Skip to content

Teaching efficacy · Dashboard

Ensemble Recruitment & Retention

Thirteen seasons of marching ensemble data — 859 unique members, 302 feeder high schools, with longitudinal cohort persistence as the metric that matters most.

Data window: 2013–2025 · compiled 2026-05-04 from a deduplicated master roster, with member identity matched on student ID, phone, and name across years.

Headline finding

Four-year cohort persistence is up +12 points under my direction.

Members who entered the program in the founding era (2013–2020) stayed for four full seasons at an average rate of 19%. The 2022 cohort — my first as sole director — stayed at 30%. Recruitment numbers tell a noisy year-to-year story; longitudinal stickiness is what tells you a program is healthy.

At a glance

Members tracked

859

unique people across 13 seasons

4-yr persistence · founding era

19%

avg across cohorts 2013–2020

4-yr persistence · my tenure

30%

+12 pts vs. founding era

New feeder schools

120

opened since 2021 · 15 out-of-region

Cohort persistence

The single most honest test of a program: of the people who entered in year X, how many were still on the roster four years later?

Cohort survival curves

Lines flatter and higher = cohorts stuck around longer.

4-year persistence by cohort

Share of each entering cohort still on the roster at year 4. Bars are colored by who was directing when the cohort entered.

Year-over-year roster

Roster size, 2013–2025

Each bar is a fall roster, split into returners (peacock) and new recruits (ochre). 2013 has no prior-year comparison; 2023 reflects a partial-data year (see notes).

Year-over-year retention

Returning members as a share of the prior year's roster. The vertical marker shows the year I joined the program.

Reach & composition

New feeder schools opened, by year

Each year's first-time feeder schools — high schools that hadn't sent a member to the program before. Bars after the marker are schools brought online during my tenure.

Geographic reach over time

Each fall roster broken out by member home region. Out-of-region recruits indicate the program's national reach.

Top feeder high schools

All-time top feeders by member count. Schools first opened during my tenure are marked.

Section composition

Brass, woodwinds, percussion, and color guard counts over time. Color guard has grown the most in proportion.

Notes & methodology

  • Source. Master deduplicated roster spanning 2013–2025, compiled from yearly Microsoft Forms exports plus archived course-grade rosters. Pre-2023 records did not capture institutional student IDs (700 numbers); identity is matched on normalized name and phone for those years.
  • Era framing. 2013–2020 is the founding era under my predecessor; 2021 is a co-direction year (when I joined as Associate Director); 2022–2025 is my tenure as sole director. Era markers in the charts reflect this.
  • Why cohort persistence and not just headcount. Total roster size is noisy: a single recruitment year can swing it by 25%. Cohort persistence — the share of an entering class still present at year 4 — controls for that noise and reflects program quality of life. It is the closest analogue to a college-retention metric.
  • 2023 data note. The master roster's 2023 export captured only 73 records, well below the ~120–140 typical, so the extractor patches in the original 2023 information form (147 members) for that year specifically. Subsequent years' returner comparisons use that patched record.
  • Data refresh. scripts/extract-mavstamp.py reads the master CSV and regenerates the dashboard's data module. Re-run after each season's master roster is updated.