First LHH Power Rankings of the Season: Islanders down, Bolts, Habs and Avs up early

[Dan Note: Apologies to Mike, as I should have run this on Friday. The good news is that his much-loved power rankings are back for another season. Let's hope the Islanders are on top soon.]

Hello and welcome to another season of minimally important numbers! This year, the formula for calculating the quality of each point a team earns in the standings, has been streamlined. But first, as always, we check in with what others are saying about the Isles...

But Where Are They?

It would appear less of the usual suspects are making power rankings nowadays. Of the five publications we were keeping track of last year, only three have rankings so far this year, and one of them (TSN) uses advanced statistics to come up with their rankings. More on that in a minute.

So, it would seem there's less and less opinion people around, and given the way they used to beat on our team, I'd say that's a good thing. I started the LHH Power Rankings to dispel the myth that the Isles were as bad as people said they were. Sure, they were basement dwellers, but they were playing in a division loaded with rich teams that always fared better than the average club. The table I'd generate, based solely on math, tried to look at each team through a totally objective lens. We continue that tradition below, but with some modifications to the formula to account for the current overtime format. But as for these opinion polls, I'll look around for more of them this week, but if I don't find any, I won't lose any sleep over it.

LHH Power Rankings - What's New

There are two general tweaks I made to the formula for this year:

1) Any game going past regulation is now considered a tie. The 3-on-3 overtime format has proven to be almost as much of a coin flip as the penalty shootout is. One weird bounce will decide a game more often than not. Teams shouldn't be rewarded for getting the full points if they have to resort to this stage. If you can't win a hockey game during 5-on-5 play, then you're fit to be tied. Therefore, whereas last year we had a complex system awarding varying amounts of points depending on victories decided in 60 or 65 minutes (or beyond), this year, it's pretty simple, and is exactly as the NHL was before regular season overtime became a thing: Two points for a win in regulation, one point for any game going past regulation (regardless of outcome), and zero points for a regulation loss. This is referred to as LHH Pts (second-to-last column in the table below). The LHH points percentage (not shown below because it's more of a behind-the-scenes number is simply the number of points earned under our system, divided by the total number of possible points a team would have earned had they won every game. If your team plays an opponent who has 4 wins, 2 losses, and 1 "tie" as described above, their points percentage is .643 (9 points out of a possible 14).

This might be more of a (1a), but a related tweak must be made when calculating the rankings score. The goal is to discern losses against good teams from losses against bad ones, as well as award teams for winning above all else. So, for each game, the opponent's points percentage is multiplied by a number. Using the above example:

• A win is five times the opponent's percentage (5 × .643 = 3.215)
• A tie is three times (3 × .643 = 1.929)
• A loss is just the percentage (1 × .643 = .643)

2) While most modifiers to the above will remain the same, the formula will keep track of how much harder it is to earn points on the road vs at home throughout the season, and adjust the scores accordingly. Last year, we had some pretty reliable data that road teams earned approximately 15% less points in the standings per game than home teams did. But this year, we're starting with a totally different reward system, so we won't know how much to weight road games until the year is mostly over. Therefore, the spreadsheet will track points earned by home and road teams, and update the percentage every time I run the numbers. As of the table below, home teams were earning about 50% more points than road teams, but it's early in the year (REALLY small sample size), and as such, I'm sticking with the 15% number until the first month of the season is over.

The other modifier, which gives a winning team a 10% bonus and the losing team a 10% penalty in the event of a blowout (game decided by more than two goals) will remain the same.

So let's see how this shakes out thus far....

LHH Power Rankings

For games through 5/25/16 and therefore does not count the loss against Montreal.

Well that's not entirely surprising. The Isles haven't earned a lot of points against good teams yet. They were 3-3 in the NHL standings, but are under .500 here as their overtime victory counts only as a tie. You would think the Oilers would be higher, with 5 wins in 6 games, but they've played weak opponents thus far. That might not sound fair, but all the numbers simply infer is that the Oilers haven't really had the opportunity to show what they can really do yet.

The rankings scores have, as always, been adjusted so that 100 is league average. It might be odd to see so many more teams below this average score, but again this is due to small sample sizes, and will normalize soon.

That's all for now. Feel free to ask questions in the comments and I'll answer when I have the chance.