|
I have completed the August update of the E-Rankings StarCraft 2 World Ranking. This month’s changes are unusually large because the results from the Esports World Cup 2025 have now dropped out of the 12-month calculation.
herO remains world #1 with 11,280 ranking points. He is the only player above 10,000 points and now leads his nearest challenger by 2,924.
The new top ten:
1.  herO - 11,280 points (-)
2.  Clem - 8,356 (+2)
3. Classic - 6,367 (-1)
4.  Reynor - 5,312 (+1)
5. Serral - 5,040 (-2)
6. MaxPax - 4,980 (+3)
7. Solar - 4,811 (-1)
8. Maru - 4,116 (-)
9. trigger - 3,665 (+1)
10. ByuN - 3,583 (+2)
Clem is probably the most interesting story of the update. Besides climbing from fourth to second, he has passed Dark for fourth place in the all-time standings:
1. Serral - 135,386 career points
2. Maru - 108,202
3. Solar - 94,386
4. Clem - 94,374
5. Dark - 93,942
Clem is now only 12 points behind Solar.
Serral loses more points than any other top player because his 2025 EWC victory has expired. However, he reached his 107th tracked tournament final in July, moving past Zest into fourth place for total final appearances. He has won 74 of those finals and needs one more title to draw level with Solar for second place in tournament wins.
Further down the ranking, two young Russian players make the largest gains. Shameless climbs from #24 to a career-high #17, while 17-year-old GgMaChine rises from #32 to #25. GgMaChine is the fifth-youngest player ever in our database to reach the top 25.
The complete update, including the biggest monthly gainers and all players reaching new career highs, is here: https://blog.e-rankings.com/starcraft-2-world-ranking-clem-climbs-to-2/
Hit me with your takes 
|
Northern Ireland27466 Posts
What’s your methodology? Looks like you’ve put plenty of work in!
|
|
|
Northern Ireland27466 Posts
Some elements of the tennis system there, but I guess one flaw there is that in SC tournaments come and go, less so the former. Serral’s dropping a ton of points on the 12 month cycle for a tournament that doesn’t exist this cycle. As opposed to say last year’s Wimbledon Champion dropping points for going out the first round this year
I think it’s a solid effort though for sure! Frankly the current scene is so fragmented and choppy it’s hard to really make any kind of ranking system that reflects reality
|
On August 02 2026 23:16 WombaT wrote:
Some elements of the tennis system there, but I guess one flaw there is that in SC tournaments come and go, less so the former. Serral’s dropping a ton of points on the 12 month cycle for a tournament that doesn’t exist this cycle. As opposed to say last year’s Wimbledon Champion dropping points for going out the first round this year
I think it’s a solid effort though for sure! Frankly the current scene is so fragmented and choppy it’s hard to really make any kind of ranking system that reflects reality
One of the reasons why I didn't work on an update...
The sample size for Premier Tournaments is too thin and changing the methodology post 2025 could be fine. But as people already disagree over much less subjective things - to me - it seemed too much of a hassle.
For example, I think broadening the sample size to the top Tier 2 Liquipedia tournaments post 2025 would widen the sample size in a sensible way. Tournaments like RSL, PiG Sty and LiuLi were massively stacked, despite "only" handing out 10k or less.
So that could be a new threshold, as super big events are ever rarer.
|
Agreed. I think we all wish Starcraft had a tournament cycle like tennis...
|
First of all, props to you for carrying out this project. Good job in setting up an interesting evaluation system and applying it consistently.
That said, I think your system heavily overrates participation in tournaments, especially in small tournaments, and it heavily underrates strong placements or wins in large tournaments. This is most evident in the results for Reynor (#4 in your current ranking) and Serral (only #5 in your ranking).
There is absolutely no way that Reynor's past 12 months can be said to be more successful than Serral's. I am certain Reynor would agree with this without a second thought. Serral has literally won multiple major tournaments in the past few months alone, whereas Reynor has not won a single one. But still, Reynor ends up ahead of Serral in your ranking because he has accumulated a larger number of middling-to-somewhat-good results in medium-sized tournaments like Wardi Cups, RSL, and so on, none of which Serral has participated in.
Of course, it all depends on what you try to measure in the first place. Your ranking sort of reflects the activity of a player in terms of tournament participation, weighted by the relative success in and size of those tournaments. But it does not so well measure the overall strength of a player, because Reynor and Classic are most definitely not ahead of Serral in this regard, and I'm pretty sure every pro would agree with this.
Just my thoughts, cool project all the same.
|
On August 03 2026 05:16 Antithesis wrote:
First of all, props to you for carrying out this project. Good job in setting up an interesting evaluation system and applying it consistently.
That said, I think your system heavily overrates participation in tournaments, especially in small tournaments, and it heavily underrates strong placements or wins in large tournaments. This is most evident in the results for Reynor (#4 in your current ranking) and Serral (only #5 in your ranking).
There is absolutely no way that Reynor's past 12 months can be said to be more successful than Serral's. I am certain Reynor would agree with this without a second thought. Serral has literally won multiple major tournaments in the past few months alone, whereas Reynor has not won a single one. But still, Reynor ends up ahead of Serral in your ranking because he has accumulated a larger number of middling-to-somewhat-good results in medium-sized tournaments like Wardi Cups, RSL, and so on, none of which Serral has participated in.
Of course, it all depends on what you try to measure in the first place. Your ranking sort of reflects the activity of a player in terms of tournament participation, weighted by the relative success in and size of those tournaments. But it does not so well measure the overall strength of a player, because Reynor and Classic are most definitely not ahead of Serral in this regard, and I'm pretty sure every pro would agree with this.
Just my thoughts, cool project all the same.
From the explanation-site:
Most important: The E-Rankings World Rankings are results-based rankings, not power rankings. They do not try to predict who would beat whom tomorrow. Instead, they measure tournament success over a rolling 12-month period, weighted by prize pool and final placement.
And considering that Serral is very picky about the tournament he plays, it isn't surprising that he isn't ranked higher. But honestly, the reason why I think this is a solid ranking is because CheeseInSpace actually did a similiar ranking for Warcraft 3 for years back in the day. The guy has experience
|
On August 03 2026 05:56 Balnazza wrote:Show nested quote +On August 03 2026 05:16 Antithesis wrote:
First of all, props to you for carrying out this project. Good job in setting up an interesting evaluation system and applying it consistently.
That said, I think your system heavily overrates participation in tournaments, especially in small tournaments, and it heavily underrates strong placements or wins in large tournaments. This is most evident in the results for Reynor (#4 in your current ranking) and Serral (only #5 in your ranking).
There is absolutely no way that Reynor's past 12 months can be said to be more successful than Serral's. I am certain Reynor would agree with this without a second thought. Serral has literally won multiple major tournaments in the past few months alone, whereas Reynor has not won a single one. But still, Reynor ends up ahead of Serral in your ranking because he has accumulated a larger number of middling-to-somewhat-good results in medium-sized tournaments like Wardi Cups, RSL, and so on, none of which Serral has participated in.
Of course, it all depends on what you try to measure in the first place. Your ranking sort of reflects the activity of a player in terms of tournament participation, weighted by the relative success in and size of those tournaments. But it does not so well measure the overall strength of a player, because Reynor and Classic are most definitely not ahead of Serral in this regard, and I'm pretty sure every pro would agree with this.
Just my thoughts, cool project all the same. From the explanation-site: Show nested quote +Most important: The E-Rankings World Rankings are results-based rankings, not power rankings. They do not try to predict who would beat whom tomorrow. Instead, they measure tournament success over a rolling 12-month period, weighted by prize pool and final placement. And considering that Serral is very picky about the tournament he plays, it isn't surprising that he isn't ranked higher. But honestly, the reason why I think this is a solid ranking is because CheeseInSpace actually did a similiar ranking for Warcraft 3 for years back in the day. The guy has experience
Yeah as I said, it's a cool project, and if it's not intended as a power ranking, that's fair enough.
But still, I think even the goal statement quoted is a bit inaccurate:
Most important: The E-Rankings World Rankings are results-based rankings, not power rankings. They do not try to predict who would beat whom tomorrow. Instead, they measure tournament success over a rolling 12-month period, weighted by prize pool and final placement.
But what the system measures is not primarily tournament sucess, weighted by prize pool and final placement (of which the latter is redundant here anyway because placement is success). It's rather tournament activity, weighted by prize pool and final placement.
And it's perfectly fine and even interesting to measure tournament activity in this way, but it's not the same as tournament success. It's misleading to say that Reynor and Classic had more success in tournaments than Serral in the past 12 months. They did not. They've just participated in more smaller tournaments and thus accumulated more minor achievements.
|
I get what you mean but i just see someone like Serral as a statistical outlier in the system. He obviously is a hugely successful player - as you can see he also leads the all-time ranking by a mile - but he is also playing much less tournaments than other top players. For the current ranking for example Serral has played 17 tournaments. In the top ten only Maru has less with 14. Everyone else has at least 28 tournaments played, so double that of Maru and 60 percent more than Serral. The amazing thing for me is thus that both players even with that little amount of tournaments still rank that high which shows how good they are. Byun for example played 50 tournaments, triple that of Serral and is still five ranks behind him. So it's not just about activity.
But even if you change the system, there will be other outliers who don't fit in perfectly and who will be ranked too high or too low compared to the general sentiment. Also the definition of "success" is subjective. One might very well argue that winning two big tournaments in a year is a bigger success than consistently playing in smaller tournaments every other week and winning 5 - 10 of them in a year. I think both views are valid and i try to balance them. But of course you can argue about how well this works. For me, it's fine.
|
On August 03 2026 11:56 CheeseInSpace wrote:
I get what you mean but i just see someone like Serral as a statistical outlier in the system. He obviously is a hugely successful player - as you can see he also leads the all-time ranking by a mile - but he is also playing much less tournaments than other top players. For the current ranking for example Serral has played 17 tournaments. In the top ten only Maru has less with 14. Everyone else has at least 28 tournaments played, so double that of Maru and 60 percent more than Serral. The amazing thing for me is thus that both players even with that little amount of tournaments still rank that high which shows how good they are. Byun for example played 50 tournaments, triple that of Serral and is still five ranks behind him. So it's not just about activity.
But even if you change the system, there will be other outliers who don't fit in perfectly and who will be ranked too high or too low compared to the general sentiment. Also the definition of "success" is subjective. One might very well argue that winning two big tournaments in a year is a bigger success than consistently playing in smaller tournaments every other week and winning 5 - 10 of them in a year. I think both views are valid and i try to balance them. But of course you can argue about how well this works. For me, it's fine.
I think big picture just need to decide what your list is. It can be a money ranking, or a skill ranking.
As of now you're trying to make an objective case on flawed data to represent arbitrary success, as you already noted.
What you have now is fine you just need to be rebranded for what it is.
If you're going for engagement a better way to go about it is make an actual ranking you believe in, and you can reinforce that with your "statistics".
Skill wise anyone that plays or follows the game knows that the current 4 best players as of now are Serral, Clem, Rogue and herO. To have any list that says otherwise without clarifying is just ragebait
Maru and Reynor are the only real unknowns because they don't play, but assuming they still touch the game in some capacity regularly my top 10 would be as follows:
+ Show Spoiler +
1. Serral
2. Clem
3. herO
4. Rogue
5. Maru
6. Reynor
7. Maxpax
8. Solar
9. Shin
10. Cure/Byun/Classic
If this entire group played multiple round robin bo5's I feel pretty confident in saying this is how they would end up.
If you want to continue as is I think the metrics you have established are fine but do not use a year's worth of data for your 'monthly' updates. I'd just do a few or 6 at the most for the month's past timeframe.
|
On August 03 2026 13:51 Agh wrote:Show nested quote +On August 03 2026 11:56 CheeseInSpace wrote:
I get what you mean but i just see someone like Serral as a statistical outlier in the system. He obviously is a hugely successful player - as you can see he also leads the all-time ranking by a mile - but he is also playing much less tournaments than other top players. For the current ranking for example Serral has played 17 tournaments. In the top ten only Maru has less with 14. Everyone else has at least 28 tournaments played, so double that of Maru and 60 percent more than Serral. The amazing thing for me is thus that both players even with that little amount of tournaments still rank that high which shows how good they are. Byun for example played 50 tournaments, triple that of Serral and is still five ranks behind him. So it's not just about activity.
But even if you change the system, there will be other outliers who don't fit in perfectly and who will be ranked too high or too low compared to the general sentiment. Also the definition of "success" is subjective. One might very well argue that winning two big tournaments in a year is a bigger success than consistently playing in smaller tournaments every other week and winning 5 - 10 of them in a year. I think both views are valid and i try to balance them. But of course you can argue about how well this works. For me, it's fine. I think big picture just need to decide what your list is. It can be a money ranking, or a skill ranking. As of now you're trying to make an objective case on flawed data to represent arbitrary success, as you already noted. What you have now is fine you just need to be rebranded for what it is. If you're going for engagement a better way to go about it is make an actual ranking you believe in, and you can reinforce that with your "statistics". Skill wise anyone that plays or follows the game knows that the current 4 best players as of now are Serral, Clem, Rogue and herO. To have any list that says otherwise without clarifying is just ragebait Maru and Reynor are the only real unknowns because they don't play, but assuming they still touch the game in some capacity regularly my top 10 would be as follows: + Show Spoiler +
1. Serral
2. Clem
3. herO
4. Rogue
5. Maru
6. Reynor
7. Maxpax
8. Solar
9. Shin
10. Cure/Byun/Classic
If this entire group played multiple round robin bo5's I feel pretty confident in saying this is how they would end up.
If you want to continue as is I think the metrics you have established are fine but do not use a year's worth of data for your 'monthly' updates. I'd just do a few or 6 at the most for the month's past timeframe.
Love how you talk about "flawed data" and then talk about "the best players", even though this list doesn't try to determine "the best player". The goals of the ranking are clearly outlined and have nothing to do with skill.
Also...Maru and Reynor do not play? Ehhh, what? Both of them are invited to the next PiG Festival, Maru just played RSL, Reynor was at HSC, both of them were in Maestros.
|
Northern Ireland27466 Posts
On August 03 2026 16:01 Balnazza wrote:Show nested quote +On August 03 2026 13:51 Agh wrote:On August 03 2026 11:56 CheeseInSpace wrote:
I get what you mean but i just see someone like Serral as a statistical outlier in the system. He obviously is a hugely successful player - as you can see he also leads the all-time ranking by a mile - but he is also playing much less tournaments than other top players. For the current ranking for example Serral has played 17 tournaments. In the top ten only Maru has less with 14. Everyone else has at least 28 tournaments played, so double that of Maru and 60 percent more than Serral. The amazing thing for me is thus that both players even with that little amount of tournaments still rank that high which shows how good they are. Byun for example played 50 tournaments, triple that of Serral and is still five ranks behind him. So it's not just about activity.
But even if you change the system, there will be other outliers who don't fit in perfectly and who will be ranked too high or too low compared to the general sentiment. Also the definition of "success" is subjective. One might very well argue that winning two big tournaments in a year is a bigger success than consistently playing in smaller tournaments every other week and winning 5 - 10 of them in a year. I think both views are valid and i try to balance them. But of course you can argue about how well this works. For me, it's fine. I think big picture just need to decide what your list is. It can be a money ranking, or a skill ranking. As of now you're trying to make an objective case on flawed data to represent arbitrary success, as you already noted. What you have now is fine you just need to be rebranded for what it is. If you're going for engagement a better way to go about it is make an actual ranking you believe in, and you can reinforce that with your "statistics". Skill wise anyone that plays or follows the game knows that the current 4 best players as of now are Serral, Clem, Rogue and herO. To have any list that says otherwise without clarifying is just ragebait Maru and Reynor are the only real unknowns because they don't play, but assuming they still touch the game in some capacity regularly my top 10 would be as follows: + Show Spoiler +
1. Serral
2. Clem
3. herO
4. Rogue
5. Maru
6. Reynor
7. Maxpax
8. Solar
9. Shin
10. Cure/Byun/Classic
If this entire group played multiple round robin bo5's I feel pretty confident in saying this is how they would end up.
If you want to continue as is I think the metrics you have established are fine but do not use a year's worth of data for your 'monthly' updates. I'd just do a few or 6 at the most for the month's past timeframe. Love how you talk about "flawed data" and then talk about "the best players", even though this list doesn't try to determine "the best player". The goals of the ranking are clearly outlined and have nothing to do with skill. Also...Maru and Reynor do not play? Ehhh, what? Both of them are invited to the next PiG Festival, Maru just played RSL, Reynor was at HSC, both of them were in Maestros.
Yeah, I know of quite a few sports that do rankings over a 12 or 24 month cycle, tennis, snooker and golf IIRC with a vaguely similar methodology
I think it’s valid, but it’s hard to translate to SC2 because it doesn’t really have a consistent seasonal structure, and indeed rarely ever has
|
On August 03 2026 16:01 Balnazza wrote:
Love how you talk about "flawed data" and then talk about "the best players", even though this list doesn't try to determine "the best player". The goals of the ranking are clearly outlined and have nothing to do with skill.
I think everything that has seen the thread knows the information it is conveying. It's more the what and why.
The 'All-time' portion functions well, but the rest leaves a lot to be desired.
If Mr. Beast hosts a random local tournament for a million dollars, we get to see that winner on the list for a year.
If there is a Qualifier for no prize pool itself, but the event is only 4 spots with a similarly massive prize pool only those four places get the astronomical injection.
Also...Maru and Reynor do not play? Ehhh, what? Both of them are invited to the next PiG Festival, Maru just played RSL, Reynor was at HSC, both of them were in Maestros.
In comparison to everyone else on the list that week get to see on a weekly basis, yes. (Hyperbole)
On August 03 2026 22:45 WombaT wrote:
Yeah, I know of quite a few sports that do rankings over a 12 or 24 month cycle, tennis, snooker and golf IIRC with a vaguely similar methodology
I think it’s valid, but it’s hard to translate to SC2 because it doesn’t really have a consistent seasonal structure, and indeed rarely ever has
Even golf uses a field strength metric in their rankings, which is amusing.
Most other head to head rankings also only consider the X best results within that year. And for the ones that stretch longer than that typically points either go up or down relative to their performance in that same tournament the previous year.
|
But it isn't a "head to head"-ranking. it specifically states that it does NOT predict who would win a match tomorrow.
You can of course argue against the Methodology. It is definetly debatable if 12 months is the appropriate length for SC2, considering that most SC2 tournaments do not happen periodically, like in Tennis for example (Tennis also has the additional bonus that all top players are forced to play all top tournaments (except for injury), so there is rarely a "skipped an event"-problem).
But attacking the Methodology because it doesn't lead to a result the ranking doesn't even want to achieve just isn't helpful. That's like saying "your recipe is trash, this burger sucks" when the recipe is for a Pizza...
|
On August 04 2026 04:01 Balnazza wrote:
But it isn't a "head to head"-ranking. it specifically states that it does NOT predict who would win a match tomorrow.
We're literally talking about the systems 1vs1 competitions (aka head to head) have that use similar rankings.
You can of course argue against the Methodology. It is definetly debatable if 12 months is the appropriate length for SC2, considering that most SC2 tournaments do not happen periodically, like in Tennis for example (Tennis also has the additional bonus that all top players are forced to play all top tournaments (except for injury), so there is rarely a "skipped an event"-problem).
But attacking the Methodology because it doesn't lead to a result the ranking doesn't even want to achieve just isn't helpful. That's like saying "your recipe is trash, this burger sucks" when the recipe is for a Pizza...
Not sure if you're genuinely lost or just trolling at this point.
Success is arbitrary by definition. For instance you might think you're "successful" in whatever it is you're trying to convey. What one person considers success, another very well might deem a failure.
Which brings us back to the the purpose of the ranking in the first place, which is -- to reflect "success".
The fact that success is arbitrary is irrelevant, the issue is that the "methodology" is based on prize pool -- which is... surprise! an arbitrary value. If you need examples of why this is an imperfect system then re-read the post above.
https://tl.net/forum/starcraft-2/645694-most-successful-sc2-players-of-q2-2026
You can read through the comments of the last thread and see that everyone is quite aware of the intention.
My final suggestion is to either keep the year and use a specific number of best results, or limit it to recent months only.
|
On August 04 2026 05:44 Agh wrote:Show nested quote +On August 04 2026 04:01 Balnazza wrote:
But it isn't a "head to head"-ranking. it specifically states that it does NOT predict who would win a match tomorrow. We're literally talking about the systems 1vs1 competitions (aka head to head) have that use similar rankings. Show nested quote +
You can of course argue against the Methodology. It is definetly debatable if 12 months is the appropriate length for SC2, considering that most SC2 tournaments do not happen periodically, like in Tennis for example (Tennis also has the additional bonus that all top players are forced to play all top tournaments (except for injury), so there is rarely a "skipped an event"-problem).
But attacking the Methodology because it doesn't lead to a result the ranking doesn't even want to achieve just isn't helpful. That's like saying "your recipe is trash, this burger sucks" when the recipe is for a Pizza...
Not sure if you're genuinely lost or just trolling at this point. Success is arbitrary by definition. For instance you might think you're "successful" in whatever it is you're trying to convey. What one person considers success, another very well might deem a failure. Which brings us back to the the purpose of the ranking in the first place, which is -- to reflect "success". The fact that success is arbitrary is irrelevant, the issue is that the "methodology" is based on prize pool -- which is... surprise! an arbitrary value. If you need examples of why this is an imperfect system then re-read the post above. https://tl.net/forum/starcraft-2/645694-most-successful-sc2-players-of-q2-2026You can read through the comments of the last thread and see that everyone is quite aware of the intention. My final suggestion is to either keep the year and use a specific number of best results, or limit it to recent months only.
My guy, no offense, but maybe don't talk about "flawed data" or "arbitrary values" when you literally stated that any list has to land on a specific result to "not be ragebait". That is really just saying "the data has to be filtered and presented in a way to get the result I already arbitrary decided is correct". A result I for one would btw heavily doubt. While Serral, Clem and herO are the clear Top 3 right now, I wouldn't put Rogue in that same league. Could change with RSL, but as of now he hasn't reached a single final this year in any of the big events. Not sure how that locks him in as No. 4 right now.
And that was my entire issue: You started right out of the gate with some big words and philosophies, only to then land on something that has absolutely no statistical basis. Don't get me wrong: Any suggestions about reducing the runtime of this particular ranking (from 12 to 6 months for example) are absolutely exempt from that, those are valid discussion points.
Lastly: A million dollar invitational would surely throw the curve off, but a)you could always exclude those (like the ATP does for example) and b)does the ranking have a cap, so at some point it doesn't matter how much you inflate the prizepool. And c)...is there literally nothing like this insight. Though maybe PiG will follow Grubby and Hera and do a Streamer Invitational, who knows.
|
|
|
Rogue is certainly not a definite top4 - though he's playing easily the best he's ever played.
Per aligulac the difference between #1 Serral and #2 Clem is 200 rating~
The difference between #3 and #10 is <200 rating.
her0 dropped off a lot because of the patch.
e-rankings is an activity ranking really.
|
|
|
|
|
|
EPT
