The most recent two-week period for Aligulac ratings (up to July15) ended with Reynor taking the #1 spot away from Serral after a 70-week run at the top.
The Italian Zerg's performances in TSL 5 and ESL Europe, along with Serral's failures in those events, gave him the boost he needed to #1, going 37-9 in matches vs top-level competition from EU and KR.
Over that time frame, Maru, Dark, and Innovation all made a challenge for #1, but surprisingly it was Reynor that rose from #10 when Serral's run started to finally catch him. Watching Reynor's results start to match his potential has been a really nice story the last 18 months.
Still, that was some amazing consistency from the Finnish Zerg and an historic achievement. Onnittelut!
Serral's Douyu Cup group results (against Armani, Rogue and PartinG) haven't been factored in yet, so Serral might very well lose first place again very soon.
I'd love to see how Serral/Reynor would do in a GSL format tournament. It's sort of a shame that there's no reason for non-Koreans to bother with GSL, so we rarely get to see them in that kind of tournament. Just the ones who do it anyway (Scarlett & Major & Astrea)...
On July 22 2020 08:10 Russano wrote: Geez, Aligulac really tells you everything you need to know about the state of race balance in SC2. Sad times to be a protoss.
Aligulac is useless for judging balance especially for Protoss. PvP is so volatile that Protoss players will always be rated lower on average. Protoss is definitely weak currently but the idea that they have been the weakest race since 2015 is just absurd.
Protoss haven’t been the weakest race since 2015, but since way back in BW. They have had the odd high point though.
It makes a lot of sense in lore that the dudes with psionic powers and robots and shield tech are a less potent military force than a bunch of dudes with guns and some tanks.
On July 22 2020 08:41 Wombat_NI wrote: Protoss haven’t been the weakest race since 2015, but since way back in BW. They have had the odd high point though.
It makes a lot of sense in lore that the dudes with psionic powers and robots and shield tech are a less potent military force than a bunch of dudes with guns and some tanks.
It's because they are so advanced that they can't get anything done. The factions spend their time arguing psionically back and forth rather than taking action. Sure they have the stronger tech, but it's really scattered. Meanwhile, the Zerg swarm act as one, and the Terran are just YOLOing it. Bureaucracies, I tell ya.
On July 22 2020 08:10 Russano wrote: Geez, Aligulac really tells you everything you need to know about the state of race balance in SC2. Sad times to be a protoss.
Aligulac is useless for judging balance especially for Protoss. PvP is so volatile that Protoss players will always be rated lower on average. Protoss is definitely weak currently but the idea that they have been the weakest race since 2015 is just absurd.
The assertion isn't that Protoss has been the weakest race since 2015, it's that Protoss has not been the strongest race since 2015. Looking at aligulac, Protoss was marked as the middle race in March 2018 for example.
On July 22 2020 08:41 Wombat_NI wrote: Protoss haven’t been the weakest race since 2015, but since way back in BW. They have had the odd high point though.
It makes a lot of sense in lore that the dudes with psionic powers and robots and shield tech are a less potent military force than a bunch of dudes with guns and some tanks.
It's because they are so advanced that they can't get anything done. The factions spend their time arguing psionically back and forth rather than taking action. Sure they have the stronger tech, but it's really scattered. Meanwhile, the Zerg swarm act as one, and the Terran are just YOLOing it. Bureaucracies, I tell ya.
But Tassadar unifying them
I do like the idea (and for all factions) of them just discontinuing things though.
‘Yeah em, I know these Reavers served us really well but we’re going to have to discontinue them. Luckily we have ancient war machines from the Aeon of Strife called collosus to use! They’re really tall and vulnerable to anti-air of our enemies but look! They can walk up cliffs’ ‘Ok I like these Collosi, they look cool and they have big lasers that shred things. You know what would be great militarily? If we had both Collosi and Reavers! Imagine it, the mobile collosus striding across the battlefield, with the Reavers scuttling around firing their scarabs, things would melt! Especially if we continued the tradition of carrying reavers in shuttles .’ ‘No.’ ‘What do you mean no? We could slay many an enemy with both these weapons of war!’ Well let me think which of these robotic tools that have served Aiur so well I’d rather have in my forces.’ ‘No we’re streamlining. It’s Collosus. That we’ve only just recovered from the Sea and don’t really know much about.’ ‘But... with minerals and gas we can manufacture anything we know how to? What’s the issue here? Also we have like thousands of Reavers sitting around already?’ ‘No it’s been decided. It’s Collosus. Destroy all the reavers. While you’re at it can you destroy all of our fleet of Corsairs and Arbiters? We can’t really see a tactical need for a machine that can trap units in space-time and teleport units to them.’ ‘You can’t? It’s extremely useful in getting past the humans tank lines!’ ‘Hm, well I suppose the ability to teleport an army is quite useful tactically, but Arbiters are yesterday’s tech. How about we give you a giant arbiter but you can only deploy one?’ ‘Well that sounds wors...’ ‘It’s giant singular Arbiter or no arbiter.’ *Sighs* ‘OK’
Reynor has definitely gotten better, it seems. His barcode MMR is the highest it's ever been, correlating with his aligulac. I guess Reynor is a very late bloomer. His brain may be rapidly developing. Better late than never. :D
I hate the fact that the S. Korea military policy has taken away, or suppressed so many talents. Serral and Reynor could use a greater pool of preys to feed on. huwahaha
On July 22 2020 07:54 Andi_Goldberger wrote: thats such a steep rise in the graph, wow. impressive
I love seeing players come from the middle of the pack to reach elite status. Over those 70 weeks, Reynor gained (2686 -> 3116) 430 points, which was a great run, but not the best ever.
I did some quick comparisons and, not to be outdone, Cure gained (2478-> 3033) 555 points in 56 weeks right in the middle of that run, coinciding with his impressive results in GSL.
In 2018, Maru went from 2496 (#17) to 3047 (#1) (+551) in less than one year.
Dream is on a solid run of late, up +500 since joining BSG.*
*Bad comparison since 2100->2650 is way easier than 2500->3000
On July 22 2020 08:07 Blargh wrote: These colors sure make it easy to understand!
You can save the graph in four different formats, but even if you put 50 players on it you only get line colors of red and blue.
On July 22 2020 08:10 Russano wrote: Geez, Aligulac really tells you everything you need to know about the state of race balance in SC2. Sad times to be a protoss.
Which is backed up by the fact that during no year of lotv has a protoss player been in the top three of earnings for that year. How incredible is that.
On July 22 2020 08:10 Russano wrote: Geez, Aligulac really tells you everything you need to know about the state of race balance in SC2. Sad times to be a protoss.
Aligulac is useless for judging balance especially for Protoss. PvP is so volatile that Protoss players will always be rated lower on average. Protoss is definitely weak currently but the idea that they have been the weakest race since 2015 is just absurd.
The assertion isn't that Protoss has been the weakest race since 2015, it's that Protoss has not been the strongest race since 2015. Looking at aligulac, Protoss was marked as the middle race in March 2018 for example.
Actually I'm making both of those assertions. At a long glance, Protoss hasn't been the best of the three since 2015. Protoss have also spent the most time at bottom fo the three. Over the last 5 years they have statistically been the weakest race. Which isn't shocking or controversial - somebody had to be - there's only 3 races after all. We definitely know it hasn't been Zerg.
Looking at the short term - PvP volatility does absolutley nothing to explain it. Stats' pvp is roughly equal rating with his vT and vZ. He wins equally consistently there as he does in other matches, so if PvP is by defintion "volatile" and thus inconsistent, then PvZ and PvT have proven to be equally volatile. As volatility increases with sample size what you get is not volatility but a conclusion of "worse on average", as that volatility is smoothed out over time.
The simple fact of the matter is that on average, the best ZvP and TvP players are performing significantly better than the top end of PvT and PvZ players. And it's not even that one player happens to be good in a matchup. Every single top player in that list is at minimum, equally good in the matchups as Stats, and then has 1-3 matchups where they smash his numbers.
The only conclusion that can be drawn is that Protoss, on average over time, is the weakest of the three. There may also be something intrinsic to the protoss race that renders it more difficult to play, leading to greater peaks and valleys of results, but overtime that simply translates to "being worse".
Note: This analysis only applies to the very topmost players. There are statistics like PvT being 54 percent in favor of the protoss during this period which point in the other direction. Gratz protoss, you are the best at making it to the round of 16.
As for Reynor, it's cool that he hit up the number 1 spot, no matter how brief. The graphs really show how far he's come. I don't think he's a better overall player than Serral, but I think that his vP and vT are sufficiently high enough to prove a challenge, and his vZ is good enough that he can beat Serral with some amount of regularity - and so far thats more than anyone else has been able to say. If he improves even more than he already has somehow, it would not be a surprise to see him as simply the best player - which is really cool to see given the stagnation in the scene - and especially Korea.
On July 22 2020 08:10 Russano wrote: Geez, Aligulac really tells you everything you need to know about the state of race balance in SC2. Sad times to be a protoss.
Which is backed up by the fact that during no year of lotv has a protoss player been in the top three of earnings for that year. How incredible is that.
And yet Protoss has still won more prize money than Terran over the life of the game. By $500k in fact. Protoss prize money is just spread out over the most players. Zerg is ahead of both by well over $2million.
On July 22 2020 08:10 Russano wrote: Geez, Aligulac really tells you everything you need to know about the state of race balance in SC2. Sad times to be a protoss.
Aligulac is useless for judging balance especially for Protoss. PvP is so volatile that Protoss players will always be rated lower on average. Protoss is definitely weak currently but the idea that they have been the weakest race since 2015 is just absurd.
The assertion isn't that Protoss has been the weakest race since 2015, it's that Protoss has not been the strongest race since 2015. Looking at aligulac, Protoss was marked as the middle race in March 2018 for example.
Actually I'm making both of those assertions. At a long glance, Protoss hasn't been the best of the three since 2015. Protoss have also spent the most time at bottom fo the three. Over the last 5 years they have statistically been the weakest race. Which isn't shocking or controversial - somebody had to be - there's only 3 races after all. We definitely know it hasn't been Zerg.
Looking at the short term - PvP volatility does absolutley nothing to explain it. Stats' pvp is roughly equal rating with his vT and vZ. He wins equally consistently there as he does in other matches, so if PvP is by defintion "volatile" and thus inconsistent, then PvZ and PvT have proven to be equally volatile. As volatility increases with sample size what you get is not volatility but a conclusion of "worse on average", as that volatility is smoothed out over time.
The simple fact of the matter is that on average, the best ZvP and TvP players are performing significantly better than the top end of PvT and PvZ players. And it's not even that one player happens to be good in a matchup. Every single top player in that list is at minimum, equally good in the matchups as Stats, and then has 1-3 matchups where they smash his numbers.
The only conclusion that can be drawn is that Protoss, on average over time, is the weakest of the three. There may also be something intrinsic to the protoss race that renders it more difficult to play, leading to greater peaks and valleys of results, but overtime that simply translates to "being worse".
Note: This analysis only applies to the very topmost players. There are statistics like PvT being 54 percent in favor of the protoss during this period which point in the other direction. Gratz protoss, you are the best at making it to the round of 16.
As for Reynor, it's cool that he hit up the number 1 spot, no matter how brief. The graphs really show how far he's come. I don't think he's a better overall player than Serral, but I think that his vP and vT are sufficiently high enough to prove a challenge, and his vZ is good enough that he can beat Serral with some amount of regularity - and so far thats more than anyone else has been able to say. If he improves even more than he already has somehow, it would not be a surprise to see him as simply the best player - which is really cool to see given the stagnation in the scene - and especially Korea.
From my understanding the matchup ratings aren't completely independent of each other the way you are implying here. Even if you only lose in a single matchup your other ratings will drop if you are gonna streak of losing in that matchup. Current ratings are irrelevant because Protoss is underpowered currently. I would like to see similar analysis in a different timeframe. I feel like for most of 2018 Protoss was completely fine vs everyone besides Serral and Maru. There's a strong argument that without Serral and Maru playing a Protoss wins every event that year besides WCS season 4, IEM, and Blizzcon.
Stats has been on a hot streak recently in PvP and PvP was just recently patched to make it less volatile. Are you actually denying that PvP has been on average the most volatile matchup in the game? I think that is near universally agreed upon and would definitely influence these types of ratings heavily. No player has ever dominated in PvP the way certain top tier Zergs and Terrans have in their mirrors.
Why would they be dependent? Losing a PvZ would not affect your PvT rating (obviously it would affect your PvAll rating)
And my initial points were 1) Protoss clearly suck the most RIGHT NOW, and by a fair margin - thus not a good time to be a protoss. and 2) That based on Aligulacs ahead/behind numbers, that it has been the worst of the three races for the last 5 years - spending the least time in their moment in the sun, and the most time in the basement. (Aligulac has been giving me some "Bad Gateway -Timeout" stuff for the last couple hours so its hard to double check stuff.)
Another of my points - and general thoughts on race balance - is that there is some sort of...let's say clunkiness or inconsistency intrinsic to the protoss race which inhibits its ability to post absurd winrates at the very very top (ala Serral/Reynor/5 fucking terrans right now). It would be interesting to see numbers with certain "best players" removed, but I'm not entirely convinced racial imbalance doesn't intrinsically limit the ability of races to have "best players".
And my point on volatility: I think PvP - based on gut feelin and general perception - is the most volatile of the matchups, and that most people would agree on that. I define volatility as "results being the least tied to skill/lack of mistakes and more towards luck/guessing." - i.e. coinflippy. Now if you take THE BEST and most consistent pvper - Stats - and rate him and his PvP win rate at 2900, then it stands to reason that less volatile matchups SHOULD have a higher rating - and significantly so. Instead what we see, is a nearly identical rating for THE BEST PvZ and PvT. This means that you are basically winning all of the matchups at the same rate as the "coinflippy matchup" - which means that either A) The matchup isn't actually volatile. Your perception is just inaccurate. OR B) The matchup is imbalanced to the point where literally nobody is winning enough that they have higher stats than a volatile matchup and - and the lower winrates that result from it.
Also personally, I think "volatility" is in some ways not a good metric. It work's in small sample sizes, but at larger sample sizes a matchup being "volatile" and your race "sucking major balance ass in this matchup" are funcitonally identical. The only real differentiation is in that of a mirror matchup, where its impossible for one side to have a racial advantage. Hence - volatility is born.
I do agree that Stat's pvp could be inflated by recent changes and is an outlier - hard to double check without current access to aligulac. My gut says that the inflation is not as much as would be necessary.
On July 22 2020 11:42 Russano wrote: Why would they be dependent? Losing a PvZ would not affect your PvT rating (obviously it would affect your PvAll rating)
And my initial points were 1) Protoss clearly suck the most RIGHT NOW, and by a fair margin - thus not a good time to be a protoss. and 2) That based on Aligulacs ahead/behind numbers, that it has been the worst of the three races for the last 5 years - spending the least time in their moment in the sun, and the most time in the basement. (Aligulac has been giving me some "Bad Gateway -Timeout" stuff for the last couple hours so its hard to double check stuff.)
Another of my points - and general thoughts on race balance - is that there is some sort of...let's say clunkiness or inconsistency intrinsic to the protoss race which inhibits its ability to post absurd winrates at the very very top (ala Serral/Reynor/5 fucking terrans right now). It would be interesting to see numbers with certain "best players" removed, but I'm not entirely convinced racial imbalance doesn't intrinsically limit the ability of races to have "best players".
And my point on volatility: I think PvP - based on gut feelin and general perception - is the most volatile of the matchups, and that most people would agree on that. I define volatility as "results being the least tied to skill/lack of mistakes and more towards luck/guessing." - i.e. coinflippy. Now if you take THE BEST and most consistent pvper - Stats - and rate him and his PvP win rate at 2900, then it stands to reason that less volatile matchups SHOULD have a higher rating - and significantly so. Instead what we see, is a nearly identical rating for THE BEST PvZ and PvT. This means that you are basically winning all of the matchups at the same rate as the "coinflippy matchup" - which means that either A) The matchup isn't actually volatile. Your perception is just inaccurate. OR B) The matchup is imbalanced to the point where literally nobody is winning enough that they have higher stats than a volatile matchup and - and the lower winrates that result from it.
Also personally, I think "volatility" is in some ways not a good metric. It work's in small sample sizes, but at larger sample sizes a matchup being "volatile" and your race "sucking major balance ass in this matchup" are funcitonally identical. The only real differentiation is in that of a mirror matchup, where its impossible for one side to have a racial advantage. Hence - volatility is born.
I do agree that Stat's pvp could be inflated by recent changes and is an outlier - hard to double check without current access to aligulac. My gut says that the inflation is not as much as would be necessary.
Losing a PvZ does affect your PvT rating. Aligulac does not treat the match-ups as completely independent, since treating them as loosely correlated helps ratings converge to where they should be faster (e.g if you haven't played a PvP in a while, but have been overperforming in all your PvTs and PvZs aligulac expects your PvP to have improved too).
On July 22 2020 11:42 Russano wrote: Why would they be dependent? Losing a PvZ would not affect your PvT rating (obviously it would affect your PvAll rating)
And my initial points were 1) Protoss clearly suck the most RIGHT NOW, and by a fair margin - thus not a good time to be a protoss. and 2) That based on Aligulacs ahead/behind numbers, that it has been the worst of the three races for the last 5 years - spending the least time in their moment in the sun, and the most time in the basement. (Aligulac has been giving me some "Bad Gateway -Timeout" stuff for the last couple hours so its hard to double check stuff.)
Another of my points - and general thoughts on race balance - is that there is some sort of...let's say clunkiness or inconsistency intrinsic to the protoss race which inhibits its ability to post absurd winrates at the very very top (ala Serral/Reynor/5 fucking terrans right now). It would be interesting to see numbers with certain "best players" removed, but I'm not entirely convinced racial imbalance doesn't intrinsically limit the ability of races to have "best players".
And my point on volatility: I think PvP - based on gut feelin and general perception - is the most volatile of the matchups, and that most people would agree on that. I define volatility as "results being the least tied to skill/lack of mistakes and more towards luck/guessing." - i.e. coinflippy. Now if you take THE BEST and most consistent pvper - Stats - and rate him and his PvP win rate at 2900, then it stands to reason that less volatile matchups SHOULD have a higher rating - and significantly so. Instead what we see, is a nearly identical rating for THE BEST PvZ and PvT. This means that you are basically winning all of the matchups at the same rate as the "coinflippy matchup" - which means that either A) The matchup isn't actually volatile. Your perception is just inaccurate. OR B) The matchup is imbalanced to the point where literally nobody is winning enough that they have higher stats than a volatile matchup and - and the lower winrates that result from it.
Also personally, I think "volatility" is in some ways not a good metric. It work's in small sample sizes, but at larger sample sizes a matchup being "volatile" and your race "sucking major balance ass in this matchup" are funcitonally identical. The only real differentiation is in that of a mirror matchup, where its impossible for one side to have a racial advantage. Hence - volatility is born.
I do agree that Stat's pvp could be inflated by recent changes and is an outlier - hard to double check without current access to aligulac. My gut says that the inflation is not as much as would be necessary.
Losing a PvZ does affect your PvT rating. Aligulac does not treat the match-ups as completely independent, since treating them as loosely correlated helps ratings converge to where they should be faster (e.g if you haven't played a PvP in a while, but have been overperforming in all your PvTs and PvZs aligulac expects your PvP to have improved too).
Thank you I was reading through aligulac and trying to find an explanation of what I stated above because I knew they weren't independent but couldn't remember how it worked.
On July 22 2020 07:54 Andi_Goldberger wrote: thats such a steep rise in the graph, wow. impressive
I love seeing players come from the middle of the pack to reach elite status. Over those 70 weeks, Reynor gained (2686 -> 3116) 430 points, which was a great run, but not the best ever.
I did some quick comparisons and, not to be outdone, Cure gained (2478-> 3033) 555 points in 56 weeks right in the middle of that run, coinciding with his impressive results in GSL.
In 2018, Maru went from 2496 (#17) to 3047 (#1) (+551) in less than one year.
Dream is on a solid run of late, up +500 since joining BSG.*
*Bad comparison since 2100->2650 is way easier than 2500->3000
On July 22 2020 08:07 Blargh wrote: These colors sure make it easy to understand!
You can save the graph in four different formats, but even if you put 50 players on it you only get line colors of red and blue.
I know that it says next to nothing. But MaxPax had his first rated match less than a year ago and has an Rating of 2200 allready (you start with 1000).
Obviously his score had to catch up to what he s actually capable of, so his score was heavily scewed the first couple of periods. It s probably still is, because his Rating is still almost always going up, but he also improved a lot lately.
Hey, new here and not too knowledgeable about Aligulac (or really top level sc2 tbh), but it seems to me that some of these Aligulac rankings are absurd.
SoO holding the #10 spot is pretty hard to believe, especially when he seems to have trouble making RO16 in GSL, which is only a part of the SC2 competition pool. I think DRG is a lot more solid than SoO. Honestly DRG is almost certainly a better zerg at this moment than Dark is (I say this as a big Dark fanboy). You can just tell by watching his games that he is happy to not take any shortcuts, because he knows as a game drags on he is going to have a big edge over his opponents as he creates opportunities for himself. And that is basically the formula for an unstoppable force of a Zerg player.
I also cannot believe that Stats is ranked as the top Protoss when it seems to me that PartinG and Trap are noticeably better players at this point in time. I feel that stats has been pretty weak for a while now, kind of since the GSL vs World where he lost to Serral in the finals.
I only watch GSL, but even amongst the GSL players there seems to be some ridiculous ranking errors. Is there something I am missing? Am I the only person that doesn't think Stats is the best Protoss in the world?
On July 22 2020 08:41 Wombat_NI wrote: Protoss haven’t been the weakest race since 2015, but since way back in BW. They have had the odd high point though.
It makes a lot of sense in lore that the dudes with psionic powers and robots and shield tech are a less potent military force than a bunch of dudes with guns and some tanks.
It's because they are so advanced that they can't get anything done. The factions spend their time arguing psionically back and forth rather than taking action. Sure they have the stronger tech, but it's really scattered. Meanwhile, the Zerg swarm act as one, and the Terran are just YOLOing it. Bureaucracies, I tell ya.
But Tassadar unifying them
I do like the idea (and for all factions) of them just discontinuing things though.
‘Yeah em, I know these Reavers served us really well but we’re going to have to discontinue them. Luckily we have ancient war machines from the Aeon of Strife called collosus to use! They’re really tall and vulnerable to anti-air of our enemies but look! They can walk up cliffs’ ‘Ok I like these Collosi, they look cool and they have big lasers that shred things. You know what would be great militarily? If we had both Collosi and Reavers! Imagine it, the mobile collosus striding across the battlefield, with the Reavers scuttling around firing their scarabs, things would melt! Especially if we continued the tradition of carrying reavers in shuttles .’ ‘No.’ ‘What do you mean no? We could slay many an enemy with both these weapons of war!’ Well let me think which of these robotic tools that have served Aiur so well I’d rather have in my forces.’ ‘No we’re streamlining. It’s Collosus. That we’ve only just recovered from the Sea and don’t really know much about.’ ‘But... with minerals and gas we can manufacture anything we know how to? What’s the issue here? Also we have like thousands of Reavers sitting around already?’ ‘No it’s been decided. It’s Collosus. Destroy all the reavers. While you’re at it can you destroy all of our fleet of Corsairs and Arbiters? We can’t really see a tactical need for a machine that can trap units in space-time and teleport units to them.’ ‘You can’t? It’s extremely useful in getting past the humans tank lines!’ ‘Hm, well I suppose the ability to teleport an army is quite useful tactically, but Arbiters are yesterday’s tech. How about we give you a giant arbiter but you can only deploy one?’ ‘Well that sounds wors...’ ‘It’s giant singular Arbiter or no arbiter.’ *Sighs* ‘OK’
Made me chuckle once or twice xD Now go make a big blogpost about this!
And btw, watch out all Terrans and Zergs, my boy MaxPax is coming to get you!
On July 22 2020 14:32 MaengDa wrote: Hey, new here and not too knowledgeable about Aligulac (or really top level sc2 tbh), but it seems to me that some of these Aligulac rankings are absurd.
SoO holding the #10 spot is pretty hard to believe, especially when he seems to have trouble making RO16 in GSL, which is only a part of the SC2 competition pool. I think DRG is a lot more solid than SoO. Honestly DRG is almost certainly a better zerg at this moment than Dark is (I say this as a big Dark fanboy). You can just tell by watching his games that he is happy to not take any shortcuts, because he knows as a game drags on he is going to have a big edge over his opponents as he creates opportunities for himself. And that is basically the formula for an unstoppable force of a Zerg player.
I also cannot believe that Stats is ranked as the top Protoss when it seems to me that PartinG and Trap are noticeably better players at this point in time. I feel that stats has been pretty weak for a while now, kind of since the GSL vs World where he lost to Serral in the finals.
I only watch GSL, but even amongst the GSL players there seems to be some ridiculous ranking errors. Is there something I am missing? Am I the only person that doesn't think Stats is the best Protoss in the world?
Aligulac isn t realy usefull to determine the best player in the game or to make an absolutly objective rating. It s just a statistic, showing how well Players perform in ALL tournament games they play. It s also somewhat precise in predicting head to head matches, if the players are from the same player pool It doesnt weigth the games between KungFu Cup, GSL or Katowice, or if it was super close or an absolute beatdown. Only the win matters and the rating difference between the two players.
It s quite good in what it tries to archive, but not that usefull otherwise. It sometimes gets used for tournament seeding proceses (see Warchest Team League, where you had to have at least 1500 Score Rating or HSC XX, where they used in additional to WCS Points). Looking at the rating alone, without also looking at big Tournament results isn t giving the full picture. It s still a realy nice database, though
On July 22 2020 14:32 MaengDa wrote: Hey, new here and not too knowledgeable about Aligulac (or really top level sc2 tbh), but it seems to me that some of these Aligulac rankings are absurd.
SoO holding the #10 spot is pretty hard to believe, especially when he seems to have trouble making RO16 in GSL, which is only a part of the SC2 competition pool. I think DRG is a lot more solid than SoO. Honestly DRG is almost certainly a better zerg at this moment than Dark is (I say this as a big Dark fanboy). You can just tell by watching his games that he is happy to not take any shortcuts, because he knows as a game drags on he is going to have a big edge over his opponents as he creates opportunities for himself. And that is basically the formula for an unstoppable force of a Zerg player.
I also cannot believe that Stats is ranked as the top Protoss when it seems to me that PartinG and Trap are noticeably better players at this point in time. I feel that stats has been pretty weak for a while now, kind of since the GSL vs World where he lost to Serral in the finals.
I only watch GSL, but even amongst the GSL players there seems to be some ridiculous ranking errors. Is there something I am missing? Am I the only person that doesn't think Stats is the best Protoss in the world?
Aligulac isn t realy usefull to determine the best player in the game or to make an absolutly objective rating. It s just a statistic, showing how well Players perform in ALL tournament games they play. It s also somewhat precise in predicting head to head matches, if the players are from the same player pool It doesnt weigth the games between KungFu Cup, GSL or Katowice, or if it was super close or an absolute beatdown. Only the win matters and the rating difference between the two players.
It s quite good in what it tries to archive, but not that usefull otherwise. It sometimes gets used for tournament seeding proceses (see Warchest Team League, where you had to have at least 1500 Score Rating or HSC XX, where they used in additional to WCS Points). Looking at the rating alone, without also looking at big Tournament results isn t giving the full picture. It s still a realy nice database, though
While aligulac doesn't distinguish between the tournaments it does distinguish between the nature of the result (how one sided it is) as that is how it measures performance. As an example if Serral is playing a few ZvPs it takes an aggregate score of those protoss player's PvZ rating and puts it against Serral's ZvP rating to infer how many maps he should take from all of them. If Serral takes more maps than inferred his ZvP rating then his ZvP score goes up in proportion to how much he's won over how much he's expected to win.
On July 22 2020 14:32 MaengDa wrote: Hey, new here and not too knowledgeable about Aligulac (or really top level sc2 tbh), but it seems to me that some of these Aligulac rankings are absurd.
SoO holding the #10 spot is pretty hard to believe, especially when he seems to have trouble making RO16 in GSL, which is only a part of the SC2 competition pool. I think DRG is a lot more solid than SoO. Honestly DRG is almost certainly a better zerg at this moment than Dark is (I say this as a big Dark fanboy). You can just tell by watching his games that he is happy to not take any shortcuts, because he knows as a game drags on he is going to have a big edge over his opponents as he creates opportunities for himself. And that is basically the formula for an unstoppable force of a Zerg player.
I also cannot believe that Stats is ranked as the top Protoss when it seems to me that PartinG and Trap are noticeably better players at this point in time. I feel that stats has been pretty weak for a while now, kind of since the GSL vs World where he lost to Serral in the finals.
I only watch GSL, but even amongst the GSL players there seems to be some ridiculous ranking errors. Is there something I am missing? Am I the only person that doesn't think Stats is the best Protoss in the world?
DRG has only started performing at a high level recently. It'll take a while for his rating to climb. That's how elo-based rating works. SoO's match history shows he's been performing relatively well. You can't just look at a blemish in a player's performance and think it's all there is. GSL results are counted equally as online results by Aligulac policy.
On July 22 2020 08:10 Russano wrote: Geez, Aligulac really tells you everything you need to know about the state of race balance in SC2. Sad times to be a protoss.
Aligulac is useless for judging balance especially for Protoss. PvP is so volatile that Protoss players will always be rated lower on average. Protoss is definitely weak currently but the idea that they have been the weakest race since 2015 is just absurd.
The assertion isn't that Protoss has been the weakest race since 2015, it's that Protoss has not been the strongest race since 2015. Looking at aligulac, Protoss was marked as the middle race in March 2018 for example.
Actually I'm making both of those assertions. At a long glance, Protoss hasn't been the best of the three since 2015. Protoss have also spent the most time at bottom fo the three. Over the last 5 years they have statistically been the weakest race. Which isn't shocking or controversial - somebody had to be - there's only 3 races after all. We definitely know it hasn't been Zerg.
Looking at the short term - PvP volatility does absolutley nothing to explain it. Stats' pvp is roughly equal rating with his vT and vZ. He wins equally consistently there as he does in other matches, so if PvP is by defintion "volatile" and thus inconsistent, then PvZ and PvT have proven to be equally volatile. As volatility increases with sample size what you get is not volatility but a conclusion of "worse on average", as that volatility is smoothed out over time.
The simple fact of the matter is that on average, the best ZvP and TvP players are performing significantly better than the top end of PvT and PvZ players. And it's not even that one player happens to be good in a matchup. Every single top player in that list is at minimum, equally good in the matchups as Stats, and then has 1-3 matchups where they smash his numbers.
The only conclusion that can be drawn is that Protoss, on average over time, is the weakest of the three. There may also be something intrinsic to the protoss race that renders it more difficult to play, leading to greater peaks and valleys of results, but overtime that simply translates to "being worse".
Note: This analysis only applies to the very topmost players. There are statistics like PvT being 54 percent in favor of the protoss during this period which point in the other direction. Gratz protoss, you are the best at making it to the round of 16.
As for Reynor, it's cool that he hit up the number 1 spot, no matter how brief. The graphs really show how far he's come. I don't think he's a better overall player than Serral, but I think that his vP and vT are sufficiently high enough to prove a challenge, and his vZ is good enough that he can beat Serral with some amount of regularity - and so far thats more than anyone else has been able to say. If he improves even more than he already has somehow, it would not be a surprise to see him as simply the best player - which is really cool to see given the stagnation in the scene - and especially Korea.
From my understanding the matchup ratings aren't completely independent of each other the way you are implying here. Even if you only lose in a single matchup your other ratings will drop if you are gonna streak of losing in that matchup. Current ratings are irrelevant because Protoss is underpowered currently. I would like to see similar analysis in a different timeframe. I feel like for most of 2018 Protoss was completely fine vs everyone besides Serral and Maru. There's a strong argument that without Serral and Maru playing a Protoss wins every event that year besides WCS season 4, IEM, and Blizzcon.
Stats has been on a hot streak recently in PvP and PvP was just recently patched to make it less volatile. Are you actually denying that PvP has been on average the most volatile matchup in the game? I think that is near universally agreed upon and would definitely influence these types of ratings heavily. No player has ever dominated in PvP the way certain top tier Zergs and Terrans have in their mirrors.
On July 23 2020 02:53 Ronski wrote: I have trouble understanding how Aligulac makes the ratings.
Serrals performance leading to negative rating adjustment
Reynors performance leading to positive rating adjustment.
Just wondering if anyone knows how they make the calculations.
Mathematical model of Aligulac is faultless. Serral was just so damn good his rating was continuously inflated by his incredible consistency and success against top level opponents. Effectively big percentage of extra points to the system by rating inflation (caused by new players introduced to the system) largely accumulated for Serral some 2 years in row.
I'm more concerned that if lists can go by without adequate correction to Serral's match history, how anyone can take other players' match histories seriously, and consequently any others' either?
By me. (when writing this comment, that match is still missing from Serral's match history, thus also Skillous' match history... impacting to both players' ratings).
Problem isn't Aligulac's mathematical and logical integrity, it's faultless, solid, and openly explained. Problem is that people do not update it enough rigorously. If there are missing data point from Serral's account, one might justifiably ask: "what else is missing from player X's or Y's match histories?"
I assume here that SKillous's match in that particular Premier tourney is just as valuable as everyone else's whom matches are properly logged, calculated, and rated... including Serral. Right?
On July 22 2020 08:41 Wombat_NI wrote: Protoss haven’t been the weakest race since 2015, but since way back in BW. They have had the odd high point though.
It makes a lot of sense in lore that the dudes with psionic powers and robots and shield tech are a less potent military force than a bunch of dudes with guns and some tanks.
It's because they are so advanced that they can't get anything done. The factions spend their time arguing psionically back and forth rather than taking action. Sure they have the stronger tech, but it's really scattered. Meanwhile, the Zerg swarm act as one, and the Terran are just YOLOing it. Bureaucracies, I tell ya.
But Tassadar unifying them
I do like the idea (and for all factions) of them just discontinuing things though.
‘Yeah em, I know these Reavers served us really well but we’re going to have to discontinue them. Luckily we have ancient war machines from the Aeon of Strife called collosus to use! They’re really tall and vulnerable to anti-air of our enemies but look! They can walk up cliffs’ ‘Ok I like these Collosi, they look cool and they have big lasers that shred things. You know what would be great militarily? If we had both Collosi and Reavers! Imagine it, the mobile collosus striding across the battlefield, with the Reavers scuttling around firing their scarabs, things would melt! Especially if we continued the tradition of carrying reavers in shuttles .’ ‘No.’ ‘What do you mean no? We could slay many an enemy with both these weapons of war!’ Well let me think which of these robotic tools that have served Aiur so well I’d rather have in my forces.’ ‘No we’re streamlining. It’s Collosus. That we’ve only just recovered from the Sea and don’t really know much about.’ ‘But... with minerals and gas we can manufacture anything we know how to? What’s the issue here? Also we have like thousands of Reavers sitting around already?’ ‘No it’s been decided. It’s Collosus. Destroy all the reavers. While you’re at it can you destroy all of our fleet of Corsairs and Arbiters? We can’t really see a tactical need for a machine that can trap units in space-time and teleport units to them.’ ‘You can’t? It’s extremely useful in getting past the humans tank lines!’ ‘Hm, well I suppose the ability to teleport an army is quite useful tactically, but Arbiters are yesterday’s tech. How about we give you a giant arbiter but you can only deploy one?’ ‘Well that sounds wors...’ ‘It’s giant singular Arbiter or no arbiter.’ *Sighs* ‘OK’
Made me chuckle once or twice xD Now go make a big blogpost about this!
And btw, watch out all Terrans and Zergs, my boy MaxPax is coming to get you!
Glad it got a chuckle, perhaps one day when inspiration hits/when I don’t have my 7 year old clinging on to me every waking minute!
On July 23 2020 02:53 Ronski wrote: I have trouble understanding how Aligulac makes the ratings.
Serrals performance leading to negative rating adjustment
Reynors performance leading to positive rating adjustment.
Just wondering if anyone knows how they make the calculations.
Because the rating adjustments of three match-ups are not independent. Result in one match-up changes the general expectation to the player which includes the other two. Therefore when only one match-up was played in a period, all three scores are adjusted simultaneously. When the general performance in a period was to bad, a single overperformed match-up could still get downvoted for the likelihood of just being lucky.
That might be the worst chart I've seen in a long time. Why aren't the lines different colors? When you have 5 different lines but only two different colors it makes it unreadable.
EPT![[image loading]](https://i.imgur.com/9jd1Q8f.jpg)
![[image loading]](https://i.imgur.com/qMaJVl5.jpg)
![[image loading]](https://i.imgur.com/J3v7idG.png)
![[image loading]](https://i.imgur.com/BdsUK51.png)
ESL Pro Tour/2020/21/Masters/Summer/EU
