(Noah Graham, Getty Images)

(Noah Graham, Getty Images)

Even as advanced statistics become more prevalent in hockey, it’s still rare to see them used in mainstream media, outside of some outliers like James Mirtle with the Globe and Mail. The doldrums of August can lead a hockey reporter off the beaten path in search of a story, however,  and Fluto Shinzawa of the Boston Globe turned to advanced statistics for a story on Sunday, interviewing Michael Schuckers, a statistics professor who does hockey analysis.

Shinzawa clearly asked Shuckers what was the oddest or most confusing free agent signing of the off-season according to his statistical analysis, which is a reasonable question to ask when looking for a story. What was odd was the answer. Shuckers skipped right past the usual punching bags from the past few months and went with Rob Scuderi, who returned to the Pittsburgh Penguins this off-season on a four-year deal.

According to Shuckers, his signing is “the one that sticks out to me this year” as the statistics show that Scuderi is “well past his prime”  and not worth what the Penguins are paying him. But is that really the case?

The main difficulty in evaluating Shuckers’ argument is that it’s difficult to separate his reasoning from that of Shinzawa. The only advanced statistic brought up in support of it is Corsi and it’s grossly misused. I sincerely hope that is because Shuckers was simplifying things for a broader audience or because Shinzawa oversimplified the argument.

Shinzawa accurately summarizes what Corsi measures (shot attempts) and what it is meant to represent (puck possession), but refers to it as “the gold standard” of hockey analytics, which unduly places it on a pedestal. He exacerbates this when he specifically talks about Scuderi’s Corsi from the 2012-13 season.

Last season, Scuderi’s Corsi (courtesy of behindthenet.ca) was 1.42. In comparison, Doughty’s Corsi was 14.84. Scuderi’s 2013 rating does not project high performance in relation to his generous $3.375 million annual payday.

This is the type of “analysis” that frustrates those on both sides of the advanced statistics debate. Using one number divorced from context makes those suspicious of advanced statistics in hockey respond with “You can make statistics say anything.” Meanwhile, the proponent of advanced statistics has to contend with the absurdity of directly comparing two completely different players who play in completely different situations.

Saying, “His Corsi was low, therefore he’s bad” is a gross and inaccurate generalization. A player tasked with shutting down top offensive players while starting most of his shifts in the defensive zone is going to have a lower Corsi rating than someone who starts most of his shifts in the offensive zone against weaker competition. That context is vitally important.

Directly comparing Scuderi and Doughty is doubly ridiculous as Doughty is one of the best puck possession defencemen in the entire NHL. Of defencemen who appeared in at least 30 games last season, Doughty had the 8th highest raw Corsi and played at least two minutes more per game and faced significantly tougher competition than all of those ahead of him. He started more in the defensive zone than all of those ahead of him but P.K. Subban, as well.

That Scuderi doesn’t compare favourably to one of the best defencemen in the league is barely a criticism.

It’s also not entirely fair to mention Scuderi’s 1.42 Corsi rating without mentioning that it is, in fact, a net positive. Corsi ratings can be both positive and negative, after all, and showing up on the positive side of the ledger indicates that the Kings outshot their opposition at even-strength when Scuderi was on the ice.

That this is Shinzawa’s interpretation rather than Schucker’s original analysis is supported by Schucker liking the Senators’ signing of Clarke MacArthur, as Shinzawa once again mentions his Corsi (minus-3.93) divorced from the context of it being the best rating on the Leafs last season.

I don’t really blame Shinzawa for this: he’s a reporter, not an analyst or a stats buff and I applaud him for making the effort. It’s just frustrating to see Corsi misused in such a way. I don’t expect anything too in-depth, but pointing out that Scuderi’s Corsi is lower than Doughty’s is disingenuous.

Disputing Shinzawa and Schuckers’ use of Corsi, however, doesn’t help answer whether Scuderi is, in fact, past his prime and overpaid, nor does it answer what the underlying advanced statistics have to say about Scuderi.

Let’s look at the context for Scuderi’s 1.42 Corsi and what it might indicate. Schucker seems to be suggesting that his Corsi itself indicates that Scuderi is past his prime, unless that is a misinterpretation by Shinzawa. In any case, that would be an odd thing to argue, as Scuderi has never posted particularly high Corsi ratings. This season and the season prior are the only two seasons of the last six that Scuderi’s even posted a positive Corsi.

This is mainly because Scuderi is used in a primarily defensive role: he started a higher percentage of his shifts in the defensive zone than any other Kings’ defenceman, excluding Robyn Regehr, whose numbers are skewed from playing in Buffalo for much of the season. Scuderi also faced the second highest quality of competition, behind only Doughty. Considering his usage, even having a positive Corsi could be considered impressive.

However, it’s useful to consider the larger context. The Kings were the best puck possession team in the league last season by Fenwick Close and only three Kings (heh) had a negative Corsi rating last season, with one of them being Regehr, who we can exclude on the basis of only playing 12 games with Los Angeles. Jordan Nolan and Matt Greene were the other two. The player directly ahead of them was Scuderi.

It’s worth questioning how much he was driving possession when he was on the ice and how much he benefited from playing for a strong possession team. Considering his role is more to throw hits and block shots, it’s likely the latter.

The statistics bear that out. Every single player on the Kings that Scuderi was on the ice with for at least 30 minutes last season when the score was close (within two goals) had a better Corsi rating away from Scuderi than they did with him. Part of that may be that when those players were on the ice with Scuderi, they were in a more defensive role, but it seems significant that the difference is so stark.

This wasn’t always the case for Scuderi: Relative Corsi compares a player’s Corsi when he is on the ice to that of his teammates when he is on the bench, showing how he performed in comparison to the rest of his team. Scuderi has had a negative Relative Corsi for years, but this past season is his worst by far. Does that indicate he’s past his prime and has nothing left to offer? It’s tough to say, given that last season was such an oddity with the lockout and the shortened schedule. 48 games is a decent sample size, but making judgements in comparison to full 82-game seasons seems ill-advised.

In any case, at 34-years-old, Scuderi is almost certainly either past his prime or reaching the end of it. A defenceman’s prime generally seems to extend from 25 to 32 or so, with some ageless wonders defying that pattern. Being past his prime doesn’t mean that Scuderi can’t be a good defenceman, by any means, though he may falter before reaching the end of his four-year contract. That term makes the deal somewhat questionable, but is the cap hit too high?

Scuderi was third on the Kings in ice time last season and it seems likely that he’ll slot in at third in ice time on the Penguins as well, at least for next season, behind Kris Letang and Paul Martin. How much do defenceman that play the third most minutes on their team generally get paid?

Thanks to Richard Pollock of Illegal Curve, we have an answer: approximately $3.12 million. That number, however, includes those on entry-level deals and contracts signed by restricted free agents, which means the average cap hit would be quite a bit higher if we only looked at contracts signed by unrestricted free agents. Scuderi’s cap hit of $3.375 million doesn’t seem the least bit unreasonable, then.

Over the course of his contract, Scuderi may slide down the depth chart to the bottom pairing, at which point his contract will be fairly high for his usage, but the salary cap is also expected to significantly rise over the next few seasons, meaning it should have negligible impact on the Penguins’ cap situation in the future.

Scuderi’s best days are in all likelihood behind him, but that judgement isn’t based on one number taken out of context or compared to one of the best in the business. While Scuderi’s underlying statistics do indicate he’s likely on the decline, that’s neither surprising — given his age — or damning — given he can still eat up minutes in a defensive role.

So, is he past his prime? Well, yeah. Of course he is. Is he overpaid? Not really. That’s about the going rate for a defenceman who plays that many minutes.

At the very least, it seems odd to single out his signing as being the biggest head-scratcher of the off-season.

Comments (12)

  1. “Considering his role is more to throw hits and block shots”

    30 hits in 48 GP. And it’s not scoring bias…

  2. From Boston, read the article Sunday and thought the same thing. Corsi needs to weighted to zone starts and situations (i.e. P-K vs. P-P time). Another argument against Corsi is that it doesn’t take into account where the shots are coming from. Some teams don’t care about shots against as long as they come from low percentage areas. A good defensemen will force a player into a low percentage area. It will negatively affect his Corsi but the coaching staff and goalie will appreciate it. You have to be able to combine both the analytics and what you actually see. The only real viable argument the article made that I agreed with is the term but you don’t know who else was offering him what and for how long.

  3. ” It’s just frustrating to see Corsi misused in such a way.”

    Pretty much exactly what I thought after reading the piece today.

  4. I like to nitpick, so bear with me… the one thing I wanted to point out is that you assume Scuderi will be a #3 D in Pittsburgh and therefor could be considered affordably within the average league range.

    However, you seem to have forgot that the Pens still have Brooks Orpik as their #3 D, which means Scuderi is slotting in at #4. And that is assuming they even use him as a Top 4 D, since prior to the signing both Bylsma and Shero had claimed that they intend to use Simon Despres as a Top 4 D this year paired with Kris Letang. That may no longer be the case, as I said it was before they had added Scuderi and most of the fan chatter I have seen assumes that Letang will slot in as his replacement for Doughty/Voynov that Scuderi was paired with in LA for the past few seasons.

    But just looking at last season, the Penguins had #1 Letang with 25:38 per game, #2 Martin with 25:19 per game, and #3 Orpik with 22:17 per game. Scuderi as the #3 D in LA was getting even less than that at 21:47 per game. And in the playoffs the Pens rely on their vets even more, with #1 Letang at 27:37 per game, #2 Martin at 26:38 per game, and #3 Orpik at 25:08 per game. Scuderi with the Kings had a #2 23:18 per game. So while adding Scuderi will allow them to spread out the ice time a little more evenly this year, it is still pretty clear that he will be coming in as a #4 D at best.

    So using the article you linked to as a reference, Scuderi as a #4/5 D should be closer to the $1.62M-$2.19M range, meaning that at $3.375M he is indeed overpaid. Of course Letang as the #1 D is underpaid (until next season when his extension kicks in), but he is already balanced out by the overpaid #2 D Martin. So it still makes the signing of Scuderi for that long and for that much money quite questionable, although I am still in favor of it despite being able to admit they overpaid.

  5. Very interesting quote:

    “It’s worth questioning how much he was driving possession when he was on the ice and how much he benefited from playing for a strong possession team. Considering his role is more to throw hits and block shots, it’s likely the latter.”

    It’s the notion of “driving possession” that I find interesting ie a player whose primary role is defence and retrieving the puck is not seen as “driving possession”, and any contribution he makes to retrieving the puck is never considered, and stops registering possession stats as soon as he changes off (which a defender tends to do after recovering the puck in his own zone).

    IOW basic, safe, smart plays, as well as defensive play in general, is entirely ignored by the so-called advanced stats.

    Scuderi is the last player on earth who should be judged by this.

    • “Basic, safe and smart plays” should, if they are actually smart, result in your team having possession. As does retrieving the puck in your zone. So yes, those things are counted in Corsi.

      • Exactly. My point there was that Scuderi wasn’t necessarily the one driving puck possession when he was on the ice for the Kings, ie. pushing the puck into and keeping it in the offensive zone, but his contributions should still be reflected in his Corsi rate.

  6. Isn’t is a little odd to say he’s worth the money based on the minutes he’s going to play, rather than how well he’s going to play in those minutes? If I have a below average third defender, I’d like to pay him below the average for third defenders.

  7. Also missed in the analysis is how desperate the Pens are for a dependable shutdown blueliner like Scuderi. Orpik might throw the highlight hits, but he finished 7th in PK ice time the last Penguin Cup. Scud lead the team in PK ice time and blocked shots that post season. Granted when the Kings won he finished a distant third in both those stats, but that is more a reflection of how different those two Cup winning teams were built than a sign Scuderi peaked in ’09.

    The Kings lost 91 out of 96 man games this past regular season to their top two penalty killing defencemen and that resulted in Scuderi leading the league in PK ice time until Regehr was added. That has got to affect your five on five play when you are carrying that heavy a load on the PK.

    I was sad to see him leave LA, especially with Mitchell’s health in doubt. I think in the short term it’s a great pick up for Pittsburgh and long term it just depends on whether Scud can play at the level he has been playing.

  8. I’d also argue that, Scuderi isn’t exactly a “physical” defender (he’s not known for his hard checks or crease clearing ability). And while he’s no Lidstrom, he plays the type of game where positioning, awareness, and smarts are his biggest strengths. He doesn’t throw a lot of checks, but his positioning and his stick work are both outstanding. I think he will slow down (already has), but I think he can still play much of the same game he’s been playing through the end of this contract.

  9. We are a bunch of volunteers and opening a brand new scheme in our community.
    Your site offered us with helpful information to work on.
    You’ve done a formidable task and our whole neighborhood
    might be grateful to you.

Leave a Reply

Your email address will not be published. Required fields are marked *