Justin Cotten
Impact
- Rating
- +1.51
- Gospel
- +1.68
- Avg Diff
- +1.63
- Proj Diff
- +1.69
Record
- W
- 129
- L
- 80
- Win%
- 61.7%
- xW
- 126.61
- xWin%
- 60.6%
Volume
- G
- 209
- Days
- 50
- G/Day
- 4.18
- % All Games
- 16.0%
- % All Days
- 27.2%
Company kept
- Team Quality
- +2.03
- Teammate Qual
- +0.29
- Opp Quality
- +0.34
- Other 9 Diff
- -0.06
- % Pos Teammates
- 62.7%
- % Pos Opps
- 57.4%
- % as Favorite
- 70.8%
- % Better Tm
- 48.3%
Honors
- MVP%
- 26.0%
- LVP%
- 4.0%
Attendance
- Mon%
- 31.8%
- Wed%
- 22.8%
- Sat%
- 26.2%
- 1st Game Rate
- 82.0%
- Last Game Rate
- 28.0%
- Last Game
- 2026-09-09
Both axes subtract out the other player's own rating, so a dot's position is about the pairing rather than about how good the other player is. Left to itself the raw version is a diagonal smear that just re-reads the leaderboard. Treat single dots gently: split any pair's games in half and the halves only agree at about r = 0.2.
What the buckets mean
- 1st — best on team
- Highest-rated of their own five.
- 2nd–4th
- Second, third or fourth option on their five.
- 5th — last option
- Lowest-rated of their own five.
- 1st–10th on court
- Same idea, ranked against all ten players in the game.
- Better than all 5
- Out-rated every opponent, whatever their own side looked like.
- Better than 4 / 3 / 2 / 1 of 5
- Out-rated that many of the opposing five.
- Worse than all 5
- Every opponent out-rated them.
- Much worse team
- Teammates rate more than 3 points below the opponents.
- Worse team
- Between 1 and 3 points below.
- Even matchup
- Within 1 point either way.
- Better team
- Between 1 and 3 points above.
- Much better team
- More than 3 points above.
- Weakest → Strongest games
- Combined rating of the other nine, in bands from under −1 to over +5.
Team advantage versus on-court quality: the first is the gap between the sides, the second is the level both are playing at. Opponents +5 against teammates +4, and opponents −4 against teammates −5, are both a −1 gap but completely different games. Across the whole dataset the two measures correlate at only about −0.1, so they really are separate questions.
Ranks use each player's rating, and genuine ties share the better rank — two players tied at the top are both 1st and nobody is 2nd. Ties are about 4% of team-games and come almost entirely from players who share a tier rating.
Reading these: Win% is what happened, Win Prob is what the model expected going in. Gospel is the gap — positive means outperforming expectation in that role. A player who wins a lot as the 1st option but posts a negative Gospel there is winning on the strength of the teams they were given, not their own.