Sun Mar 27 02:28pm EDT
(Ed. Note: Welcome to Stat Nerd Sunday, where we occasionally obsess over hockey numbers like a Dungeon Master obsessing over the level of his warrior elf. Here's Matt Barr, formerly of LCS: Guide To Hockey and Trolleytracks and now blogging hockey at Kertwang.me.)
If you draw the Bruins in the first round, shoot and crash the net. If you draw the Kings, you may want to be more circumspect about your shot selection.
Before we get to strategy in the 2010-11 playoffs, let me say a word about rebounds. To me, rebounds are, if not the Holy Grail, then at least at least the Life of Brian of hockey stats. I'm a Save Percentage skeptic for a number of reasons, and one of them is that if you want to judge how well a goalie is doing his job, you have to look beyond the two broad outcomes of a shot on goal -- that is, goal scored, or save made. I want to know whether a save resulted in a rebound or not.
Consider a save where the puck is either directed harmlessly away from the attacking team's possession, or remains in the control of the goalie. The defending team is out of immediate danger, and has a chance to begin an orderly breakout and transition up ice.
On the other hand, a dozen bad things can happen when a goalie coughs up a rebound, foremost (and easiest to record and measure) a goal against. Ken Krzywicki demonstrated that in 2009-10 shooting percentage for shots off rebounds was 32.8%, orders of magnitude higher than the overall 8.9% shooting percentage for all shots.
With comprehensive data on saves that do and don't result in rebounds, we would know in the first place which goalies control their saves the best. The ones who do are doing their job better, don't you think? We would also start to learn which teams, and which defenders and goalies, are best and worst at cleaning up after a rebound save. We could even start to recognize the forwards who score rebound goals the most. We could give them the Phil Esposito Trophy at the awards thing in Las Vegas, or something.
Even with NHL GameCenter Live, a league-wide Rebound Adjusted Save Percentage is beyond the scope of a weekly(ish) blog post. But we can take some incremental steps that might tell us something more immediately interesting or useful.
I wanted to look at every goal scored against the likeliest playoff teams since the All Star break, and count which ones were scored off rebounds.
Two caveats: I only looked at goalies who have played at least 500 minutes for those teams since the break. And "likeliest playoff teams" means I didn't look at the teams currently fighting it out for the last spot in each conference. That still gave us a collection of 17 goalies and (exactly!) 700 goals against to work with. (All stats through games of Friday, March 25.)
This wasn't an examination of every save, just every goal. Looking only at rebound goals won't tell us which teams react most effectively to rebounds given up by their goaltender, nor will it tell us how often a goalie gives up a rebound and then keeps the puck out of the net. But the data on Rebound Goal Percentage is interesting in its own right, because it does give us an idea of how much a failure to control saves could be contributing to a goalie's goals against. And if there's ever a time to know that, it's on the cusp of the playoffs.
Let's start by looking at each goalie's percentage of goals that were scored off rebounds; that is, rebound goals divided by total goals against. Of the 700 goals total, 118, or 16.86%, were rebound goals.
Three of the goalies at 25 percent or over are in the Vezina conversation. Maybe they're all world class preventers of non-rebound goals, or maybe percentage of goals scored by one method or another doesn't tell us anything about a goalie's overall level of play. However, a team facing the Boston Bruins, Pittsburgh Penguins or Vancouver Canucks (or Chicago Blackhawks) in the playoffs might profitably exploit the fact that a large fraction of the goals it'll score against them will be by forwards in position to take advantage of rebounds.
I think we get baby steps closer to learning something about a goalie's overall performance by looking at his Rebound Goals Against Average. Like GAA, it's rebound goals against per 60 minutes:
These goalies' efficiency in controlling rebounds varies a lot more widely than their overall GAAs do. Tim Thomas(notes), presumptive two-time Vezina honoree, gives up twice as many rebound goals as Carey Price(notes) does. There are very few things Brian Boucher(notes) does better than Roberto Luongo(notes), but allowing 40 percent fewer rebound goals is one of them.
To start to understand how rebound goals fit into a goalie's overall performance, I calculated the difference between each goalie's GAA and RGAA. That's the rightmost column above, "non-rebound" GAA. The difference between the best and worst GAA and the best and worst NRGAA are almost exactly the same. Different goalies just seem to get to where they are in different ways when it comes to rebound control.
How well a goalie controls his saves is one (large) factor in how many rebound goals he allows, but his teammates' ability to clean up after a rebound is a factor, as is the attacking team's ability to get stick on puck. If we can start keeping comprehensive data on saves with rebounds and saves without, and what happens next after a rebound, we'll all be a lot better informed as to how well and poorly hockey players play hockey.
In the meantime, we can guess from the rebound goal data above whose first-round opponents will be extra aggressive throwing the puck at and crashing the net.