EPT...this can't be accurate.
Introducing the SC2 World Ranking - Page 5
| Forum Index > SC2 General |
|
zokker13
Germany77 Posts
...this can't be accurate. | ||
|
Nuubie
Sweden66 Posts
8 viOLet 4204.42 9 PuMa 4064.23 10 Nerchio 3681.95 ...and this is why the ranking completely blows. End of story. | ||
|
xN.07)MaK
Spain1159 Posts
On August 08 2012 00:26 Liquid`NonY wrote: I don't think the system could ever work because so many players don't even try so many leagues. It's not fair to a ton of non-Korean players that the GSL has any weight at all because it's not a viable tournament to play without taking up residence near its location. The GSL isn't hard just because it has good players. It's hard because you have to move to a foreign country that speaks virtually no English just to participate. And then you've got to pay more and travel longer to continue to participate in NA/EU tournaments. I understand rankings in other sports work similarly in the sense that a player can forgo practice or rest to participate in more competitions in an attempt to bolster ranking. But the situation for SC2 is way more extreme and I believe impossible to solve. Ofc Korea has some unique features, but is not that different from amateurs that cannot afford to travel to X country just because they are amateur so they have not much money to go. But yes, the gap is too large even for accomplished players to go and play GSL even if they can afford it. It's very different from tennis. | ||
|
Metalteeth
United States115 Posts
On August 08 2012 00:24 Kovaz wrote: That said, I think a ranking system like this is awesome to have. It'd be really cool to have a #1 ranking that players strive for that actually means something, and tournaments could start using it for a more fair seeding system. It would be cool to see who's really the top foreigner according to the stats, and there could be interesting storylines like Roger Federer at Wimbledon a few weeks ago reclaiming his world #1 ranking. Imagine Mvp winning another GSL to get back on top after a year of struggle with his wrists. Imagine the excitement when some no-name upsets the #1 ranked player in straight sets (a la Bisu over sAviOr). That right there is exactly why I did it. Having a unified ranking system helps to create a narrative (something Day9 absolutely loves!) over multiple tournaments. I am always frustrated that GSL ranks by GSL points, MLG ranks by MLG points, NASL ranks by NASL results. Having a unified ranking (compared to TLPD, which, while very helpful, has some issues I feel) is a great thing for the scene. | ||
|
Aunvilgod
2653 Posts
| ||
|
| ||
|
Metalteeth
United States115 Posts
On August 08 2012 00:37 zokker13 wrote: Nerchio over Naniwa... ...this can't be accurate. On August 08 2012 00:37 Nuubie wrote: ...and this is why the ranking completely blows. End of story. Nerchio's score is heavily boosted by his recent win at HSC5, and two QFs at Dreamhack. Naniwa and PuMa have not had top victories in the recent events. And their results in GSL, while good, are not enough to make up a Grand Slam win from Nerchio. | ||
|
Faust852
Luxembourg4004 Posts
 | ||
|
Metalteeth
United States115 Posts
On August 08 2012 00:42 TiTanIum_ wrote: It looked fine, but  Potiguar being better than  Bomber, seems silly.Bomber has been not doing too well since Red Bull Battlegrounds. As for Potiguar, a problem with the system is that WCS South America has already ran (along with some VERY easy WCS Nationals) boosting the South American players that did well. That is why Killer (Chile), Fenix, and Potiguar are all in the Top 100. As Europe, NA, and Korea have their WCS events, I expect the South American players to start dropping. | ||
|
chris5180
198 Posts
| ||
|
Markwerf
Netherlands3728 Posts
On August 07 2012 14:19 Metalteeth wrote: Well, the only thing that is my opinion is how each tournament is weighed. Like I said, if people think that I have a tournament weighed strangely, I can easily go change it to what the community thinks is better. TLPD is good too, but since it is still separated between Korean and International, it's hard to get a sense of the two scenes together. TLPD avoids to compare the Korean and International community. Yours does but in a completely unfair way since most foreigners can't compete in GSL and few koreans travel to many tournaments. It's better not to compare than to compare in a terrible way. This method you devised is even unfair for comparing foreigners amongst eachother or koreans amongst eachother because there is no punishment or correction for difference in tournament attendence between players. ELO based systems are simply superior for trying to be get an objective as possible ranking for a game like starcraft. A point system like the ATP works in tennis because there are rules that pretty much force equal attendence, the top players are forced to attend all grand slams, 8 ATP 1000 tournaments and only a limited number of points from ATP 500 tournaments are counted. That is simply not the case in many other sports including starcraft. Starcraft unlike tennis also has many different tournament formats including round robin systems, double elimination etc which make it a bit unelegant to use a point system. For example an ELO system would differ results between dominating a MLG by 5-0'ing your group and smashing all through the winners bracket compared to getting last in group and still winning from the 'losers' bracket, a point system like this would not. Point systems are elegant in their simplicity which make it easy to make statements like: "if MC beats DRG here and goes on to win the GSL he is the new #1!!" but the conditions for it to work properly are just not in place for starcraft. You would have to contrive difficult attendence penalties etc. to make it work. The system is already a bit ugly now by implementing a regression based decay for the points which is imo unneccesary difficult since you would have to update your rankings each day then if you plan to keep this up. Just using the tennis system where results count equally till they are removed after a year is far easier as you only need to update after tournaments for the players involved that year or the previous. If neccesary you can make a two stage drop, ie points are halfed after half a year or something but regression based is just not elegant. Remember that elegancy is the primary reason to choose a point system over an ELO based system in the first place.. Finally if you implement this you would be best to do a sensitivity analysis. Rank your tournaments in order of importance which is much easier to get an agreement about, for example (1 GSL 2 MLG, IPL, WCS, Dreamhack 3 others) or just by prize money and than use some software to calculate your top 10 or so with different weights. You can see then how much the weighting really matters for the relative ranking of your top players. edit: by the way Wimbledon is also entitled to change seedings based on grass results only. So in a way they are not only seeding by overall ranking but also by a different ranking similar to how tournaments like GSL just use their own results. | ||
|
Greggle
United States1131 Posts
Stephano above MVP, MKP and MMA? I think it got worse =/ | ||
|
Metalteeth
United States115 Posts
On August 08 2012 01:38 Markwerf wrote: TLPD avoids to compare the Korean and International community. Yours does but in a completely unfair way since most foreigners can't compete in GSL and few koreans travel to many tournaments. It's better not to compare than to compare in a terrible way. This method you devised is even unfair for comparing foreigners amongst eachother or koreans amongst eachother because there is no punishment or correction for difference in tournament attendence between players. I've seen plenty of people complain that they want TLPD to be combined, so I thought about try to make something that is combined. Is it perfect? No. But I made it all in 12 hours, and already changed stuff around. As for being "compeltely unfair", I disagree. Of course a player who goes to less events is going to be ranked lower, that is obvious for pretty much any result. Besides, I would say MVP pretty much never goes to foreign events (and when he does, he tends to not do well). And yet he is top 5, because he is good in Korea. The rankings don't require travelling to every single event, and success in GSL is enough to make top 50. On August 08 2012 01:38 Markwerf wrote: ELO based systems are simply superior for trying to be get an objective as possible ranking for a game like starcraft. A point system like the ATP works in tennis because there are rules that pretty much force equal attendence, the top players are forced to attend all grand slams, 8 ATP 1000 tournaments and only a limited number of points from ATP 500 tournaments are counted. That is simply not the case in many other sports including starcraft. Starcraft unlike tennis also has many different tournament formats including round robin systems, double elimination etc which make it a bit unelegant to use a point system. For example an ELO system would differ results between dominating a MLG by 5-0'ing your group and smashing all through the winners bracket compared to getting last in group and still winning from the 'losers' bracket, a point system like this would not. The thing is, tournaments do not differentiate between running through the losers and crushing your group. Her0 at the last Dreamhack lost 2 games the entire way, but only gets a QF result because of it. While something like Leenock's run in MLG Providence, gave him a victory. A single tournament could give some results, but I'm thinking that the randomness of a year's worth of tournaments will even out the differences in tournaments, producing a single number that represents the player's average ability. On August 08 2012 01:38 Markwerf wrote: Point systems are elegant in their simplicity which make it easy to make statements like: "if MC beats DRG here and goes on to win the GSL he is the new #1!!" but the conditions for it to work properly are just not in place for starcraft. You would have to contrive difficult attendence penalties etc. to make it work. The system is already a bit ugly now by implementing a regression based decay for the points which is imo unneccesary difficult since you would have to update your rankings each day then if you plan to keep this up. Just using the tennis system where results count equally till they are removed after a year is far easier as you only need to update after tournaments for the players involved that year or the previous. If neccesary you can make a two stage drop, ie points are halfed after half a year or something but regression based is just not elegant. Remember that elegancy is the primary reason to choose a point system over an ELO based system in the first place.. Everyone else was complaining that there WASN'T regression. This resulted in HuK and IdrA in the top 10, despite not doing too well recently. And as for updating the rankings each day, it takes all of 2 minutes. I open Excel, it auto executes the new rankings, and I sort by points. Done. On August 08 2012 01:38 Markwerf wrote: That is EXACTLY what I did, if you would look at the weightings. These are the values for winning the eventFinally if you implement this you would be best to do a sensitivity analysis. Rank your tournaments in order of importance which is much easier to get an agreement about, for example (1 GSL 2 MLG, IPL, WCS, Dreamhack 3 others) or just by prize money and than use some software to calculate your top 10 or so with different weights. You can see then how much the weighting really matters for the relative ranking of your top players. edit: by the way Wimbledon is also entitled to change seedings based on grass results only. So in a way they are not only seeding by overall ranking but also by a different ranking similar to how tournaments like GSL just use their own results. GSL - 4000 MLG Championship, IPL, Dreamhack, etc (the non-GSL Grand Slams) - 2000 IEM Tour Stops, MLG Arenas - 1500 WCS Continentals - 1000 WCS Nationals, Large "Major" Tournaments (see Revenge of the Nerds, Red Bull Battlegrounds, etc) - 500 Small "Major" Tournaments - 250 As for GSL, I took a look at the official GSL Rankings, I think I will actually use those for GSL now, meaning winning GSL gives you 5000. And it will include Code A, which some people were saying needs to be in. As for Team Leagues, I think I came up with a way for it to work. Because rewardings for only All Kills are too dependent on that player's team makeup, but rewarding a team victory favors poor players on good teams, I will give just bonus points for each win by that player in a team league. I'm thinking 20 points per win in GSTL, 10 points per win in IPL TAC/IPTL, 5 points per win for EGMC. Should I then include smaller team leagues like the ISTL or STL though? I'm gonna hold off on adding Team Leagues for now, I want to see if people think those numbers need to be adjusted. I'm really just throwing them out there for now. On August 08 2012 01:57 Greggle wrote: Stephano above MVP, MKP and MMA? I think it got worse =/ Stephano in the past 5 months has won NASL, Top 6 at 2 MLG Arenas, Top 6 at MLG Spring, Top 6 at IPL4, and SF at Dreamhack Summer. There is a reason that Stephano is as high as he is, it's because he does REALLY WELL at a lot of events. MVP hasn't done much since his last GSL win. MKP in the same time span won MLG Winter (which is decaying), 3rd at MLG Spring, but only Ro16 in GSL, and no BIG wins. And come on, I would think Stephano being that high is not a surprise. | ||
|
sVnteen
Germany2238 Posts
On August 07 2012 14:17 FabledIntegral wrote: Honestly, it's far too subjective for me imo, since it's really just your opinion. I think I'll stick with TLPD, but who knows, maybe your way will catch on! well taeja is no.1 on tlpd for months now and he has certainly not been no.1 for months (I still wouldn't even consider him to be top 4 until he makes at least a final in GSL because ro8 is not enough to call him no1) he has been number one for a long time now (he became number 1 shortly before joining liquid....) and he certainly couldn't be considered even top 10 at that time he just won online stuff in koreaand they ranked him number 1 so think tlpd is far more subjective than this eventhough this is also kinda messed up imo | ||
|
1ManArmy
Netherlands895 Posts
| ||
|
Greggle
United States1131 Posts
Also, the GSL is a ~2 month tournament with the hardest qualifiers in the world, which aren't being counted at all. You can't compare that to a weekend tournament where you can plow through a pool of players which is 6x the size of the pool of players that can actually win it. If you look at a code S pool close to half of the players could have an argument made for them winning the tournament at the start of the season (and often times a player that seems hopeless to win such as Seed wins anyway). Every weekend tournament is padded with players who simply don't stand a chance, versus the GSL which has a constant pool of both champions and up and comers with few if any remaining fillers. To account for the difference in average player skill, lower frequency of events and time investment the GSL needs to be worth at least 6000-8000 for first place, and with a slower degradation of points for lower finishes. | ||
|
sixfour
England11061 Posts
On August 07 2012 14:10 Metalteeth wrote: I like to think of Starcraft 2 as similar to another 1v1 sport, Tennis. so copy a bad ranking system from one sport into another, fantastic idea | ||
|
Mrvoodoochild1
United States1439 Posts
| ||
|
Atokad
United States204 Posts
I don't agree with the results, and strongly disagree with the relation you made to tennis. In tennis the best players all compete at the same 4 grand slam tournaments every year, where I believe the only grand slam tournament sc2 has where the best all come to is the GSL. All of the European tournaments are heavily european player based and all of the american tournaments are heavily NA player based. The GSL is heavily korean based, but has been proven that only few foreigners who are good enough can come and compete at their level. This is why I think the GSL should be treated points-wise as a grand slam, compared to a smaller invitational tournament in tennis(ASUS, NASL, Dreamhack, IEM, etc.). GSL should be highest in points by a lot. Many foreigners have come to Korea with very little success. | ||
|
Metalteeth
United States115 Posts
+ Show Spoiler + GSL: 1st: 5000 2nd: 2499 Semifinal: 500 QF: 780 Ro16: 450 Ro 32: 300 Code A: Ro24->S: 250 Ro24: 204 Ro32: 163 Ro48: 130 | ||
| ||
