Building a Winner Using Advanced Hockey Analytics

Advanced hockey analytics have become very popular in the hockey world, as teams are starting to generate success from advanced stats. The NHL is starting to see a wave of new fans that use analytics to assess a player. Each NHL season, every team is seeking the same goal; capturing the Stanley Cup. All 31 teams piece together their rosters in hopes of making another run to the playoffs. Analytics are becoming more and more important to teams when constructing a roster.

Advanced Hockey Analytics and Winning

There has been a lot of pushback against analytics from some of the older minds in hockey. A lot of teams and managers still depend on the old eye-test when evaluating players. The eye test still does have merit but analytics can provide additional insights into a player that goes beyond the eye test.

Important Analytics to Know

When looking to create a perfect lineup, a general manager needs to identify the most vital stats. Some of the most important stats include OFF (total offence), DEF (total defence), and WAR (wins above replacement). These stats can all be accessed through evolving-hockey‘s website.

The RAPM chart is another important part of a player analysis, specifically GF/60 (goals for per 60 minutes) and xGA/60 (expected goals against per 60 minutes). The expected or “x” uses expected goals, eliminating the factor of luck in its respected calculation. The five bars fluctuate anywhere between 0ne to six standard deviations. The higher the deviations, the darker the bar colour turns. A good RAPM chart consists of more blue in GF/60 and xGA/60, making a “blue-chart” player a reasonable acquisition.

Using the Information

Advanced stats are a recipe for success in the NHL. Analytics can be the key to constructing a team, but only with the right formula. Advanced stats can be compared to branched-chain amino acids. They are good to use, but overuse could start to have negative effects.

As important as analytics are, a lineup needs a balance of skill, grit and intangibles like leadership and heart. Analytics is not a way to fully evaluate a player. When breaking down a player, one needs to take into account their “eye-test” and analytics in a mix. Some analysts use an even 50/50 split, while others use a 60/40 or 70/30 going either way. Other analysts don’t use analytics entirely, which can lead to a player being undervalued.

The issue with the ‘Eye Test’

The most common way to analyze a player is through the eye-test. A hockey fan will view a player through their points and highlights, using solely baseline stats like plus-minus. The problem with plus-minus is that a player isn’t 100% responsible for their plus-minus. A turnover can occur on the other end of the ice, meanwhile a forward is just stepping on the ice. If the opposition scores, the forwards end up with a -1 he can’t control.

A real eye-test would consist of evaluating a player shift by shift, game by game. Going off highlights and points is a flawed way to assess a player. The “points” argument is very flawed, as a player who can rack up a ton of secondary assists, or power-play points can still be viewed as a “good” player. If a player can’t perform well 5 on 5, what kind of value do they really bring to a team? The amount of chances a special team gets a night is unknown so being effective 5 on 5 is more beneficial.

Advanced analytics can somewhat be noticed in a proper eye-test, as there are shot-calculated stats like Corsi and Fenwick. GMs shouldn’t misinterpret analytics, as you also need goal scorers, fast players, and larger guys, which can be identified through an eye test. In other words, just looking at a player’s analytics doesn’t give you a full assessment of their game, just as watching a player using the “eye-test” doesn’t fully evaluate a player either. It’s important to be in both realms when looking to build a team.

Success Through Advanced Hockey Analytics

Many teams have found success in the past from analytics. Some of the best general managers throughout the league use analytics to construct their rosters. While it is important to bring revenue and money to a club, nothing tops winning the cup. Here are some of the best analytical teams to capture Lord Stanley.

2019 St. Louis Blues

The 2018-19 St. Louis Blues were sitting dead last among the league halfway through the season. The Blues had 11 players that posted a WAR (wins above replacement) higher than one. Some of these players include Ryan O’Reilly (4.4), Colton Parayko (2.6), and Vladimir Tarasenko (2). O’Reilly’s 4.4 WAR was good enough to put him 6th in the league for one of the most valued analytics.

During the 2018-19 season, the Blues finished with a GF/60 of 2.5, which was tied for 13th in the league. St. Louis posted an incredible xGA/60 of 2.19, which ranked them 3rd among all teams. The team also had a GF% of 53.25, sitting 10th in the league. So despite their struggles in the first half of the season, the Blues’ strong play began to get them wins in the second half of the season. They then built off their momentum all the way to a Stanley Cup win over the Boston Bruins.

2020 Tampa Bay Lightning

One of the most dominant teams to ever exist, the 2019-20 Tampa Bay Lighting were hungry after being swept by the Columbus Blue Jackets the year prior. The Lighting were loaded with talent all over the roster, with seven players posting a WAR higher than two. The team’s WAR leaders consisted of Brayden Point (3.8), Victor Hedman (3.2), and Anthony Cirelli (2.8). Point’s stellar 3.8 WAR was good enough for 4th among all skaters, only behind Artemi Panarin, Elias Pettersson, and Ryan Ellis.

Throughout the 2019-20 season, Tampa Bay finished with a GF/60 of 2.96, tied for 1st with the Colorado Avalanche. The Lightning had an xGA/60 of 2.09, good enough for 4th in the league. The team also recorded a GF% of 57.36, which finished 3rd. The numbers don’t lie, as this Tampa Bay team went on a dominant playoff run, winning the Cup in six games over the Dallas Stars.

2016 Pittsburgh Penguins

The first Stanley Cup of the back-to-back reign, the 2015-16 Pittsburgh Penguins were no disappointment. The Penguins had just acquired Phil Kessel, gearing up for an intense run. This Penguins team had 13 players record a WAR above one. The WAR leaders of the team consisted of Patric Hornqvist (3.3), Carl Hagelin (2.8), and Sidney Crosby (2.6). Hornqvist’s 3.3 WAR tied for 17th among league skaters that season.

The 2015-16 Penguins recorded with a GF/60 of 2.44, which was ranked 4th in the NHL. Pittsburgh posted an xGA/60 of 2.o2, placing them 5th among the league. The team in black and gold had a GF% of 55.4, finishing 4th overall around the league. This Penguins rostered was very experienced, filled with flashes of youth, complemented by great depth.

Advanced Hockey Analytics Matter

There’s no better way to put it, the numbers speak for themselves. The recent Stanley Cup Champions have consisted of fantastic individual and team performances. As proven, successful NHL teams don’t have two or three players with good analytics, but 10,11,12 steady analytical performers. There is a process when looking to add key pieces to a team, not just the player with the most highlights and points. GMs around the league need to become more educated about analytics if they aren’t already.  GMs need to start finding players that suit the team best and provide stable analytics for their roster.

For more details about this article, visit my page @hockeycampfire on Instagram. More team and individual stats will be provided.

Embed from Getty Imageswindow.gie=window.gie||function(c){(gie.q=gie.q||[]).push(c)};gie(function(){gie.widgets.load({id:’kBSuRabuR3tMT19MIDki5g’,sig:’0IhTVKJvpQnGFpY29Vl9tKbYGcFY83zWgvzCfJabiW0=’,w:’594px’,h:’411px’,items:’1155901668′,caption: true ,tld:’ca’,is360: false })});


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.