My post on how forward salaries have changed since the lockout, was lacking a key bit of information: the effect of age on things. In the comments, Tom Benjamin said:
Interesting stuff, Tyler, but I think the distortion created by later free agency makes it difficult to really draw many conclusions. Age was more important in than the quality of the player in determining slary under the pre-lockout CBA.
As I explained in my response to Tom, I sort of but not entirely agree with him on this point:
I agree with you to an extent. If you look at my list of the ten highest paid 1F in 03-04, it was as follows: Jaromir Jagr ($11MM), Sergei Fedorov ($10MM), Keith Tkachuk ($10MM), Joe Sakic ($9.88MM), Mike Modano ($9MM), Mats Sundin ($9MM), Bill Guerin ($8.86MM), Bobby Holik ($8.85MM), Doug Weight ($8.5MM) and Jarome Iginla ($7.5MM).
Lots of those guys started getting paid big bucks while still RFA age. Iginla was RFA age at the time. Jagr was making $10.4MM in 1999-00, when he would have been about 27. Fedorov got a $38MM/6 yr deal when he was 28 that paid him a guaranteed $14MM bonus, probably making it more like $7MM a year when the time value of money is taken into account. Tkachuk started making $8.3MM a year at 28. Sakic signed that 21MM/3yr deal with NYR in like 1997. Sundin was pulling down $7MM a year at 28. I note as well that Kariya would have been on that list but took a cut rate deal to go to COL – he was making big bucks from a young age under the old CBA.
That leaves Guerin, Weight, Holik and, to a lesser extent, Modano (although he was doing pretty well, $5MM-$6MM a year) who didn’t really hit big until UFA. Modano aside, those guys are lesser players than the ones who were doing really well at RFA age. I think it’s more sensible to say that the truly elite players were getting paid under the old CBA, generally by their mid to late 20s. It was the guys a step back who were being held back more. That seems to be rectifying itself now.
First thing’s first: you’ll recall that, in the last post, I took all NHL F who played at least 40 games in 2003-04 and all NHL F who played at least 40 games in 2011-12 and sorted them by their TOI/G. I labelled the ninety guys with the most TOI “first liner”, the next ninety “second liners” and so on. I’ve now gone through and counted how many of those guys fall into each of three age categories: 18-23, 24-30 and 31+.
An aside: In January of 2008, I wrote a post on NHL demographics. The key point from that was as follows:
The league proceeded to hit its lowest average age by 1981-82. At that point, the decrease in age stopped. If that’s all there was, I wouldn’t be looking at the Baby Boom quite so closely but the average age started to increase at that point – my suspicion is that it rose as the Baby Boomers worked their way through the league. In 18 out of the next 19 years, the league was older than it was in the preceding year. The 2006-07 year was noteworthy because the average age dropped from 27.90 to 27.79 years; the biggest drop in average age since the 1980-81 to 1981-82 season. My suspicion is that that drop will be exceeded this year.
While this data doesn’t prove anything with respect to age, it’s interesting to me that so many players aged 31+ have simply evaporated from the game. There were 26 forwards aged 18-23 who played on the top two lines in 2003-04 and 34 last year. The number of forwards aged 24-30 is basically unchanged: 100 and 102. The number of guys aged 31+ has plummeted – from 54 to 44. Probably worth following up on at some point.
That ends the vegetable portion of the post – the rest is much more fun. I divided each of the four “lines” of players into two groups, based on salary. The higher paid one I’ve labelled “a” and the lower paid ones “b”. In the table at left, you can see the breakdown. I’ve highlighted the more significant changes since 2003-04. There was only one 18-23 year old in the highest paid first liner group in 2003-04 (Marian Gaborik) – now there are four. The number of highly paid 31+ first liners has dropped significantly. The second liners are pretty interesting as well – the share of highly paid 31+ second liners has cratered, almost entirely to the benefit of the 24-30 class. With third liners, again, we see an explosion in the share of the highly paid ones who are between the ages of 24-30. The lower paid portion of the third line class has seen the number of 18-23 year olds in it double. Finally, you’ve again seen an explosion of 24-30 year olds in the higher paid class of fourth liners.
The final table there is simply how the players of each group were distributed amongst the various lines between 2003-04 and 2011-12. Things are pretty much unchanged in the 24-30 class and 31+ class. The most interesting thing to me is that in 2003-04, 64.4% of 18-23 year olds were third/fourth liners; that number’s now fallen to 55.8%, with a corresponding increase in the number of 18-23 year olds playing on the top two lines. This is a sort of different finding than the one referenced above about how the number of younger players in the league seems to be increasing – this suggests that there are more 18-23 year olds playing big roles now than there were in 2003-04. I’m not entirely sure why this is; it seems odd to think that the talent distribution in the 18-23 year old group has changed so dramatically now versus 2003-04.
Let’s talk dollars, which is the sexier part of all of this. First up, the first liners. As indicated above, I’ve created two groups, splitting my first line group into the highest and lowest paid halves. I’ve taken both averages and medians of those groups.
The last graph tells the real story here. For players who are in the upper half of first line salaries, only the 18-23 year olds have matched the growth in league wide revenues of about 5.8%. The growth in salaries for 24+ groups of the highest paid half of first liners has lagged. With the lower paid half of these players though, that trend is reversed. Keep in mind, when we’re talking about players aged 18-23 who are in the lower half of salaries for a given line, we’re basically talking about players on ELCs. The amount that a player could be paid pursuant to an ELC dropped under this CBA, which explains why they’ve seen such an decrease.
It’s the growth in salaries of the age 24+ players who are being paid in the lower half of first liners that I find fascinating. Keep in mind, I am basically comparing apples to apples to here, because I’ve created age classes. One of the effects of the CBA of the past few year has been to massively drive up the cost of players who were in the lower half of first liners in salary relative to players who are in the upper class. If you assume players who are more highly paid tend to be better than players who aren’t – and I think that this is a reasonable assumption to make – for some reason, the NHL has spent the past seven years throwing money at players who aren’t in the upper tier as far as performance.
It’s also, to me, noteworthy how salaries amongst players over the age of 24 have converged. Look at the first graph and notice how the salaries of the 24-30 and 31+ group are now almost the same. Brian Burke’s hit on this point before, the disappearance of the second contract but man. Basically, once you’re an RFA now, you seem to get paid like you’re a UFA, or close to it, when it comes to first liners.
Again, notice how the salary gap between the 24-30 group and 31+ group has crumbled. In 2003-04, a second liner in the upper half of salaries cost you $2MM while a 31+ guy cost you north of $3.5MM. Now? About $3.5MM and $4.05MM. There wasn’t nearly the salary edge for guys in the lower paid half of second liners in 2003-04 but still: the price difference between a 24-30 year old second liner in the lower half of salaries and a 31+ year old second liner of the same type is now awfully small – about $200K.
The final graph shows, again, that guys under 30 have enjoyed tremendous salary growth once they get out of the ELC system. The salaries of 31+ guys in the upper half of salaries have barely grown, while the salaries of 31+ year olds in the lower half of salaries have grown north of the league revenue growth rate. Similarly, although both halves of players aged 24-30 have enjoyed significant salary growth, salaries have grown more quickly for the guys in the bottom half. Again, it sort of seems like the growth in salaries has gone to the players who are the lesser players of a specific class.
Have a look at the median and average salaries of the 18-23 year old and 24-30 year old group of third liners being paid in the top half of salaries. I’d kind of mused before, back when Gilbert Brule was signed, that it was getting increasingly difficult to see where the surplus value in contracts for RFA third liners was. Basically, the thinking with younger players is that the team that develops them gets a deal on the price – if they’re cheaper than the equivalent player who’s UFA age, the team that develops the player benefits.
This seems to have completely disappeared with guys in the top salary half of third liners. Assuming, for the sake of discussion that these are roughly equivalent groups of players (and by sorting by ice time, I should have accomplished that), there’s now a premium that teams are paying in order to have players who they developed. I thought it was worth looking into this in a little more detail, so I put together a graph of the salaries being earned by players in the two age groups at every tenth percentile for third liners in the upper half of salaries.
You can see that the 31+ guys are, in general, more tightly grouped. I suspect that what’s happened over the past few years is that teams are paying young guys based on what they perceive as potential. The fact that they cost so much more than UFA third liners suggests that, on average, teams aren’t winning these bets. Maybe they’re getting some payoff with some cheaper second years down the road but, looking at the second liners, you see an awfully small difference between what the 24-30 and 31+ guys are getting paid.
Also keep in mind that the NHL incurs pretty big development costs to produce these 24-30 year old players, frequently by the team that’s still paying them. In a sensible world, you’d be getting that investment in player production back in the form of cheap talent in the coming years. In the pre-2005 days, you could make an argument that that was happening. I’m not so sure that you can make that argument any longer.
Aside: one wonders if the NHL wouldn’t be better off having a two round draft, limiting teams to 30 NHL contracts and letting leagues like the AHL develop talent that NHL teams can then buy from them, if the player is under contract, or sign as free agents when the player is no longer under contract. Few top two line guys spend much time, if any, in the AHL and you really have to wonder about the economic sense of having an incredibly expensive infrastructure to develop third and fourth line players.
When you look at the compounded annual salary growth, you see about what you’d expect: guys under thirty in the top half of the salary structure for third liners have gotten massively more expensive, while the price of guys playing third line minutes has fallen dramatically. 31+ guys in the lower tier have seen a much bigger salary increase than 31+ guys in the upper tier, whose salaries have gone backwards.
The fourth line’s a lot less interesting – what you’re seeing with the salary declines for those in the upper half of the 18-23 year old group is basically the impact of a lower rookie salary cap. The 24-30 group, you can see the impact of a higher minimum salary. The 31+ guys in the upper half are taking it in the neck. Nobody’s enjoyed salary growth anywhere near the growth in league revenues. Even in the NHL, in which it hasn’t been the top 1% raking in the growth in wealth, those at the bottom are still being gradually left behind.
What does it all mean?
1. You can’t refer to “the players” monolithically. I know that I harp on this but some players did quite well as a result of how the NHL’s salary structure changed in the last lockout. Some did quite poorly. If you’re a 23+ guy who would have been a cheap first line option had you played in 2003-04, you are now a far, far wealthier man that you might otherwise have been. Guys like Matt Cullen, Chris Kunitz and Mike Fisher should celebrate the “loss” in negotiations in 2003-04. (Point: it’s hard to know exactly which of these players would have benefited; some of them might have won the lottery of UFA and been second liners on better teams, with fatter contracts.)
2. The NHL has spent the past seven years rewarding the lesser players on each line. If you go through and compare, you see that the guys in the lower half of salaries for first and second liners aged 24-30 and all the way through for guys aged 31+ have had larger percentage increases in their salaries than the guys in the upper half of salaries for players 24+. It’s one of the curious absurdities of what the NHL has done in locking in a players’ share of salaries and circumscribing teams within a tight payroll range – crappy teams that used to be cheap have now become more expensive and, as a result, crappy players who used to be cheap have increased in cost more than good players. It’s a bizarre way to pay people, a consequence of having a sort of Soviet structure.
Note: this does not mean that paying lesser players much less is a way to “fix” the NHL’s “problems” – this is a consequence of a system where good players have decided that they value the chance to play with good players more than they value another couple of million dollars in annual salary. Teams like Edmonton thought that the salary cap meant that they would get a shot at the game’s superstars; what it actually meant was that they would get a shot at paying Dustin Penner or Michael Nylander far more than they would have paid him in 2003-04.
3. Roster spots have kind of become commoditized. In essence, teams are much closer to just paying players now based on where they play in the lineup. Whereas before, there was some allowance for the skill of the player, that’s been reduced. If you’re a bad first or second liner, the emphasis is now more on the fact that you’re a first or second liner than it is on you being bad.
4. Jordan Eberle is paying paid like he’s part of the NHL elite. That’s what he has to be for that contract to make sense. All this talk of Eberle getting going rate for a 60-70 point guy is nuts. He’s being paid like an elite first liner. If he’s a second line RW in two years, it’s a poor use of resources. There is a massive gulf between what an upper tier first liner is making and what an upper tier guy in a second line spot makes. It is more than a little frustrating to hear that sort of waved away – the Oilers are a team that has pissed away money on guys for years, always giving a bit too much to people. Not a problem that looks like being solved.
5. Teams should possibly be walking away from third liners more often. You can’t have everyone give up on developing third liners but with what the ones under a team’s control cost now, it’s hard to see how investing resources in developing them makes a lot of sense. I never understood the criticism that the Oilers took for not having a farm team and I’m not sure what the economic case for one would be now, at least when it comes to forwards. We’ll see what the defence numbers look like – maybe that’s where teams are making their money back.
6. UFA wins are getting cheaper, RFA wins more expensive. This is a sort of corollary to the point made above but it’s important, so I’ll harp on it. When a team signs a player to a contract, what they’re doing is buying wins (Desjardins explains in more detail here. Historically, it was cheaper to buy wins from RFA than it was from UFA and many analytical types will assume that an RFA contract contains a healthy price discount. I’m not so sure that that’s a sensible assumption to make any more. Here are 24-30 median/average salaries as a percentage in 2003-04 and 2011-12, as a percentage of the 31+ salaries paid to players in that same salary class (1a = highest paid half first liners, 1b=lowest paid half of first liners, etc.)
Trend’s obvious. Your second and third line used to be a place where you could historically generate a real competitive advantange by developing those players in house – they were just ridiculously cheap compared to players aged 31+. That advantage has either narrowed significantly or disappeared. From an analytical perspective, this probably means that we should cut way back on the salary savings we’re assuming are generated by an RFA over a UFA.Email Tyler Dellow at firstname.lastname@example.org