Esports data
Why Korea's ladder looks different
Everyone says Korean solo queue is harder. Almost nobody measures it. This is the Challenger ladder of six regional servers, side by side, counted the same way every day.
Snapshot of 2026-08-04 · collected 2026-08-04 22:00:02 KST · queue RANKED_SOLO_5x5
The ladder, side by side
| Server | Challenger | Grandmaster | Median LP | Top LP | Win rate | Games each |
|---|---|---|---|---|---|---|
| KoreaThe benchmark | 300 | 700 | 2,010 | 2,812 | 53.73% | 760 |
| Southeast AsiaSingapore-hosted regional server | 300 | 700 | 1,073 | 2,600 | 54.53% | 402 |
| VietnamSplit from SEA in 2023 | 300 | 700 | 1,856 | 3,459 | 55.5% | 598 |
| JapanSmallest ladder here | 50 | 100 | 1,683 | 2,355 | 55.42% | 473 |
| North America | 302 | 718 | 1,648 | 3,236 | 55.04% | 662 |
| Europe WestRiot's largest Western server | 304 | 745 | 2,654 | 4,617 | 54.95% | 923 |
What the numbers say
The telling column is the one you would look at last. Korea's Challenger players post the lowest win rate of the six servers — 53.73% against 55.5% on Vietnam. A win rate is the one statistic that cannot be inflated by playing more: every game has a loser, and in a closed pool it converges on 50%. The closer a server sits to 50%, the more evenly matched the people inside it are.
Read the other columns against that. Europe West leads on everything volume can buy — 923 games per player, the highest median LP (2,654) and the highest ceiling (4,617). But volume and a high ceiling are what a large server produces on its own; neither says the games inside it are close. Korea's spread is the narrow one, and that narrowness is what people are describing when they say the server is hard.
Japan is the useful control. Its Challenger tier holds 50 players against roughly 300 elsewhere, because tier size scales with the population below it. Comparing raw LP across servers of different sizes is a mistake; comparing shape is not.
Why this is a daily series
A ladder cannot be reconstructed after the fact. There is no endpoint that returns what Challenger looked like last Tuesday — the API answers only for right now. So the record only exists if somebody was writing it down at the time. We started on 2026-08-04 and take one snapshot a day, at the same hour, whether or not anything interesting happened.
That is the same reason we archive provisional trade figures the day they are published: the revised number replaces the original everywhere, and the original is gone.
What we do not collect
No player identifiers. The API returns a puuid for every entry; we discard it at collection time and keep only the distribution. We are not interested in who is on the ladder, only in what shape it is. Holding identifiers we would never use is a liability, not an asset — so we do not hold them.
We also do not collect or republish match history, chat, or anything from the professional circuit. Pro match data is not part of the developer API, and the undocumented endpoints that serve lolesports.com are not ours to use.
Method
- Endpoints:
LEAGUE-V4challenger and grandmaster leagues, queueRANKED_SOLO_5x5. - One snapshot per server per day, 22:00 KST.
- Win rate is aggregate: total wins ÷ total games across the tier, not a mean of ratios.
- Median LP is the 50th percentile of the tier, not of all ranked players.
- Missing servers are recorded as missing. A gap is never filled with an estimate.
SeoulMarkets isn't endorsed by Riot Games and doesn't reflect the views or opinions of Riot Games or anyone officially involved in producing or managing Riot Games properties. Riot Games and all associated properties are trademarks or registered trademarks of Riot Games, Inc. Data via the Riot Games API.
Nothing here is investment advice.