2021/04/23: Article updated to include a breakdown of upsets by year
After the recent upset in the NeXT 2021 S1–SC2 Masters EU qualifiers, many people including myself were wondering if there had ever been a more unexpected results to any professional game of SCII. After all 2077 Aligulac rating points separated Hyperion and Reynor, and Aligulac had given Hyperion less than a quarter of a percent of a chance of winning the bo3.
The game in question
There was a pretty straightforward way of determining if this was the biggest upset ever, so I downloaded aligulac (from http://aligulac.com/about/db/. And as always thank you to all aligulac contributors and maintainers for keeping up to date such a great tool), did some querying, and here are the results. I do have to note that Aligulac ratings are fairly inflationary, so the unadjusted results will be biased towards more recent results. As such almost all of the following rating unadjusted upsets happened in LotV.
And it goes without saying that all these players getting upset are incredibly good--the only reason that it's such an upset/achievement to beat them is that they are so good.
The first thing I noticed is that a few of these 'biggest upsets' were due to progamers offracing, and should be excluded from contention. Kudos anyways to the underdogs, since these are some big upsets, but the progamers were operating under a pretty severe handicap.
I also didn't want to include Bo1s in the final list since Bo1s are quite volatile comparatively. Here are the biggest Bo1 upsets.
There are some interesting examples here. In the case of Ptak he played the greatest Starcraft of his life that day. He beat both GSL champions ByuN and Zest during the Nation Wars IV qualifier (beating Zest was "only" a 915 rating point upset) before barely losing to INnoVation.
In the case of Hellraiser and Spatz who beat Nerchio and Reynor respectively it can be argued that they were very underrated at the time, or that their rating was still busy catching up to their rapidly improving skill. Future similarly beat GuMiho when he was only rated 1139 for a 1322 point upset.
I also did some filtering to see if there was anything to be found from some subsets of these upsets. Here are the biggest Bo5+ upsets:
These have comparatively smaller upsets, since there are far fewer Bo5s than Bo3s. MCanning defeating Neeb was quite notable though.
And here are the offline Bo3 upsets:
MeomaikA beating Maru at WESG stands out as the most notable of these results.
And here are the greatest upsets of all SCII:
As it turns out Hyperion beating Reynor is the biggest upset of all and it isn't even close. It's the only 2000 point upset with no other result coming close. Even if we include the Bo1s and the progamers offracing there isn't a result that comes within 300 points of it.
Hyperion vs Reynor had the most surprising outcome ever.
[UPDATE] Upsets adjusted for rating and broken down by year
By request I did some extra investigation into what upsets happened in the early years of SCII, and tried to account for Aligulac's rating inflation. During this extra research I found that I'd previously missed a few upsets since I'd been sorting by the difference in rating rather than the difference in per-matchup rating for the relevant rating--this didn't change any of the top upsets, but it does mean that LambNRice defeating souL slips in in 9th place for example.
As it turns out that Aligulac ratings in 2010 and 2011 are not particularly useful. Ratings simply hadn't had the time to settle, which means the biggest upsets were... rather strange.
In 2010 all the top upsets are just good players who were drastically underrated beating MorroW who was the highest rated player in the world at the time. In 2011 all the top upsets are basically just Nerchio's worse go4sc2 defeat--back then he was rated very highly and played a ton of games against much lower rated opponents (unlike most of the other top rated players).
Things start looking much more reasonable in 2012:
Honorable mentions to Forsen beating PuMa (882 points) and Nerchio (792 points), Ziktomini beating Stephano (799 points), and especially TriMaster beating Marineking (886 points) at MLG which I remember watching live.
In 2013:
KingKong is a bit of a strange case where his Aligulac rating was rather inflated from playing in SEA. And the most high profile upset of 2013, Sjow vs Life had a difference of 825 rating and so misses out from the top 5 but not by a lot.
I don't have much to say about the remaining years of upsets as many of them were already featured before the update.
Cell really impresses by having the top upset in two consecutive years.
Spy's quite young, so maybe a player to look out for in the future?
The question of how to best adjust the ratings to account for rating inflation was probably the most subjective part of this entire exercise. I ended up deciding to weigh by the rating of the 10th best player from the previous aligulac period. The reason I chose the 10th best player is that using the best player alone added quite a bit of volatility--whereas the 10th best player is usually not too far from the 9th and 11th best and as such a better way to measure rating inflation. I considered also weighing by the difference between the 10th best and 10th worst player since after all ratings have deflated a bit on the far end of the bell curve. In the end I didn't do that, since that turned out to inflate old results excessively. I think that's because the starting rating for a new player stays the same, so we don't want to double-count inflation. Additionally I didn't include results from 2010 and 2011, because of the issues mentioned above--ratings were a bit all over the place back then, and inflation was really crazy which makes adjusting for it even harder.
Using this I obtained this list of the "Greatest Upset of All Time (Rating Adjusted)" I do think it still might lean a bit towards recent results, but that's hard to say.
"Hyperion vs Reynor" is still the biggest upset, but there's a decent inflation-related argument to be made for "Cell vs Maru" and "Nefaste vs Soulkey" too.
In general there are a lot of upsets from non-Korean Asian players, and I'd tend to attribute that to scenes like the Chinese scene being a bit more insular than say the NA scene--having less contact with other regions increases the chances of some players ending up quite underrated.
Wow amazing research, thanks ! The conclusion does not surprise me, a 2000 points gap is hard to beat (especially with rating inflation). Thanks again !
Look mom, im famous! Can confirm that my game against Hurricane was a build order win, but in my head im better than Hurricane and by extension better than everyone he ever beat. Replay here: https://drop.sc/replay/18818153
aligulac isn't a great rating system to refer. Just beating bunch of below level players everyday, you will be ranked 1-5 easily, and higher than Maru who doesn't play online much.
Oh nice thanks! Really goes to show how much Reynor droped the ball on that one.
Interesting that so much of the upset seems to come from China/Taiwan. I feel like some of the upset come partly from an underestimation of some region/player. Cell in particular stand out as someone who I would say is better than a 900 rating looking at his results.
On April 21 2021 19:32 swarminfestor wrote: Most Chinese players like Hyperion, Meomaika, TIME, Firefly did has some interesting builds to counter the higher ranked opponents.
Meomaika aint Chineae.
As per the thread. Some classic upset that i've watched was sjow vs life
I m not sure if it s possible, but here is my idear to make a Inflation adjustmeant. Dividenden the point difference by the highest (or average top 5/ top 10) points. So you could compare a 2000 points upset today with a 1400 upset like 8 years ago, when no one has above 2k rating.
On April 21 2021 19:32 swarminfestor wrote: Most Chinese players like Hyperion, Meomaika, TIME, Firefly did had some interesting builds to counter the higher ranked opponents.
On April 21 2021 19:53 dbRic1203 wrote: Absolutly great research.
I m not sure if it s possible, but here is my idear to make a Inflation adjustmeant. Dividenden the point difference by the highest (or average top 5/ top 10) points. So you could compare a 2000 points upset today with a 1400 upset like 8 years ago, when no one has above 2k rating.
Yes, I agree that the inevitable inflation in Aligulac ratings should be factored in. Also, I think that adding the rank of the players at the time at the upset would be helpful.
On April 21 2021 19:53 dbRic1203 wrote: Absolutly great research.
I m not sure if it s possible, but here is my idear to make a Inflation adjustmeant. Dividenden the point difference by the highest (or average top 5/ top 10) points. So you could compare a 2000 points upset today with a 1400 upset like 8 years ago, when no one has above 2k rating.
Hmm... Making some sort of inflation adjustment is certainly possible, it's not too straightforward though. Since the worst players nowadays also have lower ratings than they did back in the day--the spread of ratings outwards also depends on the number of games being played and players in the pool which isn't constant.
I'll try to come up with something (at the very least I could get the top upsets by year). Maybe I can get the count of the players in each period and see what the biggest percentile gap is? Not sure, I'll have to try later.
I find that topic very interesting, and maybe you could readily answer the following question with the (amazing) spreadsheet work already done : which mus are the most prone to see upsets happen ? I don't know the races of every player on those lists, so I'd like to know if you could come up with something about that question.
On April 22 2021 00:32 Obamarauder wrote: I played against meomaika a few times. he plays a really tricky / gimmicky style, not really that surprised he took games off maru
If you find those matches the 2nd map was Maru's fault though. he defended the trickery and then he outplayed himself First game was a direct loss.
On April 21 2021 19:32 swarminfestor wrote: Most Chinese players like Hyperion, Meomaika, TIME, Firefly did has some interesting builds to counter the higher ranked opponents.
Meomaika aint Chineae.
As per the thread. Some classic upset that i've watched was sjow vs life
It’s a long time ago, I might be remembering it wrongly, so I’m open to correction.
For me what made that such a staggering upset was the games were relatively standard and Sjow didn’t just cheese our Life with a couple of weird builds.
I think Hyperion's aligulac rating is inaccurate. Hyperion is 5700-5800 on KR, so his aligulac should be ~1800, based on other players with similar MMR. That means the MMR gap with Reynor is closer to ~1300 than 2000, which while still huge, does not make it the biggest upset in history.
On April 22 2021 10:21 warnull wrote: I think Hyperion's aligulac rating is inaccurate. Hyperion is 5700-5800 on KR, so his aligulac should be ~1800, based on other players with similar MMR. That means the MMR gap with Reynor is closer to ~1300 than 2000, which while still huge, does it make it the biggest upset in history.
I mean Aligulac can only go off tournament results, to some degree. Alas not all tournaments are equal and there’s pretty big differences in regions and how often the best from each play.
I mean they don’t have a huge amount of games logged but it’s not nothing either. And they haven’t done amazingly well. Who are the similar players you’re cross-referencing with their MMR here?
5800 isn’t anything to scoff at on Kr by any means, but I mean you can’t extrapolate that into tournament play, it’s a different environment. It says they’re comparable to other players in that range playing a lot of games vs a rotating wheel of opponents, but playing series against known opponents is a different kettle of fish.
Intuitively though, as much as I did love the OP’s research I guess I think the ‘biggest’ upset has to have real stakes on it too. Cool thread though and some fun old memories appeared too!
On April 22 2021 10:21 warnull wrote: I think Hyperion's aligulac rating is inaccurate. Hyperion is 5700-5800 on KR, so his aligulac should be ~1800, based on other players with similar MMR. That means the MMR gap with Reynor is closer to ~1300 than 2000, which while still huge, does it make it the biggest upset in history.
I mean Aligulac can only go off tournament results, to some degree. Alas not all tournaments are equal and there’s pretty big differences in regions and how often the best from each play.
I mean they don’t have a huge amount of games logged but it’s not nothing either. And they haven’t done amazingly well. Who are the similar players you’re cross-referencing with their MMR here?
5800 isn’t anything to scoff at on Kr by any means, but I mean you can’t extrapolate that into tournament play, it’s a different environment. It says they’re comparable to other players in that range playing a lot of games vs a rotating wheel of opponents, but playing series against known opponents is a different kettle of fish.
Intuitively though, as much as I did love the OP’s research I guess I think the ‘biggest’ upset has to have real stakes on it too. Cool thread though and some fun old memories appeared too!
Hmm I think MMR is a much more accurate representation of the skill-level than aligulac rating though as aligulac rating only looks at a small subset of games vs a specific set of players (especially for lower ranked players who probably aren't as active and if they are possibly get stomped a lot by superior players). If his MMR is close to other players with ~1800 aligulac rating than I'm sure that's a much more accurate rating for his skill level.
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
On April 22 2021 20:39 dbRic1203 wrote: MaxPax played Terran vs Trifax. (According to him) The game was appearently casted by Rotterdam, so should go to the offrace section
Oh yeah I see the VOD. I'll update it. Aligulac (and Liquipedia) need updating too.
On April 21 2021 19:32 swarminfestor wrote: Most Chinese players like Hyperion, Meomaika, TIME, Firefly did has some interesting builds to counter the higher ranked opponents.
Meomaika aint Chineae.
As per the thread. Some classic upset that i've watched was sjow vs life
It’s a long time ago, I might be remembering it wrongly, so I’m open to correction.
For me what made that such a staggering upset was the games were relatively standard and Sjow didn’t just cheese our Life with a couple of weird builds.
Also at a LAN event too!
I watched the series for the first time after the VOD was linked in this thread. It‘s a great match. The audience is really into it when game 3 turns into a drawn-out macro battle in which sjow is in a winning position twice, but unable to close out the match, then at the brink of defeat, only to pull off an extremely close win.
It‘s unfortunate this is a Life match. We can‘t be sure this wasn‘t matchfixing.
On April 21 2021 18:29 NimzoBoe wrote: Look mom, im famous! Can confirm that my game against Hurricane was a build order win, but in my head im better than Hurricane and by extension better than everyone he ever beat. Replay here: https://drop.sc/replay/18818153
Dude! You're selling yourself short. This may have been a brief game, but still a finely executed one on your part. Thanks for sharing!
P.S. Hurricane once beat Serral in a Best of 3, so by extension, you know...
What a heck of a compilation! I also didn't know Hyperion is East Asian (Chinese). Nice. Seriously signature win for him as he is now embedded into SC2 history.
On April 21 2021 18:29 NimzoBoe wrote: Look mom, im famous! Can confirm that my game against Hurricane was a build order win, but in my head im better than Hurricane and by extension better than everyone he ever beat. Replay here: https://drop.sc/replay/18818153
Dude! You're selling yourself short. This may have been a brief game, but still a finely executed one on your part. Thanks for sharing!
P.S. Hurricane once beat Serral in a Best of 3, so by extension, you know...
And Serral beat Reynor a bunch of time, when are we getting Nimzo vs Hyperion for the world championship?
On April 22 2021 10:21 warnull wrote: I think Hyperion's aligulac rating is inaccurate. Hyperion is 5700-5800 on KR, so his aligulac should be ~1800, based on other players with similar MMR. That means the MMR gap with Reynor is closer to ~1300 than 2000, which while still huge, does it make it the biggest upset in history.
I mean Aligulac can only go off tournament results, to some degree. Alas not all tournaments are equal and there’s pretty big differences in regions and how often the best from each play.
I mean they don’t have a huge amount of games logged but it’s not nothing either. And they haven’t done amazingly well. Who are the similar players you’re cross-referencing with their MMR here?
5800 isn’t anything to scoff at on Kr by any means, but I mean you can’t extrapolate that into tournament play, it’s a different environment. It says they’re comparable to other players in that range playing a lot of games vs a rotating wheel of opponents, but playing series against known opponents is a different kettle of fish.
Intuitively though, as much as I did love the OP’s research I guess I think the ‘biggest’ upset has to have real stakes on it too. Cool thread though and some fun old memories appeared too!
Hmm I think MMR is a much more accurate representation of the skill-level than aligulac rating though as aligulac rating only looks at a small subset of games vs a specific set of players (especially for lower ranked players who probably aren't as active and if they are possibly get stomped a lot by superior players). If his MMR is close to other players with ~1800 aligulac rating than I'm sure that's a much more accurate rating for his skill level.
Tournaments don’t measure pure Starcraft skill, although it’s certainly a big part of it. Big test of nerves, prep may be a factor etc.
Yeah that’s a reasonable estimation of level, but we do hear about players who are ladder monsters and just can’t get over the line in tournament play.
I wonder what moments of insane skill level have been played on ladder over the years that we’ll never get to see.
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was ranked #3 on Aligulac while Sjow was around #250; Dark is currently ranked #9, uThermal is #28.
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was ranked #3 on Aligulac while Sjow was around #250; Dark is currently ranked #9, uThermal is #28.
Yeah, I mean, at the time the Korean scene was very far ahead of the foreign scene, and had much more depth that it does today. Any foreigner beating a GSL player was a huge upset. Elephant in the room and all that.
On April 23 2021 14:29 ZigguratOfUr wrote: Updated article with the top upsets for each year, and a "Rating Adjusted" list of biggest upsets in SCII history.
Interesting, man Hydra had a hell of a time in getting upset
On April 23 2021 14:29 ZigguratOfUr wrote: Updated article with the top upsets for each year, and a "Rating Adjusted" list of biggest upsets in SCII history.
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was ranked #3 on Aligulac while Sjow was around #250; Dark is currently ranked #9, uThermal is #28.
Sjow was like a top 7 foreign terran at the time? and it was a meta where terran generally did better than zergs. Don't think the #250 is accurate. Maybe the rankings were still slow too adapt from end of HOTS period.
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was ranked #3 on Aligulac while Sjow was around #250; Dark is currently ranked #9, uThermal is #28.
Sjow was like a top 7 foreign terran at the time? and it was a meta where terran generally did better than zergs. Don't think the #250 is accurate. Maybe the rankings were still slow too adapt from end of HOTS period.
Was this when the Kespa era was getting properly going or just before that talent base came through?
Either way even with just the esf boys there were absolutely tons of Koreans better than all but the most elite foreigners at that time. Plus I think the gaps within the foreign scene between the top tiers and going down have shrunk a bit too.
Plus Europe has always been choc a bloc with loads of solid if unspectacular Zerg players all the way up to the region’s best and Protoss tended to put in more consistent results too. Top 7 Terran doesn’t necessarily equal top 7 Zerg or Protoss.
Meta does factor in for sure, although favourable Terran metas outside of arguably broken ones tended to not make a massive difference in the foreign scene as compared to what the Koreans could eke out of the race. Favourable Zerg metas and styles seemed to be something that foreigners could take advantage of more readily.
Why the current era is quite exciting. Even in foreign land Terrans tended to underperform vs their counterparts, but now that gap isn’t as big a factor and they’ve begun to close the gap to top Korean Terrans too.
250 may be a tad low but wasn’t Sjow not particularly active/bouncing between part time and full time play around here? Could contribute to him being underrated by ranking systems perhaps.
Could very conceivably be outside of a top 100 though. Which isn’t to dunk on Sjowtime at all
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was ranked #3 on Aligulac while Sjow was around #250; Dark is currently ranked #9, uThermal is #28.
Sjow was like a top 7 foreign terran at the time? and it was a meta where terran generally did better than zergs. Don't think the #250 is accurate. Maybe the rankings were still slow too adapt from end of HOTS period.
also worth noting that the player pool back then was way bigger. Being #250 now would be impossible for a pro player and a player of the caliber of uthermal would be much lower ranked in 2013. Still think Sjow beating Life was a bigger upset but not by that big of a margin as the ranking difference would suggest
If soul beat Dark today I think it would be comparable
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was ranked #3 on Aligulac while Sjow was around #250; Dark is currently ranked #9, uThermal is #28.
Sjow was like a top 7 foreign terran at the time? and it was a meta where terran generally did better than zergs. Don't think the #250 is accurate. Maybe the rankings were still slow too adapt from end of HOTS period.
also worth noting that the player pool back then was way bigger. Being #250 now would be impossible for a pro player and a player of the caliber of uthermal would be much lower ranked in 2013. Still think Sjow beating Life was a bigger upset but not by that big of a margin as the ranking difference would suggest
If soul beat Dark today I think it would be comparable
The pool of pro players was bigger in 2013, it's true; there were many more koreans around, KeSpa had opened the gate s already. However, a player as skilled as uThermal is now would have most likely ranked not much lower than top 50. Sjow was effectively kind of an outlier at the time of the upset, since he stopped playing in 2012(he was ranked #211 in October) and came back in 2013 breaking the top 100 right after; he didn't go much further, in any of case.
Your example is still unfit, souL is approximately as strong as uThermal and a top 3 foreign Terran outside of Clem; even considering Dark's level then same as Life at the time of the upset, you would need BattleB or Aqueron to beat Dark in one official tournament to come close.
Life was probably hungover or something, but still :D.
Sjow was still a way more established player than Hyperion or Meomaika and the hellbats were freaking op while Life was entering a phase of relative slump.
Rewatching this game, Sjows macro is straight to master3/diamond 1. His micro could still perhaps GM today although I think there are so many more nuances to micro in todays meta than what was displayed in that series..
This patch wasnt Hellbats being OP though. Hellbats being op was when you could have 4 of them in a medivac. This was more the meta where zergs just hadnt learned to play properly against MMMM yet in terms of when to drone (and not to build roaches). A couple of months later the matchup stabilized (and then mines got nerfed and it got zerg favored, then mines got unnerfed 6 months later and the MU was fine for rest of HOTS).
I think Sjow beating Life in terms of upsets potential would be like if uThermal beat Dark today.
Life was ranked #3 on Aligulac while Sjow was around #250; Dark is currently ranked #9, uThermal is #28.
Sjow was like a top 7 foreign terran at the time? and it was a meta where terran generally did better than zergs. Don't think the #250 is accurate. Maybe the rankings were still slow too adapt from end of HOTS period.
also worth noting that the player pool back then was way bigger. Being #250 now would be impossible for a pro player and a player of the caliber of uthermal would be much lower ranked in 2013. Still think Sjow beating Life was a bigger upset but not by that big of a margin as the ranking difference would suggest
If soul beat Dark today I think it would be comparable
The pool of pro players was bigger in 2013, it's true; there were many more koreans around, KeSpa had opened the gate s already. However, a player as skilled as uThermal is now would have most likely ranked not much lower than top 50. Sjow was effectively kind of an outlier at the time of the upset, since he stopped playing in 2012(he was ranked #211 in October) and came back in 2013 breaking the top 100 right after; he didn't go much further, in any of case.
Your example is still unfit, souL is approximately as strong as uThermal and a top 3 foreign Terran outside of Clem; even considering Dark's level then same as Life at the time of the upset, you would need BattleB or Aqueron to beat Dark in one official tournament to come close.
I think there is a pretty large gap down from uthermal/soul to 5th best terran (excl special i guess). But I guess if you got a 6.5K MMR terran on EU server to beat Dark in a bo3 that might be a better example.
Did someone broke down the races played by the players who were able to pull off the upsets. At first glance it looks like it's primarily zerg players.