results. Proleague just tends to have more of them. It will probably change for coming months, since there will be almost no Proleague.
I've already suggested having a separate Power Rank for PL results, seems like the easiest and fairest solution so far although pretty inelegant.
I don't think that some of the players that get sent out on a regular basis are as good as you think they are. I don't know what your definition of "very high level" is, but the point I'm trying to make is that even though Proleague may be considered the most competitive league, we've seen Proleague regulars fall in many individual tournaments and tournament qualifiers lately to non-Proleague players, and that's what throws the rankings into contention, because if those Proleague regulars are representing their most competitive league in other events by merely trading evenly with non-Proleague players (Code A) or even outright losing to people they shouldn't really be losing to (like in WCS AM open qualifiers or to foreign team Koreans/foreigners at MLG), it just looks a little funny. And again, consider the fact that the most competitive league is a closed competition, which is problematic for reasons I've stated before.
If you want to talk about the validity of grinding games against weak opponents, then let's have a look.
Some of Flash's wins this season:
Zenio, Savage, Hope, Bbyong, Paralyze, Best, Turn, JYP, Last, Rare, Terminator, Alive, TY, Cure, Bong, Hydra, Effort, Major, and Hammer. Those are 19 wins right there (I think there are multiple wins against Effort, but to keep it simple, let's just say 19), which contributes to over half of the wins Flash has accumulated.
Now, I realize that some names are definitely weaker than others and that some players there like Effort and Hydra have actually done pretty well in Proleague, but the point I'm trying to make here is that when you talk about Proleague record and cite 37-18 it looks really damn impressive especially when you call Proleague the most competitive team league, but when you look at the quality of the opponents in a larger scope, a lot of those names are only Code A-worthy, and some of them aren't even that.
|
On July 20 2013 07:15 painkilla wrote:
We need to cut the crap about sample size and that's why PL results are weighted heavily. Those are excuses, not the real reason. Had random Kespa players like Pigbaby, Zero, Turn done as well as Jangbi, Fantasy, Flash in PL they would still not have been in those PR spots that the famous players occupied. TL writers are just BW fans and are eager to bring back the BW fan favorites to PR for good old times sake, and PL results that month gave the perfect excuse. Rain had terrible PL results that month but those were explained away because guess what, he is the OSL champion and another fan favorite (although I do think he deserved his spot). The last 3 spots are just political, given to please the esf fans, the underprpivileged group in TL.
In short, the PR is rigged
lol.
+ Show Spoiler +(to elaborate, if any player would have such a streak as JangBi had during June in Proleague, he SHOULD and I believe he also WOULD be in power rank.)
On July 20 2013 07:15 HolyArrow wrote:Show nested quote +On July 20 2013 06:48 Ammanas wrote:So then, the questions stands, how would you do the power rank then? Simply ignoring the games played in the most competitive league in Starcraft (this is a fact stated by several people, including TL and EG CEOs and their players, read the article on Polygon on that topic if you don't believe me)? Or what exactly would you do? Also, "grinding games against weak opponents" is pretty unfair - 7 players sent on regular Proleague games are mostly all very good - of course there are regulars who suck (like Zenio, Stork , Bang...) but most of them are on very high level. At the end, yes, I do think JangBi (and around 10 other SPL players - Fantasy not being one of them right now even though I am a fan, also none of the Kespa players that were eliminated in Ro16 OSL besides Flash) are better then Symbol, Maru or Supernova. And one more thing - they are not putting SUCH an emphasis on PL results, they are putting emphasis on all the results. Proleague just tends to have more of them. It will probably change for coming months, since there will be almost no Proleague. I've already suggested having a separate Power Rank for PL results, seems like the easiest and fairest solution so far although pretty inelegant. I don't think that some of the players that get sent out on a regular basis are as good as you think they are. I don't know what your definition of "very high level" is, but the point I'm trying to make is that even though Proleague may be considered the most competitive league, we've seen Proleague regulars fall in many individual tournaments and tournament qualifiers lately to non-Proleague players, and that's what throws the rankings into contention, because if those Proleague regulars are representing their most competitive league in other events by merely trading evenly with non-Proleague players (Code A) or even outright losing to people they shouldn't really be losing to (like in WCS AM open qualifiers or to foreign team Koreans/foreigners at MLG), it just looks a little funny. And again, consider the fact that the most competitive league is a closed competition, which is problematic for reasons I've stated before.
On the other hand, the bottom Proleague players (meaning, not even in Top30) like hyvaa or soO are doing pretty well in WCS KR. Yes, some players lost in WCS AM qualifiers, but they were not Proleague regulars at all. They were just B-teamers, playing coaches etc..
The MLG was a dissapointment though  + Show Spoiler [this whole section is just a joke] +(although, Stats only lost to Dear and JD, soO to Sage and Hyun (soO is pretty bad though, even though he is regular) and Dear to Polt and Naniwa (he should be ashamed ^^)
Also, Proleague is not closed competition. If you are interested in joining, you have the required 100k deposit and Kespa deems you good enough, I believe you could join. Somebody will have to replace EGTL anyway, cause I don't believe they are gonna continue in Proleague.
|
Calm yourself ESF fans, Proleague is ending soon.
|
On July 20 2013 07:32 Ammanas wrote:Show nested quote +On July 20 2013 07:15 painkilla wrote:
We need to cut the crap about sample size and that's why PL results are weighted heavily. Those are excuses, not the real reason. Had random Kespa players like Pigbaby, Zero, Turn done as well as Jangbi, Fantasy, Flash in PL they would still not have been in those PR spots that the famous players occupied. TL writers are just BW fans and are eager to bring back the BW fan favorites to PR for good old times sake, and PL results that month gave the perfect excuse. Rain had terrible PL results that month but those were explained away because guess what, he is the OSL champion and another fan favorite (although I do think he deserved his spot). The last 3 spots are just political, given to please the esf fans, the underprpivileged group in TL.
In short, the PR is rigged lol. + Show Spoiler +(to elaborate, if any player would have such a streak as JangBi had during June in Proleague, he SHOULD and I believe he also WOULD be in power rank.) Show nested quote +On July 20 2013 07:15 HolyArrow wrote:On July 20 2013 06:48 Ammanas wrote:So then, the questions stands, how would you do the power rank then? Simply ignoring the games played in the most competitive league in Starcraft (this is a fact stated by several people, including TL and EG CEOs and their players, read the article on Polygon on that topic if you don't believe me)? Or what exactly would you do? Also, "grinding games against weak opponents" is pretty unfair - 7 players sent on regular Proleague games are mostly all very good - of course there are regulars who suck (like Zenio, Stork , Bang...) but most of them are on very high level. At the end, yes, I do think JangBi (and around 10 other SPL players - Fantasy not being one of them right now even though I am a fan, also none of the Kespa players that were eliminated in Ro16 OSL besides Flash) are better then Symbol, Maru or Supernova. And one more thing - they are not putting SUCH an emphasis on PL results, they are putting emphasis on all the results. Proleague just tends to have more of them. It will probably change for coming months, since there will be almost no Proleague. I've already suggested having a separate Power Rank for PL results, seems like the easiest and fairest solution so far although pretty inelegant. I don't think that some of the players that get sent out on a regular basis are as good as you think they are. I don't know what your definition of "very high level" is, but the point I'm trying to make is that even though Proleague may be considered the most competitive league, we've seen Proleague regulars fall in many individual tournaments and tournament qualifiers lately to non-Proleague players, and that's what throws the rankings into contention, because if those Proleague regulars are representing their most competitive league in other events by merely trading evenly with non-Proleague players (Code A) or even outright losing to people they shouldn't really be losing to (like in WCS AM open qualifiers or to foreign team Koreans/foreigners at MLG), it just looks a little funny. And again, consider the fact that the most competitive league is a closed competition, which is problematic for reasons I've stated before. On the other hand, the bottom Proleague players (meaning, not even in Top30) like hyvaa or soO are doing pretty well in WCS KR. Yes, some players lost in WCS AM qualifiers, but they were not Proleague regulars at all. They were just B-teamers, playing coaches etc.. The MLG was a dissapointment though + Show Spoiler [this whole section is just a joke] +(although, Stats only lost to Dear and JD, soO to Sage and Hyun (soO is pretty bad though, even though he is regular) and Dear to Polt and Naniwa (he should be ashamed ^^) Also, Proleague is not closed competition. If you are interested in joining, you have the required 100k deposit and Kespa deems you good enough, I believe you could join. Somebody will have to replace EGTL anyway, cause I don't believe they are gonna continue in Proleague.
Lol it's effectively closed then, with such a high entry cost.
Hitman and Mini failed multiple times in WCS AM qualifiers, those aren't B-teamers or playing coaches, hitman fielded 9 times and Mini 12 times in PL. Seems regular enough to to me. I realize that they're lowish-tier PL players but again I'm just trying to illustrate the large spectrum of skill levels that we see in Proleague.
|
On July 20 2013 01:29 Baroninthetree wrote:Show nested quote +On July 20 2013 00:59 Big J wrote:On July 19 2013 22:57 Ammanas wrote:For starcraft players that means that I think every Premiere League player could achieve a top5 spot in teamleauge statistics, because the amount of wins against lowerrated opponents is mostly a result of how much they get used overall and should be very even (and high) for all of them if they get used equally often. And yet they are not. Flash is the best performer in SPL and (I guess) Ryung (or Keen) is in GSTL.. Flash, Rain, sOs, Innovation are all having the same stats. Flash was used a little more, so he has a little more wins (and more losses). Zest, Roro, Soulkey, free also have similar success when being used, but are again used a little less. It's really just a matter of how many games a player gets to play, that determines the exact proleague ranking. It as little to do with "who's better". I'm on your side. Rain had a higher Win ratio and decent amount of game played. I would argue he is the better player in Proeague than Flash. But if you think about it, innovation is far more reliable than both of the players in critical matches. So the proleague ranking is misleading a little here.
Proleague HotS win rates (rounded):
Innovation: 73%
Flash: 73%
Soulkey: 67%
Zest: 66.5%
SoS: 62.5%
Rain: 60.5%
Roro: 54.5%
So...
|
On July 20 2013 07:41 HolyArrow wrote:Show nested quote +On July 20 2013 07:32 Ammanas wrote:On July 20 2013 07:15 painkilla wrote:
We need to cut the crap about sample size and that's why PL results are weighted heavily. Those are excuses, not the real reason. Had random Kespa players like Pigbaby, Zero, Turn done as well as Jangbi, Fantasy, Flash in PL they would still not have been in those PR spots that the famous players occupied. TL writers are just BW fans and are eager to bring back the BW fan favorites to PR for good old times sake, and PL results that month gave the perfect excuse. Rain had terrible PL results that month but those were explained away because guess what, he is the OSL champion and another fan favorite (although I do think he deserved his spot). The last 3 spots are just political, given to please the esf fans, the underprpivileged group in TL.
In short, the PR is rigged lol. + Show Spoiler +(to elaborate, if any player would have such a streak as JangBi had during June in Proleague, he SHOULD and I believe he also WOULD be in power rank.) On July 20 2013 07:15 HolyArrow wrote:On July 20 2013 06:48 Ammanas wrote:So then, the questions stands, how would you do the power rank then? Simply ignoring the games played in the most competitive league in Starcraft (this is a fact stated by several people, including TL and EG CEOs and their players, read the article on Polygon on that topic if you don't believe me)? Or what exactly would you do? Also, "grinding games against weak opponents" is pretty unfair - 7 players sent on regular Proleague games are mostly all very good - of course there are regulars who suck (like Zenio, Stork , Bang...) but most of them are on very high level. At the end, yes, I do think JangBi (and around 10 other SPL players - Fantasy not being one of them right now even though I am a fan, also none of the Kespa players that were eliminated in Ro16 OSL besides Flash) are better then Symbol, Maru or Supernova. And one more thing - they are not putting SUCH an emphasis on PL results, they are putting emphasis on all the results. Proleague just tends to have more of them. It will probably change for coming months, since there will be almost no Proleague. I've already suggested having a separate Power Rank for PL results, seems like the easiest and fairest solution so far although pretty inelegant. I don't think that some of the players that get sent out on a regular basis are as good as you think they are. I don't know what your definition of "very high level" is, but the point I'm trying to make is that even though Proleague may be considered the most competitive league, we've seen Proleague regulars fall in many individual tournaments and tournament qualifiers lately to non-Proleague players, and that's what throws the rankings into contention, because if those Proleague regulars are representing their most competitive league in other events by merely trading evenly with non-Proleague players (Code A) or even outright losing to people they shouldn't really be losing to (like in WCS AM open qualifiers or to foreign team Koreans/foreigners at MLG), it just looks a little funny. And again, consider the fact that the most competitive league is a closed competition, which is problematic for reasons I've stated before. On the other hand, the bottom Proleague players (meaning, not even in Top30) like hyvaa or soO are doing pretty well in WCS KR. Yes, some players lost in WCS AM qualifiers, but they were not Proleague regulars at all. They were just B-teamers, playing coaches etc.. The MLG was a dissapointment though + Show Spoiler [this whole section is just a joke] +(although, Stats only lost to Dear and JD, soO to Sage and Hyun (soO is pretty bad though, even though he is regular) and Dear to Polt and Naniwa (he should be ashamed ^^) Also, Proleague is not closed competition. If you are interested in joining, you have the required 100k deposit and Kespa deems you good enough, I believe you could join. Somebody will have to replace EGTL anyway, cause I don't believe they are gonna continue in Proleague. Lol it's effectively closed then, with such a high entry cost. Hitman and Mini failed multiple times in WCS AM qualifiers, those aren't B-teamers or playing coaches, hitman fielded 9 times and Mini 12 times in PL. Seems regular enough to to me. I realize that they're lowish-tier PL players but again I'm just trying to illustrate the large spectrum of skill levels that we see in Proleague.
Some players only perform well in team leagues. Bisu maintained around a 70% win rate in Proleague in Brood War, yet time after time failed to even make it through the prelims in the OSL/MSL. Does this mean that he was a player with a low level of skill?
|
|
|
|
|
|
EPT
