Who’s Winning Games in Milwaukee?

The Milwaukee Admirals goal-tending tandem right now is not the breeding ground for a young prospect. With Miroslav Svoboda collecting starts in the ECHL for the Atlanta Gladiators, the Admirals are backstopped this season by veterans Troy Grosenick and Tom McCollum (who is on an AHL contract only).

Through 22 games, Grosenick and McCollum have split starts, with each having 11 despite Grosenick’s brief call-up to Nashville earlier in the season. The Admirals have allowed the sixth-fewest goals in the entire AHL this season with 58 (sidenote: San Jose has allowed a mere 35 goals this season). So, something is working well defensively in Milwaukee, but how much are the goalies contributing to this?

Public shot data for the AHL is hard to come by, so I will analyze Grosenick and McCollum’s performances the some exact data, some estimations, and, of course, some film.

Goals Saved Above Average

Traditionally, save percentage and goals-against average have been go-to statistics for goalies. These numbers are by no means useless, but they can ignore a lot of context, especially goals-against average.

For what it’s worth, Troy Grosenick is fifth in the AHL in save percentage at 0.924, and McCollum is 19th at 0.900.

To provide a better outlook of their performances, I have measured Grosenick and McCollum’s GSAA. GSAA is a statistic that details how many more or fewer goals a netminder has allowed than a league-average goalie would with the same defense. This helps differentiate goalies who benefit from shutdown defensive units from those who do not. Like most catch-all statistics, this is not a be-all-end-all measure, but it is worth a look.

There are 24 goalies who have started 11 or more games in the AHL this year. To start, we will need the average save percentage of all 24: 0.896.

Grosenick: Grosenick has faced 344 shots this season, which is the ninth most among this group. An AHL-average goalie would stop roughly 308 of those shots. Grosenick has allowed 26 goals on 344 shots, whereas a league-average goalie would allowed 36. Ten more goals allowed would put the Admirals at 23rd league-wide instead of sixth.

McCollum: McCollum has faced 310 shots this season. A league-average goalie would stop about 278 of those, leading to 32 goals-against—one more than the 31 McCollum has allowed.

2018-19 AHL GSAA [Min. 11 starts]

Of these 24 goalies, Grosenick ranks second with 10 goals-saved above average, and McCollum ranks tenth with one goal-saved above average. This is the best tandem in the AHL. Milwaukee’s defense is allowing an average amount of shots, and Grosenick—in particular—is still largely shutting the door.

You’ll notice that GSAA is a good statistic to delineate responsibility. Collin Delia has an average goals-against average of 2.35 but has saved 17 goals more than an average goalie, making up for Rockford’s league-worst 474 shots against.

GSAA is best measured at 5v5 so special teams don’t skew shot effects. Like I mentioned above, shot data in the AHL is not accessible, so my numbers above are measured in all situations. However, regardless, 11 total goals-saved above average is fantastic for Grosenick and McCollum.

Additionally—and this is much more circumstantial—Grosenick and McCollum have faced 16 shots in shootouts this season and have allowed one total goal.

Shot Locations

So, we have clearly determined that Troy Grosenick is carrying much of the load in Milwaukee when it comes to winning games. That is not to suggest that he has not had his bad outings this year. Despite having the fourth best goals-against average of this group of goalies, Grosenick has allowed more than two goals in four of his starts (not including Nov. 30).

To better understand Grosenick’s strengths and weaknesses, I charted the shot location of every goal he has allowed through November 29.

This heat map is from Emmanuel Perry at Corsica Hockey.  There is an endless amount of discussion on how to define scoring chance quality, but I am using Corsica’s, which is a basis for measuring expected-goals statistics. So, a shot from the red area is a low-danger scoring chance (expected shooting percentage of less than 3.0%), a shot from the orange area is a medium-danger scoring chance (expected shooting percentage between 3.0% and 9.0%), and a shot from the yellow area is a high-danger scoring chance (expected shooting percentage higher than 9.0%).

These scoring chances are not publicly tracked in the AHL, so there is some speculation involved below. Grosenick has allowed one (1) low-danger goal through Nov. 29, meaning an average goalie would allow the same number of goals on roughly 50.0 low-danger shots (2.0% shooting). Since Grosenick has faced over 330 shots this season, it is safe to assume that he has faced many more than 50.0 low-danger shots.

Additionally, he has allowed eight (8) medium-danger goals through Nov. 29. An average goalie would, in theory, allow the same number of goals on 133.3 medium-danger shots (6.0% shooting).

Finally, Grosenick has allowed 17 high-danger goals through Nov. 29. At 10% shooting, an average goalie would allow that many goals in 170 high-danger shots.

Again, types of chances are not publicly tracked for the AHL, so I will provide some imperfect context for Grosenick’s numbers via Natural Stat Trick. Of goalies who have played 500+ minutes in the NHL this season, the median number of shots against is 362; the median number of high-danger goals against is 16.5 (93 shots against), medium-danger goals against median is eight (94.5 shots against), and the low-danger goals against median is four (156.5 shots against).

What the eye-test and this comparison suggests is that Grosenick is facing an expected number of shots overall but more high and medium-danger chances than average.  He has almost completely shut the door on low-danger chances and is performing okay on others.  Chance quality gives important context. The lower the quality of hockey the less this is relevant, but most goalies perform similarly with low and medium-danger chances.  Grosenick has appeared to excel in his goals-saved above average mostly through those opportunities.

Goal Placement

Finally, I wanted to analyze where in his stance Grosenick is strong and weak.  This can help give context to many high-danger scoring chances.

First, it is important to note that 30.7% of Grosenick’s goals-allowed were off of rebounds, and another 11.5% were deflected in front of the net. The graphic above notes the location of Grosenick’s 26 goals allowed through Nov. 29. Most of the rebound goals are accounted for in the 5-8% areas, but you will notice his blocker side seems to be seeing a tad more action.  Our sample size is too small to draw major conclusions, but via recollection alone, Grosenick has tightened up the side a bit more lately (or is just being exploited more on his glove side).

Additionally, 34.6% of the goals Grosenick has allowed crossed what is known as ‘The Royal Road’ before being scored. This imaginary line from the middle of the crease out has become a new tool in analyzing offensive chances. Former goaltender Stephen Valiquette has led the charge on this studying and concluded that puck movement across the Royal Road provides the best opportunity for an attacking team to score (22% of goals he studied were preceded by movement across the line). You can read more about this here.

55.6% of goals allowed by Grosenick that were preceded by movement across the Royal Road went from his glove side to his blocker side.

Final Thoughts

Grosenick is obviously performing well above league average for the Admirals this season.  His impressive shootout outings have been particularly helpful.  His defense is performing okay in front of him, and he has not shown any discernible weakness: ¹⁄₃ of his goals-allowed are off rebounds, roughly ¹⁄₃ are via movement across The Royal Road, and roughly ¹⁄₃ are clean shots or deflections.

All statistics are current through Nov. 29 and are courtesy of theAHL.com, prospect-stats.com, and naturalstattrick.com.