You may not find a better crop of Free Agent Goalies than what is available this season. Let’s take a look at each available netminder.
Many like to call goaltending in hockey “voodoo”, meaning we can never predict what a goalie will go on an insane run, such as how Tukka Rask did in the 2019 Stanely Cup playoffs or on the contrary what goalie will pull a Martin Jones and put up solid results until 2018-2019 essentally costing his team putting up a SV% under 0.900%, being among the worst goalies in the league. So despite all of the voodoo associated with goalies I have ranked the 2019 UFA class in separate tiers from starter calibre to AHL calibre but using data from the past three seasons to get the best sample.
*Glossary of stats can be found at the bottom of the page.
Tier 1: Definitive Starter
Average SV% Above Expected-0.506%
Bobrovsky, the two time Vezina winner is without a doubt, a world class goaltender, and has been throughout his career. He also is the best goalie within this group available July 1st 2019. It is expected he will be given a monster contract for his consistent and elite level results. While I think he is easily a starter currently, I’m not certain that if I were a general manager of a current NHL team, I would hand him anything over four years at a maximum with his numbers clearly regressing as he ages. Furthermore, in 2016-2017, bobrovski had a 0.957 SV% above expected (SV%-xSV%), in 2017-2018 it was 0.352% and in 2018-2019 it was 0.21%. These numbers falling is probably a result of Bobrovski simply aging, here is a chart describing goaltender aging curves using GSAA (Goals Saved Above Average (League SV_perc * SA) – GA)) .
The question is, can Bobrovski sustain his Vezina calibre goaltending going into his 30’s? We will see for ourselves whether he is in the Sunshine State or somewhere in New York.
Tier 2: Probably Starter Calibre
Average SV% Above Expected-0.055%
Robin Lehner is a goaltender who is hard to predict what will happen with him next season. Historically over the past three seasons, his results did not scream “starter”, but last season he was truly a Vezina level goalie, landing third in voting and putting up respectable numbers all around with a very solid HDSV% in 2018-2019 with 82.67%. I could see Lehner being a starter next season and doing well but I don’t fully expect him to replicate a season such as the most recent one. One sense of optimism is his mental health seems to be in a lot better condition than it was in Buffalo before, which is a very important aspect of the game that goes under the radar in a sport known for the saying “just tough it out” in hockey.
Tier 3: Solid Backup
Average SV% Above Expected-0.325%
Mike Smith is probably the ideal backup on any team who has a definitive starter already, Smith has had a very consistent career of being a good goaltender whether he was behind a good or bad defensive structure. Smith is 37 years old and probably only has one year left in the NHL, but even with the age in account he is capable of being a backup goaltender for any team, posting a HDSV% of over 80% with 80.43%. Smith adds more value to the table by adding a good veteran presence in the locker room.
Average SV% Above Expected-(-0.776%)
Seymon Varmalov has been, when healthy, Colorado’s starter for the better of the past five or six seasons, putting up largely okay numbers, and in the past three seasons has been a below average goaltender which puts him within the solid backup category rather than the starter category where many would assume he would be placed. I would bet on Varmalov to continue to put up decent numbers and think any team looking for a backup able to play thirty games or more should definitely inquire on Varmalov.
Average SV% Above Expected-(-0.323%)
Petr Mrazek is about slightly below what would be you could call you average NHL goaltender which places him within the third tier. Mrazek has never had a full season of being terrible, nor has he had a full season of being amazing, just simply mediocre results across the past three seasons. Mrazek would be a good bet for a team to acquire that doesn’t have one definitive starter, to split the games 50/50 with the other goaltender, such as how Carolina just did last season with Mrazek himself. All in all I would put my money on Mrazek to continue to be the goaltender he has been throughout his career, that is, an average goaltender.
Average SV% Above Expected-(-0.161%)
Cam Talbot is an interesting case when it comes to evaluating him, after he and McDavid carrying Edmonton to the division final in the 2017 Staneley Cup playoffs and throughout the 2016-2017 season. Talbot has not put up stellar numbers since. In both the 2017-2018 and 2018-2019 season Talbot’s GSAA (Goals Saved Above Average) and SV% above expected have been negative or below the average goaltender. However these unattractive results could be as a result of the 86 games he played in 2016-2017 (playoffs and regular season combined), while facing 2015 shots, 146 shots more than the next goaltender. On the other hand, he is already 31 years old and well past his prime (a goaltenders prime is around 25) and he is set to regress. But combining all of the factors in play, I do believe Talbot could be a serviceable backup or half starter if he is behind a sound defensive structure such as Calgary, where he is rumored to be heading to. If I were a GM I would avoid paying him starter money because the results don’t show he is of starter quality.
Average SV% Above Expected-0.691%
Curtis McElhinney is among the most strange cases of goaltenders we have seen before (#GoaliesAreVoodoo), similar to this years Norris winner, Mark Giordano, McElhinney has peaked at a very unusual age. After having the worst career SV% among active goalies in the league, McElhinney has continued his three year journey of putting up fantastic results (I don’t think anyone saw this one coming). So as a result of the data from the past three seasons, McElhinney will be a solid backup for any team in need of a secondary goaltender. McElhinney probably only has a limited amount of time left in the league being 36 years old, but he would be of value for sure.
Tier 4: Mediocre Backup
Average SV% Above Expected-(-0.245%)
Keith Kinkaid is an interesting option for a backup, as he appears to be about an average NHL backup at this point in his career putting up mixed results over the past three seasons, but still NHL calibre results. Any team who misses out on the options above should take a shot at Kinkaid if they need a second goaltender.
Tier 5: Third Goalie (AHL Starters)
Average SV% Above Expected-(-0.595%)
Cam Ward is simply not the Cam Ward from 2006 when he was a rookie who won the Stanley Cup with Carolina. He hasn’t been for years and it’s extremely unlikely he will ever perform at a level as he did in 2006 again, given he is 35 years old and will only likely regress because of his age as a result. For teams who are looking for a goalie to be an emergency call up when injuries happen he is a great guy to have, but I think he will probably play in the NHL as a backup only because of his stellar play TEN YEARS AGO.
Average SV% Above Expected-(-1.313%)
Michal Neuvirth is a player whose existence as an NHL player is forgotten about. It may be because he was one of the Flyers sevens goalies this past season but don’t quote me on it. Neuvrith over the last three seasons hasn’t put up NHL calibre goaltending (was below replacement level) meaning you can bring in your best AHL goalie and he would probably put up the same results.
Index of Abbreviations:
SV_perc Save Percentage: 1 – (GA / SA)
FSV_perc Fenwick Save Percentage: 1 – (GA_ / FA)
xFSV_perv Expected Fenwick Save Percentage: 1 – (xGA / FA)
d_FSV_perc FSV_perc – xFSV_perc
GSAA Goals Saved Above Average: (League SV_perc * SA) – GA
GSAx Goals Saved Above Expected: xGA – GA_
Goals Above Average & Replacement (Goalies):
GAA_per_fenwick Goals Above Average per Fenwick Shot Against (the transformed coefficient from the regression. See below)
FA Total fenwick shots against
GAA Goals Above Average: FA * GAA_per_fenwick
GAR Goals Above Replacement: GAA + (FA * 0.007312411) // inverted replacement level GAA_per_fenwick
WAR GAR / season goals per win
Data is from, Micah Blake McCurdy 2019, https://evolving-hockey.com/, http://moneypuck.com/goalies.htm and https://www.corsicahockey.com/nhl/players/nhl-player-stats/goalie-stats