Expected versus Actual Goal Production: Why Montreal’s Offense is Underperforming
by Stephan Cooper on Apr 8, 2011 1:53 AM EDT
Email
Print
The biggest problem with the Montreal Canadiens’ offense this season has been an overall inability to convert on shots on net. The Habs average 31.7 shots per game, good for a tie for 7thin the league with the Philadelphia Flyers and Tampa Bay Lightning.
The team shooting rate is only 8.17%, among the worst in the league (better than only the Ottawa Senators with8.03%, Florida Panthers with 7.68% and New Jersey Devils with 7.27%).
The good news is that team shooting tends to be transient, with little carryover from year to year while team shooting tends to be a more permanent phenomena (for more on that look at the excellent work being done at behindthenethockey.ca). Since the NHL the lockout, Montreal has scored at 9.77%, 9.86%, 10.77%, 9.83% and 8.95%, which suggests that this year is an aberration. If the Canadiens ice a similar lineup next year they should be a relatively good offensive club.
Note that overall shooting percentage has gone down league wide this year but not nearly to the extent it has affected Montreal.
This poor team shooting percentage would explain why so many Habs have undershot their typical offensive boxcar numbers. To examine this phenomenon I have used the total shots of each player and their career shooting percentage to look at what each Canadien would have produced, if this had been a typical conversion year for each player, and what kind of total offensive year the Canadiens as a whole would have had.
A player’s shooting percentage tends to fluctuate around their career average, unless they undergo a major change in their playing style, so this could be taken as an indicator of about how much each player would have scored if they had average "puck luck."
Note that a flaw in this study is that it does not distinguish between power play and even strength shooting which tend to vary (it is easier to convert a shot into a goal on the power play).
Expected goals is the term I have elected to use for shots*career shooting percentage. Results are prorated to an 82 game season.
|
Player |
Shot # |
Career Goals to Shots Ratio |
Expected goals |
Games |
Prorated Expected Goals |
|
Gionta |
290 |
0.11 |
31.900 |
80 |
32.698 |
|
Plekanec |
222 |
0.115 |
25.530 |
75 |
27.913 |
|
Cammaleri |
186 |
0.12 |
22.320 |
65 |
28.158 |
|
Subban |
190 |
0.072 |
13.680 |
75 |
14.957 |
|
Kostitsyn |
187 |
0.124 |
23.188 |
79 |
24.069 |
|
Gomez |
152 |
0.074 |
11.248 |
78 |
11.825 |
|
Hamerlik |
129 |
0.052 |
6.708 |
77 |
7.144 |
|
Pouliot |
127 |
0.131 |
16.637 |
77 |
17.717 |
|
Pacioretty |
112 |
0.082 |
9.184 |
34 |
22.150 |
|
Moen |
99 |
0.071 |
7.029 |
79 |
7.296 |
|
Darche |
87 |
0.09 |
7.830 |
57 |
11.264 |
|
Wisniewski |
83 |
0.049 |
4.067 |
41 |
8.134 |
|
Eller |
79 |
0.103 |
8.137 |
75 |
8.896 |
|
Lapierre |
78 |
0.076 |
5.928 |
38 |
12.792 |
|
Spacek |
65 |
0.06 |
3.900 |
57 |
5.611 |
|
Pyatt |
65 |
0.035 |
2.275 |
59 |
3.162 |
|
Halpern |
62 |
0.127 |
7.874 |
72 |
8.968 |
|
Gill |
62 |
0.032 |
1.984 |
74 |
2.198 |
|
Weber |
60 |
0.014 |
0.840 |
39 |
1.766 |
|
Desharnais |
54 |
0.143 |
7.722 |
41 |
15.444 |
|
Picard |
49 |
0.068 |
3.332 |
43 |
6.354 |
|
White |
29 |
0.029 |
0.841 |
25 |
2.758 |
|
Gorges |
20 |
0.037 |
0.740 |
36 |
1.686 |
|
Markov |
20 |
0.077 |
1.540 |
7 |
18.040 |
|
Mara |
14 |
0.063 |
0.882 |
18 |
4.018 |
|
Boyd |
8 |
0.129 |
1.032 |
10 |
8.462 |
|
Sopel |
4 |
0.046 |
0.184 |
11 |
1.372 |
|
Dawes |
3 |
0.111 |
0.333 |
4 |
6.827 |
|
Palushaj |
3 |
0 |
0.000 |
3 |
0.000 |
|
O'Byrne |
0 |
0.021 |
0.000 |
3 |
0.000 |
Career goals to shots ratio is used instead of career shooting percentage is used for ease of calculation. It is the same as career shooting percentage divided by 100. Rookies and other young players do not have much of a track record to compare to, so their career average might widely differ from their true talent. Fortunately, except for P.K. Subban, their totals are not a large portion of the results, and his are not unusual for an offensive defensemen of his caliber.
Using this method we come up with a total expected goals of 226.865 compared to a real value of 207, and an expected goals per game ratio of 2.84 compared to 2.59. Here are the total differences between expected and actual goals for the entire team.
|
Player |
Expected goals |
Actual Goals |
Expected vs. Actual Differential |
|
Gionta |
31.900 |
26 |
-5.900 |
|
Plekanec |
25.530 |
21 |
-4.530 |
|
Cammaleri |
22.320 |
18 |
-4.320 |
|
Subban |
13.680 |
14 |
0.320 |
|
Kostitsyn |
23.188 |
20 |
-3.188 |
|
Gomez |
11.248 |
7 |
-4.248 |
|
Hamerlik |
6.708 |
5 |
-1.708 |
|
Pouliot |
16.637 |
13 |
-3.637 |
|
Pacioretty |
9.184 |
14 |
4.816 |
|
Moen |
7.029 |
6 |
-1.029 |
|
Darche |
7.830 |
12 |
4.170 |
|
Wisniewski |
4.067 |
7 |
2.933 |
|
Eller |
8.137 |
7 |
-1.137 |
|
Lapierre |
5.928 |
5 |
-0.928 |
|
Spacek |
3.900 |
1 |
-2.900 |
|
Pyatt |
2.275 |
2 |
-0.275 |
|
Halpern |
7.874 |
11 |
3.126 |
|
Gill |
1.984 |
2 |
0.016 |
|
Weber |
0.840 |
1 |
0.160 |
|
Desharnais |
7.722 |
8 |
0.278 |
|
Picard |
3.332 |
3 |
-0.332 |
|
White |
0.841 |
1 |
0.159 |
|
Gorges |
0.740 |
1 |
0.260 |
|
Markov |
1.540 |
1 |
-0.540 |
|
Mara |
0.882 |
0 |
-0.882 |
|
Boyd |
1.032 |
1 |
-0.032 |
|
Sopel |
0.184 |
0 |
-0.184 |
|
Dawes |
0.333 |
0 |
-0.333 |
|
Palushaj |
0.000 |
0 |
0.000 |
|
O'Byrne |
0.000 |
0 |
0.000 |
The difference between the actual and expected performance can be explained by the results of the top six shot producing forwards which correspond to the expected top two lines going into the season (Tomas Plekanec, Mike Cammaleri, Andrei Kostitsyn, Scott Gomez, Brian Gionta and Benoit Pouliot) witha total differential of -27.531 compared with the team’s -19.865.
With the new totals the Habs would have three players on pace for the vicinity of 30 goals and 5 above 20, which would be typical of a good offensive team without a dominant goal scoring forward. Gomez remains a poor goal scorer but is not quite as pathetic as this year and would likely have a higher assist total from increased linemate production, especially Gionta’s -5.9.
The biggest overproduction compared to career shooting percentage is from Max Pacioretty, Mathieu Darche and Jeff Halpern who all have an overproduction of above 4 goals this season. However Pacioretty had a very low career shooting percentage for an offensive forward due to low production in his two previous NHL seasons. His 12.5% this year is about average for an offensive forward and it is likely that he will be able to maintain it in his future career, thus his results should not be seen as particularly lucky.
Stats are from NHL.com after game 80 of the 2010-2011 season.
5 comments
|
0 recs |
Do you like this story?
Comments
Terrific post, I’ve moved it to the front page.
Follow me on twitter: http://twitter.com/andrewberkshire
by Andrew Berkshire on Apr 8, 2011 5:41 PM EDT reply actions
This is one of the most understated aspects of the Leafs second half.
They shot over 10% over the last 40 games, so to get a better gauge of where they will likely sit next season, an adjustment like this probably needs to be made.
Looking at their top players the only one significantly out preforming their shot totals for the entire year is Kulemin with 17.6 versus 13.6 (11.2 excluding this year) resulting in 23.6 goals over the year (19.4 if you use 11.2). The biggest concern for Toronto’s offense is the complete lack of scoring talent once you get past their top 4-5 forwards and the absense of strong offensive defensemen.
by Stephan Cooper on Apr 9, 2011 3:07 PM EDT up reply actions
I was talking about the second half of the season. Kulemin has had a torrid second half shooting 19%. Phaneuf who is a career 6% shot 16% since the All-Star break. As a team they have been shooting a high percentage which has lead to them playing .750 hockey.
None of us believes they are a 100+ point team, so the point was that if as a team they regress to the norm they will be between 80-90 points.
I doubt many people beyond the most rabid Leaf homers will take Toronto’s second half seriously, at this point its a meme that every year the Leafs play well after they’re effectively out of it.
As a prelude to the playoffs I ran the same analysis with the Bruins, they end up with a total result of 4.369 goals compared to expected, so they weren’t lucky compared with career shooting as individuals. Where they are lucky is in how few injuries they had, with none of their key players missing significant time, except for Savard whose absence was necessary for them to be cap compliant. As a result, a large proportion of their shots were taken by high average shooters having average or above average shooting years (i.e. Lucic, Horton, Bergeron and Ryder). As a result the Bruins over the course of the past season would expect to score 3.01 goals to the Habs’ 2.76. While this looks bad for Montreal, it is actually mostly an artifact of time missed. The Habs lineup going into the playoffs if added together would score 2.90 goals per game, which is basically a neglible difference to the Bruins over a 7 game series (.77 goals). With Pacioretty they’d expect 3.13. Markov’s sample size of 20 shots in 7 games is probably a bit too high to use, but with him and Paciorety the expected goals shoots to 3.33. Off course reduced ice time for other players when a player returns to the lineup means that the increase is likely not as strong as this (not to mention the team would be over the cap with both Markov and Wisniewski) but its nice to think that with average luck, few injuries and keeping most of the team intact the habs could be flirt with 3 goals a game team next year.
by Stephan Cooper on Apr 12, 2011 12:26 AM EDT up reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
Related FanPosts on Habs Eyes On The Prize
FanPosts
Community blog posts and discussion.
