True hits
Hit inflation has all but vanished: what changed in NHL arena scorer bias
Hits, blocks, giveaways and takeaways are judgement calls by the people logging the game. For years each arena had its own crew, and some were far more generous than others. In 2023-24, Scotiabank Arena credited 46% more hits than the same teams saw on the road, while SAP Center credited 29% fewer. When the NHL centralised event recording in 2024-25 the differences shrank to within about ±6%, and in 2025-26 they did not repeat at all: the split-half reliability for hits is -0.15, so every arena's factor is 1.00.
Is the arena effect real? Split-half reliability by season
1 = arenas that inflated in the first half of the season inflated just as much in the second; 0 = pure noise
Split-half reliability of arena scorer factors from 2017-18 to 2025-26: hits 0.77, 0.78, 0.76, 0.68, 0.63, 0.74, 0.61, 0.37, -0.15.
- Hits
- Blocked shots
- Giveaways
- Takeaways
All 32 arenas: hits factor in 2023-24 vs now
Shrunk factors: 1.00 is neutral, above 1 means the building credited more than the road. Ranked by 2023-24 hits, the last season of local scorers with a reliable effect. Gold = inflates, teal = deflates.
| # | Arena | Hits 2023-24 | Raw | Blocks 2023-24 | Giveaways 2023-24 | Hits 2025-26 | Raw |
|---|---|---|---|---|---|---|---|
| 1 | Scotiabank ArenaTOR | 1.28 | 1.46 | 0.99 | 1.24 | 1.00 | 1.08 |
| 2 | Rogers PlaceEDM | 1.20 | 1.34 | 1.01 | 1.49 | 1.00 | 0.97 |
| 3 | TD GardenBOS | 1.14 | 1.22 | 0.99 | 1.15 | 1.00 | 0.98 |
| 4 | Benchmark International ArenaTBL | 1.10 | 1.17 | 1.00 | 0.76 | 1.00 | 1.01 |
| 5 | Rogers ArenaVAN | 1.09 | 1.15 | 0.98 | 0.79 | 1.00 | 1.00 |
| 6 | Mullett Arena (team relocated)ARI | 1.08 | 1.14 | 1.01 | 1.03 | – | – |
| 7 | Canada Life CentreWPG | 1.06 | 1.10 | 1.00 | 1.08 | 1.00 | 1.02 |
| 8 | American Airlines CenterDAL | 1.05 | 1.09 | 0.98 | 1.00 | 1.00 | 1.00 |
| 9 | Capital One ArenaWSH | 1.03 | 1.05 | 1.00 | 1.23 | 1.00 | 1.05 |
| 10 | KeyBank CenterBUF | 1.03 | 1.04 | 1.00 | 0.59 | 1.00 | 1.05 |
| 11 | T-Mobile ArenaVGK | 1.02 | 1.03 | 1.00 | 0.71 | 1.00 | 0.95 |
| 12 | Amerant Bank ArenaFLA | 1.01 | 1.02 | 1.01 | 1.21 | 1.00 | 1.03 |
| 13 | Honda CenterANA | 1.01 | 1.02 | 1.01 | 1.25 | 1.00 | 1.02 |
| 14 | Canadian Tire CentreOTT | 1.00 | 1.01 | 0.99 | 1.50 | 1.00 | 0.96 |
| 15 | Enterprise CenterSTL | 1.00 | 1.00 | 1.00 | 0.86 | 1.00 | 1.03 |
| 16 | Bridgestone ArenaNSH | 0.98 | 0.97 | 1.00 | 1.46 | 1.00 | 0.98 |
| 17 | Lenovo CenterCAR | 0.98 | 0.96 | 1.00 | 1.14 | 1.00 | 1.04 |
| 18 | PPG Paints ArenaPIT | 0.97 | 0.96 | 1.02 | 0.85 | 1.00 | 0.95 |
| 19 | Xfinity Mobile ArenaPHI | 0.97 | 0.95 | 0.99 | 0.85 | 1.00 | 1.01 |
| 20 | Grand Casino ArenaMIN | 0.97 | 0.95 | 0.99 | 0.93 | 1.00 | 0.95 |
| 21 | Scotiabank SaddledomeCGY | 0.96 | 0.94 | 1.01 | 1.32 | 1.00 | 1.01 |
| 22 | UBS ArenaNYI | 0.95 | 0.92 | 0.99 | 1.33 | 1.00 | 0.98 |
| 23 | Bell CentreMTL | 0.94 | 0.90 | 1.01 | 1.62 | 1.00 | 1.02 |
| 24 | Climate Pledge ArenaSEA | 0.94 | 0.90 | 1.00 | 0.51 | 1.00 | 0.92 |
| 25 | Crypto.com ArenaLAK | 0.93 | 0.89 | 1.01 | 0.74 | 1.00 | 0.97 |
| 26 | Ball ArenaCOL | 0.93 | 0.88 | 1.00 | 0.64 | 1.00 | 1.01 |
| 27 | Little Caesars ArenaDET | 0.93 | 0.88 | 1.01 | 0.83 | 1.00 | 0.99 |
| 28 | Madison Square GardenNYR | 0.92 | 0.88 | 0.98 | 0.81 | 1.00 | 1.03 |
| 29 | Prudential CenterNJD | 0.92 | 0.87 | 1.01 | 0.83 | 1.00 | 0.97 |
| 30 | United CenterCHI | 0.91 | 0.86 | 1.00 | 1.12 | 1.00 | 1.11 |
| 31 | Nationwide ArenaCBJ | 0.84 | 0.73 | 0.99 | 0.48 | 1.00 | 0.96 |
| 32 | SAP CenterSJS | 0.82 | 0.71 | 1.00 | 1.03 | 1.00 | 1.01 |
Arenas are listed under their current home team. For arena-adjusted player totals, see the adjusted hits leaderboard.
Reliability by season
| Season | Recording | Hits r | Raw range | Shrunk range | Blocks r |
|---|---|---|---|---|---|
| 2017-18 | Arena crews | 0.77 | 0.69–1.41 | 0.76–1.32 | 0.68 |
| 2018-19 | Arena crews | 0.78 | 0.72–1.42 | 0.78–1.33 | 0.74 |
| 2019-20 | Arena crews | 0.76 | 0.60–1.56 | 0.69–1.42 | 0.79 |
| 2020-21 | Arena crews | 0.68 | 0.67–1.74 | 0.78–1.50 | 0.49 |
| 2021-22 | Arena crews | 0.63 | 0.63–1.42 | 0.77–1.26 | 0.68 |
| 2022-23 | Arena crews | 0.74 | 0.71–1.56 | 0.78–1.41 | 0.51 |
| 2023-24 | Arena crews | 0.61 | 0.71–1.46 | 0.82–1.28 | 0.19 |
| 2024-25 | Central | 0.37 | 0.84–1.13 | 0.94–1.05 | 0.18 |
| 2025-26 | Central | -0.15 | 0.92–1.11 | 1.00–1.00 | 0.00 |
How the arena factors work
- The factor
- For each arena and season: the average count per game of both teams in the home team's home games, divided by the same average in its road games. The home team's own style is in both halves, so it cancels and what is left is the off-ice scorer. A factor of 1.25 means 25% more hits were credited in that building.
- Shrinkage and reliability
- We split each season in half by date and check whether the arenas that inflated in the first half also inflated in the second. That split-half correlation r measures how much of the factor is a real, repeatable scorer habit. The published (shrunk) factor is 1 + r × (raw − 1), with r clipped to [0, 1]. When r is near zero, every arena is 1.00 no matter how far the raw numbers wander.
- What changed in 2024-25
- From 2024-25 the NHL records real-time events centrally instead of using an off-ice crew in each building. The split-half r for hits fell from 0.61–0.78 (2017-18 to 2023-24) to 0.37 in 2024-25, when the shrunk factors stayed within about ±6%, and to −0.15 in 2025-26, when no arena effect was detectable at all. Blocks, giveaways and takeaways followed the same path. One season of zero is not proof the effect is gone for good, so we keep measuring it every season.
- Adjusted totals
- A player's arena-adjusted hits divide every game's hits by that arena's shrunk factor for the season. For seasons up to 2023-24 this can move a home-heavy hitter by 10–20%. In 2024-25 the factors were small (about ±6%), so adjusted totals move by a few percent at most. In 2025-26 every factor is 1.00, so adjusted equals raw.
- Home-scorer mirage flag
- We flag a player when his home hit rate is well above his road rate, after allowing for the league-wide home edge. That means at least 40 hits, 15 hours of ice time both home and away, a z-score of 2.5+ and a ratio of 1.35+. Players with the opposite pattern are flagged as under-counted at home. A flag is a prompt to check his road numbers, not proof of scorer bias: role and opponents differ too.
FAQ
Which NHL arena counted the most hits?
Before centralised recording, the biggest inflators credited roughly 25–50% more hits than the same teams saw on the road. The table on this page ranks every arena for the last season of local recording (2023-24) and shows each arena's full history on its own page.
Is arena scorer bias still a thing in the NHL?
Much less than it used to. After the NHL centralised event recording in 2024-25, hits factors shrank to within about ±6% (split-half r 0.37, down from 0.61–0.78), and in 2025-26 they did not repeat at all, so every shrunk factor is 1.00. Seasons before 2024-25 still need the adjustment, and we keep checking whether the effect stays gone.
Should I adjust hits for the arena in fantasy hockey?
For 2025-26, no: every arena factor is 1.00. 2024-25 still had small factors (about ±6%), which our adjusted totals apply. For dynasty and draft research that leans on 2023-24 and earlier, use adjusted hits, because a home-heavy hitter in a generous building could look 15–20% better than he was.
Do faceoffs have arena bias?
No. Faceoff wins are recorded consistently: each arena's home faceoff residual is the size of binomial noise and does not repeat year to year, so we never adjust faceoffs.