About this project

Maine High School Football is the first Local project on niftiIQ Lab. It exists because the official numbers that decide Maine's football playoffs are public but hard to read, and nobody shows the math. Everything here can be traced back to a game result; every method is written down; nothing is advice of any kind.

The Crabtree Index (the official standings)

The MPA Football Bulletin says: "All divisions will use the Crab Tree Point System in determining post-season ranking." The formula, as published by the football committee and reproduced here:

Crabtree = ( wins ÷ games played  +  sum of opponents' wins ÷ sum of opponents' games played ) × 100

Rules that follow from the MPA's practice, each verified against the MPA's own published standings:

Only completed games count. Opponents' records are pooled — all their wins over all their games — not averaged team by team. An opponent you play twice counts twice. A win over a Class A team is worth exactly as much as a win over an 8-Man team. Margin of victory is irrelevant. Out-of-state and non-classified opponents (New Hampshire schools, Fort Kent, Madawaska, Valley) count in your own record but as 0-0 in the opponents' pool. The MPA carries the number to four decimals and shows three.

If the Crabtree standings tie, the bulletin's order of tiebreakers is: head-to-head, division record within the conference, overall record, Heal points, coin toss. Leagues apply those at season's end; we show ties as ties and list Heal values alongside.

Verification: on September 14, 2026, our recomputation matched all 37 team lines we checked on the MPA's pages (Class C, A South, 8-Man Large South, 8-Man Small North and South) to the third decimal. We re-check every week and log any mismatch.

Heal points (the tiebreaker)

Football in Maine does not use Heal points for seeding; they are tiebreaker #4. The football values are 70 points per win over a Class A team, 60 for B, 50 for C, 40 for D, 30 for 8-Man Large and 20 for 8-Man Small. Preliminary index = points ÷ games scheduled (a winless team gets 1.000); tournament index = the sum of defeated opponents' preliminary indexes ÷ games scheduled × 10. We follow the MPA's stated rounding (10 and 12 decimal places internally). These values come from the football committee's explainer rather than an MPA document we could locate; if the MPA confirms or corrects them, the change will be logged.

Where the data comes from, and how

Schedules and results are public facts published by the Maine Principals' Association and by Maine newspapers. We do not run automated collectors against any of those sites. Results are entered by hand from public pages, every entry is logged with what it changed, and the standings recompute from scratch each time. We have asked the MPA for written authorization to read its pages automatically; if granted, this page will say so.

Team names follow the MPA's spellings, including co-op programs (for example, "Oxford Hills/Buckfield Coop" and "John Bapst/Bangor Christian/Central").

The niftiIQ power rating

A least-squares rating on capped margins, re-solved from all games every time a result is entered, with a pull toward average that fades as games accumulate, and an uncertainty band from resampling. It is explained in full on the Ratings page. It has not been backtested this season and says so.

Weekly recaps

Recaps are written by our system from a fact sheet built out of the week's results (scores, margins, standings movement, undefeated teams). When a language model writes the prose it is restricted to those facts, and the fact sheet is published next to the text so anyone can check it. If the model isn't available, a facts-only version is published instead. No human edits the text.

Built with AI, explained by design

This site was built by a non-programmer with an AI assistant. That's a reason to be more careful, not less: every formula is documented, every result is logged, and the official numbers are checked against the MPA weekly. Corrections are welcome at niftiIQ Lab.