A “meaningful gulf” in the quality of centremen

I’ve been blogging about the Leafs for a short time, and after watching maybe about 40 games last season and looking at all the data available, there are fewer things on this world I am more sure about than “Mikhail Grabovski is the #1 centreman on this team”.

This doesn’t mean the Leafs need to look at more options to get a better second-line centreman. In fact, that ought to be their second priority behind another top-three defenceman. Unfortunately, with so many quality players locked into long contracts with their current clubs, pickings are slim.

The words “meaningful gulf” are equated to the divide between the NHL and the NHLPA during their lockout talks. After every round of meetings, Donald Fehr and Gary Bettman come out of the lroom and explain to a huddled mass of reporters that the other side has cancelled Christmas. From a Leafs perspective, “meaningful gulf” explains the difference in quality between Mikhail Grabovski and Tyler Bozak.

Advertisement - Continue Reading Below

It’s not just offensive numbers, either. A lot of discussion on newspaper columns or sports talk radio involves important things like “goals” and “points”, but they only tell the story of how a player did at one end of the ice. Half of hockey is offence, and the other half is defence. Saving a goal at one end is equally important than scoring a goal at the other.

Yet in baseball and basketball, statistics are used to provide a picture of only a player’s offensive contribution. You talk about batting averages and points per game and rebounds and little about what happens on the other side of the ball. Football is a bit difference because there are players who are brought in entirely to play defence, so you look at tackles and pass deflections and these minute numbers that, without context, mean very little.

On this blog, I like to talk a lot about a player’s “Corsi” rate, primarily in percentages. Puck-possession is an important factor since when a team has the puck, it guarantees two things:

  1. The team has a chance to score on offence
  2. The team is protecting the puck from the defence

The way Corsi is calculated is either like +/- over a 60-minute period (ex: the Leafs were -6.2 shots per 60 minutes of play with so-and-so on the ice) or by percentage, which I think does a better job at explaining the possession aspect. Hockey isn’t a game of numbers, it’s a game of ratios. You win when you score more than 50% of the goals, period.

Advertisement - Continue Reading Below

There are three websites we can use to learn of a player’s puck-possession ability, and all of them express the number in a different way. This one here, Hockey Analysis by hockey blogger and frequent commenter David Johnson, is an excellent resource because it shows us how a player performed both with and without a certain player on the ice.

So the differences between Mikhail Grabovski and Tyler Bozak are startling.

I’ve mentioned in the past that no Leaf skater was a plus-possession player without Mikhail Grabovski also on the ice. This is a fact that will continue to be true until we start the next season, when maybe some of the information will change, but it won’t prevent Grabovski from being a cost-controlled two-way force drastically underappreciated by Leafs fans.

Follow these two charts:

  With Grabovski Without Grabovski
Brown 0.509 0.471
Connolly 0.556 0.472
Crabb 0.476 0.486
Frattin 0.552 0.440
Kessel 0.635 0.483
Kulemin 0.525 0.470
Lombardi 0.441 0.464
Lupul 0.592 0.480
MacArthur 0.560 0.445


  With Bozak Without Bozak
Brown 0.450 0.474
Connolly 0.470 0.479
Crabb 0.455 0.486
Frattin 0.407 0.503
Kessel 0.488 0.508
Kulemin 0.259 0.521
Lombardi 0.445 0.468
Lupul 0.495 0.466
MacArthur 0.424 0.524

For the benefit of the class, I’ve bolded the player combinations that were plus-possession. Six of the nine regular Leafs who didn’t play centre (David Steckel was excluded) are plus players with Grabovski. Zero without. Zero of the same Leafs are plus-possession with Bozak, and four are without.

Advertisement - Continue Reading Below

Now, this double counts things, and part of the reason why Bozak’s WOWY (with-or-without-you) numbers are so brutal is because he saw 1:11 of time with Grabovski this season, who is indisputably the Leafs’ best two-way player. For all the grief we give Phil Kessel and Joffrey Lupul in this space for not being good defensively, look what happens to their possession totals when they play alongside an elite centreman.

This is an advantage of depth. The Kings had Anze Kopitar and Mike Richards. The Devils had Travis Zajac and Adam Henrique. The Bruins had Patrice Bergeron and David Krejci. The Canucks had Henrik Sedin and Ryan Kesler. Going back through Stanley Cup champions, you can see the benefit of having two quality centremen.

I’ve mused about what could happen if, say, Nazem Kadri was able to come play between James van Riemsdyk and Clarke MacArthur, or van Riemsdyk and Nikolay Kulemin, or any combination of wingers that makes logical sense. Finding a centreman out of your prospect pool is a cheap and honest way to add a quality man to your team. Kadri has good puck possession numbers, but has he played enough time in the NHL to warrant consideration as a two-way player?

If he is, it certainly makes the Leafs a better looking team up front:

Lupul – Grabovski – Kessel
van Riemsdyk – Kadri – MacArthur
Kulemin – Connolly – Frattin
Komarov – McClement – Steckel
(Brown, Bozak)

I think later I may explain my rationale for holding Jay McClement and Steckel on the same line. It’s not a perfect forward unit, but it’s perfectly acceptable. There’s the one dominant top unit and a competent second and third unit.

Advertisement - Continue Reading Below

Where does this leave Tyler Bozak? I’m just not convinced he’s a major league player. His success depends entirely on who he plays with, and with so many centremen on the Leafs right now, he doesn’t exactly fit.

Also, where is Matt Lombardi? Keep in mind that the Leafs Nation has already traded him for Keith Ballard, even if the Maple Leafs themselves are a step behind us.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  • Burke has identified and targetted JVR as the new dominant and big bodied #1C for Burke’s exciting and fast scoring style of game. People need to stop trying to do Burke’s job – he knows exactly what he is doing here.

    • Markas

      Brian burke was highered nov. 2008.

      post allstar game record 07-08 season before burke.
      16W – 12L – 3 OTL 35/62 pts .565 win%

      Go to current day.
      Post allstar game 2012:

      10W – 18L – 5 OTL 25/66 pts .379 win%
      the only team with a below .400% after the allstar break.

      .565% to .379%

      “knows exactly what he is doing”

      To quote Princess Bride,
      “I do not think it means what you think it means.”

  • Hey.

    So, I hate the Leafs. It’s late so I’m surfing the Nations’ network, but otherwise I don’t really pay any attention to Leafs-oriented media. And I watch the Leafs maybe like 5 times a season.

    But I can’t fathom how “Mikhail Grabovski” and “underrated” can rationally be used in the same sentance.

    In almost every Leafs game I’ve watched in the last 3 years, he has been the best player on the ice for the Buds. The guys just looks straight-up dangerous on every shift. A quick look at his advanced stats at behindthenet justifies this, as he looks like he leads the Leafs fwds in basically every advanced stat.

    If Leafs fans don’t appreciate him, then I don’t know what to say.

    On a side note, I’m keen to see how Lombardi does if he can manage a full healthy season in Toronto. It may be easy to forget that he was the #1C on two different playoff teams. He’s had a rough patch, but is still one of the fastest players in the NHL. Maybe converting him to wing would do the trick for both himself and the club.

    Anyways, hope to see you guys in 30th place this season,

    -The Last Big Bear.

  • “Yet in baseball and basketball, statistics are used to provide a picture of only a player’s offensive contribution. You talk about batting averages and points per game and rebounds and little about what happens on the other side of the ball.”

    That couldn’t be further from the truth. I’m not sure how familiar you are with baseball but they absolutely place a ton of emphasis on defence and defensive statistics. UZR is a huge component of WAR which looks at a player’s overall value.

    In fact I’d go so far as to say that every sport lags behind baseball in terms of advanced statistics, defence included.

    • Traditionally, no.

      There’s also quite a big debate about what UZR actually tells us. I think baseball is still ways away from quantifying defence.

      As is hockey, but we’ve got puck possession down pretty good.

  • But that’s not what you said. You said people don’t talk about defence. But they clearly do. What you just responded is that you don’t believe that people either trust UZR or that they don’t know what it represents, which is far different then suggesting that baseball doesn’t care about defensive stats.

    There isn’t a big debate about what UZR tells us. We know UZR tells us how good a player’s defence is above or below the league average relative to their position. And it is fairly accurate. It’s significantly better than any of the other available defensive metrics.

  • Great read! WOW, Lupul and Kessels numbers with Grabovski made me swoon.

    I commented previously that I think Grabovski is the obvious choice for the top line job, hopefully with the offseason to figure things out Carlyle comes to the same conclusion.

    Connolly should be given a shot manning the 2nd line, I think he could put up decent numbers with MacArthur and JVR. He might even generate interest at the deadline. Kadri could fill in when Connolly gets injured (or on the wing if either JVR or MacArthur get hurt). Either way I think Connolly on the 2nd and Kadri on the 3rd might be a better way to use them.

  • I agree with Mark – baseball’s stats are way beyond hockey’s in terms of offence and defence. A lot of that has to do with baseball being a much simpler game to track.

    I stopped reading the article after you said baseball and basketball only focus on offensive stats.

  • Markas

    I agree with Mark – baseball’s stats are way beyond hockey’s in terms of offence and defence. A lot of that has to do with baseball being a much simpler game to track.

    I stopped reading the article after you said baseball and basketball only focus on offensive stats.

  • Markas

    I agree with Mark about the emphasis of defense in baseball stats. UZR (which is incorporated in fWAR) and DRS (which is incorporated in rWAR), as well as FIELDf/x (which is yet to be released to the public; check link for more information), are all considered good measures of defense in baseball.

    Also, FWIW, WAR, which includes UZR or DRS (depending on the site you use), is becoming mainstream (you didn’t really clarify this point in your original post). This season, SI posted its Power Ranking from Fangraphs’ WAR% (although, I did not agree with its usage earlier in the season):


    I enjoyed the rest of the article, however. I knew there was a difference, but I didn’t think it was THAT large.