RATINGS FIXES (Playoff Bumps)
Admittedly a long overdue change, when projecting playoff games, we will now be acknowledging and taking into account programs which tend to "step it up" at playoff time. No sense in having our system be "surprised" when this happens year after year with certain programs. Our projection system will now expect it to occur.
California fans are likely familiar with programs such as Van Nuys Birmingham and Sacramento Grant, which tend to have somewhat lackluster performances during the regular season, only to flip the proverbial switch and "turn it on" come playoff time. The reasons why this happens of course vary from school to school. Sometimes it has to do with having raw D1 talent that takes some time to be "coached up". Sometimes it has to do with coaching emphasis-- certain programs (Birmingham most definitely being one) tend to deemphasize the importance of scoreboard-success in the regular season and openly discuss the fact that they view regular season games as essentially tune-ups for the playoffs. While one can debate the merits of the varied philosophies of treating every individual game as and end in itself vs. having a focus on the goal at the end of the season...it's obviously not our job to state a preference. It however is our job to give our site viewers the best possible prediction of what's going to happen that we can, and this change of no longer being oblivious to how various programs fare in the playoffs is a step in the right direction to do that.
Texas fans (as well as those who follow h.s. football nationally) certainly are very familiar with programs like Duncanville and DeSoto. Those two are interesting cases because they "step up" at playoff time as much as any teams in the country, but they play like elite national-level teams in the regular season as well. It's just that there seems to be "another gear" that they're able to hit even beyond that. It's important to realize that we're comparing a school to itself when attempting to analyze whether there's an increase (or decrease, or no change at all) in success at playoff time as compared to the regular season. As such, you'll see elite national programs who don't fare well at all in this metric. Las Vegas Bishop Gorman being the best example, because, no, frankly, they don't step it up at playoff time. Because they don't need to. Their regular season is effectively their "playoffs", as they play multiple games against top national teams annually. Again, the comparison is with yourself. An excellent program may just happen to be no better at playoff time than during the regular season. That's not a knock on the program.
So, how does this actually work? It's pretty simple math. Using our projection ratings, we just compare how a team's rating at the end of the season looks vs. their rating during the regular season. As an example, say a team is a 30 at the end of the playoffs, and was a 25 at the end of the regular season. That would of course be a "+5". We look at the three previous years (with more recent seasons being weighed more heavily) when determining playoff bumps (or drops) for the current season. Maybe the team in the example with the +5 had a couple other recent years which weren't quite as impressive and ends up with a weighted average of +3. Simply put, they get a 3 point playoff "bump" and any playoff games they may play this season will reflect that when we project the score for their games. So if they otherwise would have been predicted to lose by 10, for example, that would move to being projected to lose by 7. (Although of course the other team's playoff bump/drop would need to be factored in as well.)
To see this in action, check out our project a matchup screen, which now has hypothetical projections listed both ways (if it's a regular season game or playoff game).
It's important to note that this only affects projections for playoff games and has no effect at all on a team's regular season rating. Teams in areas where your association uses power ratings to determine seeding and selection for playoffs will see no change in that process.