Thursday, January 15, 2009

How The Bruins Win

There are a lot of strange stats in hockey, such as Rod Brind'Amour being near the bottom of the league in +/- or Thomas Vanek leading the league in goals with only a few assists.

Here's a stat that might really surprise you:

Boston's winning percentage when outshooting the opposition: .619
Boston's winning percentage when outshot: .864

That's right, the Bruins are 30% more likely to win a game in which they take fewer shots than the other team. But they aren't alone in having this bizarre advantage. When outshot, the Sharks have a .857 rate of victory; the Rangers see a 20% increase in their likelihood of winning; Buffalo and Columbus are both at .500 only when they see more rubber than the other team.

But Anaheim takes the cake; they are more likely to lose when taking more shots, and more likely to win when being outshot.

What's going on? How is it possible for a team to gain an advantage from being dominated by their opponent? It's beyond this humble blog to try and analyze each unique data set, but we can certainly get some valuable insight from the Bruins' results this season.

Quality vs. Quantity

In a nutshell, the Bruins win because they take better shots rather than more shots. In fact, their defensive system practically invites high-powered opponents to send volleys of rubber toward the Bruins goaltender -- so long as those shots are easily stoppable.

Earlier this season, I pointed out that the Red Wings' puck-possession system struggled against the Bruins' defensive shell. This was largely the result of the Wings settling for low-percentage chances early in the game, which had the effect of cancelling out their puckhandling creativity and rebound control. The Wings outshot the Bruins 30-19 in that game, including a dominant 15-5 during the second frame, but Boston walked to an easy 4-1 win. We saw much the same from Montreal on Tuesday -- a 17-9 shots advantage early in the game translated to a 3-1 loss for the Habs.

Finding the Sweet Spot

Teams that succeed against Boston do four things:

1) Control the puck in between the circles. Not just for lazy backhands, but for bona fide scoring chances.
2) Screen the goaltenders with heavy traffic at the front of the crease.
3) Capitalize on rebounds.
4) Prevent the Bruins from doing each of these things at the other end of the ice.

On any given play, the Bruins protect the slot with hulking defensemen such as Chara, Stuart and Ward -- not to mention the defensive prowess of forwards like Axelsson, Krejci and Yelle. On offense, slick puckhandlers like Savard and Kessel can break down individual coverage while widebodied Lucic and Ryder create havoc in front of the net.

The sum of these advantages is that the Bruins can safely allow their opponents to exhaust their offensive opportunities by settling for long-range, sharp-angled shots; meanwhile Boston will pass up such shots in order to get one quality opportunity from in front. To control the slot against Boston requires a very high level of skill and discipline, which is why the Bruins have only 7 outright losses in 43 games so far.

Crunching the numbers

All of this is fairly intuitive for those who know hockey, but it's remarkable to see it play out in the box scores. Of the Bruins' 11 total losses this season, 8 came in games in which they took the majority of shots on goal (one of the three exceptions was an SO loss to the Rangers). On the other side of the coin, the majority (19 of 32) of their wins have come in games in which the Bruins were outshot.

Let's look at a specific example: Tuesday's game against Montreal. In this analysis I'm going to use the ESPN Gamecenter feature which tracks the location and result of each shot during the game. Go ahead and click on the link to view the shot chart.

Now, time to use your imagination -- we need to define a "quality shot" (we'll call it a QS). As far as this analysis is concerned, that means a shot from the slot. We'll draw some basic boundaries to define the slot:
- Imagine a line connecting each goalpost to the nearest faceoff dot (it should run at a 45-degree angle to the goal line).
- Imagine a line from each faceoff dot to the top of the circle.
- Finally, imagine a line connecting the tops of the faceoff circles.

The resulting polygon should look a little like the "rewind" arrow on a remote control, or perhaps a sideways "house". We're going to call this the Quality Shot Zone; anything outside this area is either too far away or at too sharp an angle to pose a real threat to an unscreened goalie.

Now, click through each period of the game and look where the shots occurred. Montreal took 17 total shots in the first period, but only 6 QS; Boston had 9 total shots, with 3 Quality Shots and 3 marginally outside of the boundaries of the QSZ for a total of 6 QS. Score was 0-0 after one frame.

As the game wore on, each team pressed to get higher-quality shots. In the second period, Montreal took 8 QS, scoring once. Though outshot overall in the period, Boston took 9 QS (including Chara's borderline-QSZ shot which was deflected by a Hab at the top of the crease) and scored twice. In the final period, Boston outshot the Habs 5-4 inside the QSZ and scored another goal.

Total shots: 35-29, advantage Montreal
Quality shots: 20-18, advantage Boston
Goals inside the QSZ: 3-1, advantage Boston

This example mimics the pattern we've seen all year -- and remember, Montreal is one of the league's tougher opponents. Try the same exercise for Boston's wins against the Hurricanes and Senators; using QS analysis, the outcome of each period becomes pretty easy to predict.

More examples

At risk of overkill, here's a QS analysis for each game since the New Year:

Total shot advantage: Pens, 32-26
Bruins: 21 QS for 3 goals
Pens: 20 QS for 2 goals
Result: Bruins win 4-2

Total shot advantage: Bruins, 31-25
Bruins: 18 QS for 2 goals
Sabres: 9 QS for 4 goals
Result: Sabres win 4-2, truly a case of taking advantage of your opportunities!

Total shot advantage: Bruins, 28-24
Bruins: 15 QS for 0 goals
Wild: 12 QS for 0 goals
Result: Wild win 1-0

Total shot advantage: Sens, 30-22
Bruins: 14 QS for 5 goals
Sens: 15 QS for 3 goals
Result: Bruins win 6-4

Total shot advantage: Bruins, 32-30
Bruins: 16 QS for 4 goals
Canes: 15 QS for 1 goal
Result: Bruins win 5-1

Total shot advantage: Habs, 35-29
Bruins: 20 QS for 3 goals
Habs: 18 QS for 1 goal
Result: Bruins win 3-1


Of course, QS isn't a foolproof statistic. Above you can see a few games in which other factors were obviously at play -- Boston played a sloppy and uninspired game on 1/3, failing to take advantage of a 2:1 QS advantage, while Backstrom played a brilliant game on 1/6 to deny all 15 Bruins QS opportunities. And then there are power-play differentials, screened shots from the blue line, and of course the goaltending wizardry that we see on a nightly basis.

But if we're looking for a macro-scale explanation of how the Bruins can continue to win despite being outshot most of the time, this is it. By intentionally allowing opposing teams to control the non-QS areas of the ice, the Bruins can dictate the terms on which they play defense... and in turn they can transition quickly into a deadly attack which virtually ignores shots from the corners and points.

Stumble Upon Toolbar


Anonymous said...

great analysis. Impressive breakdown.


1) Interesting stats. Also good that the team didn't look beyond last night's game toward their game with Washington tomorrow night and took care of business

Tom said...

Agreed -- being aware of the QS-factor doesn't make me any less nervous about the Isles taking 40 shots.