Where a record and a tier disagree

UNRELIABLE: 40% of the tiers behind this scan were guessed rather than set by the committee. Tier those players before acting on it. 91 of the 227 players in this window had no committee tier.

KEEP means too few games to call, not correctly tiered. A row with few games and a large effect reads as KEEP because the evidence is thin, not because the tier is right.

Some committee names still have no account mapped, mostly C and D, so a player missing from this table has not been cleared - they have not been checked.

Window last 1y. Built 2026-09-19T06:42:46+00:00 from a model fitted 2026-09-19T06:32:21.787316+00:00 on 6382 matches.

LabelMeaningDecision
CLEARLY OVERwins more than the tier predicts; luck explains it less than 1 time in 100MOVE UP
OVERsame, less than 1 time in 20CONSIDER MOVING UP
ON TIERwithin what luck producesKEEP
UNDERwins fewer, less than 1 time in 20CONSIDER MOVING DOWN
CLEARLY UNDERwins fewer, less than 1 time in 100MOVE DOWN

OVER means the tier is too low - they beat it. UNDER means it is too high. The target tier comes off the points scale rather than the alphabet, so "up" from A is E, not S.

PlayerTierGamesPer 100Tiers offDecisionConfidenceRead with care because
L4mpjeS52-17.1-1.5MOVE DOWN: S → E1 in 159
BaczoB94+15.1+1.3MOVE UP: B → A1 in 173167% of games with SkyLine
eronnn.E76-14.5-1.2MOVE DOWN: E → A1 in 176
bodo48B68+11.7+1.0CONSIDER MOVING UP: B → A1 in 35only 68 games; too few for a 1-tier call
SkyLineB104+11.0+1.0MOVE UP: B → A1 in 16169% of games with czkk_
TerifireE194-10.7-0.9MOVE DOWN: E → A1 in 903only 194 games; too few for a 1-tier call
ddd331A77-10.6-0.9CONSIDER MOVING DOWN: A → B1 in 33only 77 games; too few for a 1-tier call
SupraB159+9.6+0.8MOVE UP: B → A1 in 15935% of games with jussi8030
mentalA214+9.3+0.8MOVE UP: A → E1 in 388only 214 games; too few for a 1-tier call
krestiS146-8.7-0.7CONSIDER MOVING DOWN: S → E1 in 67only 146 games; too few for a 1-tier call
PlainE274-8.7-0.7MOVE DOWN: E → A1 in 691
h2oS281+8.6+0.7SS TIER: already S, and beating it1 in 804
tezaXoA170+8.0+0.7CONSIDER MOVING UP: A → E1 in 6028% of games with agsor
REINAE133-7.8-0.7CONSIDER MOVING DOWN: E → A1 in 27only 133 games; too few for a 1-tier call
jussi8030E784+7.1+0.6MOVE UP: E → S1 in 10000+26% of games with devix
agsorS239+5.3+0.5SS TIER: already S, and beating it1 in 21only 239 games; too few for a 1-tier call
subakE406-5.1-0.4CONSIDER MOVING DOWN: E → A1 in 62

Per 100 is wins per 100 games above or below what the tier predicts, and is the number to argue over. Tiers off is how many tier steps that gap is worth. Confidence assumes games are independent, which they are not when one teammate fills much of a sample, so it is capped at 1 in 10000. Showing 17 of 50 players scanned.