Collective Workspace
● LIVE
Prototype · auth-free demo
Loading measured results…
0 competitions with measured results · 0 teams on the tape · scored inside their own league open team board →
Shadow scores

Canonical engine, running beside the live one.

Read-only. Every number on the right is recomputed in the browser from the same inputs the ingest uses — nothing is written back, so the stored scores your app serves today are untouched until you say promote.

Stored spread
0.0 – 0.0
band 0.0 · sd 0.0
Canonical spread
0.0 – 0.0
band 0.0 · sd 0.0
Separation gain
—×
standard deviation vs stored
Avg confidence
0%
lanes measured, not modeled
Category weights by league
ALL ROWS
LeaguePSISMESIRSFCPΣ
NFL28%22%18%14%18%100%
NBA24%28%16%16%16%100%
WNBA18%30%14%22%16%100%
MLB30%18%20%14%18%100%
NHL34%16%18%14%18%100%
MLS26%20%16%24%14%100%
NWSL22%26%14%22%16%100%
PWHL26%26%14%18%16%100%
BAL24%26%12%24%14%100%
CFB-P422%26%20%18%14%100%
CFB-G524%24%18%20%14%100%
MBB-D120%28%18%20%14%100%
WBB-D118%28%16%24%14%100%
COLLEGE18%24%20%24%14%100%
CFB23%25%19%19%14%100%
CBBM20%28%18%20%14%100%
CBBW18%28%16%24%14%100%
FCS20%24%20%22%14%100%
D2FB18%24%20%24%14%100%
D3FB18%24%20%24%14%100%
JUCOFB18%22%20%26%14%100%
DEFAULT20%20%20%20%20%100%
Gate-and-broadcast leagues weight production hardest (NHL PS 34%, MLB 30%, NFL 28%). Growth leagues weight audience and identity (WNBA IS 30% / IRS 22%, NWSL IS 26% / IRS 22%). College codes fall back to the COLLEGE row, where NIL value sits in reach, market and story rather than the box score.
Biggest rank shifts
STORED → CANONICAL
Scoring the sample…
Side-by-side ledger
0 ATHLETES · READ-ONLY
#AthleteReachStoredCanonΔPSISMESIRSFCPConf
Loading athletes and scoring the shadow run…
Greyed lane values are still modeled proxies; white values are measured from live data. Promoting the canonical engine only happens when a roster ingest or social enrichment run writes back — nothing on this page does.