Ottawa Senators · Goaltender · #30Samuel Ersson Saves: Prop Hit Rate & Last 10 Starts
Samuel Ersson went over 21.5 saves in 1 of his last 2 starts (50%) and 1 of 2 this season (50%), averaging 25.0 saves a start.
No saves line posted for Ersson yet
Books post saves lines for only a few goalies a night. Hit rates below use 21.5, the line nearest his saves per start.
Hit rate over 21.5 saves
| Starts | GS | Over | Hit rate | Avg |
|---|---|---|---|---|
| Last 5 starts | 2 | 1 | 50% | 25.0 |
| This season | 2 | 1 | 50% | 25.0 |
| 2025-26 | 29 | 16 | 55% | 20.8 |
Last 2 starts
- 19underOct 13 vs STL · 26 SA
- 31overOct 6 @ DET · 35 SA
Home, road, back-to-backs and vs FLA
| Starts | GS | Over | Hit rate | Avg |
|---|---|---|---|---|
| Home | 15 | 7 | 47% | 19.9 |
| Road | 16 | 10 | 63% | 22.3 |
| Back-to-back | 10 | 4 | 40% | 20.5 |
| vs FLA | 0 | 0 | — | — |
2026-27 and 2025-26 starts combined. Rows under 10 starts are dimmed: too few to read.
At other lines
| Line | Last 2 | This season | 2025-26 |
|---|---|---|---|
| Over 19.5 | 50% | 50% | 62% |
| Over 20.5 | 50% | 50% | 59% |
| Over 21.5 | 50% | 50% | 55% |
| Over 22.5 | 50% | 50% | 41% |
| Over 23.5 | 50% | 50% | 31% |
Against the market
We graded 120 stored saves lines from Oct 7 to Oct 20 (the last consensus line before puck drop, goalie played, 2+ books). The over hit 50% against a market average of 50%. The market's log loss was 0.693 (lower is better; a coin flip is 0.693). The last-10 hit rate scored 0.956 on its 43 lines, against 0.692 for the market on the same lines. Our projection scored 0.709 against 0.693 on 120 lines.
After the vig, flat bets at the best price: every over −3.2% on 120 bets, every under −5.6% on 120 bets, and the S1 under (our projection 1.5+ saves below the line) +1.7% on 46 bets. That is far too few lines to call anything an edge; the 95% ranges run well either side of zero.
History without prices (books' saves lines are stored only since October 2026): on 1,309 held-out 2025-26 starts, a goalie's raw last-10 hit rate at a line near his expected saves scored a log loss of 0.751, worse than one flat guess (0.692). Shrunk hard it only matched the flat guess (0.694). Expected saves from his team's shots against and the opponent's shots for did best (0.673). Hot goalies (8+ of the last 10 over) went over 44% next time (n 36): streaks don't carry.
What a saves prop measures
A saves prop is the number of shots a goalie stops in a game he starts; if he does not start, books void the bet. Saves are shots against minus goals against, so they follow his team's defence and the opponent's shot volume more than the goalie himself. Hit rates here count starts only, at Samuel Ersson's line of 21.5.
Saves swing a lot from start to start: across 3,680 starts in 2024-26 the typical miss from a goalie's expected saves was 7.1 saves, and 5% of starts ended early with the goalie pulled. Busy nights stretch the top end, so 53% of starts finished under the expected average. A goalie's season over rate at his own line repeated year to year at r = 0.11: noise.
Treat the last-10 hit rate as a description, not a forecast. On 1,309 held-out starts it scored worse than a flat guess (log loss 0.751 vs 0.692); a mean built from his team's shots against and the opponent's shots for did best (0.673), and that is information books can see too.
Frequently asked questions
How often does Samuel Ersson go over 21.5 saves?
Samuel Ersson has gone over 21.5 saves in 1 of 2 starts this season (50%), and 1 of his last 2.
Do goalie saves hit rates beat the betting market?
Not in our tests. On 120 stored saves lines this season the market scored a log loss of 0.693 (lower is better); the last-10 hit rate scored 0.956 on the 43 lines where it existed. Betting every under at the best price returned −5.6%, every over −3.2%.
Why do saves unders hit so often?
Saves have a long top end: a busy night can run ten or more over the line, so the typical start sits a little under the average, and a pulled goalie lands far under. In 2024-26, 53% of starts finished under the goalie's expected saves. Books shade the line for this; every under at the best price this season returned −5.6%.