Advanced Analytics Part IV

By JASON LOWENTHAL

Today we conclude our series on advanced analytics in hockey by putting it all together. Remember back in our first piece when we introduced Corsi and Fenwick? Let’s go back to that.

Corsi = shots on goal + blocked shots + missed shots

                                      Fenwick = shots on goal + missed shots

Generally, Fenwick is considered to be a more accurate indicator, so we’ll stick with that as our baseline possession metric. The reason is because Fenwick does not include blocked shots in the equation because it considers blocking shots a skill, whereas Corsi does not.

Fenwick is used as a way to expand possession statistics in hockey because calculating time of possession is incredibly difficult. A team can’t have 20 different guys watching each of the 20 players that dress every game. That costs time and money. So, we use Fenwick instead. Here’s an example of Fenwick using Jonathan Toews during five-on-five play.

iFenwick (Toews) = shots on goal + missed shots

iFenwick (Toews) = 141 shots + 33 missed shots

iFenwick (Toews) = 174

(iFenwick stands for Individual Fenwick)

Now that we have Fenwick down, we can take a look at Fenwick For percentage. FF% can be used both individually and for a team, and is probably the best metric that advanced analytics has to show its validity. FF% is calculated by using the following formula:

FF% = (100 x Fenwick For) / (Fenwick For + Fenwick Against)

Essentially, FF% is used to measure the percentage of shots a team has in a game to show possession. Here was the FF% for the Chicago Blackhawks last season:

FF% (Blackhawks) = (100 x FF) / (FF + FA)

FF% (Blackhawks) = (100 x 1565) / (1565 + 1285)

FF% (Blackhawks) = 156,500 / 2850

FF% (Blackhawks) = 54.91

This means that during all of their games last season, the Blackhawks had 54.91 percent of the shots (on goal and missed).

The table below shows the top ten teams from last season in the NHL for FF%:

Los Angeles Kings	56.1%
Chicago Blackhawks	55.4%
San Jose Sharks	54.6%
St. Louis Blues	53.7%
New Jersey Devils	53.6%
Boston Bruins	53.4%
New York Rangers	52.6%
Vancouver Canucks	51.6%
Detroit Red Wings	51.5%
Tampa Bay Lightning	51.3%

Green indicates team made playoffs

Red indicates team missed playoffs

As you can see, eight of the top ten teams in terms of FF% ended up making the playoffs last season. This gives significant validity to FF% to the use of advanced statistics in hockey in general.

Unfortunately, given the limited budget of USHL teams, advanced analytics are hard to come by in the league. However, as one can see from this four-part series, it can be done.

Advanced statistics are a good way for guys doing the little things to be recognized. However, it will never surpass the eye test. Although more and more teams are moving towards the whole “Moneypuck” idea, a combination of the two theories is best, because both have imperfections. Advanced stats worked for Billy Beane and the Oakland A’s, time will only tell if the same will work in hockey.