Dolphins vs Bills — Week 18, 2023
The Dolphins took over trailing 21-14 with 1:53 left, 1st & 10 at their own 37, 0 timeouts in hand, against the Bills.
What really happened
Final: Dolphins 14, Bills 21. Bills won 21-14 — the drive turnover.
Dolphins ball, down 7 — it takes a touchdown. At their own 37, 1:53 on the clock, no timeouts left.
nflverse's historical win-probability estimate for this exact situation was 9.4% for Dolphins. Our model starts at 8.6%. The two numbers are estimates, not verdicts — the gap between them is a reminder that endgame win probability is hard to pin down.
The real drive, snap by snap
- (1:53) (Shotgun) 1-T.Tagovailoa pass short left to 10-T.Hill pushed ob at MIA 44 for 7 yards (30-D.Jackson).
- (1:48) (Shotgun) 1-T.Tagovailoa pass short right to 81-D.Smythe to 50 for 6 yards (23-M.Hyde).
- (1:22) (Shotgun) 1-T.Tagovailoa pass incomplete deep right to 10-T.Hill (20-T.Rapp) [50-G.Rousseau].
- (1:17) (Shotgun) 1-T.Tagovailoa pass deep left intended for 83-C.Claypool INTERCEPTED by 20-T.Rapp at BUF 22. 20-T.Rapp to BUF 22 for no gain (83-C.Claypool). BUF-20-T.Rapp was injured during the play.
The opening call, by win probability
At the first snap, the strongest opening call in our model was Deep post (+1.06 pts), and the weakest was Spike it (-0.97 pts). A positive number means the call improved the offense's expected win probability; a negative number means it cost them.
| Opening call | Expected WP change |
|---|---|
| Deep post | +1.06 pts |
| Curl route | +0.97 pts |
| Pull RPO | +0.67 pts |
| QB keep | +0.61 pts |
| Quick slant | +0.59 pts |
| Play action | +0.43 pts |
| Deep corner | +0.35 pts |
| Deep comebacker | +0.35 pts |
| Quick out | +0.26 pts |
| Quick hook | +0.22 pts |
| Screen pass | +0.03 pts |
| Quick RPO | -0.23 pts |
| QB sneak | -0.26 pts |
| Outside run | -0.65 pts |
| Inside run | -0.89 pts |
| Off-tackle run | -0.95 pts |
| Spike it | -0.97 pts |
This is our own model estimate, not the real coach's grade and not a claim about what the “correct” call was. Football decisions depend on the opponent, personnel and game plan — the model only measures how a call changes win probability in this situation.