Score-Adjusted Fenwick and Remaining Strength of Schedule

Travis Yost
January 13 2014 02:02PM

 

http://nesncom.files.wordpress.com/2013/01/stevenstamkos1.jpg

 

As far as I'm aware, the only NHL-based strength of schedule stuff around the internet can be found through Jeff Sagarin's ratings over at USA Today. Problem, of course, is that goals and goal differentials are great to look at if you want to know why a team currently sits where they do in the standings; it does far less to assist in future forecasting.

Knowing what we know about the predictive value of Score-Adjusted Fenwick (and the fact that it's still not automated anywhere on the internet), I went ahead and compiled the number for each NHL team, accurate through January 14, 2014. Then, I went ahead and compiled strength of schedules based on the average Score-Adjusted Fenwick each club will see through the rest of the season.

First, each team's Score-Adjusted Fenwick -- the most accurate tool we have available in pinning down possession / scoring-chance differential dominance.

 

 

Some of this measures up to what we see in the Fenwick Close numbers, though I noted a week or so ago that there were a few notable discrepancies, like Philadelphia and Carolina looking better by this metric, and Buffalo/Nashville looking worse. Buffalo gets killed here -- killed even by Buffalo standards -- because of how dreadful they seem to play in the rare instances they're protecting a lead.

That group of four at the top -- Los Angeles, Chicago, San Jose, and St. Louis -- are going to provide one hell of a Western Conference playoffs.

So, using those numbers above, I was able to calculate the remaining strength of schedules of each of the league's thirty teams. These are simply averages of the opposition's Score-Adjusted Fenwick through the end of the regular season; teams at the top here are going to play tougher competition on average, and teams at the bottom are going to play weaker competition on average.

 

 

A few thoughts on this.

One, if anyone's still anticipating Tampa Bay to sort of slide out of the picture here, I'd put a stop to the breath-holding. Their underlying numbers are great. They're getting back one of the best hockey players in the world from injury at some point in the foreseeable future. And their schedule the rest of the way is, at least by Score-Adjusted Fenwick, the easiest in the league. Even with possible goaltending regression, I don't see how they miss the post-season.

On the other hand, I think it's fascinating how Minnesota, Colorado, and Toronto are going to see the toughest competition through the rest of the year. Minnesota's underlying numbers have sort of spiraled out of control since a promising start, and they're now dealing with injuries to some of their best players, including Mikko Koivu and Zach Parise.

Colorado, who is virtually Toronto West, has built a promising points-lead in the standings. But, they've been riding the percentages for forty-five games, and despite having a nice collection of young talent, they're woefully out-shot and have relied on even-strength save percentages and shooting percentages that rank in the top-five league-wide.

And then there's Toronto, who has now won all of two games in regulation in their last twenty-six. They're a woeful hockey team that's more or less leaned on the percentages (again) and winning a ton of games in the shootout to even remain in the playoff picture, and even in doing that, regression's (a) hit hard; and (b) forced them out of the wild card picture, for now. Now, a tough run of games. Barring pulling better than the league's 29th-best possession numbers, or another unbelievable Sh%/Sv% heater, I don't see how this team makes the playoffs. Same as it ever was, I guess.

 

--

 

 

73100a182a18dbb0cba9bf8f547ed466
Hockey and hoops. @TravisHeHateMe
Avatar
#1 Kacie
January 14 2014, 03:40PM
Trash it!
0
trashes
Props
1
props

What about home games versus traveling? Wouldn't that have a huge impact on the difficulty level of the remainder of the season as well?

Avatar
#2 JCDavies
January 13 2014, 08:22PM
Trash it!
0
trashes
Props
0
props

@Travis

Have you done, or are you going to do, one of these for the first half too? It would be interesting to see which teams had an easier/more difficult first half to the season.

Avatar
#3 van Malmsteen
January 14 2014, 03:08AM
Trash it!
0
trashes
Props
0
props

Presumably for that, the numbers will be inverted (assuming you're using the same statistics to rank teams) seeing as the first half of the season is the opposite of the second half of the season.

Avatar
#4 JCDavies
January 14 2014, 09:39AM
Trash it!
0
trashes
Props
0
props

@van Malmsteen

I thought of that (that the rankings for each conference would simply be inverted) but I think that there are enough scheduling differences between and within the divisions that it wouldn't be so straight forward.

Avatar
#5 Woodguy
January 15 2014, 08:51AM
Trash it!
0
trashes
Props
0
props

Cool stuff Travis, thanks for the work.

Have you ever thought about using Home Score Adjusted Fenwick and Road SAF to figure out strength of schedule?

Some teams are much different home vs. road.

A good example is WIN.

Home Fen 53.2% (8th in NHL) http://www.extraskater.com/teams/on-ice?sit=5v5home&type=total

Road Fen 47.8% (22nd in NHL) http://www.extraskater.com/teams/on-ice?sit=5v5away&type=total

I realize that these numbers are not score adjusted, but I'd be surprised if it didn't hold for the most part.

If a team has one game vs WIN, the strength of that game is very dependent on where its played.

Thoughts?

One downside is that it makes the samples smaller (splitting SAF in 2 samples instead of one) which may create more noise, but I think doing it will also add some nuance.

Comments are closed for this article.