A Second Look at Save Percentage

There's been a lot of talk about the Atlanta Thrashers goalies this season, and for good reason. Goaltending has been a frequent concern for the Thrashers, but this year, the tables have turned. Even with Kari Lehtonen out, the Thrashers have seen a goaltending renaissance this season. Johan Hedberg is experiencing nothing short of a rebirth in his 9th year in the NHL, and young Ondrej Pavelec has finally made the jump from Prospect to NHL Goaltender.

For a look at this season's goaltending performances, see The Falconer's post here.

What really interests me this season is that both Thrashers goaltenders have seen improved numbers this season. There's some variance in save percentage from year to year, but seeing the upward trend across the board got me thinking. Perhaps it's been a reduction in high-percentage scoring chances?

Well, apparently, no. At least according to forum members WINGZ_25 and Tom Awad, who have taken a look at the shot charts. So if it's not quality of shots, what could it be?

With a bit of research, I may be able to shed some light on the subject. Follow beyond the jump for more.

Let's start by taking a look at the save percentage for Lehtonen, Hedberg, and Pavelec since the lockout.

Both Hedberg and Pavelec are having career seasons so far this year. We don't have Lehtonen's stats to compare, but my assumption is that he would also have seen a nice boost. The question remains: is there something to account for the improved stats other than improved play by the goalies?

For that, we need to take a deeper dive into the data. We're going to take the table from above and split each goalie into three rows: even-strength, penalty kill, and power play.

We can probably ignore the PP line - it's still early in the season to talk about something like short-handed goals allowed. Here's a few things that I'll point out.

1. With only one exception, Kari Lehtonen has had a better save percentage every year in each situation than Hedberg and Pavelec. The exception was at even strength in the 2007-2008 season, though Pavelec played just 7 games that season. This is important: with the exception of the small data point, Lehtonen has always had a higher save percentage than our current set of goalies.
2. Pavelec's even-strength save percentage went up this year (.898 to .906), but his save percentage on the penalty kill has skyrocketed. Folks, that's not a typo: Pavelec has a better save percentage on the penalty kill than he has at even-strength or on the power play. If you're looking for a reason that Pavelec's save percentage has gone up, look no further. (It is worth suggesting here that these numbers are almost certainly unsustainable over the course of an entire year: Pavelec's PK SV% should drop a bit to be more in line with his overall save percentage.)
3. Hedberg's save percentage on the penalty kill has also gone up significantly, from .820 to .876, but his play at even-strength has pulled his stats up as well. If you consider the number of shots in each situation, his improved play at even-strength and on the penalty kill have both been significant.

As a stats guy, I can say that the quality of shots against is as bad as it’s ever been. At even-strength, the Thrashers are giving up shots with an expected shooting percentage of 8.7% versus 7.8% league average. On the penalty kill it’s 14.5% vs 12.7% average. It’s really thanks to the goalies that you guys are doing so well.

I don't have year-to-year stats for this - Tom, feel free to link to your stats in the comments - but it wouldn't surprise me if the quality of shots is down year-over-year, especially on the kill. 

Both Johan Hedberg and Ondrej Pavelec deserve a round of applause for stepping it up in the absence of Kari Lehtonen, but I think a fair bit of credit should go to our penalty killers as well. Perhaps more interesting than the question of "why are our goalies stopping more shots" is the unanswerable "what would Kari Lehtonen's stats look like this year if not for his injury/surgeries?" Speculate at will below.