Minnesota Boys' High School Hockey
Technical Info on PS2 (PageStat2)
FAQ
This FAQ (Frequently Asked Questions) is the result of questions which have been asked about PS2 during the past two years. Hopefully, if you have questions about how the system works and the results, you'll find the answers there and save the necessity for you to email me about your questions and concerns.
Before emailing me with questions about the ratings, read this information first! Questions which are covered here won't be responded to in email.
Q: How often are the ratings updated?
A: During the 2006-07 season, PS2 will usually be updated each Monday starting on December 4.
Q: A team hasn't done very well early in the season. How could they be rated so high?
A: PS2 not only looks at how each team has done this season, but also looks at their results from the previous season. As a team plays more games, last season's results are gradually phased out. This is done because you can't, in many cases, derive much from how a team does in their first few games. By New Year's, teams' results from last season are completely phased out.
Q: We have a better W/L record than another team in our conference. Why are we rated lower?
A: A team's W/L record is meaningful only when also taken into account along with the strength of the teams they have played. If you play weak teams, and don't do as well against those teams as other teams do, your rating won't be as high.
Q: But we beat a team handily which is rated above us. What gives?
A: The obvious answer to this is that one game does not a season make. You can't rate one team ahead of another team simply because they beat them in head-to-head competition. There are 22 regular season games, and up to seven more in section and State tournament games. No single game is given more weight than any other game played.
Also, by using the "our team beat your team" argument, it can lead to a "circular logic" of sorts. Team A beats Team B which beats Team C which beats Team A. See how that works?
Q: We should be rated higher, but we've had a player or players injured. Isn't that taken into account?
A: No. How could this possibly be programmed into a system? Different players have a different impact on a team's results. Assigning a value to a player who is injured is nearly impossible. It could be done, but it would be a logistical nightmare. I'd have to know, at any given time, every player who was injured in Minnesota, and what their stats were prior to being injured.
Q: Lately our team has been doing very well. We just started slowly. Isn't that taken into account?
A: It is taken into account. The system give a heavier weight to recent games played than to games played earlier in the season.
Q: Why do the PS2 ratings for Class A teams usually lower than for Class AA teams?
A: This issue is also addressed below on this web page. Simply put, Class AA teams are mostly superior to Class A teams. Results over the past several seasons point out that in games pitting a Class A team against a Class AA team, the class AA team has won 65-70% of those games.
There have been a number of issues raised by people which have led to changes in how PS2 works. Appreciation is expressed to those whom have raised such issues!
Overview
PS2 has its roots in the 70's, when it was developed as a means to rate NFL teams. There were never any real objective studies to see how well it worked; my memory was that it had some success in predicting winner of upcoming games. You wouldn't get rich in Las Vegas using it, however.
I imported it in 1996 to Minnesota high school hockey. It was in reponse to a basic question I had of how all teams in Minnesota stacked up against each other. Tom Hawley's ratings, which used to be posted weekly in the Duluth News-Tribune, really piqued my interest. And so the project was born.
The system has undergone a number of revisions, and two huge overhauls. By overhaul I mean that the system was torn completely done and redone from scratch. The basic information it was derived from hasn't changed, however; results of games, and strength of schedule.
After a lot of experimentation, I've come to an obvious conclusion; it's impossible to develop a system that's perfect or is even 100% objective. After all, the weight given to each data piece is really determined by the person creating the system. What I believe to be important may be considered more or less so by someone else using the same data.
The goal has always been to come up with some kind of ranking system for Minnesota high school hockey teams that looks reasonable. In reality, however, even that is open to all kinds of debate. Accuracy is in the eye of the beholder. I recognize this along with the implication that some will think this system isn't worth the web page it's posted to. Some will like it. Hey, I crunch numbers for a living and this project has been a fun one to work on. It it wasn't fun, or seemed like a lot of work, I wouldn't do it.
Nuts and Bolts
NOTE: Beginning January 16, 1999, PS2 has gone a total overhaul. It's dramatic enough that I was considering renaming the system PS3 or something like that. Sort of like upgrades to the Microsoft Windows (Microsloth Windoze) system. Special thanks go out to Mitch Hawker, Tom Hawley, and several others who have proposed new ideas and different ways of looking at the numbers and coming up with various calculations.
Two new factors were looked at during the off-season; most recent scores, and home ice advantage. The system has now been revamped to give a heavier weight to more recent scores. Programming for home ice advantage is a possibility in the future, but it will not be used in 1999-2000. [update 12/3/200: Research into the home/away factor during the past off-season revealed that adjusting for home/away would result in no appreciable change in any team's rating by the end of the season.
The following things are taken into account when determining a team's PS2 ratings:
The goal differential in each of their games
The PS2 rating of each opponent a team has faced.
The following things are not taken into account:
Injuries a team has suffered. This would be nearly impossible to program into the system, so it is ignored.
Tradition. How can you put a numerical value on the fact that a team is the defending state champion in either class? Even if you could, what value would you place on that?
How the pollsters and other so-called experts would rate a team. Humans are subject to bias and are limited to only the games they've seen. Unless you are able to see every game which every team in Minnesota has played, your expertise is limited. Even if it were humanly possible to see every game, you still have bias. Count on it.
How good a team is "supposed" to be this season. PS2 doesn't care how many players you have returning, how you are expected to compete for the Conference title. The results on-ice are all that matters, not how things are supposed to be.
Each team starts with a pre-determined number. An average is assigned to teams who didn't have a previous PS2 rating, and for other teams, their previous PS2 rating is used. Then the process begins of assigning a "point value" to each game every team has played, using their rating, their opponent's rating, and the goal differential.
Once this is completed, the ratings are used to predict the outcomes of each game. Then the ratings are run again, with each team's start point being the rating from the previous "iteration". Again, the percentage of games the higher rated team actually won is determined. This continues on until the teams' ratings have reached a point where further iterations wouldn't result in a higher percentage of games being accurately predicted.
last updated december 3, 2006
|