On Wednesday night in Buffalo, the Philadelphia Flyers came out of the first period of their game against the Sabres down 3-1, the sixth time in their last eight games that they have trailed after 20 minutes. It was at that point that Claude Giroux put the entire team on his back, recording points on all four of the Flyers remaining goals, including scoring the game winner in overtime, his third gamewinner in as many games. When I say that he put the team on his back, I mean it: no other Flyer recorded more than a single point in that game.

Those four points vaulted him past Phil Kessel for the league lead in points and put him on pace for 113 points. This is pretty remarkable for a player whose career high in points, recorded just last season, is 76. Kessel, for his part, is on pace for 103 points, a big step up from last season’s career-high 64.

They’re not the only two surprising names among the top scorers. Jonathan Toews may have a Conn Smythe trophy, but his career high in points is 76. This season, he’s on pace for 94. Joffrey Lupul’s career high of 53 points came six seasons ago, but he’s also on pace for 94 points this season. Then there’s the Panthers’ top line of Stephen Weiss, Kris Versteeg, and Tomas Fleischmann, with career highs of 61, 53, and 51 points respectively, who are each on pace for point totals in the high 80s. Heck, Johan Franzen, who has never reached 60 points in his career is on a point-per-game pace.

It makes me wonder why these players are scoring so much. In what context are they scoring?

I am a blogger of modest means who is currently attending grad school. I can’t afford to watch every single game in the NHL, either in terms of time or money. I watch what I can and catch the highlights of everything else while trying to write papers and help take care of a newborn baby. That’s not the best way to do analysis. This is why advanced statistics are fantastic. They allow me to take a whole bunch of information compiled by people paying very close attention to every single game and use it to build a context.

Weiss, Versteeg, and Fleischmann have pleasantly surprised Panthers fans.

So I took the top 30 scorers in the NHL as of Wednesday night from NHL.com and took a look at four advanced statistics from Behindthenet.ca to provide some context to their performance. I will provide the complete chart at the bottom of the post for perusal, but I will highlight a few interesting results first.

The four advanced statistics I’ll be using are Offensive Zone Start Percentage (OZone%), Corsi Relative Quality of Competition (Corsi Rel QoC), Relative Corsi (Corsi Rel), and PDO, which doesn’t actually stand for anything for those of you who were wondering.

OZone% is the percentage of times that a player’s shifts start in the offensive zone as compared to the defensive zone, with neutral zone starts omitted. Two types of players tend to get higher percentages of offensive zone starts: fourth liners who are a complete liability in their own zone and first liners getting prime opportunities to score. Starting a shift in the offensive zone has obvious advantages for offensive-minded players. Alain Vigneault in Vancouver is the coach whose zone starts are the most skewed. Other than George Parros, Alex Burrows and the Sedins have the highest percentage of offensive zone starts, while four Canucks, including defensive centres Manny Malhotra and Maxim Lapierre have the lowest percentage in the NHL.

Corsi is a measurement of puck possession that is essentially a plus/minus of attempted shots. Relative Corsi compares the team’s Corsi rating when a player is on the ice to the team’s Corsi rating when that player is off the ice. Corsi Rel QoC takes the Corsi Rel ratings of the opposition’s players to give an indication of whether a player is facing offensive dynamos or pylons.

I covered PDO extensively in a previous post - it’s essentially a measurement of whether or not a player is getting the bounces – but I’d like to clarify one thing. PDO trends towards 1000 on a team level, but individual players will frequently have PDO numbers that are a little bit higher or lower based on their level of skill. Certain players tend to create opportunities for higher shooting percentages when they are on the ice. Once again, the Sedins are a good example as they frequently leave the shooter with a wide open net at the end of a passing play.

One final caveat is necessary: all four of these statistics are at even-strength only. A number of forwards on this list scored a significant number of their points on the powerplay, such as the Sedins, Giroux, Nugent-Hopkins, Backstrom, and Selanne.

Seriously, Claude Giroux is fantastic by any metric.

The first thing that jumps out at me is that Claude Giroux starts more often in the defensive zone than he does in the offensive zone. Giroux has started 47.1% of his shifts in the offensive zone. This makes his Corsi Rel and scoring pace particularly impressive: Giroux has been able to consistently push possession into the offensive zone while starting in his own end and he’s been able to turn those offensive zone possessions into goals for the Flyers. And it’s not because he’s been enjoying the bounces: his PDO is a perfectly normal 1005. While his personal shooting percentage is well above his career average – meaning he won’t continue scoring goals at his current pace – it seems entirely likely that he will in fact top 100 points this season.

The only player with a lower percentage of offensive zone starts in the top 30 is Ryan Smyth and it’s worth looking at him with his teammate Ryan Nugent-Hopkins. Smyth has started only 46.1% of his shifts in the offensive zone compared to defensive zone, while Nugent-Hopkins is behind only the Sedins for the top OZone% in the top 30 in league scoring. Smyth is also facing the second toughest competition in the top 30, while Nugent-Hopkins has faced the easiest competition, just ahead of Jordan Eberle.

Nugent-Hopkins and Eberle (along with Taylor Hall) are being extremely sheltered, while Smyth and Shawn Horcoff take the tough defensive zone starts and tougher competition. With that said, it’s hard to argue with the results for the Oilers. It’s been a while since they’ve had three players in the top-30 in scoring. Heck, it’s been a while since they had one. Smyth isn’t likely to stay in this elite company, however: his PDO of 1060 is unsustainable judging from previous years.

Speaking of unsustainable PDOs, let’s come back to the Florida Panthers’ top line. Cam Charron previously posted about how impressed he has been by the Panthers, and they are performing remarkably well as a team. Their individual numbers, however, are a bit concerning. Weiss, Versteeg, and Fleischmann have the second, third, and fourth highest PDO ratings among the top 30 scorers. Weiss and Fleischmann are also two of only seven players in the top 30 with negative Corsi Rel numbers, indicating that they aren’t that good at moving the puck up ice with possession.

Of the three, Weiss is the only one with a consistent history of a high PDO, suggesting that his performance might not fall off quite as much over the rest of the season, but it seems likely that Versteeg and Fleischmann won’t be able to keep up their scoring pace. Will the Panthers falter if their top line can’t produce at the same clip?

There are five players in the top 30 who have abnormally low PDO numbers, however, and will potentially perform even better throughout the rest of the season. Backstrom, Spezza, Datsyuk, Neal, and Kopitar all have PDO numbers below 990, with all but Kopitar falling below 980. Backstrom is already scoring at over a point-per-game, but has a minus-6 plus/minus thanks to a low on-ice save percentage. His on-ice shooting percentage is also lower than his previous seasons, and his scoring pace could pick up, particularly if Ovechkin gets his game back on track.

Spezza has been lauded for his improved two-way play, but he’s still seeing mostly offensive zone starts and has a minus-8 plus/minus because of some mediocre goaltending when he’s on the ice. Datsyuk and Kopitar have struggled at times this season, but some of those struggles have simply been because they’re not getting the bounces. Quite frankly, Kopitar is very impressive: his OZone% is only 48.1%, but he pushes possession into the offensive zone with great effectiveness, with a Corsi Rel of 14.0, fourth among players in the top 30. He does this while facing the toughest Corsi Rel QoC of any player in the top 30 in scoring. Kopitar is drastically underrated.

Is James Neal's success more surprising than that of Giroux or Kessel?

The most interesting name among those five is James Neal, as he’s not a player that is necessarily expected to be among the top scorers in the league. His career high in points is 55, but he’s currently on pace for 73 points despite having a surprisingly low PDO. With Crosby back in the lineup (at least, once he returns again to the lineup), Neal could potentially have a massive season.

Then there are the Sedins, who have the highest Corsi Rel in the top 30. In fact, it’s not even close. What also isn’t close, however, is their percentage of offensive zone starts, which greatly influences their Corsi Rel. This isn’t too surprising: the Sedins had similar numbers last season. What is surprising is that their scoring at even-strength isn’t anywhere near as high as might be expected from their underlying numbers. Daniel and Henrik have 18 and 17 points respectively at even-strength, with 15 and 13 points each on the powerplay helping them round out their point totals.

The leader in even-strength scoring is Phil Kessel, whose PDO has significantly regressed, indicating that he may in fact be able to continue at his current pace and finish near 100 points. The future is not so bright for Lupul, however: while his PDO is no longer unsustainably high, he has the worst Corsi Rel among the top 30 scorers. At some point, that will catch up to him. On the plus side, he has a healthy head start on setting a career-high in points.

Here is the complete chart of top 30 scorers with their advanced statistical context. Does anything stand out to you that I haven’t mentioned?

 

  Player Team Pos GP G A P +/- P/G Ozone% Corsi Rel QoC Corsi Rel PDO
1 Claude Giroux PHI R 26 16 20 36 6 1.38 47.1 0.668 2.7 1005
2 Phil Kessel TOR R 28 17 18 35 1 1.25 56.8 1.653 2.3 1021
3 Daniel Sedin VAN L 27 12 21 33 10 1.22 76.1 0.693 18.8 1039
4 Jonathan Toews CHI C 28 17 15 32 11 1.14 59.6 1.486 7.2 1011
5 Joffrey Lupul TOR R 28 13 19 32 3 1.14 56.0 1.647 -7.2 1021
6 Steven Stamkos TBL C 27 16 14 30 8 1.11 52.6 1.106 8.7 1052
7 Ryan Nugent-Hopkins EDM C 28 13 17 30 5 1.07 69.5 0.210 2.4 1039
8 Patrick Kane CHI R 28 8 22 30 11 1.07 63.6 1.220 4.9 1039
9 Henrik Sedin VAN C 27 8 22 30 11 1.11 78.6 0.728 18.2 1038
10 Stephen Weiss FLA C 27 11 18 29 15 1.07 50.2 0.938 -5.6 1075
11 Jordan Eberle EDM R 28 10 19 29 4 1.04 67.3 0.263 4.6 1041
12 Nicklas Backstrom WSH C 27 10 19 29 -6 1.07 52.7 0.735 8.6 971
13 Jason Pominville BUF R 27 9 20 29 1 1.07 63.0 0.882 -2.5 1038
14 Thomas Vanek BUF L 27 13 15 28 4 1.04 62.0 0.616 -4.5 1048
15 Kris Versteeg FLA R 26 12 16 28 14 1.08 51.3 0.756 4.8 1068
16 Tomas Fleischmann FLA L 27 12 16 28 15 1.04 52.3 0.896 -3.4 1080
17 Patrick Sharp CHI L 28 13 14 27 10 0.96 64.5 0.789 13.5 1017
18 Teemu Selanne ANA R 27 9 18 27 -1 1.00 50.5 0.489 11.9 1021
19 Johan Franzen DET R 26 13 13 26 8 1.00 58.5 1.533 -4.4 1034
20 Ryan Smyth EDM L 28 12 14 26 4 0.93 46.1 1.678 -4.4 1060
21 Marian Hossa CHI R 27 11 15 26 11 0.96 60.8 0.955 6.7 1020
22 Anze Kopitar LAK C 27 10 16 26 -4 0.96 48.1 1.820 14 986
23 James Neal PIT L 28 14 11 25 0 0.89 62.5 0.913 10 978
24 Tyler Seguin BOS C 25 13 12 25 19 1.00 53.8 1.588 11.6 1083
25 Jason Spezza OTT C 28 9 16 25 -8 0.89 60.3 1.130 11.1 974
26 Radim Vrbata PHX R 27 13 11 24 12 0.89 56.1 1.570 14.2 1047
27 Jeff Skinner CAR C 30 12 12 24 -6 0.80 50.2 0.560 1.6 994
28 Ray Whitney PHX L 27 10 14 24 11 0.89 55.0 1.239 11.3 1048
29 Evgeni Malkin PIT C 21 9 15 24 3 1.14 62.9 0.790 10.1 991
30 Pavel Datsyuk DET C 26 8 16 24 0 0.92 56.8 1.383 13.7 975

 

Comments (4)

  1. Great Post!

    I love when advanced metrics can be applied to player performance. I’ve been following the metrics movement in baseball a little more closely than any sport but only because that is the one sport that advanced metrics work well to explain performance.

    Baseball is the most static of the major sports with expected starts and stops to the play. There is very little fluid movement to baseball and rarely do the offensive performances of specific players dictate the future strategy of the defense against another player.

    For example, people don’t walk Josh Hamilton because Elvis Andrus is on second. They would walk Josh Hamilton regardless of who was on second.

    Getting back to hockey, the advanced metrics movement has hand a harder time solidifying itself in Basketball and Hockey because of two main reasons:

    1. The sport is dynamic and fluid which means plays don’t necessarily have a beginning and an end. The play can go into transition but the play doesn’t stop and allow defense to reset once the offense creates a turnover (like football). A player who is seemingly uninvolved directly in a play in the offensive zone may make a decision that, upon a quick turnover and transition, may put his team at a disadvantage on defense. I call this the Dion Phaneuf rule (as a Calgary Flames fan).

    Your post is awesome in recognizing certain things and praising certain players. The baseball movement has the BABIP stat which measures the batting average of balls that are put into the playing field (fouls that are caught as outs included). It’s almost a measure of how “lucky” a batter is. I think PDO is somewhat similar to that. Those players below the 1000 PDO are relatively unlucky with the bounces they are getting while above are relatively lucky. Mix this stat with the Ozone% stat and you can see how many hurdles some scorers are overcoming just to be in the top 30. I find it interesting that only 3 players in the top 30 are below 50% Ozone% (although it should be mentioned that Jeff Skinner, Teemu Selanne, and Steven Weiss are very close). Mix that with 3 of those players being at or below average in the PDO category and you can see that Giroux, Kopitar, and Skinner are all overcoming impressive odds to be in the top 30 in scoring.

  2. Trust me, the sabres defensemen did their damndest to make Giroux look good in that game. I think that Giroux is a phenomenal forward and absolutely capitalized on the turnovers, but just look at his game winning goal. I’m pretty sure all of his points came off of plays like that by the Sabres defensemen.

  3. Is there a way to compensate for penalty kill in the OZone%? Giroux plays like 2:30 per game on the penalty kill, which probably skews that statistic a bit to the lower end. Giroux is still a fantastic player though, as the stats show pretty well.

Leave a Reply

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