Thursday, June 11, 2009

Odds? Not in Pittsburgh's favor

The Vegas line for the Pittsburgh Penguins winning Game Seven tomorrow night in Detroit has been edging ever slowly against them. Twenty-four hours ago, a six-sports book average had them at 37.4% to win; they're now at 36.0%.

This seems like an underestimate to me. If we use their regular season goals for and against as an estimate of their true talent (.594 Detroit; .550 Pittsburgh), Pittsburgh has a neutral-site winning percentage of 45.5%.

Of course, there is a very strong home-ice advantage in the playoffs. Since 1979-80:

Round 1 - Home Team WPCT = .576 vs Neutral-Site WPCT = .509 (Adv = .067)
Round 2 - Home Team WPCT = .548 vs Neutral-Site WPCT = .506 (Adv = .042)
Round 3 - Home Team WPCT = .546 vs Neutral-Site WPCT = .504 (Adv = .042)
Round 4 - Home Team WPCT = .563 vs Neutral-Site WPCT = .505 (Adv = .058)

Using the Round 4/Finals Home-Ice Advantage, Pittsburgh should have almost a 40% chance of winning Game 7, which means the oddsmakers are underestimating them by more than 10% - provided you don't think Pittsburgh is actually better than their GF/GA predicts. Of course, it's been 38 years since the visiting team has won Game Seven of the Stanley Cup Finals, so perhaps the home-ice advantage is larger in a deciding game than it is normally. It has only happened six times, so it's hard to know. Regardless, Pittsburgh is not the favorite to win the cup tomorrow night...

Labels:


Thursday, April 23, 2009

Swallowing the Whistle?

Since the early 1990s, many fewer penalties have been called in playoff games than during the regular season, with the ratio dropping as low as 74%. But for the vast majority of seasons before 1990, many more penalties were called in the playoffs:

The reason, of course, is fighting - lots and lots of bench-clearing brawls. Once they were eliminated, the total penalty count dropped significantly. But that doesn't tell us if hooking and tripping and interference penalties have gone down - the second curve on the plot shows the penalty ratio only for players who had even penalty totals (mostly players who had zero fights...but obviously also two, four, six or more). The ratio drops for such players, but it still doesn't answer the question. We'd need to find minor penalty totals to figure this out...

Labels:


Playoff Performance by Age

There's no question that scoring rates drop during the playoffs, but James Mirtle wonders if it's fair to count "every game as being equal, whether those postseason games were played as an 18-year-old rookie, 27-year-old sniper or a 37-year-old greybeard..."

At first glance, it seems like a reasonable thing to do. The ratio between playoff and regular season scoring is shown below for ages 18-40:

And it's incredibly rare to play in the playoffs at either an old or young age - 94% of players in the playoffs are between 20 and 34 on January 1st of a given season. A linear fit indicates no relationship between age and scoring ratio, while a polynomial fit indicates a slight difference. Counting all games as equal seems like the right thing to do.

Labels: ,


Saturday, April 11, 2009

Playoff Offensive Levels 1968-2008



One unsurprising result here: there are fewer goals scored in the playoffs than during the regular season. If you're down by three in the second during the regular season, you'll probably rest your best players to keep them from getting hurt in a meaningless game. In the playoffs, there's no point, and no giving up.

But the OT scoring levels are surprising: throughout the 1970s and early 1980s, teams apparently played more offensively in overtime than they did during regulation. But the opposite has now been true for more than 20 years.

Labels: ,


Thursday, April 12, 2007

Playoff Stats are up!

It took a little while, but all of the Behindthenet stats are available for the playoffs. I'm hoping to update at least every other day.

One sad note: the NHL isn't going to post shift charts from last night's Vancouver-Dallas marathon. So that game won't be reflected in the data...

Labels:


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

Subscribe to Posts [Atom]