|
On July 22 2024 02:36 Charoisaur wrote:Show nested quote +On July 22 2024 02:14 PremoBeats wrote:On July 22 2024 02:01 Charoisaur wrote:On July 22 2024 01:53 WombaT wrote:On July 22 2024 01:41 Charoisaur wrote:On July 22 2024 01:16 PremoBeats wrote:On July 22 2024 01:05 Charoisaur wrote:But in my opinion, this whole reasoning is ludicrous to begin with. As stated above, Serral played many of the best the 2015-era had to offer. 3x World Champion sOs, 1x World Champion Zest, 3x World Champion Rogue, 1x World Champion PartinG, 1x World Champion TY, Classic, TaeJa, Trap, Creator, soO as well as INnoVation. Serral is still is battling World Champions ByuN, and Dark as well as herO, Cure, Stats, Solar and Maru on a regular basis who all were relevant in 2015 already. On top, Serral was able to fend of new talents such as World Champion Reynor or Clem and MaxPax.
And how many of those were in their peak when they faced Serral? The versions of sOs, PartinG, TaeJa, Zest, ByuN, INnoVation and soO that Serral faced were nowhere near their peak. That's like taking Serrals negative record against Dayshi and Jaedong as an argument. Yeah, Serral is the most dominant player ever and probably has the best case to be the Goat due to no player really sticking out during the most competitive period but you can't derive from him beating players way past their peak that he would beat them at their peak too, that argument makes 0 sense. TY was still ranked 3 near the end of 2018. INno took ranks of Serral at times and was still ranked 2 mid 2020. His win rates still exceeded 60%. Zest had similar win rates to his prime when facing Serral, the same is true for PartinG. Why exactly do you think they got worse and not that the new talents simply were better which dropped these players win rates, tournament rates and overall glow like it was always the case? The bulk of the players you mentioned did not drop in skill because of old age. For most, their 2018-versions - from my estimation - would have easily kicked the asses of their 2015-versions, if I look at gameplay and strategy. Plus, I never made the claim that "Serral would beat these players at their peak". I simply stated that he beat players that were already around in the 2015-era. Well, I can prove "they got worse" just as little as you can prove "they didn't get worse, others just surpassed them." But I think it's rather unlikely that the number of serious championship contenders decreasing at the same time as the competitiveness of the scene decreasing is a mere coincidence. Also looking specifically at the players I mentioned Inno was rather vocal about losing motivation and practicing less, TaeJa and PartinG took a long break from sc2 and never really caught up after that, for sOs and soO I think LotV never resonated with their strengths (hard to directly compare skill when they are playing a completely different game now). For Zest and ByuN it's less clear but their gameplay just didn't convince me as much anymore as when they were in their peak. If your claim wasn't that Serral would beat these players at their peak I'm not sure what the point in mentioning that he's beating them now is. As I see it it has no relevance on the argument In a sense, minus players impacted notably by injury like Taeja, that is kind of on those players though. I mean I realise I just throw out the ‘other side of the coin’ arguments perpetually in this domain, but it Serral gets a few minus points for the era he’s in and the drop off, then surely you have to take points off all those players who couldn’t keep up their peak levels? Byun is a patch Terran :p Agreed on SoS for sure. Less so perhaps with soO although I can see it to be fair: partly given his biggest result was in Legacy, partly given how strong he is mechanically and how fast Legacy is and how demanding in that capacity. I think soO is a player who’s waxed and waned a bit with Legacy metas, whereas $o$ it’s the game itself that didn’t suit him. Yeah, it's on the players. The problem is that due to lack of new talent there's nobody who takes the spots as top championship contenders of declined soO/Inno, making tournaments objectively easier to win. Regarding soO I just think that his big weakness (lategame and spellcaster control) is something HotS Zerg could get away with but definitely not LotV Zerg. I know he won IEM Katowice but that was more of a flash in the pan result compared to his consistent excellence in HotS Well, there is Serral 2018, Dark 2019, Rogue 2020, Reynor 2021, Serral a 2nd time in 2022 and Oliveira 2023. That is 6 world champions for 7 years, a similar rate to WCSs 2017 and before with 5 champions in 6 years. Plus, you got MaxPax who only plays online, Clem who is getting better and better. Maru wins a lot, but herO, Shin, Solar, GuMiho, Cure, Bunny and Zest all pop up in Premier Tournaments as winner or runner up in the last two years. In my opinion we saw a lot new talent (MaxPax, Clem, Reynor, Serral) to replace them, as well as veterans still going strong or becoming even stronger (Oliveira, Maru, GuMiho, Shin, Cure, hero, Solar). As a matter of fact, one talent and one veteran that are extreme outliers. The thrones of Premier Tournaments would be much more fought for, if those two didn't exist. But in my opinion that isn't an issue of new talent (as Serral is the new talent post-2018) but simply a matter of dominance portrayed by Maru and Serral. In terms of serious championship contenders: Today: Serral, Reynor, Clem, Maru, Dark, herO Everyone else you would be extremely surprised to see win -I wouldn't include Oliveira, because he only ever won 1 premier tournament in 1 finals appearance (otherwise you could also include Flash etc in the 2015 list). 2015: PartinG, Rain, Classic, Zest, sOs, herO, Maru, Inno, soO, Life, ByuL None of them would've been a big surprise to win a big tournament. In terms of 2nd tier players who could cause an upset to the top contenders the gap would be even bigger
Agreed on Oliveira. It is just something to point out though. That there are still players who are able to penetrate through Serral and Maru AND on top all the other high tier contenders.
GumiHo had 2 runner ups last year, Cure 3. If you add MaxPax who only falls away because of him not playing offline, you get pretty similar numbers. At least if you compare it to the image people paint of that era where you could get the idea that nowadays no players are left in the StarCraft 2 scene 
|
Bisutopia19229 Posts
I don’t disagree with the results, but I am still one of those people that removes 3 years of results in my head from when Zerg was so broken in LoTV. I’m honestly surprised any Protoss besides Trap kept playing the game. And to be fair, he had to endure the highest level beat downs of anybody.
Thanks for the great post OP. I love the effort and data.
|
Excellent and highly commendable work! o7
It would be nice to see how much one should handicap and penalize Serral using these same metrics to actually make Serral/Maru appear numerically dead-even, thought I suspect that even the most hardcore nostalgia-driven Korean Elitist Maru Fanboi couldn't agree with such absurd weightings, as they would effectively render the analysis meaningless and only offer an upside-down view to the same ridiculousness of Serral's dominance that is apparent already.
It's clear why leaving Reynor/Clem/ vs foreigner -part out from Serral's body of work is fitting and easy simplification for analytic purposes in this context, but nobody can really evaluate Serral's "GOATNESS" without it, as intangibles directly or indirectly relating to it are also important factor why so many consider him as The Goat. We can call it "growing the scene" or "Let's singlehandedly extend the longevity of the competitive SC2 scene at it's highest levels" -factor, for examples, as innumerable as it ever can be. Handicapping this aspect of Serral's resume to nul feels most arbitrarily unfair aspect of this analysis, much more than heavily undervaluing era/tournaments etc. of Serral's career of full time Pro.
Anyway, great job, man!
|
On July 22 2024 04:52 UnLarva wrote:
Excellent and highly commendable work! o7
It would be nice to see how much one should handicap and penalize Serral using these same metrics to actually make Serral/Maru appear numerically dead-even, thought I suspect that even the most hardcore nostalgia-driven Korean Elitist Maru Fanboi couldn't agree with such absurd weightings, as they would effectively render the analysis meaningless and only offer an upside-down view to the same ridiculousness of Serral's dominance that is apparent already.
It's clear why leaving Reynor/Clem/ vs foreigner -part out from Serral's body of work is fitting and easy simplification for analytic purposes in this context, but nobody can really evaluate Serral's "GOATNESS" without it, as intangibles directly or indirectly relating to it are also important factor why so many consider him as The Goat. We can call it "growing the scene" or "Let's singlehandedly extend the longevity of the competitive SC2 scene at it's highest levels" -factor, for examples, as innumerable as it ever can be. Handicapping this aspect of Serral's resume to nul feels most arbitrarily unfair aspect of this analysis, much more than heavily undervaluing era/tournaments etc. of Serral's career of full time Pro.
Anyway, great job, man!
You have a point! At the moment I am pretty done with working on this, but getting these guys in the bigger picture might seem like a worthwhile idea (perhaps without having to analyze each tournament they won/placed 2nd as that was the most work... I think I am now able to grab a tournaments value by simply looking at the structure and players pretty well after all this).
Making Maru dead-even, except for the tournament score, would take ridiculous nerfs especially in the Aligulac Rank Analysis, Hall of Fame, as well as match win rates and tournament win rate. I can look at this superficially in a couple of weeks if you want.
|
On July 22 2024 01:29 Mizenhauer wrote:
No Mvp, no goat list.
I had some time, so I evaluated Mvp's data:
1. His Aligulac ranking went as expected. He was 106 lists in the top 10. Around 30 times less than Serral was on rank 1. But when he was in the top 10, he was on rank 1 71,30% - the best score after Serral. This is countered by his low representation in the top 10, with only 84 times, the second lowest score.
2. On the Aligulac HoF he is placed 8th, which gives him the number 4 spot on the list. His PP score versus INnoVation, Maru and Serral is pretty bad though.
3. Mvp's overall match win rate is 52,38%, by a large margin the lowest score. His best year wouldn't even put him in the top 5 years, as Serral's and Maru's best years are better.
4. 2011 saves him a 2nd place in the Tournament win rate analysis. But if he had continued like in 2012 and 2013 he would have dropped down to last place as well in 2014.
5. Average place also sees him in the last place, as 2012 and 2013, drag his hyper successful year 2011 down.
6. As it is obvious by simply counting that Mvp can't win in the tournament score and I was too lazy to actually calculate all the tournaments he participated in, I simply gave him a tournament multiplier of 1 for each placement (massively boosting his score), except his world championship, which of course was multiplied with 1,1. He still has the lowest tournament score.
7. Efficiency-wise he is placed 3rd behind Serral and Maru, as 2011 is his only year where he collected many points, which in relation made his efficiency go down. This score is also boosted as it is a dividend of the tournament score.
Mvp is the player with the least consistency and duration and the trend implies that more years would have simply made his statistics look worse. He only has 1 hyper dominant year, which is not even the best year in this whole comparison and he sports match win rates of under 50% (2012), 38% (2013) and 31% (2014) .
It is safe to say that my pre-screening was correct in not letting him in the final contender list.
|
Northern Ireland25037 Posts
On July 22 2024 06:00 PremoBeats wrote:I had some time, so I evaluated Mvp's data: 1. His Aligulac ranking went as expected. He was 106 lists in the top 10. Around 30 times less than Serral was on rank 1. But when he was in the top 10, he was on rank 1 71,30% - the best score after Serral. This is countered by his low representation in the top 10, with only 84 times, the second lowest score. 2. On the Aligulac HoF he is placed 8th, which gives him the number 4 spot on the list. His PP score versus INnoVation, Maru and Serral is pretty bad though. 3. Mvp's overall match win rate is 52,38%, by a large margin the lowest score. His best year wouldn't even put him in the top 5 years, as Serral's and Maru's best years are better. 4. 2011 saves him a 2nd place in the Tournament win rate analysis. But if he had continued like in 2012 and 2013 he would have dropped down to last place as well in 2014. 5. Average place also sees him in the last place, as 2012 and 2013, drag his hyper successful year 2011 down. 6. As it is obvious by simply counting that Mvp can't win in the tournament score and I was too lazy to actually calculate all the tournaments he participated in, I simply gave him a tournament multiplier of 1 for each placement (massively boosting his score), except his world championship, which of course was multiplied with 1,1. He still has the lowest tournament score. 7. Efficiency-wise he is placed 3rd behind Serral and Maru, as 2011 is his only year where he collected many points, which in relation made his efficiency go down. This score is also boosted as it is a dividend of the tournament score. Mvp is the player with the least consistency and duration and the trend implies that more years would have simply made his statistics look worse. He only has 1 hyper dominant year, which is not even the best year in this whole comparison and he sports match win rates of under 50% (2012), 38% (2013) and 31% (2014) . It is safe to say that my pre-screening was correct in not letting him in the final contender list.
Data alone cannot express the greatness of Mvp. That intangible kind of greatness that one feels in the heart, or possibly balls.
That said I mean has Serral ever had a sub-70% win rate year since 17/18? Serral’s basically managed that for a ridiculous amount of time. Longer than Mvp’s entire career, hell at this stage longer than the entirety of WoL and HoTS combined.
Enjoy the TL+, albeit I fucked up and awarded it to the wrong post :p
|
On July 22 2024 06:21 WombaT wrote:Show nested quote +On July 22 2024 06:00 PremoBeats wrote:On July 22 2024 01:29 Mizenhauer wrote:
No Mvp, no goat list. I had some time, so I evaluated Mvp's data: 1. His Aligulac ranking went as expected. He was 106 lists in the top 10. Around 30 times less than Serral was on rank 1. But when he was in the top 10, he was on rank 1 71,30% - the best score after Serral. This is countered by his low representation in the top 10, with only 84 times, the second lowest score. 2. On the Aligulac HoF he is placed 8th, which gives him the number 4 spot on the list. His PP score versus INnoVation, Maru and Serral is pretty bad though. 3. Mvp's overall match win rate is 52,38%, by a large margin the lowest score. His best year wouldn't even put him in the top 5 years, as Serral's and Maru's best years are better. 4. 2011 saves him a 2nd place in the Tournament win rate analysis. But if he had continued like in 2012 and 2013 he would have dropped down to last place as well in 2014. 5. Average place also sees him in the last place, as 2012 and 2013, drag his hyper successful year 2011 down. 6. As it is obvious by simply counting that Mvp can't win in the tournament score and I was too lazy to actually calculate all the tournaments he participated in, I simply gave him a tournament multiplier of 1 for each placement (massively boosting his score), except his world championship, which of course was multiplied with 1,1. He still has the lowest tournament score. 7. Efficiency-wise he is placed 3rd behind Serral and Maru, as 2011 is his only year where he collected many points, which in relation made his efficiency go down. This score is also boosted as it is a dividend of the tournament score. Mvp is the player with the least consistency and duration and the trend implies that more years would have simply made his statistics look worse. He only has 1 hyper dominant year, which is not even the best year in this whole comparison and he sports match win rates of under 50% (2012), 38% (2013) and 31% (2014) . It is safe to say that my pre-screening was correct in not letting him in the final contender list. Data alone cannot express the greatness of Mvp. That intangible kind of greatness that one feels in the heart, or possibly balls. That said I mean has Serral ever had a sub-70% win rate year since 17/18? Serral’s basically managed that for a ridiculous amount of time. Longer than Mvp’s entire career, hell at this stage longer than the entirety of WoL and HoTS combined. Enjoy the TL+, albeit I fucked up and awarded it to the wrong post :p
I utterly feel you... that is exactly why this list was necessary for me. We all suffer from biases, subjective ideas and certainties that can very fast turn to uncertainties. But learning that you were wrong about something is always a great thing... as now you are wrong about one less thing :D
And to answer your question: These are Serral's win rates since 2018:
85,71
76,67
85,71
70,31
73,68
85,11
92,31
Mind you, only against the top Koreans of the world.
|
Wow I'm really surprised Rogue pales so heavily when it comes to the metric of how many times he was ranked #1, #2, or #3 on aligulac compared to the other top 4. I knew how he hit "above his skill level", but damn i didn't think it was that drastic.
It's really incredible how many premiers and WCs he won despite that. That is a big part of what makes him great too.
I also think the players you chose to examine for the list was great, the cutoff is perfect at those 4 players.
Aligulac PP and stuff is cool too, gonna check it out.
Thanks for all the hard work!! It was an enjoyable read, and I like how it often considered counter arguments, and overall it was very logically written and unbiased. It's incredible how handicapped Serral was in your measurements, yet he still came out #1 (or tied for #1) in all 7 categories.
I would love to see who the top 10 would be given your methodology, I wonder if you would be able to give a rough list off of what you remember or if you'd have to do a lot more research. I'm especially curious about how Taeja, Life, sOs, and MVP would compare.
|
On July 22 2024 07:17 Yoshi Kirishima wrote:
Wow I'm really surprised Rogue pales so heavily when it comes to that metric (times ranked #1, #2, or #3 on aligulac) compared to the other top 4. I knew how he hit "above his skill level", but damn i didn't think it was that drastic.
It's really incredible how many premiers and WCs he won despite that. That is a big part of what makes him great too.
Aligulac PP and stuff is cool too, gonna check it out.
Thanks for all the hard work!! It was an enjoyable read, and I like how it often considered counter arguments, and overall it was very logically written and unbiased. It's incredible how handicapped Serral was in your measurements, yet he still came out #1 (or tied for #1) in all 7 categories.
I would love to see who the top 10 would be given your methodology, I wonder if you would be able to give a rough list off of what you remember or if you'd have to do a lot more research. I'm especially curious about how Taeja, Life, sOs, and MVP would compare.
That surprised me as well...
Mvp's statistics are explained in my answer to Mizenhauer somewhere in this thread... I will do Life next, when I have the time, although I probably won't go as deep into the tournament analysis as grinding out those Ro16 and Ro8 was an absurdly time-consuming task and a real pain in the ass.
Thanks for your kind words!
|
“If you torture the data long enough, it will confess to anything”.
|
On July 22 2024 15:23 MJG wrote:
“If you torture the data long enough, it will confess to anything”.
Not sure how serious your comment is, but in which of my 7 metrics that were analyzed did you see me "torture data"?
Did you read the article? Did you understand that the data regarding the person which came out on top was penalized or "tortured" the most? By far?
|
On July 22 2024 15:29 PremoBeats wrote:Show nested quote +On July 22 2024 15:23 MJG wrote:
“If you torture the data long enough, it will confess to anything”. Not sure how serious your comment is, but in which of my 7 metrics that were analyzed did you see me "torture data"? Did you read the article? Did you understand that the data regarding the person which came out on top was penalized pr "tortured" the most? By far?
That's exactly the point, you only analysed seven metrics! I'm sure I could find seven metrics that would give Has as the correct answer if I tried hard enough!
I was being facetious and I thought it was pretty obvious...
No amount of statistical analysis is likely to change my opinion that Legacy of the Void is a balance/design shitshow. It's the videogaming equivalent of getting 100m sprinters to wear clown shoes. Fun and entertaining? Sure. Provides useful data for measuring accomplishments? Definitely not.
|
France12790 Posts
On July 22 2024 07:17 Yoshi Kirishima wrote:
Wow I'm really surprised Rogue pales so heavily when it comes to the metric of how many times he was ranked #1, #2, or #3 on aligulac compared to the other top 4. I knew how he hit "above his skill level", but damn i didn't think it was that drastic.
It's really incredible how many premiers and WCs he won despite that. That is a big part of what makes him great too.
I also think the players you chose to examine for the list was great, the cutoff is perfect at those 4 players.
Aligulac PP and stuff is cool too, gonna check it out.
Thanks for all the hard work!! It was an enjoyable read, and I like how it often considered counter arguments, and overall it was very logically written and unbiased. It's incredible how handicapped Serral was in your measurements, yet he still came out #1 (or tied for #1) in all 7 categories.
I would love to see who the top 10 would be given your methodology, I wonder if you would be able to give a rough list off of what you remember or if you'd have to do a lot more research. I'm especially curious about how Taeja, Life, sOs, and MVP would compare.
He (Rogue) didn't particularly rise above his skill level though? It's mainly that he was a big tournament player, and not an aligulac stat padder like other players
|
On July 22 2024 01:05 Charoisaur wrote:Show nested quote +But in my opinion, this whole reasoning is ludicrous to begin with. As stated above, Serral played many of the best the 2015-era had to offer. 3x World Champion sOs, 1x World Champion Zest, 3x World Champion Rogue, 1x World Champion PartinG, 1x World Champion TY, Classic, TaeJa, Trap, Creator, soO as well as INnoVation. Serral is still is battling World Champions ByuN, and Dark as well as herO, Cure, Stats, Solar and Maru on a regular basis who all were relevant in 2015 already. On top, Serral was able to fend of new talents such as World Champion Reynor or Clem and MaxPax.
And how many of those were in their peak when they faced Serral? The versions of sOs, PartinG, TaeJa, Zest, ByuN, INnoVation and soO that Serral faced were nowhere near their peak. That's like taking Serrals negative record against Dayshi and Jaedong as an argument. Yeah, Serral is the most dominant player ever and probably has the best case to be the Goat due to no player really sticking out during the most competitive period but you can't derive from him beating players way past their peak that he would beat them at their peak too, that argument makes 0 sense.
That argument of the supposed peak in skill of someone does not make sense. In a game/sport based only on 1 vs 1 confrontation you can just compare players using their results as data. There is no correlation between numbers of competitors and skill. More than that, I would say that continuously training made the players raw skill great with the time passing (trying to clarify: if we had a time machine and we took a middle of the pack player of today and, for example, MVP from 2011, the latter would be destroyed because of the way better knowledge of the have of the former). So the only data we should take into account when comparing players should be their results (probably weighted on the importance of tourneys from which that results came) not the year in which that results were achieved.
|
On July 22 2024 15:56 MJG wrote:Show nested quote +On July 22 2024 15:29 PremoBeats wrote:On July 22 2024 15:23 MJG wrote:
“If you torture the data long enough, it will confess to anything”. Not sure how serious your comment is, but in which of my 7 metrics that were analyzed did you see me "torture data"? Did you read the article? Did you understand that the data regarding the person which came out on top was penalized pr "tortured" the most? By far? That's exactly the point, you only analysed seven metrics! I'm sure I could find seven metrics that would give Has as the correct answer if I tried hard enough! I was being facetious and I thought it was pretty obvious...No amount of statistical analysis is likely to change my opinion that Legacy of the Void is a balance/design shitshow. It's the videogaming equivalent of getting 100m sprinters to wear clown shoes. Fun and entertaining? Sure. Provides useful data for measuring accomplishments? Definitely not.
I am heavily inclined to believe that your proposal (although again you are not being serious about it, I presume) of portraying Has as GOAT in objective metrics is not possible.
I analyzed the metrics that show important qualities of a GOAT the most. Would it make Has a GOAT contender if he was the best... I don't know... the best cannon rusher in the history of the game? Or that he was the best at playing StarCraft 2 with one hand strapped to his back? No, of course not. I contemplated 7 metrics that show consistency, dominance and efficiency. What other qualities do you think a GOAT needs of not these? And which metrics would you further include to measure them, if these 7 don't satisfy you?
In my opinion you are displaying a hyper-skeptic attitude or even an incredulity fallacy, as the evidence that is presented is overwhelming.
But I am open to criticism... thus please tell me what kind of metric it is, that you are possibly missing here.
Why would you say that LotV is a balance shitshow? What exactly do you mean by it? There is extensive data about balancing out there.... but as you said: You seem to have pre-formed opinion that isn't likely to change.
|
On July 22 2024 16:27 Parser wrote:Show nested quote +On July 22 2024 01:05 Charoisaur wrote:But in my opinion, this whole reasoning is ludicrous to begin with. As stated above, Serral played many of the best the 2015-era had to offer. 3x World Champion sOs, 1x World Champion Zest, 3x World Champion Rogue, 1x World Champion PartinG, 1x World Champion TY, Classic, TaeJa, Trap, Creator, soO as well as INnoVation. Serral is still is battling World Champions ByuN, and Dark as well as herO, Cure, Stats, Solar and Maru on a regular basis who all were relevant in 2015 already. On top, Serral was able to fend of new talents such as World Champion Reynor or Clem and MaxPax.
And how many of those were in their peak when they faced Serral? The versions of sOs, PartinG, TaeJa, Zest, ByuN, INnoVation and soO that Serral faced were nowhere near their peak. That's like taking Serrals negative record against Dayshi and Jaedong as an argument. Yeah, Serral is the most dominant player ever and probably has the best case to be the Goat due to no player really sticking out during the most competitive period but you can't derive from him beating players way past their peak that he would beat them at their peak too, that argument makes 0 sense. That argument of the supposed peak in skill of someone does not make sense. In a game/sport based only on 1 vs 1 confrontation you can just compare players using their results as data. There is no correlation between numbers of competitors and skill. More than that, I would say that continuously training made the players raw skill great with the time passing (trying to clarify: if we had a time machine and we took a middle of the pack player of today and, for example, MVP from 2011, the latter would be destroyed because of the way better knowledge of the have of the former). So the only data we should take into account when comparing players should be their results (probably weighted on the importance of tourneys from which that results came) not the year in which that results were achieved.
That is my point too.
Further, looking at careers, most players have gotten better over time in their strategy and understanding of their sport. Most hit their peaks in athleticism from mid to late 20s. Except for injuries or motivation, there is simply no reason to assume anything else that would drop skill.
Actually, as I am thinking about it right now... one could compare the win rates of the 2015-era-players among themselves in 2015 and 2018. If they are stable and only the win rates versus the new talent is dropping their match win rates, it is obvious, that the new guys simply were better than the old ones. But if a given players goes down versus both groups, it is him who lost skill either through demotivation or perhaps injury.
This might be a future project.
|
On July 22 2024 16:31 PremoBeats wrote:Show nested quote +On July 22 2024 15:56 MJG wrote:On July 22 2024 15:29 PremoBeats wrote:On July 22 2024 15:23 MJG wrote:
“If you torture the data long enough, it will confess to anything”. Not sure how serious your comment is, but in which of my 7 metrics that were analyzed did you see me "torture data"? Did you read the article? Did you understand that the data regarding the person which came out on top was penalized pr "tortured" the most? By far? That's exactly the point, you only analysed seven metrics! I'm sure I could find seven metrics that would give Has as the correct answer if I tried hard enough! I was being facetious and I thought it was pretty obvious...No amount of statistical analysis is likely to change my opinion that Legacy of the Void is a balance/design shitshow. It's the videogaming equivalent of getting 100m sprinters to wear clown shoes. Fun and entertaining? Sure. Provides useful data for measuring accomplishments? Definitely not. Why would you say that LotV is a balance shitshow? What exactly do you mean by it? There is extensive data about balancing out there.... but as you said: You seem to have pre-formed opinion that isn't likely to change.
Luckily for me, I've already posted about Legacy of the Void's design problems elsewhere, so I'll quote it here:
We had two versions of the game where a general overview of Starcraft 2 strategy would tell us that Zerg is the race that benefits most from being able to take faster expansions, from having larger maps, from having more open maps, and from the game being less "deathbally". Legacy of the Void was specifically designed so that we need to take faster expansions, so that the maps are generally larger, so that the maps are generally more open, and so that the game is significantly less "deathbally". Zerg has gone on to win more money than any other race in Legacy of the Void. This is going to read as balance whine, but I don't think it's a balance issue. No amount of tweaking unit stats is going to make a significant difference to core design flaws that result from Blizzard's decision to move to a 12 worker start and fewer minerals per base...
EDIT:
I'm definitely sceptical that statistics are capable of telling us who the GOAT is. They can definitely weed out people who shouldn't be in the conversation, but there are so many subjective factors that also deserve to be part of the conversation. Here are some examples:
- How much harder are prep-tournament than weekend-tournaments? How do you factor this into the statistics?
- How much harder are team-leagues when snipers can be prepared?
- How important are team-league statistics compared to individual-league statistics?
- How much more pressure does a player feel when playing in the GSL studio than when playing at a LAN?
- How much more important than everything else was GSL in the early days?
- How quickly did the importance of GSL wane with time?
- How important is sustained dominance in a period where barely any new talent is coming through? Does the fact that the NesTea award has been achieved many more times in Legacy of the Void demonstrate that the lack of new talent coming through has made sustained success easier?
- How did losing Korean team-houses post-Proleague affect talent development in Korea?
- How much has balance impacted on win-rates at various points in time? Or game design choices? Or map-pools?
It's admirable - and I said this about Mizenhauer's attempt as well - that you're trying to only use statistics. But it's ultimately a losing effort when there are so so so so so many subjective factors that should also be considered.
A purely statistical analysis can give Serral or Maru or whoever else as the answer all it wants, I couldn't care less because it isn't the whole picture.
|
France12790 Posts
On July 22 2024 17:08 MJG wrote:Show nested quote +On July 22 2024 16:31 PremoBeats wrote:On July 22 2024 15:56 MJG wrote:On July 22 2024 15:29 PremoBeats wrote:On July 22 2024 15:23 MJG wrote:
“If you torture the data long enough, it will confess to anything”. Not sure how serious your comment is, but in which of my 7 metrics that were analyzed did you see me "torture data"? Did you read the article? Did you understand that the data regarding the person which came out on top was penalized pr "tortured" the most? By far? That's exactly the point, you only analysed seven metrics! I'm sure I could find seven metrics that would give Has as the correct answer if I tried hard enough! I was being facetious and I thought it was pretty obvious...No amount of statistical analysis is likely to change my opinion that Legacy of the Void is a balance/design shitshow. It's the videogaming equivalent of getting 100m sprinters to wear clown shoes. Fun and entertaining? Sure. Provides useful data for measuring accomplishments? Definitely not. Why would you say that LotV is a balance shitshow? What exactly do you mean by it? There is extensive data about balancing out there.... but as you said: You seem to have pre-formed opinion that isn't likely to change. Luckily for me, I've already posted about Legacy of the Void's design problems elsewhere, so I'll quote it here: Show nested quote +We had two versions of the game where a general overview of Starcraft 2 strategy would tell us that Zerg is the race that benefits most from being able to take faster expansions, from having larger maps, from having more open maps, and from the game being less "deathbally". Legacy of the Void was specifically designed so that we need to take faster expansions, so that the maps are generally larger, so that the maps are generally more open, and so that the game is significantly less "deathbally". Zerg has gone on to win more money than any other race in Legacy of the Void. This is going to read as balance whine, but I don't think it's a balance issue. No amount of tweaking unit stats is going to make a significant difference to core design flaws that result from Blizzard's decision to move to a 12 worker start and fewer minerals per base...
Playing the game right now, it feels freaking "snowbally" in terms of macro, and I think it would be torture to play it without all the rapid fire tweaks and stuff. Having a good setup is becoming more important than before. But yeah playing versus zerg on larger maps feels like hell if you want to be "active", and if you like babysitting units type of games there aren't many maps where you can do that anymore
|
On July 22 2024 17:08 MJG wrote:Show nested quote +On July 22 2024 16:31 PremoBeats wrote:On July 22 2024 15:56 MJG wrote:On July 22 2024 15:29 PremoBeats wrote:On July 22 2024 15:23 MJG wrote:
“If you torture the data long enough, it will confess to anything”. Not sure how serious your comment is, but in which of my 7 metrics that were analyzed did you see me "torture data"? Did you read the article? Did you understand that the data regarding the person which came out on top was penalized pr "tortured" the most? By far? That's exactly the point, you only analysed seven metrics! I'm sure I could find seven metrics that would give Has as the correct answer if I tried hard enough! I was being facetious and I thought it was pretty obvious...No amount of statistical analysis is likely to change my opinion that Legacy of the Void is a balance/design shitshow. It's the videogaming equivalent of getting 100m sprinters to wear clown shoes. Fun and entertaining? Sure. Provides useful data for measuring accomplishments? Definitely not. Why would you say that LotV is a balance shitshow? What exactly do you mean by it? There is extensive data about balancing out there.... but as you said: You seem to have pre-formed opinion that isn't likely to change. Luckily for me, I've already posted about Legacy of the Void's design problems elsewhere, so I'll quote it here: Show nested quote +We had two versions of the game where a general overview of Starcraft 2 strategy would tell us that Zerg is the race that benefits most from being able to take faster expansions, from having larger maps, from having more open maps, and from the game being less "deathbally". Legacy of the Void was specifically designed so that we need to take faster expansions, so that the maps are generally larger, so that the maps are generally more open, and so that the game is significantly less "deathbally". Zerg has gone on to win more money than any other race in Legacy of the Void. This is going to read as balance whine, but I don't think it's a balance issue. No amount of tweaking unit stats is going to make a significant difference to core design flaws that result from Blizzard's decision to move to a 12 worker start and fewer minerals per base... EDIT: I'm definitely sceptical that statistics are capable of telling us who the GOAT is. They can definitely weed out people who shouldn't be in the conversation, but there are so many subjective factors that also deserve to be part of the conversation. Here are some examples: - How much harder are prep-tournament than weekend-tournaments? How do you factor this into the statistics?
- How much harder are team-leagues when snipers can be prepared?
- How important are team-league statistics compared to individual-league statistics?
- How much more pressure does a player feel when playing in the GSL studio than when playing at a LAN?
- How much more important than everything else was GSL in the early days?
- How quickly did the importance of GSL wane with time?
- How important is sustained dominance in a period where barely any new talent is coming through? Does the fact that the NesTea award has been achieved many more times in Legacy of the Void demonstrate that the lack of new talent coming through has made sustained success easier?
- How did losing Korean team-houses post-Proleague affect talent development in Korea?
- How much has balance impacted on win-rates at various points in time? Or game design choices? Or map-pools?
It's admirable - and I said this about Mizenhauer's attempt as well - that you're trying to only use statistics. But it's ultimately a losing effort when there are so so so so so many subjective factors that should also be considered. A statistical analysis can give Serral or Maru or whoever else as the answer all it wants, I couldn't care less.
So the data to suggest this fiasco you are portraying, is prize money earned, did I understand you correctly?
I might not even agree that prep-tournmanets are harder, but (and I know that doesn't help the core issue) they are simply a different quality of hard.
Anything else you raise are more or less valid points (except the notion about barely new talent coming in... it simply isn't coming as much from Korea as in the past, where GSL is played and the NesTea award is handed out) but each of them can be discussed on their impact on the metrics. You make it seem like all data simply is superfluous and arbitrary, to which I heavily disagree. In the end, these matter the most:
1. Are the race win rates in an area of 45-55% most of the time?
2. Are there measurable differences between players?
3. Can these differences be compared by adding context and factors?
4. Even if we disagree on factors: is the end result the same?
And in my opinion all 4 questions can be answered with yes.
That is not to say that your concerns are not valid.. I simply don't think that they are that important. In the end, the best and most dominant player will win the most tournaments given that #1 is true.
We could also argue that Scarlett was the best player in the world in 2014 when was in the top 10 because we know that she never had the support of team houses or the million dollar machine of Korean pros in the background... we just have to stop somewhere.
All things are influenced by contexts, but that does not mean, that they are less real or immeasurable.
|
On July 22 2024 17:58 PremoBeats wrote:Show nested quote +On July 22 2024 17:08 MJG wrote:On July 22 2024 16:31 PremoBeats wrote:On July 22 2024 15:56 MJG wrote:On July 22 2024 15:29 PremoBeats wrote:On July 22 2024 15:23 MJG wrote:
“If you torture the data long enough, it will confess to anything”. Not sure how serious your comment is, but in which of my 7 metrics that were analyzed did you see me "torture data"? Did you read the article? Did you understand that the data regarding the person which came out on top was penalized pr "tortured" the most? By far? That's exactly the point, you only analysed seven metrics! I'm sure I could find seven metrics that would give Has as the correct answer if I tried hard enough! I was being facetious and I thought it was pretty obvious...No amount of statistical analysis is likely to change my opinion that Legacy of the Void is a balance/design shitshow. It's the videogaming equivalent of getting 100m sprinters to wear clown shoes. Fun and entertaining? Sure. Provides useful data for measuring accomplishments? Definitely not. Why would you say that LotV is a balance shitshow? What exactly do you mean by it? There is extensive data about balancing out there.... but as you said: You seem to have pre-formed opinion that isn't likely to change. Luckily for me, I've already posted about Legacy of the Void's design problems elsewhere, so I'll quote it here: We had two versions of the game where a general overview of Starcraft 2 strategy would tell us that Zerg is the race that benefits most from being able to take faster expansions, from having larger maps, from having more open maps, and from the game being less "deathbally". Legacy of the Void was specifically designed so that we need to take faster expansions, so that the maps are generally larger, so that the maps are generally more open, and so that the game is significantly less "deathbally". Zerg has gone on to win more money than any other race in Legacy of the Void. This is going to read as balance whine, but I don't think it's a balance issue. No amount of tweaking unit stats is going to make a significant difference to core design flaws that result from Blizzard's decision to move to a 12 worker start and fewer minerals per base... EDIT: I'm definitely sceptical that statistics are capable of telling us who the GOAT is. They can definitely weed out people who shouldn't be in the conversation, but there are so many subjective factors that also deserve to be part of the conversation. Here are some examples: - How much harder are prep-tournament than weekend-tournaments? How do you factor this into the statistics?
- How much harder are team-leagues when snipers can be prepared?
- How important are team-league statistics compared to individual-league statistics?
- How much more pressure does a player feel when playing in the GSL studio than when playing at a LAN?
- How much more important than everything else was GSL in the early days?
- How quickly did the importance of GSL wane with time?
- How important is sustained dominance in a period where barely any new talent is coming through? Does the fact that the NesTea award has been achieved many more times in Legacy of the Void demonstrate that the lack of new talent coming through has made sustained success easier?
- How did losing Korean team-houses post-Proleague affect talent development in Korea?
- How much has balance impacted on win-rates at various points in time? Or game design choices? Or map-pools?
It's admirable - and I said this about Mizenhauer's attempt as well - that you're trying to only use statistics. But it's ultimately a losing effort when there are so so so so so many subjective factors that should also be considered. A statistical analysis can give Serral or Maru or whoever else as the answer all it wants, I couldn't care less. All things are influenced by contexts, but that does not mean, that they are less real or immeasurable.
You haven't tried to account for any of that context though. You've simply assumed that all things are equal across all eras and expansions, and then treated the numbers as absolute. It should be obvious to everyone here that all eras and expansions are not equal, and so treating the numbers as absolute is somewhat foolish.
Mvp is an obvious example where handwaving away all the subjective context is silly. Mvp was dominant in a time period were simply reaching GSL was a massive achievement, and new stars were frequently bursting onto the scene. In comparison, Maru's first GSL win was in a tournament that NoRegret qualified for. Not to dunk too hard on NoRegret, but those are not comparable data sets, and I'm sure there are lots of examples of other data sets that don't deserve to be taken into equal consideration. I'm also sure that every one of those examples is going to be incredibly subjective.
So torture the data all you like, you can't answer the GOAT question through data alone.
|
|
|
|
|
|
EPT
