Wednesday, September 30, 2009

Percentages

There are two ways to win hockey games. Option A is outshooting the other team and winning through a greater volume of scoring chances. Option B is playing even or getting outshot and relying on superior percentages to end up ahead on the scoreboard.

In the long run, teams that outshoot generally have a better chance to achieve success. Over the course of a single season, however, it can be the team that rides a hot streak in terms of percentages that ends up on top of the standings.

The problem with outshooting numbers are that they are often skewed by the score (see JLikens' preview of the Colorado Avalanche for a more detailed explanation of how the score affects shot results). This effect is especially large in the third period, where teams like Colorado and Toronto had strong outshooting results because most of the time they were losing the hockey game. On the flip side, top teams like Detroit and San Jose noticeably shut it down in the third period, trading off some of their outshooting advantage to lower the overall number of scoring chances and increase the likelihood of preserving their lead.

To avoid being misled by the score effects, it is likely better for predictive purposes to look at the data set from the first and second periods only. Using the data from the Hockey Summary Project, I separated out each team's performance in the first and second periods only over the last two seasons. The teams are ranked in order of the projected number of points they could be expected to put up if they record the same shot for/against ratio, shooting percentage and save percentage in the first two periods in 2009-10.

RankTeamProj PtsSF/SASh%Sv%
1.Detroit1131.428.8%.902
2.San Jose1091.268.8%.910
3.Boston1040.979.6%.920
4.Montreal1030.9110.6%.916
5.New Jersey1021.098.4%.920
6.Chicago1001.059.7%.909
7.Pittsburgh990.9310.3%.914
7.Dallas991.0410.1%.905
9.Washington971.149.0%.904
10.Minnesota960.889.7%.920
10.Florida960.918.5%.928
10.Buffalo961.049.3%.909
13.N.Y. Rangers941.147.6%.916
14.Vancouver920.969.4%.912
15.Ottawa910.989.5%.908
16.Philadelphia900.899.9%.912
16.Carolina901.108.6%.905
18.Anaheim890.989.0%.911
19.St. Louis880.979.2%.908
20.Calgary871.019.3%.903
21.Edmonton830.8210.3%.909
22.Phoenix820.938.5%.914
22.Colorado821.038.7%.901
24.Columbus811.107.7%.906
24.Tampa Bay810.969.1%.904
26.Nashville770.978.3%.907
27.Los Angeles750.978.6%.901
28.Toronto731.058.5%.891
29.N.Y. Islanders720.947.7%.911
30.Atlanta670.799.5%.903

These numbers are not meant to be exact predictions, just historical data. Many teams have turned over much of their rosters and/or coaching staffs since 2007-08, which may mean that some of the numbers aren't very useful for assessing the 2009-10 squads. However, the SF/SA ratio should still have some value in predicting results. The correlation between SF/SA over the first two periods in 2007-08 compared to 2008-09 was 0.79. If a team is returning most of its key players and has the same coach or a similar coaching philosophy there is good reason to expect a similar SF/SA in 2009-10.

The numbers for Detroit and San Jose show the advantage of majorly outshooting the opposition. Detroit demonstrated last year during Chris Osgood's early-season struggles that they could win despite weak goaltending. As long as the Red Wings and Sharks continue to pepper opposing goalies with shots they will win games even with mediocre percentages. In contrast, Edmonton and Atlanta will have almost no chance of making the playoffs unless they drastically improve their territorial play, because it is very hard to compensate for the handicap of facing 20% more shots against.

This chart allows us to see the teams that are relying on percentages rather than outshooting for their success. These teams are more likely to be susceptible to large swings in the standings depending on the luck of their shooters/goaltenders. For example, Montreal's fall from Conference Champs in 2008 to 8th seed in the 2009 playoffs was almost entirely driven by percentages (while Boston's ascent was also pretty much the same story, in reverse). Along with the Habs, Pittsburgh, Minnesota, Florida and Philadelphia are the teams that have been most dependent on having the percentages in their favour.

I'd expect Pittsburgh to be a better outshooting team this year, based on the way they performed late in the season and in the playoffs under Dan Bylsma. If they are they should finish at or near the top of the Eastern Conference.

Philadelphia could also see their outshooting numbers rise in 2009-10 because of additions and the continued maturation of some of their young stars, although I expect a drop in team save percentage might very well wipe out some of those gains. The Flyers are a popular pick among many pundits heading into this season, but I'm not sure they are an elite team yet. However, these numbers might also be selling Philadelphia a bit short, given that they are for the first two periods only and the Flyers were a strong third period team over the last two seasons.

On the other hand, there are several teams that have been outshooting the opposition but have not had as much luck with the percentages. These teams include the New York Rangers, Carolina, Columbus, and Toronto. We saw with both Carolina and Columbus in the second half of 2008-09 what happens when an outshooting team goes from poor percentages to good percentages. The improved goaltending of Cam Ward and better Blue Jacket shooting took both teams into the playoffs. If Columbus continues to improve their scoring ability they could move into the second tier of good teams in the Western Conference, behind Detroit and San Jose.

The Toronto Maple Leafs are an interesting team, since goaltending was the glaring problem that played a big role in sinking their last two seasons. The Leafs were .010 behind every other team in save percentage, which naturally meant they usually were playing from behind. League average goaltending this season probably makes Toronto a playoff contender. It appears the hopes of Leaf fans are riding on the shoulders of Vesa Toskala, or, perhaps more likely, Swedish rookie Jonas Gustavsson.

The other team that was held back by goaltending in 2008-09 was the Dallas Stars. Marty Turco is past his prime but he is still likely to be better this season, and the Stars should be back in the playoff picture.

The Buffalo Sabres have looked like a playoff team in terms of their underlying numbers for the past two seasons. They are slightly above average in all three of shot ratio, shooting percentage and save percentage. Unless somebody has a better explanation I'm inclined to attribute their lack of playoff qualification to bad luck, and the Sabres should have a good shot at returning to the postseason.

I'd say that Edmonton, Phoenix and Colorado are unlikely to make the playoffs in the West. In the East it should be safe to write off the Islanders, and Atlanta is facing a steep uphill climb unless they get some major improvement from their young players. Other than that the picture looks pretty wide open. The elite teams should end up at the top of their respective conferences again (Detroit, San Jose, Chicago, Pittsburgh, Washington), but percentages will play a big role in determining which of the rest of the teams get a chance to play for the Cup.

Friday, September 25, 2009

Conn Smythe Trophies

I think the worst commonly used argument for ranking Patrick Roy over Dominik Hasek is the oft-invoked "4 Cups, 3 Conn Smythes vs. 1 Cup, 0 Conn Smythes". I've dealt with the issue of team success pretty much all along in this space so you've heard those arguments before, but I also don't like it when players are evaluated based on the number of Conn Smythes they win.

There isn't a more situational award in all of hockey than the Conn Smythe Trophy. I'm not trying to diminish the importance of playing great in the playoffs, I just don't believe that one's Conn Smythe haul is a good way to evaluate playoff success or to compare the relative performance of different players. The binary nature of the award (winner gets 1, all others get 0, only 1 award handed out every year) makes it more arbitrary. If you ranked all the playoff performances of the 1990s for every single player, for example, you might have three players from 1993 end up in the top 10 (Patrick Roy, Wayne Gretzky, Doug Gilmour). Obviously only one of them can win that year's MVP award, meaning that the others end up, by "Smythe only accounting", ranked below players from weaker postseasons that weren't at their level of play. Was Cam Ward in 2006 better than Dominik Hasek in 1999? John Vanbiesbrouck in 1996? Martin Brodeur in 2003? Olaf Kolzig in 1998? Ward's the only one of those guys with a Smythe, even though I wouldn't rate his playoff as even close to any of the others.

The rest of the team is obviously a big factor, since in almost all cases the Conn Smythe Trophy winner is from the Cup champion. To win the Stanley Cup, a player needs to have great teammates around them. There have been a number of absolutely terrific playoff performances by players whose teams ended up losing in the second round or the Conference Finals. Gilmour in 1993 and Peter Forsberg in 2002 are two of the most commonly cited examples, and we can also point to the more recent efforts of Alex Ovechkin and Jonas Hiller in the 2009 playoffs. As such, by simply counting Smythes you are at least partially giving double credit for team success. Being the best player on a Stanley Cup winner is a notable achievement, but that by no means signifies that player was the best player on any team.

A player's Conn Smythe chances depend a lot on who else happens to be playing great in that particular playoff season. In the 2009 playoffs, Sidney Crosby scored 15 goals and 31 points, which is more than anyone has recorded in either category since Joe Sakic in 1996. That sounds pretty deserving of some hardware, but unfortunately for Sid the Kid his teammate Evgeni Malkin scored 36 points in the same postseason, which means Malkin gets a Smythe while Crosby gets nothing. In other years there have been no singularly outstanding candidates and the writers more or less picked somebody out of a hat (2007 is a good example, the Ducks likely split the votes among a number of good candidates). If Crosby had the same playoffs as he did this year in 2006, 2007, or 2008 he would have easily won the Conn Smythe, probably even if his team lost the Final. Therefore, giving no credit to the second place guy makes little sense.

I also dislike the excessive focus on the Stanley Cup Finals in Smythe voting. Over a short series, luck has a big impact. A player can be driving puck possession and creating scoring chances and yet not have his shots or his linemates' shots hit the net for a few games in the Finals, throwing him out of Smythe contention. Again Crosby is a good example, as in the 2009 Finals he scored 1 goal on 16 shots while linemates Bill Guerin and Chris Kunitz combined for 0 goals on 32 shots. Sometimes the puck luck is against you, and that's pretty much the only conclusion you can make when a top line shoots 2% and hits more posts than they score goals over an entire series. Crosby's 3 points in 7 games cost him the Conn Smythe, but it looked to me like he was doing the same thing he had been doing all playoffs. The difference was that he was playing against tougher competition and the shots weren't going in.

The last thing is that, like any awards voting, sometimes the voters make debatable choices. I don't agree with the voters' choice of Cam Ward over Rod Brind'Amour or Chris Pronger in 2006, or Joe Nieuwendyk over Hasek in 1999, to give just a couple of examples.

If I was running the league, I would make the Conn Smythe voting results public so we could see who was finishing in 2nd and 3rd place. I would also adopt all-star voting for the Stanley Cup playoffs, like they do in international tournaments and at the end of the regular season. This would allow credit to be given to the players who were the best at their position but who may not have ended up with a ring because their teammates weren't as good.

Here is my attempt at putting together postseason All-Star teams for each season since the lockout:

2009:
G: Jonas Hiller
D: Nicklas Lidstrom, Brent Seabrook
F: Evgeni Malkin, Sidney Crosby, Alex Ovechkin

2008:
G: Marc-Andre Fleury
D: Nicklas Lidstrom, Sergei Gonchar
F: Henrik Zetterberg, Sidney Crosby, Johan Franzen

2007:
G: Roberto Luongo
D: Nicklas Lidstrom, Chris Pronger
F: Daniel Alfredsson, Sami Pahlsson, Dany Heatley

2006:
G: Dwayne Roloson
D: Chris Pronger, Scott Niedermayer
F: Eric Staal, Rod Brind'Amour, Patrik Elias

I doubt the Conn Smythe voters would have agreed with me on those teams (I left 2 of the 4 trophy winners off of my teams entirely). If you disagree with my selections, by all means take a shot at it in the comments. However, regardless of disagreements over a few individual cases, awarding All-Stars would give players like Crosby and Lidstrom the credit they deserve for their outstanding playoff performances.

Monday, September 21, 2009

Don't Always Believe Your Lying Eyes

How do sportswriters vote for the year's best goalie? I bet they'll tell you they evaluate lots of different things, like how good he is at stopping the puck, how good he is at controlling rebounds, his mental toughness, how much he helps his team win and a bunch of other criteria that sound pretty good. They'll watch lots of hockey and check out a lot of different goalies, enough that they should be in a pretty good position to be able to evaluate them. And then who do they often end up picking? The goalie on the best team, or at least the best defensive team. That's what they've been doing for years.

If you asked sportswriters the same question back in the 1960s they'd probably give you a similar rundown, describing in detail some technical aspects of Jacques Plante that made him great, or how competitive Terry Sawchuk was in the clutch, or how much Glenn Hall helped his teams win. Some of them no doubt simply copied down the wins leaders, but I bet that a majority of the journalists sat down every year with their detailed opinions and jot notes and went through their memories of the games they had watched and tried to evaluate a dozen different subjective things. They then mailed off their ballots to league headquarters, certain that they had taken everything into account and identified the league's best goaltender.

Were they sincere? I think they were. Yet give me a random stat sheet from one of the seasons back then, with the names of the goalies blacked out, and from that alone I can predict with over 80% accuracy who the voters thought was the best goalie that season. If it wasn't for one particular outlying goalie it would be almost a sure bet.

From 1935 to 1970, there is a very simple algorithm to determine the First Team All-Star goalie. Look at all the goalies who played in at least 75% of their team's games, rank them by goals against average, and take the guy at the top of the list. That's the First Team All-Star. That solution worked in 30 out of 36 seasons, all of them except for 1957, 1958, 1960, 1964, 1968 and 1969. Here are the goalies who were named First Team All-Star ahead of the high-minutes GAA leader in those seasons:

1957: Glenn Hall
1958: Glenn Hall
1960: Glenn Hall
1964: Glenn Hall
1968: Gump Worsley**
1969: Glenn Hall

**-Led the league in GAA by a wide margin but did not play in 75% of his team's games during a season where all of the top 4 guys in GAA were in platoon situations.

Glenn Hall was quite obviously the exception to the rule. That he was the only guy to get named the league's best goalie without playing on one of the league's top defensive teams reflects well on him, and it's even more impressive that he did it 5 times. That shows that he was considered a difference-maker and is evidence of why he is considered one of the best goalies to ever play the game.

Other than Hall, though, it was pretty much a succession of GAA leaders that were named the season's best. Whether or not the voters made a conscious, stat-based decision to rely on GAA in their rankings or whether their subjective views were impacted by the fact that the goalies they watched were less likely to allow goals than everyone else is unclear, I'd guess it was a combination of the two but in the end the result is the same.

That certainly does not mean that the voters never got it right, of course. There were some seasons, perhaps even quite a few seasons, when the best goalie played on the best team or where strong goaltending made the difference between a good and a great defensive team. It remains quite obvious to anyone with a reasonable view of a goalie's role within a hockey team that the best goalie does not always allow the fewest goals against.

The voters were no less predictable in the 1970s, but what had changed was that it became common to platoon goalies. As a result, very few goalies played in 75% or more of the team's games. The voters still, whether consciously or subconscously, voted in the GAA leaders. In every season from 1971 to 1979 the same goalie won the Vezina (which under the old definition went to the starting goalie on the team that allowed the fewest goals) and was named the First Team All-Star.

I'd say the voting has improved at least somewhat in recent years, perhaps partly due to the introduction of the modern Vezina Trophy that allowed GMs to vote on the year's best goalie as well as the sportswriters. Yet even the game's insiders are not immune, as the starting goalie on the team with the fewest goals against is still overrepresented in both recent All-Star and Vezina voting. If you are the starting goalie on the team with the fewest goals against and you play in at least 50 games, you're still almost guaranteed to be named First or Second All-Star at the end of the season. The only post-expansion exceptions to that rule both came in the last decade (Ed Belfour in '99 and Roman Cechmanek in '03, although Cechmanek shared the Jennings with First Team All-Star Martin Brodeur so it's debatable whether that one should even count as an oversight). These results suggest that the bias towards goalies on strong defensive teams is still alive and well.

Some people will tell you that all goalie statistics are flawed because the rest of the team affects them, which means that the best way to evaluate goalies is to watch them play. It is true that a goalie's statistics are affected by his teammates, but what they are failing to take into account is that the rest of the team also affects how a goalie is viewed. This is especially true in the NHL where the margins between goaltenders are so narrow. I'd bet for many sportswriters and most hockey fans this effect is considerably more significant than any team-to-team differences in shot quality. A goalie on a good team that doesn't let in many goals and makes a lot of saves with his team in the lead usually looks better than a goalie on a bad team who faces more shots, allows more goals, and spends more time with his team trailing.

I think that at the NHL level there is some value to goalie scouting, but that there is likely more value in numerical analysis. That's especially true if fans or teams are not able to watch and grade the goalie in every game, which means they will end up seeing certain guys good and certain guys bad. There are other issues with evaluating goalies by watching them, such as the impact of a goalie's style and whether a viewer can accurately assess the importance of various goalie skills (I'd suggest things like rebound control and puckhandling tend to get overrated since they are more obvious and easier to grade than skills like positioning, post play, a goalie's ability to block screened or close-in shots, etc.).

If I was forced to pick only one of the two methods of evaluation (stats vs. watching), I'd prefer to take the numbers. Of course I'd rather watch a guy as well to see if there are any nuances that can be added to what the statistical record is telling me, but the moral of the story is this: Always be aware that your eyes can lie.

Sunday, September 13, 2009

Investigating Special Teams Skill

Is penalty kill goaltending a distinct skill from even strength goaltending? We know that skaters show differences in ability on special teams compared to at even strength, which suggests that 4 on 5 hockey is a different game than 5 on 5. If so, maybe there are differences in goaltending results as well. Just anecdotally, there are a few goalies that seem to do better at even strength than on the penalty kill. J.S. Giguere comes to mind as someone I've run across recently who fits that profile. However, we would expect a few goalies in the entire population to have some weird results just by chance or because of possible team factors, and even if a few guys show a difference in ability it does not mean that is typical around the league. We need to look at some overall numbers over a larger sample size to get some better evidence for or against this potential effect.

I checked the correlation between even strength save percentage and penalty kill save percentage at the team level, using the 10 seasons' worth of special teams numbers from NHL.com. The correlation coefficient was 0.71, which is a high result that suggests a fairly close relationship.

With a correlation that high, we should be able to use EV SV% to predict PK SV% with some reasonable accuracy. I used Excel to generate a linear equation to predict PK SV% from EV SV%, and compared each team's results with the expected. The predictions were pretty good. Only 7 out of 30 teams had a difference between actual and expected PK SV% greater than .006. They were the following:

Minnesota: .886 PK, .878 EXP (+.008), .925 EV
Detroit: .875 PK, .866 EXP (+.009), .917 EV
Washington: .875 PK, .864 EXP (+.011), .915 EV
Columbus: .872 PK, .862 EXP (+.010), .914 EV
Boston: .861 PK, .869 EXP (-.008), .919 EV
Phoenix: .858 PK, .869 EXP (-.011), .919 EV
Atlanta: .849 PK, .859 EXP (-.010), .911 EV

There's Minnesota again, with yet another result that suggests they are doing something abnormal shot quality-wise. Another familiar team with some shot quality outlier evidence behind them are the Atlanta Thrashers, who also make their usual appearance at the bottom of the table.

It could be that some teams were excellent or weak for a few seasons, but over a longer time frame that tends to get mostly washed out in the average. A few teams appear to have persistently higher or lower results than expected, which suggests that the penalty kill situation is very similar to the one at even strength: For most of the league there is little shot quality effect, although there are apparently a few outliers at each extreme. It seems that for the most part goalie performance is similar at even strength and on the penalty kill. Either that or teams tend to have similar shot quality against at EV and on the PK. One other factor that would impact PK shot quality is the number of 5 on 3 chances against. I suspect this is one of the reasons Minnesota ranks so high. With their excellent team discipline they were less likely to have to face 5 on 3 situations.

I also checked the correlation between EV SV% and PP SV%, and it was essentially zero (0.01). The sample size for power play shots is a lot smaller, as the most shots any team faced over the last 10 seasons while on the power play was 1,096, which is about half a season's worth of shots for a 70+ game starting goalie. I'm not sure a goalie's save results while his team is on the power play are particularly meaningful.

What these results suggest to me is that we should rate goalies based on their performance both at even strength and on the penalty kill. Focusing on EV only would be underrating the Lundqvist/Luongo types that are likely adding a lot of value through their play on the penalty kill. It does not appear that there is much of a skill difference by game situation for goalies. It is more difficult for all goalies to make a save when their team is down a man, but a better goalie is still generally more likely to make that save than a weaker goalie. We certainly need to take into account the number of power play shots against for each goalie, to level the playing field for goalies on teams with higher or lower penalty totals. Just as at EV, it is likely that in today's NHL most teams around the league are pretty similar in terms of shot quality against on special teams, although there are probably a few outliers that we need to keep in mind. Looking at situational save percentages for even strength and 4 on 5 play only would probably be the best way to go.

Wednesday, September 9, 2009

Trending Up: Carey Price

Down Goes Brown has a great post on why hockey predictions are almost always wrong. I generally try to avoid making predictions for similar reasons, especially at the team level where there are just too many variables and too much randomness to worry about. The chances get a little better when we start looking at individual players, although even then the percentages over a single season have a lot of impact on those results (hands up, everyone who thinks Loui Eriksson is going to score 36 goals on 178 shots again). However, I am going to throw caution to the wind anyway and make one prediction for this upcoming season, because I think that it has a very high likelihood of being correct. That prediction is that Carey Price will have a bounceback year in 2009-10.

In case anyone has forgotten over the long offseason, Carey Price burst onto the scene in 2007-08, putting up a 2.56/.920 line for the Conference champion Montreal Canadiens and performing admirably for the first round of the playoffs against Boston before hitting a hiccup or two against the Flyers. Price's numbers then took a significant step backward last year, dropping to 2.83/.905. Price then went 4.11/.878 in the playoffs as the injury-riddled Canadiens were completely outclassed by the top-seeded Bruins.

Based on that two season sample, the trend doesn't look particularly good for Carey Price. Some put a lot of weight on his playoff sample especially, and as a result are skeptical of Price's ability to become a top goalie in the NHL. I think that opinion is likely misguided. There are a number of indicators that I like to look when trying to estimate future performance, and they seem to be positive across the board for the Habs' young netminder.

The biggest factor is Price's age. He just turned 22 on August 16. When we look at guys like Price and Steve Mason, we can sometimes forget that few goalies play well at the NHL level in their early 20s. I looked at how all the goalies in the league in the post-lockout period did at different ages (excluding Price himself):

age 21: .896
age 22: .899
age 23: .907

Most goalies aren't in the NHL at age 21 or 22, and the ones that are there don't usually put up good numbers until they are 23 or 24. Even with his struggles in 2008-09, Price is well ahead of the development curve. It is probably reasonable to expect him to be improving for the next couple of seasons, just as we have seen Marc-Andre Fleury and Cam Ward take steps forward in recent years. This isn't guaranteed, of course, since not every 21 year old goalie improved in his age 22 year old season (the most spectacular example being Jim Carey). It is, however, the most likely scenario.

If we look at Price's situational results from 2008-09, we see that he was actually above average at even strength (.920). What hurt him the most was an .843 save percentage on the penalty kill. Penalty kill results tend to be more random, since the sample sizes are smaller. There is little reason to expect a very good or a very bad PK number to carry forward to the next season, unless the goalie is especially strong or weak (or if we have reason to believe that the team's PK unit is an outlier). In 2007-08 Price had an .885 PK save percentage, which suggests there is little reason to think that he isn't able to perform at 4 on 5. He is very likely to improve on that .843 this season, which will boost his overall numbers.

Price's even strength performance so far in his career has been quite strong. Over the last two years his EV SV% ranks 9th in the league among goalies with at least 80 starts. The only goalies in front of him are pretty much the cream of the crop at the NHL level: Thomas, Luongo, Vokoun, Brodeur, Fleury, Giguere, Bryzgalov, Backstrom.

It is difficult to predict how well the team in front of him will play, given Montreal's offseason overhaul. Montreal did hire a new coach, Jacques Martin, who is known for his focus on systems and defence. The team also improved its blueline depth and should be stronger at even strength. Whether the Habs are a better team next year I'm not sure, but I think there is a good chance that they will be better defensively, which would make Price's job easier.

Finally, streaks can have a big impact on the success or failure of an individual season. One terrific or one awful month can put a goalie either in the running or out of contention for the Vezina. Extreme results like that are less likely to occur the next season. For Carey Price this is a good thing, since his season wasn't really all that bad outside of two months. He had a strong start, was awful in January and February, and then recovered towards the end of the year.

Here is the breakdown of Price's career so far, splitting out January and February of 2009:

Jan and Feb '09: 2-8-1, 3.82, .866
Rest of career: 45-20-12, 2.54, .918

Price was coming back from an injury, which was probably mostly responsible for his sudden regression. Jaroslav Halak also had a pretty forgettable January (3.53, .888), which suggests that there were more things going wrong in Montreal at that time than just the goaltending. It seems unlikely that those factors will converge in the same way in 2009-10. I think the "rest of career" line is a lot closer to Price's true ability than that two-month sample returning from injury on a slumping team, and therefore it is more likely to predict what Price does in the future. Other than his nightmarish February, Price has been at .909 or better in every month of his career in which he has played 300 or more minutes.

One thing I find particularly interesting is that Price's 2009 struggles led to widespread criticism that he was not mentally tough enough to handle the pressure in Montreal. This despite most scouts agreeing that Price's unflappability and mental strength were among his best attributes as a prospect. It is possible that some of Price's playoff performances were adversely affected by the pressure, but it seems unlikely that his biggest strength would suddenly become his biggest weakness almost overnight.

In my opinion much of what is said about a goalie's mental abilities is merely an attempt to rationalize randomness. That is not to say that there aren't any goalies who are mentally tough or mentally weak, just that a few bad months or a playoff series here or there are not nearly enough evidence to make that claim. Especially when the goalie in question is playing in Montreal, where the "he can't handle the pressure" narrative is the journalist's go-to rationalization for poor performance. Over the last 10 years, Montreal ranks 4th in the league in total save percentage, which doesn't really suggest that its goaltenders have been stumbling under the intense pressure. That's still apparently not enough to dissuade journalists from writing sparkling copy about Montreal's goalie being the "The Loneliest Man in Sports".

The signs seem to point to good things for Carey Price in 2009-10. Nothing is certain, of course, since goalie performance is variable and most young goalies don't develop in a straight line, but I think it's a pretty good bet. I doubt Price's winning percentage will be quite as high as it was in 2007-08, unless I'm misjudging Gainey's moves, but his other numbers could easily return to a similar range (~.915) this coming season.

Friday, September 4, 2009

Rebound Control

Another Behind the Net post I have been meaning to comment on is this comparison of rebound shot rates for goalie tandems. Gabe Desjardins posted the data more or less without comment, and since I think the numbers are worthy of being discussed in some more detail I'll take a stab at it here. There are a few things that can be done with this data, but the most obvious one is to use it to estimate the range of individual goalie rebound skill. There are also numbers for two teams that have often been discussed at length on this blog (New Jersey and Atlanta), which gives us more evidence of the relative skills of those goaltenders.

There is an average of about a 1% difference among goalie teammates in rebound percentage at even strength. The average shots faced was 845, so that means that a typical difference in the number of rebound shots faced between two goalies who split starts right down the middle would be about 8. Figure that about a quarter of those shots go in, and that's a 2 goal difference in a platoon scenario. That's not a whole lot of margin, and given the small sample size those numbers are likely to be subject to a high degree of randomness and luck.

The biggest gap observed was between Jonas Hiller and J.S. Giguere. Of the shots against Hiller, 7.9% were rebounds (shots within 4 seconds of the previous shot), compared to 5.78% of Giguere's. That's a difference of about 18 rebound shots or likely around 4-5 goals, assuming both would have a similar rebound save percentage. If Hiller is giving away 4 extra goals per season from extra rebounds and Giguere has a .920 save percentage, than Hiller needs to be at just .924 or better to make up the gap.

There are a few factors that need to be considered before we go ahead and claim that the goalie with the lower rebound shot percentage is better at controlling shots. We would expect to see the largest rebound differentials when comparing goalies with high save percentages to goalies with low save percentages (e.g. Hiller vs. Giguere in 2008-09). The reason is that the goalies with high save percentages make more saves, and in particular will tend to make more tough saves. Not every shot has an equal likelihood of a rebound. Every goalie in the league is able to deal effectively with easy shots. It is more difficult to control the rebound on a difficult save, and therefore rebounds are more likely to come from tough saves. If rebounds come primarily from tough saves, and good goalies are more likely to make tough saves than bad goalies, then we would also expect the good goalies to give up additional rebound opportunities after stopping pucks that weaker goalies wouldn't have touched in the first place. This seems to be supported by the data, as the goalie with the higher save percentage generally also has a higher percentage of rebound shots against.

This factor likely explains the Hiller/Giguere differential. Both play a similar goaltending style and are not noted for their ability to control rebounds. Another potential factor is whether the team adjusts its defensive coverage to reflect a goalie's rebound control skill. I doubt that was the case in Anaheim, but it might be a possibility for some other teams where there is more of a difference in goalie skill sets.

Finally, there is likely to be differing abilities to stop rebound shots. Dominik Hasek, for instance, was an innovator in dealing with second-chance opportunities, and would certainly have been better than average at getting in front of rebounds. The problem is that the rebound shot sample sizes are so small that we would likely rarely know with any certainty if one goalie is actually better than another. I also suspect that in most cases the differences in rebound saving skill is relatively minor.

Taking all aspects of rebound control into account (both the number of rebound shots against and the number of rebound goals against), the overall effect is likely to be worth only a couple of thousandths to a goalie's save percentage. Rebound control becomes a useful tiebreaker for goalies with very similar abilities, but most of the time it appears that the goalie who is better at making the first save should play. This is especially true if the goalie's defensive teammates are able to make a defensive adjustment to cover the areas where rebounds are likely to end up.

Rebounds also seem to have a very limited effect on the number of shots a goalie faces. If Hiller faces 18 more rebound shots than Giguere in half a season, that equates to an additional 0.4 shots per game, and that's the highest observed value. For most goalie tandems, the difference is more like 0.1 or 0.2 shots per game. Assuming that my estimate of +/- 1 shot as the typical range of goalie shot prevention is correct, rebound control is apparently a very minor variable in the shot prevention equation.

Let's look at New Jersey and the results for Martin Brodeur and Scott Clemmensen, two goalies generally seen as on opposite ends of the spectrum of rebound control ability:

Martin Brodeur: 629 SA, 32 shots within 2 sec (5.1%), 42 shots within 4 sec (6.7%)
Scott Clemmensen: 869 SA, 33 shots within 2 sec (3.8%), 44 shots within 4 sec (5.1%)

What we see, somewhat surprisingly, is that Clemmensen faced fewer rebound shots than Brodeur, by both definitions of rebound shots. I don't think this means that Brodeur is actually worse than Clemmensen at controlling his saves, but merely that there are other factors involved. It is possible that Brodeur made more tough saves that were likely to lead to subsequent rebound chances, but given that they had similar overall save percentages that is not likely to account for the difference.

What I would speculate is more likely to be a significant is New Jersey's defensive play. It makes sense that the Devils would take a more active role in clearing rebounds with Clemmensen in the net, knowing that there were likely to be more second-chance opportunities. Looking over the pairings in Desjardins' post, it seems that the effect of team context is generally larger than the effect of the individual goaltender, i.e. there are generally larger differences between teams than between goalies on the same team. The ability for the team to compensate could explain why there appears to be little margin in goalie rebound control.

Brodeur was not only outperformed by Clemmensen, he was actually among the leaders in highest percentage of rebound shots faced. Could it be that his rebound control is somewhat overrated, possible team effects notwithstanding? It is possible that his teammates have done a good job over the years of allowing easier shots against (including fewer power play shots against) and clearing pucks and made him look better than he actually is. Or it might be possible that his age is having some effect on his reflexes and reaction speed. Or maybe Brodeur just had an unlucky season, or the numbers are a result of making more tough saves or his team's defensive play. Either way, I think rebound control has little to do with Brodeur's success in assisting his team's shot prevention.

The numbers for Atlanta are also worth discussing, given that the Thrashers employ the two goalies with the largest shots against gap in the league over the last few seasons (Kari Lehtonen and Johan Hedberg).

Lehtonen: 1068 SA, 40 shots within 2 sec (3.7%), 66 shots within 4 sec (6.1%)
Hedberg: 621 SA, 22 shots within 2 sec (3.5%), 38 shots within 4 sec (6.1%)

The Atlanta duo had almost identical numbers, which does not support the theory that the reason that Lehtonen faces more shots against per game is because of additional rebound shots. Lehtonen's numbers actually appear to be quite strong, given that his save percentage is so much higher than Hedberg's.

Looking at the big picture, these numbers seem to confirm my contention that rebound control is an overrated skill. Because of its visibility it is something that observers often focus on, but the actual rebound shot frequencies don't suggest that it is important as many claim.

This analysis does not consider any possible indirect effects of rebounds controlled or allowed, such as for example whether rebound control might be something that impacts a shooter's decision to shoot from a bad scoring location, or whether some goalies have an effect beyond the 4 second mark by doing a better job of directing pucks to their teammates instead of putting rebounds back out in areas where the other team can get to them and maintain possession in the offensive zone. I wish the NHL still tracked the time the puck spent in each zone, which would allow us to investigate possible indirect effects. I'd guess that these factors may have a slight impact but not enough to make much of a difference in overall terms.

Wednesday, September 2, 2009

More on Minnesota and Shot Quality

Gabe Desjardins at Behind the Net has had a strong summer, posting a number of interesting items. There's a couple of them that I want to highlight in case anyone hasn't seen them and since they might lead to some good discussion.

One of them is his look at Minnesota's shot quality against. He illustrates the shot locations graphically, and also presents quotes from the Wild's director of hockey operations that show that according to the team's own evaluators they give up less dangerous chances than most teams do. This speaks directly to one of the counterarguments by Vic Ferrari (quickly paraphrased: "Why would the Wild pay Niklas Backstrom $6 million a year if we know they track scoring chances and other things and since they seem to know what they are doing?"). I agree that the Wild seem to know what they are doing, but Desjardins' quotes show that they seem to be aware of shot quality effects as well.

Desjardins also recently posted shot quality numbers for various coaches. Not surprisingly, Lemaire and Hitchcock show up at the top of the list, while someone like Quenneville, who I've identified before as a coach who apparently has a negative save percentage effects on his goalies by using a before/after analysis, shows up near the bottom. The list seemed to have a pretty close similarity with the subjective perception. That suggests there is something real behind the numbers. In fact, this makes it look like coaches might be quite significant in terms of a team's defensive system. If so, that could be one of the reasons why I found little difference between goalies who switched teams. A goalie playing on the same team with a new coach could be in a environment that is just as dissimilar to the previous year as if he was traded to another team.

There was also this post by James Mirtle where he looked at overall team save percentages for the last 10 years and post-lockout. Minnesota was .005 clear of the field in both periods. The case can certainly be made that Minnesota had good goaltending, but I think it is pretty clear they did not have goaltending that was that much better than the rest of the league. Part of this figure is related to power plays against, of course, and as a disciplined team Minnesota has an advantage in this regard.

I think there's enough good evidence to suggest that Minnesota had a significant shot quality effect under Lemaire. That doesn't mean that Backstrom, Roloson, et al are mediocre goalies who lucked their way to the top, just that they got some help on the way. Strong goaltending was also a contributing factor to the Wild's stellar defensive record over the last decade. It will be interesting to see how Minnesota plays post-Lemaire, and whether Backstrom's numbers will continue to be very good.

In the EV shot quality debate, I'm currently on the side that says that for most of the population it doesn't appear to be a significant variable but that there are a few outliers at both ends of the scale. I'd guess that there are something like 6-8 teams with shot quality effects that are big enough to worry about, either positive or negative.