Showcase:

Showcase: our popflash evening of 16 Dec 2024

  • Real: the ten of us, the teams we played, every map score and every K/A/D/ADR line.
  • Real engine: skill S from each player's FACEIT history before that evening (ELO is today's).
  • Made up: join times, votes (a 4 : 4 tie on purpose) and who was late. Buttons do nothing yet.
  • Leetify card: live, the last 30 days before today (Leetify keeps no 2024 matches), never stored.

Mix #1 · Mon 20:46

10/10
Voting · 9 of 10 voted · waiting for Nefcio
Win chance A : B
50 : 50
Avg S
1535 vs 1534
Most even
Player · tap for detailsBalance score S
Team A1535
fontek38510
2208
czopo9
1778
stannn7
1414
loownn7
1362
Nefcio4
914
Player · tap for detailsBalance score S
Team B1534
sm1leyz10
2052
jawOla217
1536
botjacek7
1429
J4neX7
1414
roevs6
1241
How was this calculated? czopo

S = E + F + 0.5·M + 0.5·A = 1797 − 13 − 14 + 8 = 1778

EE contribution1797

1,797 ELO at lineup · group admin's manual value because FACEIT ELO was unavailable

FFACEIT form, last 30 days−13

39 matches 16.11–16.12 · rating 1.17 vs 1.25 over 171 older ones · ratio 0.94 → 0.95 after shrinkage (k = 10) · raw 500 × -0.048 = −24 · × 0.54 for a slump at 1797 ELO

MMix form · weight 0.5−14

10 mix maps · Mixer Rating 0.99 vs group 1.07 → 1.02 after shrinkage · raw −50 × 0.54 for a slump at 1797 ELO

AActivity, last 30 days+8

16 sessions (evenings) 16.11–16.12 · 3 a month is neutral, fewer costs up to −60, 6+ adds +15.

Variant cost: 0.1 pp imbalance + 0 pp rule penalties = 0.1. Rank 1 of 126 possible splits; the three cheapest distinct ones are shown. Teammates in all three variants: czopo + Nefcio, J4neX + botjacek (splitting them would cost more than 3 pp of balance).

Variant 14
  • fontek385
  • loownn
  • roevs
  • botjacek
Variant 24
  • stannn
  • jawOla21
  • J4neX
  • You
Variant 31
  • sm1leyz

Votes are public. Not voted yet: Nefcio

Admin

Voting closes when you close it, or by itself 60 minutes after the last vote once all ten have voted.

Nefcio is late. As tradition demands.