Category: NHL

3v3 Overtime is Working

By Antonio Papa

This season, the NHL has initiated a rule change to create more overtime goals and fewer shootouts. Now, overtime play will be 3-on-3, instead of 4-on-4. A quick statistical analysis shows us that the new rule has – and will continue to – increase overtime scoring.

Shootouts were added after the 2005-06 lockout as an alternative to ties in the regular season, but they have been criticized as essentially flipping a coin to decide the winner. 3-on-3 play, in contrast, gives stronger teams an increased chance of scoring goals. In the early 1980s, the Edmonton Oilers even told their defenders to get into mutual roughing penalties on purpose so that the game would become 4-on-4 or 3-on-3. Then, Wayne Gretzky, Jari Kurri and Mark Messier would take over on the open ice. This was effective because players with superior skating ability gain an upper hand in 4-on-4 and 3-on-3 situations, resulting in more goals scored.

The NHL instituted the “Gretzky rule” in 1985 as a direct response to these shenanigans. The “Gretzky Rule” created the concept of coincidental minor penalties and allowed full strength play for offsetting penalties. A few years later, the NHL reversed the change in an attempt to reclaim some of that high-scoring open play. Expect this year’s 3-on-3 overtime to benefit top-heavy teams, like the Pittsburgh Penguins, who are sure to take advantage of the situation with skaters like Sidney Crosby, Evgeni Malkin and Phil Kessel.

Over the past eight seasons, 43% of overtime games had a goal and the other 57% needed a shootout (2,227 games). In this preseason’s overtime games with the rule change in place, 72% of overtime games had a goal and only 28% needed a shootout (24 games). Even with the small sample size, we can use a T-test (difference of means) to determine whether this change is statistically significant. The standard binomial error is σ = .0105 for the regular season set and σ = .0926 for the preseason set. The result is that we are 99% confident that the new rule decreases the proportion of shootouts in overtime by 40%-58% (about half) and should lead to high-octane teams winning more games in overtime.

[Editor’s Note: The last paragraph was edited post-publication to clarify the statistical test used]

Shootout for the Ages

By Antonio Papa

How unlikely was the twenty round shootout between the Capitals and the Panthers?

Last Tuesday, the Florida Panthers defeated the Washington Capitals in the longest NHL shootout ever. It was a grueling, twenty-round battle that dwarfed the previous record. The previous record for longest shootout, which only lasted fifteen rounds, took place in 2005 between the New York Rangers and the Capitals.

Just how incredible was the shootout marathon between Washington and Florida?

Shootouts themselves aren’t terribly rare; there have been 1409 since the league instituted them for the 2005-06 season. Between the 2005-06 and 2013-14 seasons 13.3% of games have been decided by shootouts. The majority of those shootouts have been resolved in the first three rounds. If the score is tied after the first three rounds then more rounds are added until one team scores and the other misses.

Here is a table that counts the number of shootouts that reached up to fifteen rounds (the previous record). The number of shootouts drops rapidly as the number of rounds increases, and barely any last more than eight rounds. Again, this data includes all games between the 2005-06 and 2013-14 (previous) season and thus does not include this season’s data.

Pic1

*Note: A shootout can end after two rounds if one team leads by 2-0 at the end of the second round (leaving no room for a comeback and resulting in an automatic shootout win).

This relationship of the number of shootouts that reach a particular round can be better described graphically. The following graph only shows shootouts that went to at least the fourth round, because the shootout rules change to sudden death after the third round. The y-axis is on a logarithmic scale.

Pic2
A function of natural log fits the plot quite well. Based on the trend line, the probability that any particular game will end in a shootout lasting twenty or more rounds is 0.00112%. If we extend this relationship, we predict that, with 1230 games in a season, we can expect a shootout like the one on between Florida and Washington to take place about once every seventy-two seasons.

Pic3

So, how rare was Tuesday’s shootout marathon?

Once in a lifetime.

There is No Place Like Home

By Jeffrey Gleason

Nine weeks into the NFL season, no teams remain unbeaten. This could’ve actually been said after eight weeks, after seven weeks, and after six weeks as well. Week 5 was the last time an unbeaten team remained, when both the Cardinals and Bengals were sitting at 3-0.

However, after these same nine weeks, five teams remain unbeaten at home. The Patriots, Broncos, Eagles, Packers, and Cardinals have yet to lose on their own turf.

Home field advantage is a phenomenon that gets a lot of traction in sports. Experts often use it to justify their predictions and betting lines usually reflect the perceived advantage of the home side. However, people often generalize home field advantage with a “one size fits all” approach, acknowledging its presence, but assuming it displays a constant impact across different situations.

With five unbeaten NFL home teams and the recent impetus of a road team finally winning Game 7 of the World Series (the Giants topped the Royals on October 29th to capture their third championship in five years), I was interested in how home field advantage was quantitatively different in different situations. How does it vary across sports? Do both good teams and bad teams experience the same advantage? Is it magnified in the postseason? What about differences in earlier eras? These are the questions I set out to resolve.

Continue reading