June 18 2013 02:39PM
Our pal Steve Dangle wrote a post on Monday afternoon about David Clarkson and why the Toronto Maple Leafs shouldn't sign him. His primary reason was that unrestricted free agents usually end up being disappointments. It's tough to disagree with that assertion. Generally, players peak in scoring between ages 24 and 27, not later, so signed UFAs on average are, well, disappointments.
But I want to get into Clarkson a little more as a commodity, and specifically, isolate his key characteristics—size and shots per game rate—and see how players that produced at his level from ages 26 to 28 produced ages 29 through 32.
Also, just to be clear before we get into it, I went into this bit of research firmly in the anti-David Clarkson on the Leafs camp, but mostly because I think Toronto has enough wingers to compete. I don't think their Top Six is too small and I think Jonathan Willis has proved that good teams aren't necessarily good because of size. Toronto's organizational strength lies in wingers.
However, I was open to the possibility that Clarkson's traits, particularly his 2.86 shots per game rate over his last three seasons, may make him more valuable perhaps than others.
My first step was to go to Hockey Reference and find comparable scorers that had similar scoring levels between ages 26 and 28 as Clarkson. I looked between the years of 2001 and 2009, with at least 200 games played, with a goals per game rate between .24 and .30. Clarkson in his years played 210 games and had a .27 goals per game rate.
The list had a few interesting names, and you can check it out here. Alexei Ponikarovsky and Darcy Tucker are already familiar with Leafs fans, a notable power forward in Ryan Malone, and former big UFA target Brad Richards and another upcoming one in Mike Ribeiro. Also, Scott Gomez.
Five of the players on that list all missed their age 29 season due to the 2004-2005 lockout (oddly, no other years were affected) and Mike Comrie left the NHL after age 30, but otherwise we're generally looking at four seasons. In the spirit of "an honest mess over a clean lie", I should point out that scoring rates are wonky in this period of time because in the 2006 and 2007 seasons the NHL called a zillion penalties, resulting in more goals. Hopefully, it balances out in the end and the only ridiculous number was Scott Gomez's 33-goal 2005-2006 campaign, although 24 of his goals that year were at even strength.
Here's a chart showing how those 15 players aged:
|Aged 26-28||Aged 29-32||Change|
Shot rate was unsurprisingly the most consistent talent, but players dropped across the board in goal-scoring, shots per game rate, shooting percentage, the amount of games played per season*, the number of minutes and average ice-time per game.
* - I omitted Mike Comrie's 31- and 32-year-old seasons completely, but the games played and minutes numbers drop even further if you count for the fact that Comrie retired after season-ending hip surgery at age 30.
In the case of Clarkson, or predicting his future, you have to account for the fact that he's going to lose some of his production in the coming years. If you sign him for four years, you're probably getting a player running at 91% capacity compared to the one we've seen for the last three seasons and has been very good in New Jersey.
But how do players who exhibited Clarkson's talents age?
Clarkson's shot rate was 2.86 between 26 and present day and at 6'1", 200 lbs, he's on the higher end size-wise when looking at all this list of players.
Shots Per Game Rate (SPG)
So how do players who take the most shots per game do? I looked to see how goals per game and shots per game changed between the two age division, for those under the average SPG rate of 2.18 and those above it:
|SPG > 2.18||SPG < 2.18|
The players that have the higher shots rate do lose more of their scoring touch than those players with lower. This could be a problem with sampling (almost assuredly is) but if players with really high SPG rates continued scoring, it would probably show up in the data somewhere.
Brad Richards, who was a 3.31 in the first segment, is probably going to be bought out, even though he's maintained an SPG of 3.20 in the last four seasons. Scott Gomez, while a 3.12 from ages 26 to 28, kept a 2.44 over his next four seasons, but fell off the map offensively and also was bought out. Mike Fisher went from a player who had four out of five seasons with 20 or more goals and has just one out of three since.
Maybe this is different for Clarkson, and I'd expect his career path to be different than Richards', Gomez' or Fisher's and it already is. But you're looking at two comparable players in both scoring and shooting, and neither was worth the contract they signed.
Shooting Percentage (SH%)
While Clarkson shot 13.2% in 2012 and had a 30-goal season, he's actually only at 9.5% in the last three seasons, which puts him on the lower rung of shooters in this sample. The mid-range of shooting percentage, again, was 12.5% so I looked at players above and below that to see how each group aged:
|SH% > 12.5%||SH% < 12.5%|
Like shots rate, the players in the upper echelon fell further than those in the lower level. Unlike shots rate, shooting percentage itself dipped as well. It occurs to me that I've never seen a study looking at age and shooting percentages.
The major point though is that even though Clarkson has a low shooting percentage compared to other players, it probably won't regress towards a higher mean. Rather, it will, but the mean appears to be dropping with age as players lose power off of their shots.
This is the kicker. At 6'1", 200, Clarkson is by no means the biggest rhinoceros in the shed, but he's a powerful player.
To create a "size index" I averaged out the height and mass of each player and weighted each component 50-50, then sorted them and took the seven "big" players: Ponikarovsky, Malone, Bonk, Fisher, Legwand, Tanguay and Prospal, and the seven "smaller" players to see how both groups would age. Since Scott Gomez was right in the middle, I left him out of both groups. Our "bigger" group averaged 6'2" and 207 lbs, while our "smaller" group was 5'11" and 186.
First, the "skill" components sorted by the bigger and smaller guys on the chart. Again, if bigger players are less likely to lose goals and shots, it should show up in the data:
…but it doesn't really. The smaller players actually had a slight bump in goals, shots rate and a lighter dip in shooting percentage. I don't think that this should be meant to say that "smaller is better" because this is a real small sample here, but it should be taken to mean "it doesn't look like size is an advantage".
I took a look at longevity as well, sorting games played per season*, penalty minutes per season, ice-time per season, and average ice-time per game to see if anything interesting popped up. If big players had a notably larger amount of game time, it may be worthwhile to note. I think longevity and endurance are very underrated qualities in a player, and it's best to get a player who can get you 50 points per 82 over a full 82-game season than a player who can get you 60 points per 82 over a season with a pair of separate injuries.
* - Since five of the players on my list played the 2013 season, I counted the 48-game season as 0.59 of a season rather than 1 full campaign as a way to not skew the numbers against the older players.
Anyway, here's the chart:
Again, tough to generate any conclusion because our sample is smaller. The smaller group lost games, but they also lost less ice-time, and if we're using penalty minutes as an indication of toughness, the smaller group actually recorded more of those during their older period.
But it's pretty close as to be pretty well inconclusive. If the smaller players are slightly better in their older age group than the bigger players, they lose it due to endurance, but the numbers aren't far enough apart that I'm convinced there's much of a difference.
David Clarkson will almost assuredly "fall off" between ages 29 and 32. That much is certain. The question is to what degree, and whether he's still worth paying top dollar for on the UFA market.
The problem is that when you get into free agent bidding wars, managers tend to put a lot of stock into what players have done rather than what they can be realistically expected to do. If you sign Clarkson for four years, you have to understand that you're going to get him at about 90% of what he is now and reflect that.
So, bringing back to Steve's point from Monday… the reason why so many UFAs seem to disappoint is because so many of them are signed in their late 20s, right when they're around the period of their careers where they stop being primary offensive contributors. That doesn't apply to every player necessarily, but it happens with astonishing regularly and yet every July, managers open their wallets for players on the wrong side of 27 with lavish contract proposals.
David Clarkson is a good hockey player, but he will probably be slightly less good next year, and if not next year, the year after, and again until his contract ends. Or maybe he's different, but his two notable qualities being shot rate and size, neither of those traits is shown to keep players at his level playing at a higher level for longer.
You know a big winger the Leafs have already? Nik Kulemin. When Clarkson was coming off his 26-year-old season, he had never had a 30-goal campaign (Kulemin has one). When Clarkson was 26, he had 52 NHL goals (Kulemin has 75). Clarkson had 100 NHL points (Kulemin has 175).
Rather than try and *get* David Clarkson, perhaps the Leafs could re-create him with a current piece they already have, locked up next year for a good deal.