How the Regionals numbers are made

1 · Who runs

Only runners on the official regional entry lists (as of Thu Oct 22, 2:28 PM). Each school runs its 7 fastest entered runners by our projection; a school with fewer than five entered runs as individuals with no team score. A runner we have no 2026 race for gets a cautious back-of-the-field time and the widest range, marked "no 2026 race yet".

2 · Each runner's projected time

Every 2026 varsity 5K result counts. We take out what isn't the runner: how hard each course is, the weather, the day itself, and the fitness runners of each grade gain through the season. What's left is each runner's level, which we put on the regional course (Belfast for the North, Cumberland for the South) on October 24. Races that aren't varsity 5Ks (like a 4,000 m race) are left out.

3 · 4,000 simulated races

We run each race 4,000 times. In every run, each runner has a good or bad day as big as runners really have, bigger for runners with only one or two races this season (we know less about them), and each team also gets a shared good or bad day, because whole teams often run well or poorly together. Each run is scored like the real meet: five score, the 6th and 7th push other teams back, a tie goes to the better 6th runner.

4 · The chances

Chance to win, top 3, to States = the share of the 4,000 runs where it happened. We never show 0% or 100%: anything can happen on race day.
State meet line = half of the teams with five or more runners entered, rounded up, at least four, and the top 30 runners (the MPA's 2026 rule). The MPA decides who qualifies.
Likely finish / likely place = where it lands in 8 of 10 runs.
Bad day / doesn't run (Plus, on school pages; runners numbered) = the same runs again with that runner on their slowest day in 10, or out of the race, so you see how much each runner carries the team.

5 · How well it works

We tested the exact same steps on races we already know the answer to, using only results from before each race:

2026 Festival of Champions: using the entry lists cut "phantom" teams (teams projected that didn't really field five) from 25 to 9 and halved the projected runners who didn't run; the typical runner was off by about 8 places (boys) and 6 (girls); our projected top 30 held about 84-85% of the real top 30.
2024 and 2025 regionals: we had the right team winner in 20 of 24 races; the "likely place" range held 85-86% of runners (it is built to hold 80%).

6 · Every input: Festival of Champions vs Regionals

Brett asked for certainty that Regionals uses everything the Festival of Champions used. Here is every input, side by side.

Input	Festival	Regionals	Notes
Every 2026 varsity 5K result we have	IN	IN	KVAC and PVC each split into their two divisions; Red Riot (4,000 m) left out: varsity 5K only.
Times scaled to 5K (2-mile and 4K races)	IN	IN	Same rule.
Course difficulty for every course, 2023-2026	IN	IN	Same four-season course model, with re-routed course versions.
The race course itself	IN · Belfast	IN · Belfast + Cumberland	Troy Howard (North) and Twin Brook (South) each have their own rating.
Past races' weather taken out (heat + humidity)	IN	IN	Same measured table.
Race-day extras taken out (mud, long course, a wrong turn)	IN	IN	Same model.
Time of season: each grade's improvement to race day	IN	IN	Freshmen, sophomores, juniors and seniors each have their own curve to Oct 24.
Race-day forecast (heat cost)	IN · each section's start time	IN · 10 AM-1 PM at each site	2026 start times aren't posted; we switch to each race's time when they are.
Rain and wind	not in	not in	We measured them: no consistent effect on times (see the Weather page), so neither meet uses them.
Calibration (gaps between runners pulled to what's really run)	IN · Festival factor	IN · championship factor	Each fitted on its own kind of meet (past Festivals vs past regionals and states).
Official entry list only	IN	IN	Runners not entered are left out.
Name matching (spelling, word order, near spellings)	IN	IN	The same five steps.
Coach scratch list	IN	IN	None reported for Regionals yet.
Entered runners with no race in our data	seed time from the entry list	cautious time + widest range	The regional entry lists have no seed times, so we can't use them; these runners are marked 'no 2026 race yet'.
Each runner's good or bad day	IN · one size for all	IN · sized by races run	More market margin for runners with 1-2 races (the Festival lesson).
A shared team good or bad day	not in	IN	Added after the Festival showed whole teams run well or poorly together.
An extra freshman boost	not in	not in	Not yet tested on past regionals; freshmen already have their own (bigger) season curve.
7 runners per team, scored 5 + 2 displacers	Festival rules	IN	MPA and national high school rules; incomplete teams run as individuals.

It is a projection from the data, not a promise. After the race we grade it in public, the same way we graded the Festival of Champions.

We don't show this to promote or demote any athlete. Results are entered by hand from public meet result files. Team projections and every runner (shown by school and number) are free; each runner's impact on the team is part of Plus; runner names are for coach accounts only.