As of Wednesday, every NHL team hit the 20-game mark. We are finally starting to see what teams are really made of and the standings are starting to separate. After 20 games, the Wild sat at 24 points and were just outside a playoff spot.
So what do the first 20 games say about this team, and have they shown enough to earn a playoff birth? Let's see what the stats say.
One of the best stats to use to predict season-long outcomes is Score-Adjusted Fenwick (S.A.F.). This is a metric invented by Eric at Broad Street Hockey back in 2012 that accounts for Fenwick percentages at every score situation.
Before Score-Adjusted, bloggers tended to use Fenwick Close, which looks at data when the score is within 1 goal in the first two periods or tied in the third. The problem with this metric is that you are constraining yourself to a smaller sample size, since teams don't always play within this score situation.
The problem with just using Fenwick % at all score situations is that it doesn't take into account score effects. Teams who are up by two or more goals tend to sit back defensively and possession numbers drop. Score-Adjusted Fenwick takes care of both the small sample size problem and the score effects problem.
Score-Adjusted Fenwick looks at a team's performance within a score situation and then compares that percentage to what the league average is. Here's an example from Eric's original article:
"Over the last four years, the average Fenwick for a team that's behind by two goals is about 56%. So if a team gets 57% of the shots when they're trailing by 2, that's 1% better than average -- just like if they had 51% of the shots in tied situations. Similarly, the average for a team that's behind by one goal is about 53.9%, so a team that gets 52.9% when down by one is 1% worse than average, like a 49% Fenwick Tied."
The Score-Adjusted metric takes each score situation and weighs them together.
So how well does S.A.F. predict outcomes? Eric calculated this as well by looking at the correlation between Fenwick percentage and a team's point total. He looked at three metrics: Fenwick Tied, Fenwick Close and S.A.F., and also four different points of the season: after 20, 30, 40 and 60 games.
Score-Adjusted proved to have the best correlation to point totals of the three measurements. Eric also found that the best correlations to point totals happened after the 20-game mark. S.A.F. had a correlation of .46 after 20 games, meaning about 46% of a team's point total at the end of the year can be explained by their Fenwick value after 20 games. Obviously 46% isn't a great percentage for predicting outcomes, but it is possibly the best metric we have at this point in hockey analytics.
Using this information, I looked at the last six years of S.A.F. percentages after 20 games and compared them to this year. (S.A.F. can now be found at puckon.net)
At the 20-game mark, the Wild sat in second place with a 56.3 S.A.F. percentage. This is a great sign for the team going forward. Over the last six seasons, only 13 teams have posted S.A.F. percentages over 55 at the 20-game mark. All 13 made the playoffs. In fact, teams who merely posted a percentage over 50 made the playoffs 76.1% of the time.
Obviously S.A.F. doesn't account for everything. Bad possession teams make the playoffs and good teams miss them. However, the really good possession teams almost always qualify. Looking at the top-10 teams for all six years, 88.3% made the playoffs. Of the top-5, 96.7% made it.
Does this mean the Wild are going to win the Stanley Cup? Not exactly. As you can see from the chart, Cup winners come from all over. Los Angeles had very average numbers in 2011/2012 before making some key trades and switching coaches before the playoffs. What it does mean is that the Wild are playing in a way that should put them in the playoffs. Once they get there, who knows what could happen.
Stats collected from Puck On Net