The Gate We Apply to One Sport, Skip in Another
PlayerGem's gate system was built sport by sport, and the inconsistencies between them turned out to be the most inst...
Read itLatest Articles
The Gate Soccer Passed That Hockey Retired
PlayerGem's participation gate works differently in soccer and hockey — one sport kept it active, the other quietly m...
Read itSoccer and Basketball's Shared Stat, Kept by One
PlayerGem tracked touch-efficiency across soccer and basketball, found the same structural stat in both, and watched ...
Read itThe Tennis Gate That Told Us Nothing About Tennis
PlayerGem's availability gate cleared tennis players reliably for months before the shop realized it was measuring sc...
Read itShrinkage We Applied Evenly and Shouldn't Have
PlayerGem's uniform shrinkage rule pulled basketball and tennis estimates toward the same base rate — a method error ...
Read itShrinkage Floors Set Differently Per Sport
PlayerGem's shrinkage floor is not a single number — it is calibrated separately for each sport, and the reasons why ...
Read itThe Gate Soccer Kept Failing That Baseball Never Reached
A single availability gate worked cleanly in baseball for three seasons before soccer exposed a structural flaw the s...
Read itThe Corpus Entry That Survived Every Gate and Lied
PlayerGem traces how one player record cleared every methodological gate, held a clean sample, and still produced rat...
Read itWhen Instinct Held and the Corpus Caught Up
PlayerGem traces a stretch when a colleague's instinct outpaced the corpus across two sports — and what that gap reve...
Read itRotation, Rest, and the Limits of Prior Performance
PlayerGem's cross-sport analysis of how rotation and rest patterns break prior-performance models differently in bask...
Read itIndividual Sports Break Our Method
PlayerGem traces how its player-rating method transfers from team sports to individual ones — and exactly where the c...
Read itA Sport We Gave Up On, and Why
We covered a sixth sport for two seasons, then stopped. Not because we were inaccurate — because we could not grade it cleanly.
Read articleThe Method Transferred, the Thresholds Did Not
Adding a second sport was supposed to be a configuration change. It broke in a way that told us which parts of our method were actually general.
Read articleThe Same Model Failing in Two Directions
PlayerGem's cross-sport rating model produced opposite calibration errors in basketball and hockey — and the same shrinkage rule caused both.
Read articleWhat Baseball Taught Us About Our Hockey Gates
PlayerGem rebuilt its hockey candidate gates using a baseball availability model — and found out, slowly, what the two sports do not share.
Read articleWhat Each Sport Punishes You For
The same method applied to five sports fails in five different ways. Each sport has a specific weakness it exploits, and none of them are the obvious one.
Read articleAll Across Articles
- The Gate We Apply to One Sport, Skip in Another
- The Gate Soccer Passed That Hockey Retired
- Soccer and Basketball's Shared Stat, Kept by One
- The Tennis Gate That Told Us Nothing About Tennis
- Shrinkage We Applied Evenly and Shouldn't Have
- Shrinkage Floors Set Differently Per Sport
- The Gate Soccer Kept Failing That Baseball Never Reached
- The Corpus Entry That Survived Every Gate and Lied
- When Instinct Held and the Corpus Caught Up
- Rotation, Rest, and the Limits of Prior Performance
- Individual Sports Break Our Method
- A Sport We Gave Up On, and Why
- The Method Transferred, the Thresholds Did Not
- The Same Model Failing in Two Directions
- What Baseball Taught Us About Our Hockey Gates
- What Each Sport Punishes You For