What Is Corsi in Hockey? A Simple Explanation of the Stat Everyone Argues About

Twenty-five Corsi events happen for every goal scored in an NHL game. Think about that. For every puck that actually crosses the line, there are 25 other moments — shots, misses, blocked attempts — that tell a much richer story about which team was truly controlling the game. That’s exactly why Corsi exists. And once you understand it, you can’t watch hockey the same way again.

The Basic Idea: Shots Tell a Story Goals Can’t

Goals are rare. In a tight 2–1 game, the losing team might have outshot the winner by a wide margin and spent 60% of the night camped in the offensive zone. The final score doesn’t capture that. Corsi does.

At its simplest, Corsi measures total shot attempts — not just shots on goal, but also shots that miss the net and shots that get blocked — and compares what one team generates versus what it gives up. If your team is producing more shot attempts than the opposition at even strength, you’re controlling the puck. That’s the entire idea.

The number you’ll see most often is CF% (Corsi For Percentage). The formula is straightforward: divide your team’s or player’s shot attempts for by the total shot attempts in all situations where they were on the ice (for and against combined), then multiply by 100. A CF% above 50 means your squad generated more than half of the shot attempts. Below 50, and you’re being outworked. Above 55% over a full season? That’s a legitimate possession powerhouse.

Who the Heck Is Jim Corsi?

Here’s where the origin story gets genuinely weird. Jim Corsi was a longtime NHL goaltending coach — most famously with the Buffalo Sabres. He started tracking total shot attempts, including misses and blocks, because he wanted to understand the true workload on his goalie. That makes complete sense: a goalie might face 28 shots on goal in a game, but if 15 pucks were also blocked or sailed wide, that’s a completely different level of pressure than 28 easy shots.

Corsi didn’t name the stat himself. The actual name came from a blogger called Vic Ferrari (real name Tim Barnes), a financial analyst from Chicago who had heard Sabres GM Darcy Regier discussing shot differential on the radio. Ferrari built out the formula, went looking for a Sabres staff member whose name fit the concept, landed on a photo of Jim Corsi — whose mustache, reportedly, sealed the deal — and the rest is hockey analytics history.

It sounds absurd. But here we are, in 2026, talking about the most widely used possession metric in professional hockey, named for a man’s mustache.

What CF% Actually Tells You

Most NHL players cluster between 45% and 55% CF% over a full season. Once you understand this range, the numbers start meaning something real.

A player or team sitting at 57–58% CF% is genuinely dominant. They’re winning the possession battle in nearly every game. Historically, NHL.com analytics data shows that teams in the top five for CF% nearly always make the playoffs, while teams in the bottom five rarely do. It’s not a perfect predictor, but it’s a powerful one.

There’s also CF% Relative (CF% Rel) — which compares a player’s CF% when they’re on the ice versus when they’re off it. This strips away team context and asks a cleaner question: does this player make his team better at generating shot attempts? A positive CF% Rel means yes. A negative number means the team actually controls the puck better without him.

For coaches and GMs, CF% Rel is the real lever. It helps identify players who are quietly dragging possession numbers down despite counting stats that look fine on the surface.

Why Not Just Use Goals?

Because goals lie — at least in small samples. About 8% of all even-strength shots go in over a full season. That means in any single game or short stretch, a team can go cold on conversion and lose games they dominated, or get hot and win games they had no business winning. Corsi smooths that noise out. Since there are roughly 25 Corsi events for every goal, you get a statistically meaningful sample size much faster.

That’s the core value proposition: more data, faster. A team’s Corsi numbers stabilize after about 20–25 games and become predictive. A team’s goal differential might not stabilize for 40 games or more. If you’re trying to figure out whether a team is actually good or just running hot, CF% gives you the answer weeks earlier.

The Real Limitations

Corsi isn’t perfect, and anyone selling it as such is doing you a disservice. Three specific limits matter.

First, not all shot attempts are equal. A snap shot from the perimeter is not the same as a one-timer from the slot. Corsi doesn’t distinguish. This is exactly why expected goals (xG), which weights shot attempts by their location and type, has grown alongside Corsi as a complementary metric — not a replacement.

Second, zone starts matter. A player who begins most shifts in the offensive zone will naturally generate more shot attempts than a defensive specialist who starts in his own end. When comparing players across different roles, you have to account for this or the numbers mislead you.

Third, Corsi measures what happens at even strength with the player on the ice, not their true individual impact. Two mediocre players with great linemates can post excellent CF% numbers. Context always matters.

How to Use Corsi Without Driving Yourself Crazy

The best way to use Corsi is as a first filter, not a final verdict. If a player has a CF% Rel that’s been consistently negative across multiple seasons and multiple teams, that’s a real signal worth taking seriously. If a player posts a 60% CF% on a powerhouse team, that tells you less than if he posts it on a rebuilding club.

Use it alongside xG, scoring chance rates, and good old-fashioned zone time percentages. No single number wins the argument. But Corsi, for all its quirks and limitations, is still the cleanest, fastest way to ask whether a team or player is actually controlling hockey games — and that question never goes out of style.

In a sport where puck possession drives everything, a stat that measures possession better than the box score will always have a home.

Think Corsi is overrated or underrated? Drop your take in the comments or find me on X — I want to hear which advanced stats you actually trust when evaluating players.