Phil Kessel vs. the NHL's Top Scorers - A Statistcal Analysis

by clrkaitken on Sep 20, 2009 11:11 PM EDT

Editor's Note: clrkaitken's been working on some interesting calculations about Phil Kessel and his comparables in the NHL. The numbers paint quite the picture.

Long story short...

In the aftermath discussion of the announced Kessel trade, "say plan the parade one more time" mused about how Phil Kessel's season stacked up against other top scorers in the league last year.

Over the past few hours, I've pored over the stats of the top scoreres from the 2008-09 season, and created a huge spreadsheet that served to try and accurately compare Kessel's season, apples to apples, to those of the league's top guns.

Seeing how it's past midnight, and I'm barely coherent from the double-barrelled shotgun of the Kessel trade reaction and doing this for the past two hours, I'll put this up for now, and then tomorrow I'll post my findings and go into detail about how our new boy stacks up. The comparative players might put a new spin on the trade for some people.

First off, allow me to apologize for sleep and real-life getting in the way of me finishing this. Hopefully much like the Kessel deal itself, this will be worth it once it finally is ready to go.

After compiling all of the relevant statistics from NHL.com, I've tried to determine who is a comparable offensive player to Phil Kessel. This was primarily based on three statistics:

G/A Ratio - A measure of the number of goals a player scored for every assist. For instance, Phil Kessel's G/A ratio was 1.5 last season. For every 3 goals, Kessel would register 2 assists.
GPG - A measure of the average number of goals a player scored in each game.
PPG - A measure of the average number of points a player scored in each game.

To be considered "comparable", a player needed to have a number that is within 25% of Kessel's statistics. 10 players had comparable numbers to Kessel for those three stats.

Jeff Carter, Eric Staal, Marian Hossa, Thomas Vanek, Dany Heatley, Patrick Marleau, Loui Erikkson, Johan Franzen, Jason Arnott, Alexander Frolov

First, let's put everyone on equal terms. Using the GPG and PPG for each individual, I've projected what we could expect their stats to look like if they all played 70 games, which is what Kessel played last season.

Kessel 70GP 34-26-60
Carter              39-33-72
Staal                34-30-64
Hossa             38-29-67
Vanek              38-23-61
Heatley            33-28-61
Marleau           35-30-65
Eriksson         31-23-54
Franzen           34-24-58
Arnott               36-25-61
Frolov              29-25-54

Based on 70 games, Kessel would have been outperformed by Carter, Staal, Hossa, and Marleau. He would have been superior to Eriksson and Frolov, and pretty much in-line with Vanek, Heatley, Franzen and Arnott However, these statistics aren't entirely representative, because almost everybody played a significantly higher number of minutes than Kessel did. Kessel played about 16 and a half minutes a night, while most played 18 or 19, and some up to 21 minutes a night.

Let's factor in another statistic that I calculated from the information available, ATOI/G. The calculation determines the approximate amount of time a player is on the ice before scoring a goal. For instance, based on 16:33 minutes a night, and 70 games, and 36 goals, Kessel scored a goal every 32:11 of ice time. This figure is significantly better than all of the comparitave players (except Hossa & Vanek).Let`s adjust everybody`s 70 game statistics, to simulate that they all played an average of 16:33 a night, statistics that are exactly identical to Kessel`s.

Kessel 70GP 34-26-60
Carter              32-26-58
Staal                27-23-50
Hossa             35-28-63
Vanek              37-22-59
Heatley            27-24-51
Marleau           27-24-51
Eriksson         26-19-45
Franzen           31-22-53
Arnott               31-23-54
Frolov              24-21-45

Now, all of a sudden Kessel's production seems very favourable. If you take away some of the additional minutes these guys played, suddenly their production starts to pale in comparison to Kessel, and he's right on line with guys like Jeff Carter, Marian Hossa, and Tomas Vanek, people who are considered elite goal scorers in the league.

Now here's where the naysayers start to talk about "the Savard effect"; primarily, the fact that Phil Kessel's production was largely dependent on the quality of his teammates. So let's look at the Quality of Competition and Quality of Teammate statistics from Behindthenet.ca for these players.

Kessel QOC 0.014 QOT 0.225
Carter             0.039           0.039
Staal               0.053          (0.037)
Hossa            0.023           0.481
Vanek             0.016           0.032
Heatley           0.039          0.003
Marleau          0.018          0.334
Eriksson        0.044         (0.115)
Franzen          0.057         0.235
Arnott             (0.002)        0.132
Frolov             0.037          0.013

So, there may be a little bit of truth to the effect that Boston's great squad had on his numbers.But two people's numbers jumped out at me; Marelau and Hossa.

Patrick Marleau had a greater QOT, and a pretty comparable QOC. Given the same amount of ice time, we would have expected him to score 10 points less than Kessel. Marleau was also 29 last year, and making $6.3M.

Marian Hossa, however, played against stiffer competition, but was surrounded by immense talent in Detroit. If we project similar ice time to Kessel, his production would likely be consistent with Kessel's. He turned 30 last season, and Chicago has seen fit to pay him $7.9 million for the next seven years (cap hit being $5.275M because of the 12 year contract).

I'm not suggesting this is in any way conclusive, and it makes a ton of assumptions. But I think what I've managed to demonstrate is that the raw numbers for a lot of these guys are masked by the fact that they log pretty high minutes for a forward. Playing far less minutes, Kessel was a significantly more effecient player than all of these "elite" scorers. The other thing is that most of these players are already in their prime, while one could argue that at 21, Kessel hasn't even reached his yet.

If Kessel can click with one of our good young centres (Grabovski, Kadri, hell, maybe even Stajan) and match his breakout season, I think that $5.4 million might be a downright bargain.

Below, is the spreadsheet I compiled to do my initial analysis.

Player	GP	G	A	Pts	G/A	GPG	PPG	ATOI	ATOI/G
Phil Kessel	70	36	24	60	1.50	0.514	0.857	16.55	32.18

Alex Ovechkin	79	56	54	110	1.04	0.709	1.392	23	32.45
Jeff Carter	82	46	38	84	1.21	0.561	1.024	20.57	36.67
Zach Parise	82	45	49	94	0.92	0.549	1.146	18.75	34.17
Ilya Kovalchuk	78	43	48	91	0.90	0.551	1.167	21.8	39.54
Rick Nash	78	40	39	79	1.03	0.513	1.013	21.15	41.24
Eric Staal	82	40	35	75	1.14	0.488	0.915	21.03	43.11
Marian Hossa	74	40	31	71	1.29	0.541	0.959	17.78	32.89
Thomas Vanek	73	40	24	64	1.67	0.548	0.877	17.18	31.35
Dany Heatley	82	39	33	72	1.18	0.476	0.878	20.1	42.26
Mike Cammalleri	81	39	43	82	0.91	0.481	1.012	17.55	36.45
Patrick Marleau	76	38	33	71	1.15	0.500	0.934	21.35	42.70
Loui Eriksson	82	36	27	63	1.33	0.439	0.768	19.82	45.15
Evgeni Malkin	82	35	78	113	0.45	0.427	1.378	22.52	52.76
Jarome Iginla	82	35	54	89	0.65	0.427	1.085	21.6	50.61
Jonathan Toews	82	34	35	69	0.97	0.415	0.841	18.62	44.91
Simon Gagne	79	34	40	74	0.85	0.430	0.937	19.02	44.19
Johan Franzen	71	34	25	59	1.36	0.479	0.831	18.1	37.80
Alex Semin	62	34	45	79	0.76	0.548	1.274	19.23	35.07
Brad Boyes	82	33	39	72	0.85	0.402	0.878	19.13	47.54
Sidney Crosby	77	33	70	103	0.47	0.429	1.338	21.93	51.17
Jason Arnott	65	33	24	57	1.38	0.508	0.877	18.9	37.23
Alexander Frolov	77	32	27	59	1.19	0.416	0.766	19.92	47.93
Pavel Datsyuk	81	32	65	97	0.49	0.395	1.198	19.2	48.60
Corey Perry	78	32	40	72	0.80	0.410	0.923	18.6	45.34
Jason Spezza	82	32	41	73	0.78	0.390	0.890	19.68	50.43
Shane Doan	82	31	42	73	0.74	0.378	0.890	20.23	53.51
Daniel Sedin	82	31	51	82	0.61	0.378	1.000	18.78	49.68
Patrick Elias	77	31	47	78	0.66	0.403	1.013	18.57	46.13

PensionPlanPuppets.com is a fan community that allows members to post their own thoughts and opinions on the Toronto Maple Leafs and hockey in general. These views and thoughts may not be shared by the editor of PensionPlanPuppets.com.

Tweet Comment 13 comments | 4 recs |

Do you like this story?

Comments

KESSEL #1

in everything

Except for cap friendly contract-ness

Boo hoo!

by Chris Stoikoff on Sep 19, 2009 12:26 AM EDT reply actions

E-mail me when you get it up because I’ll likely front-page it.

Pension Plan Puppets: A Toronto Maple Leafs blog and a group therapy session.

by PPP on Sep 19, 2009 9:36 AM EDT reply actions

Great stuff as usual. The only bone of contention I might throw in is in the ATOI figures with certain players. I looked at the group you selected as comparable to Kessel and figured out PPTOI/game and SHTOI/gm. I chose these stats because I feel that not all ice time is equal and that this may colour the overall analysis. I think my math is relatively accurate. On the other hand, some of my assumptions for choosing these stats and using them this way are based on little more than gut feeling. I was surprised by a some of the results.

PPTOI/g SHTOI/gm
Carter 2.45 3.01
Staal 4.16 1.30
Hossa 3.04 1.05
Vanek 3.51 0.35
Heatley 4.05 0.23
Marleau 3.54 2.33
Eriksson 3.00 1.57
Franzen 2.49 0.49
Arnott 4.21 0.10
Frolov 3.23 2.26
Kessel 2.23 0.14

While most players are in a semi-reasonable range of each other in PPTOI, there are far greater discrepancies in SHTOI. Kessel had the least PPTOI of the whole bunch, but also had almost the lowest PKTOI of the whole bunch.

Correct me if I am wrong in this assumption, but it seems to me that the scoring of points goes up for PPTOI at a lesser factor than it goes down for SHTOI. Scoring of points on the PP occurs about 20% of the time for a maximum of three guys on the ice. Scoring of points on the PK occurs maybe 1 or 2 % of the time? Less than 1%? 4%? [I really don’t know. You may well have this info close at hand.] The rarest of offensive points are scored shorthanded.

So a player with lots of PK time, when it is extremely unlikely he will score a point at all, should not have that ice time used to calculate ATOI/G or ATOI/P in the same way as a player who seldom kills penalties.

A couple of half-baked conclusions:

I was surprised Kessel had so little PP time (the least of all players on the list) and not surprised a bit he had so little PK. While his 14 seconds a game of PK vs, say, Carter’s 3.01 per game will skew the comparative points/ATOI for the two players, the good news is that he may well score far more PP goals if given PP time similar to others on the list.

Only Staal and the relatively unidimensional Arnott had less PPTOI than Heatley. And Heatley cried about it. Statistical evidence of even greater levels of suckdom.

Arnott really is not that good anymore. Avoid him like the plague in your hockey pool.

"I'd walk into the Leafs dressing room to get ready for the day and Harold would be there in his boxer shorts shaving. King Clancy would drop by a little later, play the fool, and then head off to the racetrack." John Brophy

by Mike Pelyk's Hairdo on Sep 20, 2009 7:17 PM EDT reply actions 2 recs

That’s a rec’n to both the article and this last comment. Good point, MPH.

by JP Nikota on Sep 20, 2009 11:29 PM EDT up reply actions

Seconded

And he actually pointed out something that I missed. Another factor that would impact the ATOI numbers and each players’ offensive efficiency is PP and PK time.

Thanks to MPH for coming up with those stats; looking at those numbers, I would definitely increase the numbers for Jeff Carter, because 3 of the 4 extra minutes he played over Kessel were on the PK, where his offensive output would be reduced.

Resident Capologist

by clrkaitken on Sep 20, 2009 11:48 PM EDT up reply actions