1. Making tweaks as needed to get this to the level that feels confident. I feel making the tweaks on a weekly review helps instead of waiting until after the season.  2. Very much the system is adjusting as teams win and lose games. It is detecting the weaker teams better at this point. First 5-6 weeks are really a crap shoot if we want to be honest because of only so much data. Last two weeks hitting nearly 79% says a lot.  3. Thought is for 2026 is to use at least two years worth of season data, maybe three depending on what it looks like to have the preseason base to help start it off. Once we hit that point of so many games we are fully away from preseason data and focused just on the in-season data basically.

74 percent for a new product is outstanding. It’s hard to get 74 percent as a human. Lol.    Do you think the winning percentage is higher in these later weeks because there is so many mismatch games, or do you think it’s adjusting week by week and getting better accordingly?

Amen.  Play out-of-classification games in weeks 1-3, and district games in weeks 4-11 with 7-/8-/9-team districts.  Examples below.  Too easy... 6A, District 5 Wiregrass Ranch Wharton Tampa Bay Tech Armwood Riverview Bloomingdale Durant Bartow 5A, District 4 Sebring Lakeland Lake Gibson Kathleen Spoto East Bay Braden River Manatee

Maybe they aren’t getting the talent they used to get before it started going to Chaminade, AHP, etc.

So crazy. STA can play any FL team and look like an all star professional team. But the second they play a non-FL team they can’t block, can’t catch or can’t hold on to the football.