|
The hockey sabermetric community has tweaked many statistics to account for context—such as quality of opposition, quality of teammates, game situation, score—but there has been little, if any, such work regarding faceoffs.
Measuring actual faceoff skill is complicated by several factors:
1. In many cases, it is a very subjective decision as to which player/team controlled the faceoff. It's logical to believe that different scorers will have different tendencies in making these subjective decisions, leading to different biases at different rinks, skewing results for the home team players.
2. In man advantage situations, the team with more players on the ice tends to win more faceoffs.
3. Players may face stronger or weaker opposition in the faceoff circle. In particular, opponents tend to play their best faceoff men in special teams situations.
4. Teammates contribute to controlling the puck. It's logical that some teams and teammates are better or worse at this skill.
Ultimate Faceoff Percentage (UFO%) is an attempt to mitigate the most significant of these factors, to get a better measure of who are the best and worst faceoff men in the NHL.
While conscious of significantly diminishing sample size, we've had to bite the bullet, limiting the data to road faceoffs only, to eliminate rink bias. We've taken even strength and non-empty net situations only—also decreasing sample size—to eliminate any need for estimating the effects of special teams situations. And we've adjusted for strength of opposition in the faceoff circle. We haven't adjusted for all of the other players on the ice, but that's expected to be a much less significant factor.
How much of the overall faceoff sample did we pare down? Typically, about 35% of faceoffs were used to calculate UFO%. But you can look at it as a necessary evil—conventional faceoff percentage remains too murky otherwise.
So who was the "real" top faceoff man in the NHL?
Top 20 faceoff performers by UFO%, 2010-11
Rank Player Team FO% B-Rank OFO% UFO% Difference
1 Manny Malhotra VAN 61.7% 2 49.4% 62.3% +0.6%
2 Paul Gaustad BUF 59.8% 3 49.8% 60.4% +0.6%
3 Zenon Konopka NYI 57.7% 4 49.0% 60.3% +2.6%
4 Jerred Smithson NSH 57.4% 6 49.7% 59.0% +1.6%
5 Tim Brent TOR 52.0% 36 50.8% 57.5% +5.5%
6 Ryan Kesler VAN 57.4% 7 49.5% 57.4% +0.0%
7 Steve Ott DAL 56.6% 9 50.5% 57.0% +0.4%
8 Antoine Vermette CBJ 55.6% 15 50.0% 56.5% +0.9%
9 Jarret Stoll LAK 57.5% 5 49.9% 56.4% -1.1%
10 Marty Reasoner FLA 54.5% 21 49.0% 56.1% +1.6%
11 Jason Spezza OTT 56.3% 11 49.0% 56.1% -0.2%
12 Rich Peverley BOS 55.9% 13 50.6% 56.1% +0.2%
13 David Steckel NJD 62.3% 1 50.1% 55.7% -6.6%
14 Vernon Fiddler PHX 53.9% 24 50.2% 55.7% +1.8%
15 Paul Stastny COL 53.2% 28 50.5% 55.6% +2.4%
16 Travis Zajac NJD 55.3% 17 50.6% 55.5% +0.2%
17 Sidney Crosby PIT 55.6% 14 49.5% 55.4% -0.2%
18 Eric Belanger PHX 55.3% 18 49.9% 55.1% -0.2%
19 Pavel Datsyuk DET 54.6% 19 50.4% 54.8% +0.2%
20 Jonathan Toews CHI 56.7% 8 50.4% 54.8% -1.9%
Rank: Ultimate Faceoff Percentage, rank among players that qualified
for NHL faceoff leaders
FO%: Faceoff Percentage (conventional)
B-Rank: Base rank, i.e. conventional ranking among NHL faceoff leaders
OFO%: Opposition Faceoff Percentage at even strength/non-empty net situations
UFO%: Ultimate Faceoff Percentage – faceoff percentage at even strength/non-empty net
situations, modified by quality of competition (OFO%)
Difference: Difference between UFO% and FO%, i.e. how much better (+) or worse (-)
the player is than perceived by conventional FO%
It wasn't Washington's David Steckel, who had the greatest negative swing of all 86 players with enough faceoffs to qualify among the NHL's faceoff leaders (based on a relatively poor road faceoff performance). UFO% confirms that Steckel was elite (55.7%), but drops him down from first to 13th overall.
Nope, it was Vancouver's defensive faceoff whiz Manny Malhotra who topped the list at 62.3% in Ultimate Faceoff Percentage, coming in solidly ahead of the Sabres' Paul Gaustad (60.4%) and the Islanders' Zenon Konopka (60.3%). The biggest gainer on the entire list was Toronto's Tim Brent, who jumped all the way up from 36th to fifth, with a UFO% of 57.5%. The 26 year old played a full campaign with the Maple Leafs after having played in only 19 games over the previous four seasons.
We'll take a further look at Ultimate Faceoff Percentage in the upcoming Hockey Prospectus 2011-12 annual.
Many thanks to Mike Rovito for his assistance in developing UFO%, including helping me find an appropriate "U" for the name. And thanks as well to the expertise of statistical sounding board Tom Awad.
Timo Seppa is an author of Hockey Prospectus.
You can contact Timo by clicking here or click here to see Timo's other articles.
| 
Plugging Holes (06/19)
2 comments have been left for this article. 
Very interesting stuff - what about a list of the worst 20 who saw significant time in the circle?