David Legwand's Comparables Based Upon On-Ice Performance

Power play production is a fair criticism. But I never put much stock in such a small segment of the game. Nashville spent 5.2 minutes per game at 5-on-4, or less than 9% of the season. Add in 5-on-3 and 4-on-3 time, and you still only have 5.4 minutes per game with the man advantage.

Further, Legwand isn’t being paid to score. That’s almost certainly difficult to swallow – most people see $4.5 million and expect 65 points – but for a player like Legwand, expectations are defense. If he isn’t a power play guy – as Trotz shows – the answer is to up his PK time. You don’t need a point-producer to provide value at a $4.5 million cap hit.

….at even strength, penalty kill and power play. I recall that he had ungodly numbers on the PK.

Can you find that in the archives and post a link?

I tried, but its like finding a needle in a haystack.

haha, I just couldn’t stand the comparisons to Mike Richards. It’s not fair to Richards. :)

Nice counter-argument.

Thanks. It’s something I did for fun with Andreas Nodl earlier today.

I’m aware of the shooting percentage luck factor, which is why I included PDO. Now, that accounts for on-ice shooting percentages as opposed to individual shooting percentages, but I used it as a proxy for luck.

You’re right in that it affects the goals per 60, but this also affects Ryan Getzlaf. He’s a career 12.6% shooter, but this year it spiked to a career-high 16.2%. That difference increases his goals per 60 by about 20%.

I do tend to agree with you in that Legwand’s true talent is below his performance this year, but above his performance last year. Factor in his PK skills, and I think a fair assessment is that Legwand is fairly paid, if at the higher end of his pay-scale.

Use it in a sentence: That was a colossal reply failure.

Then yes, I’d object because “colossal” allows for a lot of room above “barely overpaid”.

I’d accept: “Legwand is paid fair value for what a player of his skills would achieve on the open market, which is known to overpay.”

When you come up with only 4 or 5 comparisons to Legwand, it seems to me that such comparisons would be flawed by weeding out players that really do have a similar profile to him, but are just better (or worse).

This is a valid criticism. The difference is that I purposefully made the criteria to be one with a small amount of comparables rather than one with a large amount. The one area I was most skeptical about, but decided to make based on success, was the CorsiRel cut.

What’s important to remember, though, is that this isn’t a “who is similar in profile to Legwand”, but more of a “Who performed similarly to Legwand each year, given similar situational use”. Surely a three-year average or something similar would work as well, almost certainly returning different people. This may even be a better, or more telling, method. As is, I avoided the extra – though simple – math.

4. It’s hard to believe that Legwand’s role on the Preds is similar to that of Getzlaf and Selanne to the Ducks. I think of them as being very different players. And they’re both grossly underpaid.

Selanne, yes. Getzlaf, not so much. Both players do bring more offense, but Getzlaf has seen his quality of opponents increase each of the last 4 years while he has spent a decreasing amount of time in the offensive zone. Through it all, he’s still scoring at a high rate and driving the play phenomenally well. I don’t think there’s any denying that Getzlaf is a better player than Legwand, but the two were remarkably similar this season.

Selanne was a big surprise for me personally, which is why it’s again important to note that I merely compared their use and production this year, not their playing styles in general.

How much do you think Rinne, Weber and Suter (especially Rinne, whose PK numbers are amazing if I recall previous posts about the topic correctly) influence Leggy’s PK numbers? That’s what makes me think we can do without him.

This is a phenomenal question. Unfortunately, the answer isn’t so clear. Recently, a study was done looking at what effect skaters have on goalie’s PK save percentages, and the answer was “minimal”. As Eric T. (commented above) wrote in response to that, “it’s a fair approximation to say for now that on the PK, the skaters need to stop shots from going on net and the goalie needs to save the ones that get through to him.”

What this means for Legwand? Not too good, actually. Opponents got a lot of shots off when he was on the ice this year and in 08-09, but not last year. So while he’s giving up more shots on the PK than his teammates, he’s giving up fewer goals. This is probably luck though.

6. The Preds’ payroll is 22nd in the NHL, ~$52M as opposed to Boston’s $66M per year cap. So "overpaid" as a descriptor for Leggy could be magnified by the team he is playing for.

This is actually something I was talking with Dirk about. I’m sure a lot of the criticism comes from Legwand’s draft expectations and him receiving the highest salary in team history. It’s all relative. And while it may be true that Legwand’s performance this year was worth $4.5 million on the open market, that doesn’t mean he is worth that much, nor that he is worth that much to the Predators. Simply, these are things I can’t account for or even begin to put in perspective. The Flyers tend to throw money around without thinking. It’s both nice and infuriating.

7. I wish that Don’tFeedtheBelak was here to make a counter-arguement. He would love the opportunity for sure.

Haha, as long as the counter-argument was as nicely put as yours. I’m all for counter-arguments, so long as they’re reasonable. And you hit on a lot of them. There are definitely other ways to look at this, and you could easily get different results. I just think looking at on-ice performance is the better starting point than salary, because you’re getting all sorts of players used in all sorts of roles just looking at salary.

I think he’s grossly overpaid and we could cut him and not miss a beat, but that’s just me.

Well, I respect that opinion, mostly because you backed it up. I think the same of Scott Hartnell, so I’d love if the Flyers pulled a 1-for-1 trade. haha

Thanks for responding.

What’s wrong with Corsi numbers? Outshooting your opposition is a good sign, no?

As far as putting up offensive numbers, I think it all depends what that means. But since Legwand only hit 20 goals once in the eight seasons prior to signing his contract, I don’t think that is a benchmark people should expect. He didn’t do it prior to the contract, so why expect it after? He’s also proven to be a 0.6 point-per-game player, which would result in 49 points over an 82 game season.

As far as $3.5 million goes, Matt Stajan, Matt Cullen, Matt Lombardi are the only three players to make $3.5 million. Cullen scores less, against weaker competition, with a worse Corsi. Stajan faces weaker competition, with more favorable zone starts, driving the play worse, but scores as much. Lombardi is Ville Leino.

Legwand is better than the three guys making $3.5 million, proving he’s worth more than $3.5 million.

Why exactly is that a reason not to believe in Corsi?

To me, believing in Corsi means believing that shot differential is a meaningful measure of who controlled the play. You can believe that and still believe that per-game calculations are more useful than per-60-minutes calculations.

And if by “Corsi”, you mean “all non-traditional stats in general”, then I’d point out that a number of stat-focused analysts said that they look closely at time on ice per game when judging a player, so you may not be in disagreement with the stat community as much as you think.

And finally, you’re wildly exaggerating the ice time Getzlaf gets. He’s never played 22 minutes per game in any season. He gets about 10% more ice time than Legwand, and he got paid about 15% more this year. Is that really a reason to dismiss Geoff’s claim that the comparable numbers of Getzlaf, Erat, and Selanne suggest that Legwand earned his pay this year?

… they are massaged into norms to compare players.

I don’t believe I did that at all. By separating out year to year performance is not a norm, it’s looking at the sample size of one NHL season.

My comparison? See what that player did per 82 games (with a minimum amount played of course, then compare)

This is what I did. I didn’t extrapolate it out to 82 games, but I took a minimum of 50 NHL games played, and took the numbers as per-60 rates. If you think a per-82 game comparison is better, go right ahead. Do the math and see if there’s a difference.

For a player who plays 25-26 minutes a game in Getzlaf to be compared to a 18-20 minute player in Legwand is laughable at best.

As Eric said, Getzlaf has never averaged even 22 minutes a game in his career, let alone 25 minutes. For their careers, Legwand averages 17:33 per game and Getzlaf 18:35 per game. I’d agree that a 25 minute player being compared to an 18 minute player is laughable, but that’s simply not the case here.

And I have no idea what any of this has to do with Corsi. Nothing in your comment even relates to what Corsi is – a plus/minus statistic of shots used to show how well a player controls the play. Maybe if you explained why you don’t believe in Corsi numbers better, we can understand, but as you left it, you haven’t criticized Corsi other than to suggest I did used it as a norm, which I didn’t do.

Did the guy produce to match his salary? No.

I proved this isn’t true. Repeating yourself without evidence contradicting an analysis doesn’t make it so. It proves you have an opinion that won’t be changed despite facts proving your opinion wrong.

You can really make numbers look like whatever you want them to look like (oh boy during a five minute stretch Legwand was the equivalent of Wayne Gretzky’s greatest 5 minute stretch EVAR!), but those two "stats" above are all that matter.

This is funny, because it shows just how closed-minded you are. Rather than debating the facts, you ignore the facts. Rather than arguing your point, you resort to sarcasm and hyperbole. The analysis that showed Legwand similar to Getzlaf was done after looking at 950 minutes. Rather than argue that one season’s worth of play is not enough to properly evaluate a player, you resort to comparing 950 minutes to 5, and Ryan Getzlaf to Wayne Gretzky.

You’ve proven it’s a lot easier to twist emotion and opinion into whatever you want to show than it is to twist stats to show what you want. Which is precisely why advanced stats are catching on: To remove the emotionally charged, biased statements made by those who refuse to believe anything but their memory.