April 24, 2009
Faceoffs:
This One Goes To Eleven

by Timo Seppa

Game One : Early in the first period, the Penguins line up a formidable unit for their first power play, including Sidney Crosby, Evgeni Malkin, Chris Kunitz, Kris Letang, and Sergei Gonchar. Although Sidney Crosby (51.3 percent regular-season faceoff percentage) is tossed from the offensive zone draw, Evgeni Malkin (42.4 percent faceoff percentage) manages to win the faceoff from Jeff Carter (48.3 percent faceoff percentage). The Pens move the puck, looking for a weakness in the Flyers’ penalty kill. Defenseman Sergei Gonchar’s wrist shot is blocked by Carter, but the Penguins retain possession. Moments later, Malkin takes a point-blank shot at netminder Martin Biron. The rebound is put home by Crosby. Elapsed time: 15 seconds.

That power-play goal put Pittsburgh up 1-0 in a contest that they would never trail. We frequently hear about the importance of faceoffs in playoff hockey, particularly on the man advantage. Is that notion based in statistical truth, or is it just an old hockey myth?

By definition, winning faceoffs gains possession of the puck for your team. A better faceoff percentage infers greater puck possession time, more shots on goal, and more goals than your opponent, all other things being equal. That said, faceoffs are only one means of a team winning or losing possession, though faceoffs are the only such means with an objective measure. In contrast, gaining or losing the puck by tracking down missed shots or rebounds from saves are not recorded in hockey as rebounds are in basketball. Even murkier is measuring giveaways and takeaways. An excellent concept, the game-to-game variation of what constitutes a giveaway or takeaway is at least as subjective as that of errors in baseball. To be fair to the folks frantically keying in the play-by-play logs, the speed at which events occur in hockey makes it extremely difficult to measure every nuance of the game.

Since 1997-98, the best regular-season faceoff percentage achieved by any one team was 56.1 percent by the 2001-02 Carolina Hurricanes, while the worst was 44.1 percent by the 1998-99 Tampa Bay Lightning. This season’s top mark was posted by the Detroit Red Wings (55.1 percent), while the young Phoenix Coyotes trailed the rest of league by posting a dreadful 44.8 percent. The best individual efforts of the campaign were by Carolina’s Rod Brind’Amour (61.0 percent) and Detroit’s Kris Draper (60.3 percent). The worst regular was Andrew Cogliano of Edmonton, as he achieved just 37.2 percent over 702 faceoffs.

During the regular season, there is good correlation between faceoff percentage and shot differential and between faceoff percentage and goal differential. From 1997-98 to present, the difference between a very good 55 percent faceoff percentage and a very poor 45 percent faceoff percentage amounted to 6.7 shots on goal (SOG) per game, and 0.7 goals per game on the average.

What happens during the playoffs, though, seems to be an enigma. While higher quality defensive play can explain a decreased 5.8 SOG per game advantage, you may be surprised to hear that the difference in goals per game drops to literally 0.0 goals per game. To repeat, over the course of roughly one thousand playoff games since 1997-98, having a 55 percent to 45 percent faceoff advantage over your opponent statistically gains you zero goals and zero wins.

How in the world is that possible? Take two statistically similar teams with identical winning percentages and goal differentials that tie for the last playoff spot in their conference. Now consider that their greatest statistical difference is a 55 percent faceoff percentage for one team, and a 45 percent faceoff percentage for the other team. It makes perfect sense that the team with the 45 percent faceoff percentage will have be marginally better in every other facet of the game to have made up for that deficit in faceoff skill. On average, teams that have gone through the qualifying process of making the playoffs are by definition equals in goal difference per game regardless of varying faceoff skills, even though the better faceoff teams will tally more shots on goal. Therefore, don’t ignore the faceoff strength of a playoff team when analyzing them, but know that they will likely have other deficiencies compared to lesser faceoff teams.

Taking the subset of Stanley Cups finalists since 1998-99 yields the same result—zero goals per game difference, no matter what their faceoff percentage. Recent Stanley Cup winning teams have varied from the faceoff-dominant 1998-99 Dallas Stars (55.6 percent playoff faceoff percentage) down the relatively poor 2002-03 New Jersey Devils (47.0 percent). Instructively, the extremes of the runner-ups were their respective opponents, but while the 2002-03 opponent Mighty Ducks of Anaheim (54.9 percent) outclassed that season’s champions in faceoffs, the reverse was true of Dominik Hasek’s 1998-99 Buffalo Sabres (45.2 percent).

This of course does not mean that teams should punt the faceoff. So far in the 2008-9 postseason, 132 goals have been scored. Though 56 of those goals occurred within the first 30 seconds after a faceoff, a surprising 23 of the 56 were scored by the team that lost the faceoff. The takeaway from all that? In hockey time, 30 seconds is an eternity; dialing back the time threshold, we have a result that makes more sense: Over the first 11 seconds following a faceoff, 14 non-empty net goals have been scored this postseason by the faceoff winner versus one goal against (a defensive zone giveaway by Anaheim). Everything is fine for the first 11 ticks, but when the clock strikes 12 seconds, the statistical advantage has turned into a pumpkin. For 12-30 seconds after a faceoff, it has been 18 goals for, 22 goals against. Though this is a small sample, it correlates well with a study by Gabriel Desjardins looking at data from the 2003-04 regular season.

Clearly, offensive-zone faceoffs are the more advantageous than neutral-zone or defensive-zone faceoffs. This postseason, 12 offensive-zone and two neutral-zone faceoff wins resulted in goals within the key 11-second threshold. It is also not surprising that a large portion of these goals have been scored on the man advantage. Of the 14 goals scored, seven were on the power play, and seven were at even strength. As there have been 105 faceoffs on the man advantage, this corresponds to healthy 6.7 percent of power play faceoffs yielding a goal within 11 seconds, or one in 15. For the 1506 even strength faceoffs, the rate drops to a mere 0.5 percent of draws yielding a goal, or 1 in 215.

A version of this story originally appeared on ESPN Insider .

Timo Seppa is an author of Hockey Prospectus. You can contact Timo by clicking here or click here to see Timo's other articles.