Wednesday, August 30, 2006

 

Shoooot!

In my torturously long post on the Oilers Powerplay in the Stanley Cup Finals, I came across what was to me a disturbing piece of information. It looked something like this:

Oilers Powerplay Summary: Games 1-7
  
   
Powerplay Shots  
 
Total
% of Total PP Shots
Shots Missed
27
24%
Shots Blocked
37
33%
Shots Through
48
43%
Total Shots On Net
112
100%


Call me crazy, but missing 57% of your powerplay shots in a series is ridiculously ineffecient. Stupendously so. The problem is, I didn't and still don't know the following things:

• Combining what the NHL marks as missed shots with what the NHL marks as shots blocked, what is the NHL average for missed shots on net on the powerplay?

• Is that average any different in the playoffs as compared to the regular season?

• Combining what the NHL marks as missed shots with what the NHL marks as shots blocked, what is the NHL average for missed shots on net during 5 on 5 play?

• Is that average any different in the playoffs as compared to the regular season?

• Does it really matter, or am I actually just crazy?

My logic in all of this is pretty simple: an increase in shots equals an increase in goals. If you are missing over 50% of your shots in a game or a series, and especially on the powerplay, you are seriously hurting your team's chances of winning. Forget the adage that "if you don't shoot, you don't score." It's more like, "if you don't shoot and hit the net, you don't score." You don't allow the goaltender or defence to make a mistake, you don't force the goaltender to use up energy in making that save, you don't force the defence to make a play on a forward around the net, and you eliminate the odds of a scoring chance off of a rebound. In short, you decrease your team's chance to win hockey games.

Of course, I have no numbers to indicate whether the Oilers shot inefficiency is the norm in the NHL. Nor do I have any numbers to support my theory that an increase in shots equals an increase in goals. I thought I would throw this out there, however, because a) I know there are people who can find those numbers in a timely and efficient manner, b) I'm actually interested to see if people think it is an important issue (wouldn't basketball people be pulling their hair out if NBA teams were having over 50% or their shots blocked or airballed?), and c) there is nothing else to talk about right now. Personally, I'm not horned up by the news that Toby Petersen has signed a contract with us, although it is good to hear that Mikhnov is finally in North America. So there it is. Have at 'er.

Comments:

JavaGeek could probably answer these questions without even getting out of bed.

One note, though. It's tricky measuring regular season vs. playoffs, as the teams theoretically are "better" and thus generate naturally different results. It's always a tricky question, that one.
 


Oh, just for fun I ran some questions against the dataset I pulled before, which only covers Anaheim's 82 regular season games.

Anaheim's PP:
70% Shots on net (shots or goals)
27% Missed Shot
3% Blocked shot by opponent

Opponents' PP:
73% Shots on net
25% Missed shots
2% Blocked shot by Anaheim

Anaheim's EV*:
59% Shots on net
21% Missed shots
20% blocked shots

Opponents' EV*:
60% Shots on net
23% Missed shots
17% blocked shots

* I haven't separated out any 5-on-5 from 4-on-4 time, nor anything similar for PP situations.

Interesting? Believable?
 


Crap, I blew that last one. I was looking at shots blocked when the opponent was on the PP, when I should have been looking at the shots blocked when the opponent was SH!

Here's the more correct PP percentages:

Anaheim's PP:
55% on net
21% missed
24% blocked

Opponent's PP:
62% on net
21% missed
17% blocked

Even-strength numbers are the same as above.
 


Hmm. Substantial, but less than the Oilers. I'm guessing most teams miss in the 35-55% range, which just seems ludicrous.
 


I mean, it seems your numbers should be skewed in that it's all against one team, Carolina, that demonstrated a keen desire to block shots.

The Duck season numbers include 8 games against the porous Kings.
 


REG-Season:
BS: 8362 (24.6% ± 0.6%)
MS: 7323 (21.6% ± 0.4%)
SH: 18258 (53.8% ± 0.5%)
---------
33943

Playoffs:
BS: 585 (25.6% ± 2.2%)
MS: 488 (21.4% ± 1.7%)
SH: 1208 (53.0% ± 2.1%)
--------
2281
(Intervals are 95% CI)

Hypothesis tests:
H0: playoffs = reg-season
HA: playoffs <> reg-season
BS: p-value: 20%
MS: p-value: 40%
SH: p-value: 23%
Generally you reject the Null Hypothesis (H0) for the "better" alternative (HA) when p-value is < 5%.
Those are pretty high values, so playoffs are not that different from the regular season.

More info here: Statistical hypothesis testing

The reason I show all the "range" is so you can look at your example (~112 shots at net) With a plus minus of about 9.2% If you do the hypothesis test on TST, you get a Z-score of 2.32 or a p-value of 1%. (which is less than 5%)

So can conclude that Edmonton in the finals got statistically fewer shots through than average.
 


Okay, I don't know if all of that was in English, but what I got from it is that last season, the league average for Shots Not Through (Missed Shots + Blocked Shots) was somewhere arround the 45-46% mark.

Does that seem right? And if so, doesn't that seem really, really bad? That is what I don't understand. Shouldn't this be a big deal? Or am I fussing over nothing? Wouldn't a team who could cut that down have a better chance of scoring more goals? How much would that increase by? And are there players who consistently miss fewer shots that others? To me, it's like OBP in baseball. I would want to find guys who consistently get shots on net over guys who don't. Where is the Scott Hatteberg of hockey?
 


I don't think so Andy. A few reasons why...

A shot directed towards the net is still dangerous (unlike in basketball). It can still be deflected, it can still be controlled by one of your forwards, and you can still regain possession of the puck on the rebound off of the backboards if worst comes to worst. There are times where the pointmans best play is a (soft) shot wide of the net.

Also, how does the NHL deal with deflections? If Stoll takes a point shot that's going on net, Torres deflects it and it goes wide, I'd assume they'd mark it down as a missed shot by Torres - but that would skew these stats. What happens if Iginla takes a shot and it goes off Moreau's stick and into the seats? Is that blocked or missed?

Having said that, I'm interested to see what kind of numbers Detroit/Ottawa had? I'm also thining that these stats are very hard to compare from team to team due to the schedule. Just as an example, a team in the NW I'd expect to do worse in the season than a team in Central.

Comparing Vancouver and Detroit for example; Vancouver plays 24 games against Minnesta, Edmonton and Calgary - 3 teams known for outstanding shotblocking, while similarily Detroit plays Columbus, Chicago and St. Louis. It has to work into these stats, does it not?
 


A shot directed towards the net is still dangerous (unlike in basketball). It can still be deflected, it can still be controlled by one of your forwards, and you can still regain possession of the puck on the rebound off of the backboards if worst comes to worst. There are times where the pointmans best play is a (soft) shot wide of the net.

Also, how does the NHL deal with deflections? If Stoll takes a point shot that's going on net, Torres deflects it and it goes wide, I'd assume they'd mark it down as a missed shot by Torres - but that would skew these stats. What happens if Iginla takes a shot and it goes off Moreau's stick and into the seats? Is that blocked or missed?


These are shots that just plain miss the net, or are blocked. If you look at the Play-By-Play from Game One, for example, you'll see that Ryan Smyth had a powerplay shot on net at around the 15 minute mark of the 1st. The shot is labelled a "tip-in", and no other Oiler is noted before that as having a blocked shot or a missed shot. So the stats aren't skewed, assuming that the NHL is being consistent in how they approach tip-ins, and that JavaGeek is getting that consistent info.

I also disagree that the blocked shot and missed shot is more dangerous in hockey than it is in basketball. If the ball is still in play, it's as dangerous as a puck still in play. Plus, we haven't looked at how many of those Shots Not Through lead into play going the other way. I'd guess there was a large amount.

I do agree that players sometimes take shots that are really more like passes. They are thrown at the net, sometimes wide of the play, to allow a forward near the net the chance to tip. But players make passes like that in the NBA, too. A guard will dump in a high pass for a centre, or a wide pass that he thinks only one of his guys can reach. So maybe a better way of doing things were if the NHL started counting Missed Shots and Blocked Shots as Turnovers, if that is indeed what they are.
 

Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?