Islanders 'Real' Power Ranking: Better than polls, barely

I have spoken to numerous NHL executives about advanced statistics and one of their biggest complaints is that the scoreboard is deceptive. When they win despite playing badly -- which happens all the time -- they have limited means to convince their players of their poor play.

Does that not sound like a variation of Scott Gordon's concern after the Florida swing, when he felt the team was feeling good thanks to (perhaps fortunate) early wins and falling into bad habits as a result? While part of my intention at this site is to live up every win and mourn every loss, if I could somehow ordain one orientation for every discussion it would be this: In single-game bites, the scoreboard can be deceptive. (Figuring out how deceptive, of course, is fodder for our endless debates.)

If you are at all intrigued by Corsi or "advanced stats" or any measure of hockey beyond "they won and looked good," I encourage you to read this post at Behind the Net from which the opening quote stems, to get a quick view of why statheads bother. Gabe's post is a takedown of a grumpy MSM writer's lament of all things numeric, but even if that's not your thing there are some great, very accessible observations there.

Because while everyone seeks their own basket of fun in this game -- from the calculatingly analytical to the "Just go troops, go!" -- there is something to be learned, or to at least think about, from this advanced stats perspective on the game.

It's not about treating players like unchanging autobots with finite values (although there are interesting attempts to peg a "wins" or "money" value to each player). It's not about trying to take the fun out of watching the game or enjoying a win streak. Rather, I'd argue at its heart it's about that quote above: An attempt to understand when a player or team is truly good, and when it's riding luck (puck luck, not injury luck) that's due to end -- the old, "Is this sustainable or do we need changes?" question..

To that end, we know shot totals (including blocked and missed shots, upon which Coris is based) are situational, context dependent. Nonetheless, in aggregate totaling those up -- Corsi or Fenwick, essentially -- can provide insight:

Point taken: it's tough to figure out what an individual shot is worth. But isn't the same true of individual goals?  Do teams never score lucky goals? Fortunately, we have a little thing called "the law of large numbers" - essentially, if you count up enough shots (or goals), the type of gun they came from tends to not matter.

I don't try to make this a stats site -- I'm not qualified even if I wanted to. But I do think everyone can benefit from the philosophic/conceptual side of seeing the game in this way. That's why I recommend anyone who's remotely interested in this angle to go read Gabe's post.

Survey of Media Power Rankings, Nov. 2

What did we say last week about the lag in national media opinion? While the Islanders were already entering the first week of what has been a two-week poor stretch, media rankings were still catching up to their hot start -- pushing them up in every media poll. This week, finally the warts are showing up in the opinion polls in a big way, with the Isles plummeting in every poll and landing in 20th in the true "power ranking."

"Real" Power Rankings

One of our many "Mike"s around here, LHH reader ICanSeeforIslesandIsles, has created his own true "power" rankings by weighing teams' results against strong and weak opponents. A heavy strength of schedule component.

He was kind enough to share it with us, so hopefully it's something we can use throughout the season to get a real ranking beyond standings, beyond points percentage, and certainly beyond opinion polls.

I'm butchering it some, so here is his explanation of how these are calculated (and he'll be around later to answer questions):

How the table is calculated:

• When Team A beats Team B, Team A earns Team B's percentage points (Column "Pct" in the table). Team B earns none.
• If Team A wins in OT or a shoot-out, "A" earns "B"'s pct points, and "B" earns half of "A"'s points.
• All Pct points earned are added up, and divided by the team's total number of games played. That's the "Earned/Game" column above.
• Some people consider this "Earned/Game" column as the indication of how well a team has done, because it calculates how well a given team has done, while taking into account its strength of schedule.
  
• I, however, wanted to go further. For any given team, I wanted to base their strength not only on their own strength of schedule, but also the strength of schedule of the teams they defeated. It's better to get a win against a team with 5 wins and 1 loss when that team has played against mostly first-place competition, versus a team with an identical record, who have played against teams who all have losing records.
  
• Therefore, I then go back and calculate everything again as described above, only this time I use the "Points Earned / Game."
• So, if Team A beats Team B, I add Team B's points earned per game to Team A's Power Rankings totals. After adding all of the points Team A earned, I again divide by number of games played to arrive at the final column, the "PWR" (Power Rankings points).

Here's how it looks as of last night's games (pardon some of roughness...I'm still working on ideal formatting). The Isles rank 20th by this measure, boosted by having played almost exclusively teams in the top half of these rankings. New Jersey and Buffalo look positively awful. Oh, and go Blues.