Please visit Puckerings for more hockey stuff by me   

As regular readers may recall, last season I made my official predictions for the orders of finish in both NHL conferences the subject of a series of articles here on HockeyZonePlus. I compared the accuracy of my predictions to the accuracy of The Hockey News’ (THN) predictions. In the end, my predictions were slightly better. Perhaps THN was paying attention, and took the defeat to heart; after all, the words “Bible of Hockey” are nowhere to be seen on the cover of this year’s THN Yearbook, unlike last year.

Bearing in mind that I have no illusions about last year’s results being anything but blind luck, I hereby submit my predictions for the 2003/04 NHL season. Once again, the purpose will be to compare my accuracy to THN. To do this I will use the coefficients of correlation between predicted conference standings (based solely on points, ignoring the division-leaders-must-be-1-2-3 rule) and the actual conference standings. A coefficient 0.7 or higher is very high, but really I’m only interested that my coefficients be higher than THN’s. I will also calculate the coefficients of this year’s standings compared to last year’s standings, as a sort of control. I will update the results at the one-quarter, halfway, and three-quarter marks, before finally releasing the year-end results many months from now. Now, the drum roll please…

Eastern Conference

Western Conference

You may be wondering how I came up with my predictions. Well, I tried to be as objective as possible, starting with each team’s results last year and factoring in new player additions and deletions, injuries, age, and that sort of thing. I used only the same information that THN had access to at the time; that is, any transactions or injury news that was not included in THN Yearbook was not considered. For some teams, this will have little or no effect (e.g., Pittsburgh’s multitudinous signings of meaningless space-filler veteran free agents), but for others it may have a significant effect. But in the end, THN and I are in the same boat, so it shouldn’t affect the results of this contest in any significant way.

There are only two really significant differences between my predictions and THN’s. One is San Jose. THN has predicted them to finish as they did last year, 14th in the Western Conference. I have them at 10th. The key here is Pythagorean analysis. San Jose significantly underperformed last year compared to their Pythagorean winning percentage. Pythagorean winning percentage is usually a better predictor of future success than is actual winning percentage. That, coupled with no significant roster moves, leads me to ranking them four spots higher than THN.

The other main difference is Columbus, who THN have at 10th in the West, and I have at 14th. THN’s “quick and dirty” reasoning for their ranking (they finished 15th last year) was that they “got tougher”. Great, that’s true. They also lost their leading scorer, Ray Whitney, who had 76 points last year. Now, I do believe Columbus will improve this year, one spot up to #14. But THN has seriously overvalued “toughness” if they believe that it will both make up for the loss of their top scorer and also propel the team five spots up the standings. They have acquired some good scoring players to replace Whitney, but if all they do is replace, they should be 15th again. Therefore it seems “toughness” is the only reason THN sees the Jackets finishing 10th instead of 15th. We’ll see.