Tuesday, October 21, 2008

Predicted 2008-09 Save Percentages

As requested by one of the commenters in response to my post-lockout goalie ranking, here are some predicted save percentage numbers for 2008-09, based on a weighting of each goalie's results over the last 3 seasons (the weights were 50% for 2007-08, 33% for 2006-07, and 17% for 2005-06). I have included both a predicted shot-quality neutral save percentage number, as well as an expected raw save percentage number based on predicted team shot quality (the team numbers were weighted and calculated in the same way as the individual save percentage numbers). There is no subjectivity in these results, I expect several of these goalies to do better or worse than predicted, but it gives a sense of how some of these guys are trending and how their results could be impacted by the teams around them. Here is the list (sorted by predicted save percentage):

RankGoalieSQNSV%Save %


Anonymous said...

i also see carey price finishing the season with a high save percentage since montreal goalies seem to get pretty decent defensive support, compounded by the fact that he will likely split games with halak. so carey price will be this years poster boy for the i had a .925 sv% because i only played 55 games and had a good defense in front of me award

Number31 said...

I don't agree with that, anonymous. Both he and Halak face way more shots than they should be... When defense works, it stays in the mid-twenties. But in the Habs case, they usually allow too many quality shots and get outplayed or forget to play a few periods.

The season so far:
10/10 vsBuffalo Price 35/36 (+shootout 2 goals)
10/11 vsToronto* Halak 24/25
10/13 vsPhilly Price 29/32
10/15 vsBoston Price 31/34 (+shootout 3 saves)
10/18 vsPhoenix* Price 28/29
10/20 vsFlorida Halak 35/36
*considered weaker team

Normally, both have a trend of allowing 2-3 goals per game. (Based on previous season). There is a situation the coach should be addressing (especially after that Florida game). Only good thing is defense is usually good at clearing rebounds and share in shotblocking. The bad, allowing too many quality shots, turnovers, overall PK with a few 5-on-3s are very goalie dependent.

If you want a goalie with a good percentage thanks to a lock-down defensive system, I'd go more with the usual suspects: Brodeur, Backstrom (for the most part the Wild are pretty systematic), Osgood...

Anonymous said...

completely disagree, judging from your blogger name i can see where your bias would come from and it is understandable. but montreals defense doesnt limit shots, they limit chances. they also have adopted the hew nhl's version of the trap, the "puck possession" game.
although price has good numbers, so does halak. now are you really going to tell me that both these guys are 2 or the top 5 goalies in the league based off their save
%? that would be absurd. cristobal huet had the same numbers as carey price last year, and halak had better numbers, which again would indidcate that the team is a large reason why every goalie to play for the habs over the last 2 years has put up numbers well above the league average.
because the canadiens are a better team than most of their opponents, they can successfully play their system and their goallies will put up good numbers. only when montreal plays teams stronger than them are the goalies really tested and as seen last year with price, he really isnt as fundamentally sound as everyone wants to think.

Number31 said...

Eh, I never suggested they're the top goalies in the league. I'm just saying what is typically known as a Habs game being very goalie dependant, especially when they have a tendency to play only 1 period. For instance, if Marc Denis were doing the backup duty during that Florida game on Monday, I believe the outcome would have been 5-3 Panthers. If they get their act together maybe we'll see them notch a shutout with under 20 shots instead of 38...

overpass said...

Thanks for this list, CG. I think this list is interesting because as a prediction of the coming year, this list should roughly correspond to lists of top goalies that THN and other sources put out around this time.

I understand these aren't your predictions and the weights you are using aren't empirically derived, but the predicted SQNSV% number can at least serve as a summary of how the SQNSV% method would rank goalies currently, and in that way can be compared against the conventional wisdom.

For example, compare the SQNSV% list against this list from THN.


Unfortunately this list is from March, so it hasn't been updated for playoff results, but it's a pretty good proxy for the conventional wisdom in most cases.

If we take Predicted SQNSV% to be the "correct" or "most correct" ranking, who are the most underrated and overrated goalies on THN's list, as measured by differences in rank?


Name ..... SQNSV% THN Diff
Lehtonen .. 9 ..... 30 .. -21
Huet ........ 3 ..... 22.. -19
Backstrom. 4 ..... 18 .. -14
Legace .... 6 ..... 19 .. -13
Mason ..... 16 ..... 29 .. -13


Name ..... SQNSV% THN Diff
Nabokov ... 27 ..... 2 .. 25
Leclaire ... 25 ..... 4 .. 21
Miller ...... 21 ..... 8 .. 13
Osgood .... 17 ..... 5 .. 12
Kiprusoff ...23 ..... 11 . 12

I was also surprised to see Giguere at the top for predicted SQNSV%. I would have thought Luongo would be at the top, as he always seems to come out well from SV% analysis, but I guess he hasn't kept up his level of play.

The Contrarian Goaltender said...

Re: the Montreal debate, I think the answer is somewhere in the middle. The shot quality numbers seem to suggest that Montreal allows slightly easier than average shots, but they certainly aren't one of the best defences in the league. BehindtheNet had their expected save percentage at .909 last year. I follow Montreal well enough to know that they do give up a number of outside shots, but I think this has only a mild positive effect on save percentages. Huet, Halak and Price have had good numbers because they are all solid goalies.

Lehtonen .. 9 ..... 30 .. -21

What do you know, I was right!

Anonymous said...

tonights game against carolina was the perfect examply of montreals system making price look better then he is. he faced an above average number of shots, and the only two goals he gave up were on weak shots, i.e on from the point, the other from a sharp angle about 15 feet out. the "difficult shots he faced, were chances were carolina was still rushed, and of the 4 saves i would qualify as difficult, 3 of them were cases in which the puck was shot into him, and he would of stopped it had he not moved at all. realistically, price could of easily had a shutout tonight, and everyone would have wrongfully been praising him for a "dominant 31 save performance". this is what i meant earlier. you could have stuck any goalie in the nhl in the net tonight, and likely he would not have given up more than 3 goals at most, while most wouldnt have even given up 2. this is why guys like price and henrik lundqvist are overrated.