Advanced NHL Betting Stats: Corsi & Fenwick

Hockey analytics had its revolution quietly. While baseball’s Moneyball moment played out in bestselling books and Hollywood films, hockey’s equivalent happened in blog comment sections and spreadsheets shared on forums. At the center of that revolution were two metrics that measured something deceptively simple: shot attempts. Corsi and Fenwick became the foundational language of modern hockey analysis, and for bettors willing to learn that language, they remain two of the most useful tools for evaluating team quality beneath the surface of win-loss records.

The core idea behind both metrics is that shot attempts are the best available proxy for puck possession and territorial control. A team that consistently out-attempts its opponents at five-on-five is spending more time in the offensive zone, generating more chances, and forcing the opposing team to defend. Over time, that territorial dominance translates to goals, and goals translate to wins. The relationship is not perfect on a game-by-game basis, but across a full season it is one of the strongest predictive signals in all of professional sports.

Corsi Metrics in Hockey Betting

Corsi counts all shot attempts at even strength, regardless of whether they hit the net, miss the net, or are blocked by a defender. If a player or team directs the puck toward the goal from a shooting position, it counts as a Corsi event. The metric is named after Jim Corsi, a former NHL goaltending coach, though he was not the one who developed the statistic in its modern form.

Corsi For (CF) is the number of shot attempts by a team. Corsi Against (CA) is the number of shot attempts by the opposing team. Corsi percentage (CF%) is the share of total shot attempts belonging to one team: CF / (CF + CA). A CF% above 50% means a team is generating more shot attempts than it is allowing, which indicates positive territorial control. A CF% below 50% means the opposite.

The reason Corsi uses all shot attempts rather than just shots on goal is sample size. In a typical NHL game, each team registers 25-35 shots on goal, but the total shot attempts including misses and blocks run much higher, often 50-70 per team. A larger sample stabilizes the metric more quickly and provides a more reliable signal. A team might have a bad night in terms of shots on goal because of fluky blocks and wide misses, but its Corsi number still captures the underlying volume of offensive activity.

At the team level, Corsi percentage correlates strongly with standings points over a full season. Teams with a CF% above 52% consistently outperform their expected win totals based on goal differential alone, and teams below 48% tend to underperform. This makes Corsi a leading indicator: it tells you where a team is headed before the scoreboard catches up.

What Is Fenwick?

Fenwick is a close relative of Corsi but excludes blocked shots from the count. It measures only shots on goal and missed shots, filtering out the events where a defender stepped into the shooting lane. The metric is named after Matt Fenwick, who proposed it as a refinement of Corsi based on the argument that blocked shots carry less information about offensive quality than shots that actually reach the goaltender or miss the net unimpeded.

The logic is that a blocked shot might reflect the shooting team’s offensive pressure, but it also reflects the defending team’s positioning and willingness to sacrifice the body. By removing blocks, Fenwick isolates the shot attempts that more purely represent offensive generation without the defensive noise of the opposing team’s block rate.

In practice, Corsi and Fenwick tell similar stories. Their correlation is extremely high, often above 0.95, meaning teams that rank well in one almost always rank well in the other. The differences tend to appear at the margins, in matchups where one team has an unusually high or low block rate. A team that blocks a lot of shots will look better in Corsi Against than in Fenwick Against, which could cause Corsi to overstate their defensive quality. Fenwick corrects for that potential distortion.

For bettors, the practical distinction between Corsi and Fenwick is small but occasionally meaningful. If you are evaluating a team that is known for an aggressive shot-blocking scheme, checking both metrics helps you determine whether their defensive numbers are genuinely strong or artificially inflated by blocks. Similarly, a team with a high Corsi but lower Fenwick might be generating a lot of shot attempts that are getting blocked, which could indicate that their offensive quality is lower than the raw attempt volume suggests.

Score Effects and Why Context Matters

Raw Corsi and Fenwick numbers are useful but incomplete without adjusting for game context. The most important contextual factor is score effects. Teams behave differently depending on whether they are leading, trailing, or tied. A team protecting a two-goal lead in the third period will naturally concede more shot attempts as the trailing team pushes aggressively. That concession inflates the leading team’s Corsi Against without indicating any decline in their actual quality.

This is why analysts and bettors rely on score-adjusted or close-game Corsi rather than raw totals. Score-adjusted Corsi applies a weighting formula to account for the known behavioral changes that occur at different score states. Close-game Corsi restricts the sample to situations where the score is within one goal during the first two periods or tied in the third, which captures the most competitive and representative minutes of play.

For betting purposes, close-game Corsi is the more actionable version of the metric. It strips away the garbage time where teams are either coasting with a lead or desperately chasing a deficit, and it shows you what happens when both teams are genuinely competing. A team with a 54% close-game CF% is a team that controls play when the game is on the line, and that control is the foundation of sustainable winning.

Using Corsi and Fenwick for Betting

The most direct application of Corsi and Fenwick for bettors is identifying teams whose underlying process does not match their results. In any NHL season, some teams outperform their Corsi profile by riding unsustainably high shooting percentages or elite goaltending that is likely to regress. Other teams underperform their Corsi profile because their shooting percentage is suppressed by bad luck or their goaltender is performing below career norms.

When a team has a strong Corsi percentage but a mediocre win-loss record, the market often prices them based on the record rather than the underlying metrics. That creates value on the moneyline and the puck line. The market is saying the team is average; the Corsi data is saying the team is good but unlucky. If you trust the process metrics over the results, you can bet into that discrepancy before the results catch up, which they tend to do over the course of a season.

The reverse is equally valuable. A team riding a hot streak with a weak Corsi profile is likely to regress. Their wins are built on unsustainable shooting luck or goaltending that is performing above its true level. Betting against these teams, especially when the market has inflated their odds based on the hot streak, captures the regression that Corsi predicts.

Combining Corsi with totals markets adds another layer. Teams with high combined Corsi rates play high-event hockey that produces more total shot attempts and, by extension, more total goals. Games between two high-Corsi teams tend to go over the posted total more frequently than games between two low-Corsi teams, because the sheer volume of attempts creates more scoring opportunities on both sides.

The Limits of Corsi and Fenwick

Corsi and Fenwick are powerful but not omniscient. They measure volume, not quality. A team can generate 60 shot attempts in a game, but if 40 of those are low-danger shots from the perimeter, the raw Corsi number overstates their offensive threat. This is why expected goals models, which weight shot attempts by location and type, have emerged as the next evolution of possession metrics. Corsi tells you a team is controlling the puck. Expected goals tells you what they are doing with that control.

Goaltending is another blind spot. Corsi is a five-on-five metric that measures skater-driven territorial control. It does not capture the quality of goaltending, which is the single most impactful variable in any given NHL game. A team with a 55% CF% and a below-average goaltender might still lose more games than it wins, because the goals they concede on the 45% of attempts they allow are not being stopped at a high enough rate.

Despite these limitations, Corsi and Fenwick remain essential starting points for any analytical approach to hockey betting. They are freely available on multiple public analytics sites, they are updated in near real-time during the season, and they provide a baseline measure of team quality that filters out the noise of shooting percentage variance and goaltending fluctuations. Think of them not as the final answer but as the first question: is this team actually controlling play, or are they surviving on luck and goaltending? The answer shapes every bet that follows.