Statistics and Bounce Back Situations
If you’ve looked at the stats located on this website about the three Canadian Hockey Leagues (OHL, QMJHL, and WHL) and the American Hockey League you probably noticed at the end an abbreviation, BBS%. At the end of 2016 more stats appeared with that “BB” involved in the abbreviation somehow. If you haven’t made your way over to the Stats Glossary you’re probably wondering what you are looking at. Even if you have seen the stats glossary page than I wouldn’t blame you for wondering more about the theory behind tracking bounce back related statistics.
Where did I come up with this crazy idea to create these Bounce Back Stats? If you didn’t know I myself am a goalie and have been a goalie since Novice (was a forward in Tyke). I attended various goalie camps put on by different levels of goalie coaches (levels as in what they worked so a level such as Jr. C). The camp I learned the most from is Steve McKichan’s Future Pro camp. If you don’t know who he is he’s not only a former NHL goalie, he’s also a former goalie coach for the Toronto Maple Leafs who worked with Pogge, Belfour, etc.
When we had the 30 minute lectures that took place once a day there are two aspects that have stuck with me permanently. The number one aspect you learn at goalie camp is how important the mental game is. When I was at camp the quote was: “goaltending is 90% mental and 10% physical”. I’m sure it’s different with whatever goalie or goalie coach you talk to but the premise is the same, a strong mental game is vital in order to be a goalie.
So what I look at when going through game logs or just looking at statistics such as Quality Starts % I’m looking for consistency in goaltending. For me that’s priority number one. If you’re looking for a goalie you want to try and find a goalie that is consistent. Even if that goalie is consistently average that’s still valuable.
If a goalie is struggling to be consistent are they a write-off? To me no because then I go to priority no. 2 which is how does a goalie bounce back after a bad game? I ask this question because of a quote from the Future Pro goaltending camp where McKichan said (not exact quote but its close enough) “I don’t go view a goalie when they are having a great stretch of games, I want to see a goalie after they have had a bad game because you learn the most about a goalie after they’ve had a bad game”. It’s a good point because the best goalies are the ones who stop themselves from having one bad game after another. Case in point is Sergei Bobrovsky. After having a statistically bad game Bobrovsky has responded with a Quality Start seven out of eight times in this 2016/17 NHL season.
With Bounce Back games being important the question then becomes how can you tell if a goalie is strong at bouncing back after a bad game? First you have to decide what a “bad game” statistically is. For me a “bad game” is defined by posting a start where the save percentage is lower than the replacement SV% for that respective league. Now what do you track in the game to determine a goalies ability in bounce back situations? I am currently tracking three statistics that you can see on the stats page: Bounce Back Starts, Quality Bounce Back Start % and Bounce Back Save Percentage.
In preparing these statistics for my website I chose these three stats to track because knowing how many bounce back starts is important due to sample sizes. I view Quality Starts % as an important stat to gauge consistency so I applied it to bounce back situations which gives us Quality Bounce Back Start %. Lastly is the Bounce Back Save Percentage. I’m not the biggest fan of raw save percentage as it can be inflated or low amount of shots making the SV% look worse than it is. Tracking the SV% in different situations though can be useful as a back-up when the sample size for Quality Starts % is too small or just tracking SV% in situations such as High Danger shots, 5v5, etc.
It’s key to keep in mind though that you avoid using just one statistic at all times. When talking with a NHL analytics consultant he brought up a very good point that it’s not important how many stats you have, rather how you interpret them. What I try to avoid doing is purely looking at just one statistic in any category whether it’s a regular situation or a bounce back situation. Think of it like painting a picture on a canvas only in this case you’re painting a picture about a goalie. You wouldn’t use one colour of paint just as you shouldn’t use one statistic when looking at a goalie.
Let’s take Avalanche prospect and San Antonio goalie Spencer Martin for example. Starting with Games Started we see that Martin has started 21 games which is the 4th most in the AHL so we can assume that Martin is trusted by his coach. Martin has also posted a 0.619 QS% which shows us he’s an above average AHL starting goalie who is more consistent than your average goalie. We can also see that Spencer Martin has a 0.500 QBBS% which is average but he’s only had two Bounce Back Starts so we need more context and when we add his 0.925 bounce back SV%. From this we can gather that after Martin has a bad game he’ll at least provide an average game so the coaches won’t have to hesitate giving him the start right after a bad start.
This is just a general snap shot of one goalie almost halfway through the hockey season. We could go deeper into Spencer Martin and make his portrait more detailed. For instance we could compare his last AHL season to his current AHL season. In the 2015/16 season Spencer Martin had a 0.588 QS% and as mentioned above he now has a 0.619 QS%. That’s improvement you want to see especially taking into account that Martin only started 17 games in the AHL last season. Now he’s taken a starter role in the AHL with 21 starts so far with more to come and Martin is doing this at the age of 21 years old with the rookie tag still on him.
Of course this not the complete picture of Spencer Martin. This statistical portrait is no substitute for taking in a game and getting a live view of Spencer Martin’s (or any goalies) play. What these interpretations do though is help target goalies that might be undervalued or just underrated. Plus we can start doling out awards properly to goalies as well rather than relying on Goals Against Average, the W-L record and raw SV%. For instance Stephen Dhillon is looking like one goalie who should be considered to be drafted in 2017 or Jake McGrath who is having one of the best statistical draft eligible seasons for a first time 2017 draft eligible goalie.
This is the first season I’ve been tracking these bounce back statistics so using the statistics for long term projection is not possible yet. It is certainly a future project I would like to move forward on in the future. Unfortunately time is not my greatest ally right now as that time goes towards making money to pay for rent and food. Despite the lack of time I will continue to work on finding ways to properly evaluate goaltenders and make sure statistics along with analytics keep evolving. I’m excited though about these statistics and another project I’m currently working on.
As always if you have any questions or comments please feel free to contact me on Twitter @CreaseGiants.