Do The Thrashers Have Large Talons?

Monday, April 09, 2007

Predicting the Playoffs: What Matters Most in the Post Season

Last year I got my head handed to me when it came to predicting the playoffs. (It didn't help matters that in the West all the home ice teams were upset.) So I decided to do some number crunching to see what regular season numbers were most strongly associated with teams that won their playoff matchups.

Over the years I've heard a lot of broadcasters and coaches bandy about their personal beliefs: a good defense, a strong power play, great goaltending and so forth. So I decided to test a lot of different things and see which proved to be most useful.

I went through my handy NHL Guide and Record Book for the past 5 playoff seasons and recorded each playoff match. Each year there 8 1st round matchups, 4 2nd round, 2 3rd round and just 1 Finals matchup. Since I looked at 5 years worth of playoffs my dataset includes a total of 40 1st round, 20 2nd round, 10 3rd round and 5 finals pairings for a grand total of 75 playoff matchs. OK on to the findings.

Home Ice
This is one of the most important factors in determining who wins a series. The home ice team won 68% of the 75 series I looked at.

Offense (Regular Season Goals For Average)
The team with the better offense won just 57% of the time.

Defense (Regular Season Goals Against Average)
The team with the best defense won 60% of the matchups.

Goal Differential (Regular Season GFA-GAA)
The team with the better goal differential won 63% of the time. The team with the best goal differential usually has home advantage as well so they overlap significantly.

Power Play Percentage
This surprised me a bit, but the team with the better regular season power play won just 53% of the time. I would have expected it be higher than that.

Penalty Kill Percentage
The team with the better regular season penalty kill won just 55% of the time, I anticipated this being higher.

Drawing Penalties
Some teams receive more power play opportunites because of their players or style of play. My study showed that the team that drew more power play opportunties in the regular season won their playoff match 56% of the time.

Avoiding Penalties
Every coach preaches about the need to stay out of the box, some teams do it better than others. This seems to REALLY matter, more than a lot of other factors as the team that had to kill fewer penalties per game in the regular season also won their playoff series 64% of the time.

Special Teams Goal Differential
Some teams have a great PP but a bad PK or vice versa, what if a team excels at both aspects of special teams? I looked at specical teams goal differential (PP and SH Goals Scored - PP and SH Goals Allowed) and found this too is an important predictor of playoff success in 64% of the series.

Even Strength Goal Differential
Since I just looked at Special Teams Differential, I was curious to see how Even Strength Goal Differenial would compare. 61% of series winners had the better ES Goal Differential, which is lower than the 64% for Special Team Differential.

Great Goal Tending
Everyone says goal tending matters (and it does) so I was shocked by this one. When you look at playoff matchups the team with the better regular season Save Percentage won just 53% of the series. I really expected that to be a better predictor than it is. Perhaps if I replaced the team SV% with the SV% of the playoff starter it might perform better--if I get some time I'll have to check that one out.

Great Team Shooting
This was perhaps the single most surprising thing to come out of the entire study. I threw team shooting percentage into the mix and what do you know--IT IS THE BEST PREDICTOR. I have to say that this really surprised me at first, but the more I thought about it makes a certain amount of sense. If ST% measures a team's ability to finish off scoring chances--and scoring chances are harder to come by in the playoffs--then it makes sense that the most opportunistic team will triumph in the series.

OK to sum up all these factors I've ranked them in the table below based up how well they predicted the series winner in the past. Remember that these are all based upon regular season stats.

Factors That Predict Series Winners
72% Team Shooting Percentage
68% Home Ice Advantage
64% Special Teams Goal Differential
64% Short Handed Less Often
63% Total Goal Differential
61% Even Strength Goal Differential
60% Better Defense
57% Better Offense
56% More Power Play Opportunities
55% Better Penalty Kill Percentage
53% Better Power Play Percentage
53% Team Save Percentage
53% Shots on Goal Differential

How about Predicting Upset Winners?
One of the great things about playoffs is that your team might win even if they are not favored, that is what makes them interesting and exciting to watch. Everyone loves an underdog (unless your team is the overdog). So if you are going to pick some upset winners in your playoff pool are there any regular season numbers that might guide us?

In my study I looked at 75 playoff series and in 24 of them the non-home ice team won. That's a very significant number of upsets (1/3) so I looked at those 24 upset matchups to see which factors mattered most. Here it is-->focus on teams with strong special teams. Upset winners were teams that received more PP chances and avoided being on the PK and tended to execute when given man advantage opportunities. It makes sense that good special teams plus a hot goaltender could spell upset victory.

Series with Upset Winners (Regular Season Stats)
63% More PP Opportunities
54% Fewer SH Opportunities
54% Special Teams Differential
46% Better PK%
42% Better PP %
42% Better Team SV%
42% Better Team Shooting Percentage
38% Better Defense
29% Better Offense
29% Shots on Goal Differential
25% ES Goal Differential
17% Better Goal Differential
00% Home Ice Advantage

Coming up Next: Eastern and Western Conference Breakdowns

0 Comments:

Post a Comment

Links to this post:

Create a Link

<< Home


 
Who links to my website?
View My Stats