Wednesday, December 14, 2011

Sixth Defenseman Shuffle

[Further to the post Get There From Here, The Theory will be doing an irregular series on methods of general management. The goal of these pieces is not to focus on the specific details of individual moves and decisions, but rather to use specific cases as a jumping-off point for the analysis of the principles and values by which managers operate. Eventually, we would like to get to the evaluation of those principles against one another, but honestly I’m not sure we’re ever going to get that far. It is my most fervent of hopes to be able to have an ongoing discussion about these ideas, so please, if you have a comment in your head, take the time to type it out.]

I wasn’t planning on dealing with trades until further along in this project, because the trade is a very sophisticated and complex piece of general management technique. A trade usually involves numerous contending factors, each of which is itself a massively intricate subject also comprised of myriad smaller contending factors, which in turn are comprised of many other complicated things, and frankly just thinking about the whole mess is enough to make a girl want to give up on hockey and take up alcoholism as a pastime. Trades for picks require an understanding of draft efficacy; trades for prospects raise questions about development trajectories and organizational needs; trades involving multiple players at multiple positions involve issues of tactical structure and the apples-to-oranges frustrations of comparing a second-line winger and a fourth-line center against a first-pairing defenseman and a goalie prospect. This complexity is what makes it so hard for even the most expert of experts to correctly identify the eventual ‘winner’ of a trade at the moment of trading- I would guess that most hockey-pundits’ record on trade evaluation is only slightly better than 50/50, with more outliers to the worse than the better. Before one can speak with authority on any complex multiplayer trade, one has to know pretty much everything else about all the other elements of GMing, and that is more than most of us can rightly claim to know.

However, the Kaberle/Spacek deal is actually a terrific starting point for this work because it’s about as simple as a trade can be: a one-for-one exchange of two players occupying the same role. There are basically only four questions this trade raises, and two of them can be set aside for the time being. This leaves only two points of contention, which is a good number of points of contention to handle in one blog post.

Firstly, so that we’re all on the same page, let’s look at the players concerned at the time of the trade. I don’t love the comparison because Spacek has played less than half the games, and as I’ve said before I think a guy deserves 20 before folk start passing judgment, but sometimes it can’t be helped:




With all due respect to Jaro Spacek as a hockey player and a human being, this deal was all about Kaberle: the Canes wanted to get rid of him and the Habs wanted him. Spacek being sent the other way is mainly a matter of convenience and bodies and the fact that a trade is not a trade without at least one thing going each way. Presumably what the Canes like about Spacek is that he’s a UFA at the end of this season and less than three million until then. What the Canadiens like about Kaberle … is a bit more complicated.

It is possible that Gauthier believes Kaberle is going to get better. For many years, in Toronto, he was considered a most excellent defenseman, and he might again recover that power. We do not deny this possibility, but neither will we speculate on it, because right now there is no way to know, and debates based on speculation just leave everybody involved more deeply entrenched in their preexisting positions. There are real questions to be asked about the fluctuations in players’ performance throughout a career, why they do better with this team or worse with that one and how to know in advance which is more likely, but that is a huge issue for a different day, so for now, we will just proceed under the assumption that Kaberle is what he is and will not play dramatically better or worse for the Canadiens than he did for the Hurricanes.

It is also possible that Gauthier believes that the new CBA will somehow absolve him of the burden of Kaberle’s contract by mitigating either the value or the term, and this is the other thing we will not speculate about. We don’t know what the new CBA will or won’t be, and assuming it will be in some way convenient for the Habs is creating pseudo-facts in support of a position. Having been raised by a very nice man with an unhealthy interest in cryptozoology, we have taken a rather dim view of pseudo-facts, so therefore we are going to assume that the Habs are now on the hook for all of Kaberle’s remaining time and money unless they offload him to another team or demote him to Hamilton, according to the terms of the current CBA.

So leaving aside those two issues, here are the questions this deal raises:

Under what circumstances is it good general management to overpay for a player?

One of the first principles of general management under a salary cap structure has to be getting value for money. If one can’t spend infinitely, one better spend intelligently, or face the unenviable prospect of ending one’s seasons with neither wins nor dollars. Therefore, we might say that, as a general rule, a good GM is one who avoids overpayment.

However, it is not always possible. Top-tier players are well-identified and highly-paid, and sometimes the arrangement of the market at the moment they go UFA means that they can ask more than their worth and get it. Gomez was one such UFA- a good player who was able to get far more than his worth- and although he does not now nor will ever really earn his contract, he’s still good enough to have substantive value. He’s good in a way that the Canadiens have not had much success acquiring at any price, and as such, the sacrifice they had to make for him was his bloated contract. Sometimes a good GM will absorb an overpayment to get a needful player.

But overpaying Kaberle is not the same as overpaying Gomez. Gomez-as-he-is is a solid two-way second-line center who contributes on both kinds of special teams, stays positive in all the underlying possession metrics, and applies offensive pressure even though, yes, he hasn’t gotten himself a goal in forever. It would be very difficult to get another Gomez without sacrificing real asset players in addition to money, and even then, a man might wait a long time for one to become available. Kaberle-as-he-is, though, is a bottom-pairing defenseman and power play specialist, which is not all that scarce or essential of a role. The way he plays at evens, you could get that for less than $1 million. You could get it out of a newly minted prospect with a higher potential ceiling and an ELC salary. Okay, the PP production is important, but even so, powerplay specialists are not exactly the orchids of the hockey world. I would think with a little nosing and a little digging, Gauthier could have found a piece equivalent to Kaberle-as-he-is for half the price of Kaberle.

All of this would make no matter, except overpaying for players is something a GM can only do so many times, and it seems logical to save those times for positive players. If it is worth it to pay Kaberle (approximately) 330% more than Spacek because he’s (generously approximately) 15% better, then why not do the same for ALL one’s bottom-ranked players? One could easily find a player 15% better than Darche if one was willing to spend $2.5 million on the position. Same for Diaz, same for Nokelainen, same for Emelin. But overpaying for marginal improvements at perennially weak positions will eventually lead to madness, or at least spending to the cap for a mediocre team- which, lest we forget, is exactly the hole the Habs are trying to get out of.

The Canadiens’ power-play struggles were in part structural but also partially luck-driven. Their PP SV% against was unsustainable and bound to fall eventually, even with no changes whatsoever to the team. Kaberle helps, but once he’s made his difference and the regression has set in, the power play will no longer be the crisis of the moment and the Canadiens will still be paying $4.5 million for two more years to a guy who comes out negative at ES while taking soft zonestarts against soft opposition. Is this really the guy one wants to use up an overpayment opportunity on? Without assuming some future massive change in either his performance or his contract, is there any way this is a good long-term decision? I don’t think so, but I’m open to countervailing views. In the meantime, on to another question:

Is it ever good general management to make a move for symbolic/psychological reasons?

In his press conference after the trade, Pierre Gauthier explained the move in part by asserting that he had to do something to help his players. The Canadiens’ power play troubles had become a major talking point for the hockey media and a flashpoint for fan hostility. Power play units had been shuffled and reorganized a dozen different ways, often pressing players into positions outside their comfort zone. Early in the season there had been plenty of great power play shifts that simply failed to capitalize, and the Habs were actually getting more shots on goal with the man advantage than League average, but as the sense of futility lengthened and deepened, it is true that their PPs were starting to look weaker and weaker.

Gauthier asserted that he could see the players becoming frustrated and demoralized, and I see no reason to disbelieve him. It was a frustrating and demoralizing situation. But there is no doubt that it was also in large part an issue of bad luck, and one would like to think that a GM is not going to take on nearly ten million in additional salary just to shake off bad mojo.

However, choking is a real problem and it happens to professional athletes far more often than anyone wants to admit. Runs of bad luck are particularly likely to trigger it, and many of the common reactions to bad luck- additional pressure from coaches, rearranging of customary roles, harsher self-evaluation- are pretty much the definition of choke-inducing events. Bad results, paradoxically, can be the cause of bad performance. It’s an easy vicious cycle to fall into and a hard one to crawl out of.

There are, then, two things here that help the Habs’ PP: Kaberle and the idea of Kaberle. The Kaberle placebo, if you will. Bringing in a long-standing and even famed PP specialist- more than that, an expensive PP specialist- is pretty much the perfect gesture to counteract a choking cycle: it shows the players that the GM doesn’t blame them for the trouble, and gives them another person to offload the pressure onto.

Considering this, Kaberle is still a long-term solution to a short-term problem- he’s still gonna be there when this shitty PP run is just a dim memory- but I must confess that I’d never really considered before that the GM might occasionally have to make symbolic moves in order to create the best possible working conditions for the players in whom he is really, deeply invested- your Subbans and Plekaneces and Prices. If Gauthier sees an ugly situation developing that’s hurting the performance of players individually and/or their ability to work well together, does he have a duty to do something about it, even if that something is not necessarily the shrewdest dollars-and-cents-and-years decision? In short, does a GM have to help his players or does his responsibility end at the composition of the roster?

4 comments:

wychwood said...

That is a great question! I don't know that that's the GM's job as such, but it definitely seems like a factor that is likely to be relevant when the GM is making decisions.

Compare with trades made because the room has gone bad? The whole Dry Island Thing (tm), for instance. That's trades not being made for the sake of the return, nor for pure quality level / salary cap / contract duration reasons, but again, for the balance of the team and the interactions of the human beings that make up the mystical entity that is The Habs or The Flyers.

E said...

see, historically i've been really skeptical of the idea that the gm ought to care too much about personality factors and interpersonal relationships, just because i think that's one of those workplace problems that tend to get worse and worse if indulged. if you don't want your players to act like thirteen-year-old girls with their cliques and bffs and whatnot, then don't acknowledge those type of behaviors. but, on the other hand, the kind of lives that hockey players lead through their childhoods and teenage years probably leads to a lot of socially maladjusted people, so perhaps a gm has to do a little bit of daddying from time to time. i dunno. but i thought this case is interesting because paradoxical performance effects are such a well-documented psychological phenomenon, which puts this issue a little bit ahead of 'bad room' insofar as tangibility goes.

Doogie2K said...

Presumably what the Canes like about Spacek is that he’s a UFA at the end of this season and less than three million until then.

Also, I suspect they may have seen him good in the Finals in '06. (I didn't think he was a world-beater even then, but Chris Pronger made Marc-Andre Bergeron look like a top-pairing D that season, FFS.) A few years ago, someone (Matt Fenwick?) went through and looked at a bunch of weird trades and found that the players acquired were guys who'd played very well against the team that acquired them. Sometimes it works out (Glencross in Calgary springs to mind), but most of the time, they just kind of are what they are. Call it a form of sampling bias.

Having been raised by a very nice man with an unhealthy interest in cryptozoology

Ha ha, that's awesome. Any particular Ogopogo "sightings" you want to share? (Or World Champion Cubs, or whatever mythical creature lurks in the streets of Chicago.)

It would be very difficult to get another Gomez without sacrificing real asset players in addition to money

You mean like Ryan MacDonagh and Chris Higgins? ;)

I would think with a little nosing and a little digging, Gauthier could have found a piece equivalent to Kaberle-as-he-is for half the price of Kaberle.

I don't think Steve Yzerman is willing to part with Marc-Andre Bergeron at this point, though. (Oh, damn, two MAB references in one comment. Go me.)

However, choking is a real problem and it happens to professional athletes far more often than anyone wants to admit. Runs of bad luck are particularly likely to trigger it, and many of the common reactions to bad luck- additional pressure from coaches, rearranging of customary roles, harsher self-evaluation- are pretty much the definition of choke-inducing events. Bad results, paradoxically, can be the cause of bad performance. It’s an easy vicious cycle to fall into and a hard one to crawl out of.

I've long believed this, and been yelled at by a few stats bloggers along the way for it along the way, but fuck it, it makes a lot more sense than the clattering of dice alone as an explanation. Professional athletes are generally mentally tougher than the average person, but they're still human, and given that the stakes are much higher for them than for the average person, it probably mostly evens out. That the vagaries of the human mind, in concert with bad bounces, inconsistent reffing, inconsistent physiological execution, etc., all seem to cancel out for the most part over the course of a season is immaterial: in the small samples, all that stuff matters. Just because you can't predict or account for it doesn't mean it isn't real.

Besides, V&T's found evidence of streakiness, which would seem to support this notion, and score effects are nothing if not psychological. So I think this is slowly coming to be a more-accepted notion: yes, shit happens, but shit can happen for a really fucking long time, and contribute to stuff like Scott Gomez's goal-scoring "prowess," or the Minnesota Wild's stubborn refusal to lose like the horrific EV team they are. Sure, eventually, it'll fall off, but when? No one knows.

So indirectly, I guess that's my answer to your ultimate question. It's not something I'd do as a GM unless literally everything else had failed, and only if the deal made at least half-assed sense, but maybe well-timed reinforcements can snap the funk sooner than chance alone would provide. Nothing's guaranteed, but like firing your coach, sometimes it's a gamble that needs to be taken, rational or not.

Doogie2K said...

And to be clear, I understand the disdain for using psychological explanations for some of these things: it goes back to the narrative thing from the previous post. It's an easy narrative, and too often a crutch for those who don't want to dig a little deeper and see what's really going on. But there are certain phenomena, like streakiness, that go well beyond what one might reasonably expect from chance alone, and I think it's fair to invoke psychological explanations, with the caveat that a) they're never the only thing going on, and b) they don't preclude eventual regression to the mean; getting "on a roll" or "in a funk" can merely delay the inevitable, for both primary (doesn't play as well/plays better, because the brain is connected to the rest of the body) and secondary (coach/teammates don't trust him/trust him more) reasons.