Alan Hili at Planetary Qualifier Wynnum

melee.gg

Planetary Premier April 26, 2026 Record 4-2-0

Rating after 1,529

+65

Rating before 1,464 RD 151

Rating after 1,529 RD 131

Effective multiplier 1.0×

Performance 1,719

Field strength Mean 1,564 · Median 1,551 · 48 rated

// rating math · glicko-2 phase replay

How this rating change was computed

Glicko-2 doesn’t reward record — it rewards surprise. Winning a match the system expected you to win is worth almost nothing. Losing one is expensive. Losing to someone rated below you costs the most. The Planetary tier multiplier (1.0×) amplifies every gain and every loss.

Shape D top-cut adjustments QF SF F Wins +0.12 +0.18 +0.25 Losses -0.18 -0.27 -0.35

Making cut never costs you net rating — the made-cut floor pins the tournament delta to ≥ 0.

Start 1464 → Swiss 1529 +64.6

Top Cut Bracket Full bracket on the event page

Quarterfinals

Rookie2222 0

Doovelle 2

BDST_Zerolimit 2

Lily Hill 0

ThatsFine 0

BDST_Tivey 2

cayliss 2

brettc 1

Semifinals

Doovelle 2

BDST_Zerolimit 1

BDST_Tivey 2

cayliss 1

Finals

Doovelle 0

BDST_Tivey 2

See full bracket on the event page →

Swiss · 6 matches 4-2 +64.6 1.0× +0.0 +64.6 1529

Total 4-2-0 +64.6 1529

Biggest upset Beat BDST_Tivey 1686 at 32% odds / surprise +0.68

Costliest loss Lost to brettc 1452 at 51% odds / surprise -0.51 /

Matches (6)

R1 BDST_Tivey Win 2-0 1.0× +50.2 Swiss

Glicko-2 predicted only 32% odds for you. Upset wins move the rating model strongly because the system learns a lot from results it didn't expect.

R2 dylan solomon Win 2-1 1.0× +20.7 Swiss

R3 brettc Loss 1-2 1.0× Swiss

Glicko-2 predicted 51% odds in your favor. Unexpected losses carry the most rating signal — the system learns more from one upset than from many predictable wins. Their rating wasn't well-established (RD 137, true skill could span ±274).

R4 BDST_Clint Loss 1-2 1.0× Swiss

R5 Brian Evans Win 2-0 1.0× +28.7 Swiss

R6 Goon_Clown007 Win 2-1 1.0× +42.0 Swiss

Glicko-2 predicted only 42% odds for you. Upset wins move the rating model strongly because the system learns a lot from results it didn't expect. Their rating wasn't well-established (RD 132, true skill could span ±264).