During the first period of tonight's Rangers/Devils game, they showed a graphic for Henrik Lundqvist's stats for his last seven starts. The topic brought on this exchange between the two commentators. These aren't exact quotes. I'm just trying to type what I heard, but it's close enough. Also, I don't know the names of the commentators, so we'll use some completely generic names.
Pilot Inspektor: "Is there a save percentage level that really defines the good goalies from the so-so goalies?"
Paradorn: ".920 is really the minimum save percentage to have to really say you're giving your teammates a chance to win. A really good save percentage level is around .940."
Ok. So three goalies in the league so far this year are posting "really good" save percentages. Are you ready? They are Anton Khudobin, Mike Brodeur, and Justin Peters. This trio has a combined 8 starts this year. I believe it might be worthwhile to eliminate Mr. Tremendousson's brother from another mother, small sample size. Let's try that.
Using a 20 game minimum as a requirement(which is still pretty conservative, since that's less than 1/3 of the club's games thus far), there are zero goalies that have a "really good" save percentage. In fact, Ryan Miller's league-leading .932% isn't even close to being "really good." Scoring must be up this year, because these lousy goalies just don't stack up. Vokoun? .930%? Don't make me laugh. Ha ha ha...too late!
Here is a list of the top save percentages in the NHL since save percentage became a stat, with a minimum of 30 games. Hasek's .937% in 1998-99 tops the list. Apparently we have yet to see a "really good" goalie. Can't wait! I know this guy Shane, and he's looking really promising. I don't know his last name, but I'm just saying, if there's a Shane in the NHL in a few years posting .940+ sv%, you read it here first. Also, there's this one kinda small guy who always wears a blue jersey...he's a beast when it comes to ball hockey. Speaking of small guys, there's another one who doesn't like to pay league fees. I heard he was interested in Matt's DVD player though.
Now, let's take a look at how many goalies are posting a .920+ save percentage, using the 20 game minimum as a guideline. Here's the list.
Ryan Miller, Tomas Vokoun, Evgeni Nabokov, Tuukka Rask, Jaroslav Halak, Jimmy Howard, Miikka Kiprusoff, Craig Anderson, Henrik Lundqvist, and Ilya Bryzgalov. Ten goalies are doing the "minimum to really say they're giving their teammates a chance to win," and two of them(Rask and Halak) aren't even leading their team in games played. How do teams even win with these lousy goalies? Bring back the trap! We need to get some decent goalies out there!
To be honest, I'm actually kinda surprised there are ten on this list. I expected it to be closer to five or six.
This comment has been removed by the author.
ReplyDeleteOriginally, I was going to be an assdick and point out that the commentator may've meant save percentage on a per-game basis. It would appear the question asked is looking for a seasonal total. The answer "chance to win" sounds to me like per-game you want to give up less than 3 goals.
ReplyDeleteBut then I realize I'm giving analysts and commentators credit and as surveys show, they always know less than we do. So yeah, that's a ridiculous answer.
He was right though. .940 would be a really good save percentage level. Credit where credit is due.
Also, Dacque, tags and labels are a fun way to spice up any article or essay!
I disagree. Isn't it pointless to talk about save percentage when referring to giving your team a chance to win a single game? In that case, wouldn't it be easier to just say it's a good idea to keep your goals against at two or lower, with the occasional three goal against performance being sufficient? And, to be honest this is why it's pointless to use a per-game basis at all for goalies. I'm giving the guy the benefit of the doubt that he's smart enough to realize this(although maybe I shouldn't).
ReplyDeleteSave percentage is a stat that you can look at throughout a group of games, but if you look at single game percentages, they fluctuate like crazy. Someone's save percentage for consecutive games could go from .788 to 1.000. Neither of those are even close to the guy's .912 on the year, and in no way represent what the goalie does on a nightly basis.
Here's a chart of what would be needed to maintain around a .940 save% in a single game.
16/17=.941.
31/33=.939.
47/50=.940.
17 total shots faced in a game is a bit low, but possible. 33 total shots is a bit high, but also possible. 50+ shots faced in a game is very rare.
If you want to argue that a goalie may have no chance on a few goals, then giving up 3 goals wouldn't be considered a bad night, but in order to maintain a 94.0% on the night, he would have to face somewhere around 50 shots. Completely unreasonable. Plus, the graphic was for Lundqvist's last seven starts(he posted a 93.9% in those seven starts) so I honestly think he was referring to a season total, or at least a total for a certain amount of consecutive games. I hope, anyway. Using save percentage to talk about a single performance is asinine. That would be like using batting average for a single game to determine how good a player is. "Holy crap, Ryan Raburn batted .750 tonight! He was 3 for 4! That's the best batting average of all time! A very good batting average is only around .320!"
This comment has been removed by the author.
ReplyDeleteOr like a bowler who throws 174, 264, 129 to arrive at his 189 average?
ReplyDeleteBatting average is a mediocre-to-lousy stat to use when determining how good a player is no matter the games played. Not your point, I know, but I just wanted to get that dig on batting average in.
Also, eliminate the 3 goal nights. I originally said less than 3 goals. NHL average for one team is 2.76 goals/game this season. 3 goals would actually be more than average. Plus, he says .920 is what you need to "say you're giving your teammates a chance to win" not .940.
Now, do I think this guy (who is probably Joe Micheletti) did the leg work you or I just did? Added up all the goals and GP? Looked at a list of this years SV% leaders? No, of course not. And the difference between a .920 and .940 is less than 1 save on 30 shots.
It doesn't matter that SV% can fluctuate from game-to-game wildly. Having a .778 in one game means your team is probably in some trouble. Having a 1.000 means you won, barring a shoot out loss. If you had a .920 in the game you played, your team was probably was at least in the game and probably won.
Since they flashed a stat up about this, he was probably saying at any time. Which for a season has never happened, for a 7-game stretch is excellent, and for a game is probably good enough to win. He never specified and that's why he's an idiot. Learn how to communicate before you take a position as a communicator.
Or a bowler with a 210 average bowling an 803 series and an 823 series in one season, but bowling a 440 series at the state tournament. Yeah, just like that.
ReplyDeleteOk, but for some nights, giving up three goals is still enough to give your team a chance to win, which was his point. The bottom line is whether it's a per-game basis or for an entire season, it either doesn't make sense or it's obvious and not worthy of even mentioning.
But I think it does matter that save percentage fluctuates so much, because that's the basis of his entire argument. Don't use something like that when referring to a single game. And don't use a save percentage that's never been reached in the history of the NHL when referring to an entire season.
I also meant to mention in my reply that a goalie has no control over how many shots he faces. If he gives up three goals that he had absolutely no chance on, he has no control over whether he ends up facing 50 shots or 24 shots.
I really enjoy those labels. You're one funny guy, Dacque. Not two, though; just one. You're not funny enough to be two funny guys. Maybe if you would keep coming up with awesome labels, you could be as funny as two funny guys, perhaps even three funny guys!
ReplyDelete