Looking at NHL Hockey Stats. Are Advanced Stats Really that Advanced? by Jamie Gilcig – Nov 9, 2014

jg2CFN – Stats stats stats.   In this cap age the NHL is going advanced stats crazy; but many of these stats really are meaningless.     A stat, when viewed to ascertain data should come up with clear conclusions.

For simple examples we know that scoring more goals than you give up leads to more wins.

We  know that the more you win generally the better the chance you will win Lord Stanley’s Cup.

Are more penalties better or worse?  Are Corsi numbers really an indicator of future results?   I think the results clearly are mixed at this stage.

Back in 1995-96 when I became a NHL team’s first video statogragpher I created a set of tools that were exclusive to the Habs and many still haven’t been used in the NHL because I left hockey.

Many talented minds have filled the void, but I think that many are mistaking results for the real purpose of what stats are for and that’s to be able to reasonably predict results and to put tangent numerical data to the essence of what good hockey scouts and people know instinctively.

Are more blocked shots good?   Do teams that hit more lead to more wins?   Do teams that shoot more win more?    These are simple questions; b the answers can be complex.   A player has a set of skills; but those skills are impacted by the players he plays with and against.   That’s why sometimes when a player gets moved to a new team they seem to explode statistically. (or implode)

What smart NHL teams look for are those little bits of information that can reinforce the direction and path they are choosing.  Back in 1995 I would crunch players to see how they would fit on the Habs vs the team they were playing for.  All you need is that one gem sometimes to make a season’s work worth it.    That season my numbers predicted strong futures for Tomas Vokoun and a very young Jarome Inginla who Dallas would trade to Calgary before he made his NHL impression.

It also showed fan favorite Kirk Muller could not produce at the level the team would have needed.  He ended up signing with Dallas and played a defensive lower line role.   Of course scouting and coaching can’t force players to perform or adjust.   Prior to the Habs first series against the Rangers that year, after NY had just won the cup, I crunched the team and had a series of pieces of info that helped result in the Habs stealing the first two games of the series.    But like any great team the Rangers adapted the Habs could not fast enough.

While Adam Graves was bottled up it was shown that he took nearly 70% of his shot attempts from the same piece of ice.

So let’s look at some very early season data this NHL season.

Goals Per Game

1) Pittsburgh 4.15

2) Tampa Bay 3.86

3) Philadelphia 3.21

4) Washington 3.14

5) Vancouver & San Jose tied with 3.07

30) Buffalo 1.12

29) Florida 1.67

28) Winnipeg 1.87

27) Montreal 2.20

26) Arizona 2.36

Two of the top GPG  teams are currently in the Top Ten Overall in the standings.    Three of the bottom GPG teams are in the bottom 10.

You would think that the teams that score the most would win the most.  Montreal is 27th in goals for, but is 4th overall in the league.

Goals Against Per Game

1) Chicago 1.86

2) St. Louis, Nashville, Los Angeles, and Anaheim  1.93

30) Columbus 3.64

29) Dallas & Edmonton 3.50

27) Buffalo 3.31

26) Arizona 3.29

Now this appears to be more indicative.  Four out of the five top teams are in the top ten in wins.  Ironically Chicago which is first isn’t in the top ten.

All 5 teams that gave up the most goals per game were in the bottom ten.  That’s a clear indicator; but then most hockey fans know that if you give up more goals you lose more.

Shots On Goal Per Game

 1) Chicago 38.1

2) Minnesota 34.6

3) NY Rangers 32.1

4) San Jose 32.0

5) NY Islanders 31.6

30) Buffalo 22.2

29) New Jersey 26.6

28) Calgary 26.9

27) Florida 28.1

26) Montreal 28.2

An interesting point is that not one of the top five shooting teams is in the top 10 overall.    Three of the team that shoot the least was in the bottom ten.   Montreal clearly has bucked some trends to achieve it’s overall ranking.  Calgary also has won a few more than its low shooting numbers would normally achieve.

Shots on Goal Against per Game

1) Minnesota 23.0

2) Washington 26.4

3) St. Louis & Detroit 26.6

5) Chicago 27.2

30) Buffalo 37.4

29) Ottawa 35.6

28) Colorado 35.1

27) Philadelphia 33.2

26) Toronto 33.8

 Two out of our five make the top ten; but only one of the bottom five made the top ten.

Looking at the numbers so far it looks like shooting percentage could answer some questions?

Looking at Special teams.

Power Play

1) Pittsburgh 37.5

2) Washington 28.9

3) Tampa Bay 27.1

4) San Jose 26.0

5) St. Louis 24.5

30) Buffalo 4.0

29) Minnesota 5.0

28) Montreal 7.7

27) Winnipeg & LA 26.6

Three out of the top 5 teams were in the Top Ten teams overall.  Two out of the bottom five ranked as well.   In this category clearly some teams had many more opportunities than others so percentages should even out over the course of the season.

Penalty Kill

1) Chicago 92.9

2) Detroit 89.6

3) Pittsburgh 88.5

4) Winnipeg 87.9

5) Vancouver & Colorado 87.7

30) New Jersey 65.5

29) NY Islanders 71.1

28) Nashville 71.8

27) Arizona 74.5

26) Dallas 76.0

Three of our top 5 at the top are in the top ten.  Three of our bottom 5 are in the bottom ten.

Face Off %

1) Montreal 54.9

2) Boston 54.2

3) Arizona 53.6

4) Carolina 53.1

5) Chicago 53.0

30) Buffalo 42.5

29) Calgary 45.8

28) Nashville 27.1

27) Vancouver 47.2

26) NY Islanders 47.3

Only Montreal was one of the teams in the top ten which is ironic as the addition of face off specialist Manny Malhotra is a key factor in improving their face off and penalty kill numbers.

And that’s the magic of hockey at the NHL level.  A hot goalie or streaking Sydney Crosby can tilt the machine on any given night or for stretch of games.   Finding those little pockets where you can get an edge is worth the price of gold; maybe even the price of your lowest priced roster player; but with more and more NHL fans and people gaming on hockey those edges will count more and more.

What do you think hockey fans?  Are you stat freaks?  Are you living and dying by what are called advanced stats?  You can post your comments below.

(Comments and opinions of Editorials, Letters to the Editor, and comments from readers are purely their own and don’t necessarily reflect those of the owners of this site, their staff, or sponsors.)

Comment policy reminder 

CFN suggests you post comments using your real name. If you wish to post with a pseudonym you can register that user id by emailing info@cornwallfreenews.com with your name, address, phone number and user id you wish to register.

Leave a Reply