|
Recall the theme music from your favorite Clint Eastwood western and think: An NHL netminder is like a gunslinger. He wins, he survives, he earns his glory for a day. Yet alas, there is always another gunslinger. Winning today never is a guarantee of winning tomorrow. Listen, my friends, to the distant wail of the Coyotes. Gather a fading glimpse of the distant Lightning. Here is the lamentable tale of Ilya and Mike: For some poor unfortunates, winning the next day has been particularly unlikely this season.
Rob Vollman recently introduced the Quality Start (QS) metric for goalies.
Start Type Save Percentage Winning Percentage
Quality Start 0.913+ 0.777
Average Start 0.900-0.912 0.536
Average Start 0.885-0.899 0.503
Poor Start 0.884- 0.246
As shown in the table above, not only does QS show the strong correlation between Save Percentage and Winning Percentage, but it identifies key ranges where relatively small changes in Save Percentage can make a big difference between winning and losing games. A Quality Start of .913 or higher Save Percentage by a goalie yields a greater than 75% chance of winning the game, while a Poor Start (PS) of .884 or lower Save Percentage yields less than a 25% chance of winning the game.
Looking through the lens of Quality Starts, what does an NHL team lose by starting a goalie in consecutive games? After all, one of the main uses for the backup goalie is in starting one of the games in a back-to-back situation to spell the number one goalie. Yet with teams wanting to get as many games possible out of their top netminder, there are many instances throughout the season where he will play on consecutive nights.
Effect of back-to-back games on Quality Starts, 2008-2009
# of days rest Quality Start % Poor Start %
0 47% 36%
1+ 53% 31%
The table above illustrates the slight detrimental effect of starting the same goalie back-to-back, perhaps to the level where teams should reconsider the perceived benefit of starting the better goalie twice. The effect of playing with no days rest is a decrease in QS of 6% and a similar increase in PS of 6%. This amounts to a significant, but not overwhelming factor against winning the second game. With a typical team starting their number one goalie in back-to-back situations 4 to 10 times per season, this inefficiency might cost some teams one win.
That said, all teams and all goalies are not created equal. Over the course of this season’s data sample, many goalies showed little or no detrimental effect in QS from playing consecutive days. These goalies included Ryan Miller, Tomas Vokoun and Roberto Luongo, among others. Putting aside QS, whether high workload or playing on consecutive nights has been a factor in any goalie injuries is for another discussion and a complex one at that.
Among the goalies that have been negatively affected in performance by playing two days in a row this season are Vesa Toskala, Steve Mason and Scott Clemmensen. In Toskala’s case, increased workload may have exacerbated a preexisting injury. Dealing with lingering hip and groin problems, the Toronto backstop’s last game was a 52 shot effort on March 3rd.
Now let’s turn our attention to the alluded to heroes of this column, Ilya Bryzgalov and Mike Smith, two backups in their late twenties that changed cities last season to become the number one goalie of their new teams:
Ilya Bryzgalov – Effect of rest on QS, 2008-2009
# of days rest Quality Start % Poor Start %
0 25% 75%
1+ 41% 22%
You might be surprised to find out that the 29 year old Bryzgalov is leading the NHL in back-to-back starts with 12, perhaps thinking that the leader might be a more established workhorse like Marty Turco or Henrik Lundqvist. Unfortunately, Bryzgalov’s 25% QS and 75% PS rates for consecutive games are the worst in the NHL (for goalies with more than 5 such starts). While the former Duck never exhibited this tendency in the past, it seems that coach Wayne Gretzky went to the well too many times against what has been an unmistakable trend this season and this is when he has acknowledged in previous seasons the need for rest to keep Ilya sharp:
Phoenix coach Wayne Gretzky is disappointed in his
goaltender’s recent play, expressing concern that
Bryzgalov might be getting tired. “I thought the
first goal wasn’t a very good goal…I know he’s
much better than this. He was below average in
Edmonton and he was average the other night in
Vancouver and again tonight.”
The Russian goaltender’s performance in consecutive games declined over the course of the season, starting at a Save Percentage level of .913 at the beginning of the season and dropping to a Save Percentage level of .825 level over the course of the year:
Ilya Bryzgalov – Consecutive games, 2008-2009
Date Opponent Result Save % Start Type
10/12 ANA 4-2 W .943 QS
10/18 MTL 1-4 L .879 PS
11/9 SJS 4-2 W .938 QS
11/13 MIN 0-4 L .875 PS
12/7 CHI 1-7 L .750 PS
12/11 MIN 3-1 W .968 QS
12/27 LAK 0-4 L .867 PS
1/18 EDM 3-6 L .821 PS
2/4 DET 4-5 L .848 PS
2/12 VAN 3-4 L .875 PS
3/6 BUF 1-5 L .839 PS
3/22 ANA 2-6 L .800 PS
The game log is a nightmare. It doesn’t take a statistician to point out that Ilya was not cut out for consecutive games this year after giving up 4 to 6 goals in his last six attempts and never breaking an .875 Save Percentage. Aside from staying the heck away from starting him on no rest, it’s up to the Coyotes’ management to figure out whether conditioning, injury or effort is compounding the situation.
Mike Smith – Effect of rest on QS, 2008-2009
# of days rest Quality Start % Poor Start %
0-1 37% 42%
2+ 77% 9%
Before Mike Smith’s season ended with a concussion in late January, the former Star was another poster child for requiring additional rest. In his case, at least 2 days. Unlike Bryzgalov, the 27-year-old Smith had exhibited a need for rest in previous seasons, but never to this degree. When pushed to start twice within 2-3 days, you saw a completely different goalie in 2008-2009. Smith’s dropoff from 77% to 37% in QS was even bigger than Bryzgalov’s. With PS increasing from 9% to 42%, he was 4-5 times more likely to have a PS on 0-1 days rest.
Another disturbing factor was that Smith’s Save Percentage steadily declined from a .940 to a .900 level over the course of the season. In addition to not handling short rest well, was he wearing down in general? The question going forward –once the much more critical concussion issue is resolved– is how the Lightning can build Smith’s stamina to a level where he can handle more frequent starts and a greater workload. Until that point, it would be advisable for Tampa Bay to split starts between Smith and Kari Ramo much more evenly, to maximize the quality of Smith’s starts.
One commonality between the two goaltenders in question is their 6’3” height, towards the upper end of typical goalie heights. While not every 6’3”or 6’4” netminder exhibits the need for additional rest, it is worth looking at whether factors like height or goaltending style may influence consecutive days’ performance.
There is much to be learned about the effects of rest and workload on a goalie’s effectiveness and health. Aside from trying to avoid consecutive days’ starts for their number one goalies, NHL teams would do well to monitor how their specific goalies handle back-to-back starts over the course of their career as well as over the course of a specific season.
As for this season, Ilya and Mike are riding into the sunset.
Timo Seppa is an author of Hockey Prospectus.
You can contact Timo by clicking here or click here to see Timo's other articles.
| 
0 comments have been left for this article. 
Conn Smythe Watch: Rask Leads Race