League player scrapes 2.27 million accounts and claims Emerald rank is flooded with smurfs

The data shows fresh accounts climbing fast, but critics say the method might create its own problems.

OP.GG logo on blue background
(Image via op.gg)
TL;DR
  • A player scraped 2.27 million EUW accounts from OP.GG and used a scoring model based on account level, win rate, and games played to estimate smurf prevalence.
  • The analysis claims Emerald rank shows an unusual spike in likely smurfs because fresh high-MMR accounts tend to land there after placements.
  • Critics say the methodology might create an artificial spike since the scoring system uses different weights at Emerald 4, and the player acknowledged this as a fair criticism worth addressing in future work.
Community Reactions
How do you feel about this story?
👍
0
👎
0
😂
0
😡
0
😢
0

A League of Legends player by the handle jack37512 on Reddit has published a data project claiming to expose just how bad the smurf problem really is in Emerald rank. They scraped roughly 2.27 million accounts from OP.GG on the EUW server and built a point-based scoring model to estimate which accounts look like smurfs climbing quickly through the ranks.

The analysis aimed to move beyond gut feelings about smurfing. Using Python and database tools, the player created a scoring system that flagged accounts based on three main factors: unusually high win rates for their rank, low account levels, and small numbers of ranked games played. The idea was to catch the profile of fresh accounts rocketing through lower ranks.

Their central finding was striking. Emerald rank showed a pronounced spike in accounts that scored as likely smurfs compared to neighboring ranks like Platinum. The player argued this makes structural sense because Emerald is where high-MMR fresh accounts tend to land after placements or rapid climbs.

But the methodology immediately drew criticism. The player admitted they used different scoring weights at Emerald 4 versus lower ranks. Critics pointed out that changing the formula at the exact rank where you claim to find a spike could mechanically create that spike rather than detect a real phenomenon.

The player defended the approach by noting Emerald is genuinely different from a matchmaking perspective. It’s a relatively new tier inserted between Platinum and Diamond in 2023, and fresh accounts with strong performance often place directly into this bracket. They said they ran checks applying each scoring method across all ranks and the Emerald increase still appeared, though they acknowledged the transition point was “a valid methodological criticism.”

Some of the data got people talking in unexpected ways. A portion of Challenger accounts got flagged as smurfs, which the player clarified doesn’t mean smurfing inside Challenger but rather alt accounts that climbed all the way to the top. The model was designed to catch accounts that looked like they climbed fast, regardless of where they ended up.

Explore More
Meet the Editor
mm
Head of Spilled