I was looking at the most recent Maxpreps State of California high school football polls. Folsom gets pummeled by DLS and drops 2 spots to #8 in the state, while DLS moves up 3 spots to #12, just behind St. Francis. Help me understand why Folsom stays ahead of DLS, St. Francis and Serra. It has become obvious now that Folsom fattens up on their weak league opponents while St. Francis and Serra play in a tougher league, with St. Francis beating the big dog this year!
Maxpreps needs to rethink their polling strategies.
Ok, I guess we need to go through this every couple of years to keep people updated.
1. Maxpreps does not actually generate the list at which you are looking. They purchase it from Calpreps.com.
2. The Calpreps system is not a poll. A poll is where people are asked for their opinions. That's not what happens here.
3. Calpreps is not a ranking system. It is a rating system. No, that's not the same thing.
4. All teams are rated in the preseason mostly based off of the previous year, with adjustments due to certain players leaving or returning, schedule improvement/regression, etc.
5. As the year goes on the results of each week are evaluated for game "performance". The performance value is based on the result in comparison to the ratings of the teams of each game.
6. There is a minimum value for a win/loss and a maximum in cases of blowouts (i.e., teams can't just run up the score and get infinite credit for it).
7. Minimum value for a win is currently 15 points + whatever the losing team's rating is. Minimum value for a loss is winning team's rating - 15 points. Max win value is 31 + losing team's rating. Max loss value is winning team's rating - 31.
8. Games that involve teams with a ratings gap over more than 31 points and have a result of more than 31 points count less than those within the min/max windows. This is to say that if DLS clobbers someone like Acalanes, it doesn't really help them all that much. If they were to struggle against Acalanes, it will hurt them quite a bit.
9. As a team plays more and more games with significance (i.e., within the min/max window), the more the preseason rating goes away. The idea is that eventually, the rating will generally be the average of all of the significant games a team plays within the current year only. I believe there are cases where this won't actually happen, but that's the idea.
10. Playoff games are rated higher than those of the regular season (roughly 2.1x more), so teams can make significant jumps in the post-season.
11. While a 'bad loss' can have an impact, because the rating is an average, it can definitely be negated by one or more 'good wins'.
12. 'Good' losses can actually increase ratings. In the above Acalanes-DLS example, if the Dons were able to stay within 30 points of the Spartans, then the current value of the game performance would be roughly between 47-32, which is higher than their current rating of 10.9. This would boost the rating.
13. While human polls generally will try to daisy chain wins (i.e., Team A beats Team B, who beats Team C, therefore the rankings are 1. A, 2. B, and 3. C). Calpreps doesn't really do this with one exception. If an undefeated team beats someone, they
must remain rated higher than that losing team. Even if the results of all other games don't necessarily support this, the winning team of that game will remain at least 0.2 points higher. Once the winning team of that game loses, all bets are off and the ratings can go back to being mostly an average of significant games.
So, now that we have all of that out of the way...
Folsom didn't drop below DLS probably because of the Spartan loss to St. Francis. With that loss, comment #13 above does not apply to Folsom-DLS, but it does to St. Francis-DLS.
Folsom's drop wasn't severe because they have two significant wins that are boosting them, against Pittsburg and Jesuit. These are countering the loss to DLS for the most part.
Oakdale did not lose six ranking spots despite a big win. Six teams' ratings increased above theirs. The Mustang's 55-point win fell out of the min/max window, therefore it is not considered a significant win in terms of impacting their rating (i.e, their opponent was rated too low). Teams cannot infinitely boost their rating by beating significantly lower-rated teams.
Any questions?
