NHL Betting: Using Advanced NHL Statistics to Improve Your Handicapping Experience

SBR Staff

Tuesday, April 28, 2015 3:48 PM UTC

Tuesday, Apr. 28, 2015 3:48 PM UTC

As the NHL Playoffs continue on, I wanted to discuss how the new wave of advanced NHL statistics can help you create better prices in your NHL handicapping.

Over the past decade (and specifically in the last few years), there has been an emphasis on advanced metrics that better explain why players and teams are better than others.  Sound familiar?  I'm sure you've heard of Sabermetrics and Moneyball and how it has been applied to the MLB. 

Before the new wave of advanced hockey statistics, there were not many hockey statistics to use to gauge performance.  Shots, goals, and saves would be the obvious stats to look at, but these metrics are so common that NHL odds makers already factor these into their prices.  You'll start to hear Corsi, Fenwick, Zone Starts, GVT, and PDO as generally accepted advanced hockey stats that can better explain a player’s or team’s value than the traditional hockey statistics.  Furthermore, oddsmakers are just learning how to apply these into their prices.  This presents an opportunity to find mispriced NHL games if we know how to use these advanced stats.

The purpose of this article will be to explain Corsi and Fenwick and how these metrics can be translated into Shots on Goal...which can be translated into Goals.  Before Corsi and Fenwick existed, there were no stats that measured possession of the puck without watching a game with a stopwatch.  And if you think about it, possession is pretty important in a time-based event like NHL.  Obviously, you need possession of the puck to shoot and score!  This is different than sports like basketball and American football where possessions are typically equal for both participating teams.  And that is why Corsi and Fenwick can be so valuable.  Simply put, both are advanced metrics that can now measure how well a hockey team can possess a puck.

Simply put:

Corsi = Shots on Goal + Missed Shots + Blocked Shots

Fenwick = Shots on Goal + Missed Shots

A Corsi event is recorded when any player directs a shot toward their opponents net.  A Fenwick event is recorded when a shot on net is successful or a shot misses the net (but is not blocked).  As these concepts start to sink in, you’ll notice the more Corsi/Fenwick events a team records, the longer they must have possessed the puck.  Please note that these two statistics are recorded separately.  A Corsi event may be a Fenwick event and vice versa.  The main difference is the blocked shot....Corsi accounts for this while Fenwick ignores.

So how can these be useful handicapping a NHL game? Here is a simple outline of how my basic handicapping works:

By using season-to-date Corsi/Fenwick statistics for shots, I will do the following:

- Predict Corsi (total shots directed toward opponent's net)

- Predict Fenwick - what % of Corsi is actually Fenwick (uses a conversion rate that factors how often a defending team will block a shot)

- Predict Shots on Goal - what % of Fenwick actually converts to SOG (uses a conversion rate that factors how often a shooting player/team will miss the net)

Let’s use a real life example to use Corsi/Fenwick to predict shots and goals between 2 teams:

NY Islanders vs. Washington 5 vs. 5 Corsi Fenwick Stats:


Corsi For

Corsi Against

Corsi Average

Fenwick For

Fenwick Against

Fenwick Average















Here's another NHL betting method you can use to enhance your NHL betting profit!

To start computing how many goals the Islanders will score against the Capitals, you start by comparing Washington’s Corsi Against vs. the Corsi Average.  You’ll notice Washington’s defense (Corsi Against) is lower than the league average.  This signifies that New York is facing a defense that is better than an average NHL team.  That is important to calculate since we need to adjust New York’s Corsi For as a starting point to determine how many shots we can expect from New York.  Simply stated, we want to multiply New York’s Corsi For by the % difference between Washington’s Corsi Against and the league average.  In formulaic form:

NYI Projected Corsi For = 49.4 * (41.7/43.7) = 47.1

The next step will be to convert New York’s projected Corsi into Fenwick.  In other words, what percent of the time are shots being blocked?  Similar to how we calculated Corsi, we want to determine a Corsi to Fenwick ratio for New York.  As their stats indicated, New York’s raw Corsi to Fenwick ratio is 73.7% (36.4/49.4).  If we want to project New York’s ratio against Washington, we will calculate the below formula:

NYI Projected Fenwick = 47.1 * (73.7% * (29.9/41.7)/(32.3/43.7)) = 33.7

Since Fenwick equates to projected shots and missed shots, we will need to deduct a certain percent of shots directed at the opponent’s net that will miss the net.  Based on my research, about 28% of all shots will miss the net.  Because of the randomness of missed shots, I apply this linear percentage to all teams.  With that said, here are New York’s projected shots on goal:

NYI Projected Shots on Goal = 33.7 * 72% = 24.3

In short, New York is projected to accumulate 24.3 shots on goal against Washington during 5 vs. 5 situations.  Since 80% of an average NHL game occurs with a 5 vs. 5 situation and the fact that penalties are unpredictable, this can be valuable information for projecting a final score, win probability, etc.  If you assume an average team will play 10% of a game on the Power Play and 10% on the Penalty Kill, you can use the calculations above with the Corsi/Fenwick statistics for those situations to project total shots for an entire game.  Furthermore, these projections can be “fine-tuned” by applying adjustments for home/away, rest, and any of trends or situational factors that may influence a team’s performance.

As the NHL playoffs continue, broaden your NHL handicapping to include advanced statistics like Corsi and Fenwick.  As I said earlier, these are fairly new statistics that odds makers are just learning how to use to develop optimal prices for an NHL game.  Whenever we can use any factors or statistics that odds makers are not using, we can develop an advantage for our NHL picks!

comment here