Randy Carlyle and the faceoff paradox
By Cam Charron
11 years ago
The bad news, as pointed out by James Mirtle, is that Tyler Bozak is averaging more than three minutes a game more than Mikhail Grabovski this season. That can mean that Randy Carlyle actually-factually thinks Bozak is a better player than Grabovski.
Worse, as I’ve seemed to notice, it seems as if Grabovski has been held to such a low number based on a statistic. Like has been pointed out to me many times before in comment sections, you can’t build a team based on a spreadsheet, whether it’s Corsi, Offensive Zone Start %… or face-off percentage.
From Jonas Siegel’s Leaf Report last night. When I first saw the quote from Carlyle after the Jets game, my interest was piqued:
“If you’re not starting with the puck, then you’re defending… any line, I guarantee you, would rather start with the puck than have to chase it and recover it.” Carlyle noted that any such improvement did not fall on the centreman alone, but required (as per usual) secondary aid from teammates on the ice. Grabovski has been hit or miss on the draw, a sampling of his work in the past five games revealing such a trend: 7-13 (54%) vs. the Penguins, 4-14 (29%) vs. the Jets, 9-14 (64%) vs. the Penguins 2-11 (18%) vs. the Bruins, 12-21 (57%) vs. the Senators.
Puck possession is important. It’s extremely important, in fact, and the correlation between winning teams and teams that possess the puck is extremely high. That said, face off percentage is a terrible way to determine which team controls the puck.
The NHL used to record “zone time”, or the amount of time a team started in the offensive end. A vintage Oilers blogger under the blogname Vic Ferrari noticed that teams often talked about possession, but nobody tallied up to see which team had the most offensive zone time. It’s something that appears at EA Sports, but no longer NHL.com.
I thought it would be worth looking at a season to see how shots, and shots directed at net, meshed with the zone times for the teams. Now obviously some teams play more of possession game than others, and unbalanced scheduling means that some teams played this type of opponent more than others. Still, the connection between shots and zone time will still shine through, as will the link between the ability to outscore opponents and these two things.
His chart still has zone time numbers from 2002. You should read that whole post from above, but I have my own chart here with the pertinent information: Zone Time differential, Shot Attempt differential (Corsi) and a stat I took from NHL.com for that season, Faceoff Win Differential:
Faceoffs are very important for determining which team has good zone time. I went into Excel to calculate the relationship between Faceoffs and Zone Time. It was pretty high, with an r-squared value of .24585. An r-squared of 1 implies perfect correlation, an r-square of 0 implies zero, so you can say that faceoffs explain about a quarter of team variance in puck possession ability.
Lots of good puck-possession teams are good at faceoffs. The 2008 Detroit Red Wings were first in the league. The 2002 Carolina Hurricanes were first in the league. The 2010 Chicago Blackhawks… third in the league. The elite possession teams throughout the years have been really good at controlling the puck off the draw.
That said, you need to be a good player to turn faceoffs into meaningful possession. The 2008 Detroit Red Wings had a player named Pavel Datsyuk whose faceoff percentage was slightly lower than that of another player they had named Kris Draper. Draper was the best faceoff man on the team, winning an incredible 58.6% of the 944 draws he took that season. But Datsyuk played 21:23 of ice-time per game, and Draper played just 15:37.
When Pavel Datsyuk was on the ice, the Red Wings got 63.2% of the overall shots for both teams. That is an incredibly high number. When Kris Draper was on the ice, the Red Wings took just 50.5% of the shots. Naturally, this needs to be qualified. If you check the usage chart, Draper played slightly better competition than Datsyuk and took slightly more of the defensive zone draws. Still, Datsyuk, despite being not as good at faceoffs as Draper, was relied upon more because he was so much better at doing everything between the faceoffs.
The correlation between shot attempts and zone time is r-squared of 0.81386, or was in the 2002 season. That concept I doubt has changed much. The teams that are the best at controlling the shot clock have produced the most elite teams in the NHL.
Relationship between Zone Time (x-axis) and Faceoff Win Differential (y-axis):
Relationship between Zone Time (x-axis) and Corsi (y-axis):
The shot differential numbers are much, much more indicative of zone time than faceoff numbers could ever hope to be. This is why the explanation Carlyle gives about faceoffs, and the deductive reasoning that Carlyle sends Grabovski out for fewer faceoffs because he isn’t as good as Bozak, is absurd. He’s looking at a small statistic that influences a fraction of the overall puck possession picture.
I forget where I read that many traditional sports statistics don’t represent the big picture. Simply put, baseball, basketball and hockey have primitive box scores because when the box score was invented, teams didn’t count relevant information, just what was easy to count. It’s a modern era, where every game is televised, multiple official statisticians record every game and we have piles of new information relevant to what we need.
Grabovski is a better player than Bozak between the whistles. Despite starting almost exclusively in his own end against the toughest competition the NHL has to offer, his Corsi % is almost what Bozak’s is. Bozak’s is higher, but Grabovski influences it better. Besides, the current difference between Bozak’s faceoff percentage (53.8%) and Grabovski’s (50.3%) yields a difference of 0.70 per game over 20 faceoffs. Even if there were a legitimate difference between the two, say 55% and 45%, we’re looking at a difference of two wins per game. Two lousy, marginal, ten-second long possessions.
Grabovski has just as many even strength points as Bozak. Grabovski does more in harder competition than Bozak. Grabovski is more talented than Bozak, Grabovski’s synergy with Phil Kessel is more than that of Bozak and Kessel, and Grabovski is a better all-around hockey player than Bozak.
The only explanation as to why Grabovski gets less ice-time than Tyler Bozak at this point is Randy Carlyle’s belief in the faceoff statistic. It’s a backwards, backwards world.
Recent articles from Cam Charron
