4identity shift
Massachusetts: the fanbase's identity shifting.
Joe Harasymiak begins his tenure with a 0-12 losing record to confront — UMass is playing for the version of itself that gets invited to the conversation.
Massachusetts
2-0Mid-American · Joe Harasymiak, year 2 (2-12) · Warren McGuirk Alumni Stadium · 17,000 · 14,895 avg in 2026
The nationunrankedAP / Coaches
The model#137 CFB ZeitgeistSP+ #135 (-25.8) · FPI #117
The marketno line yetvs Stonehill · our model 77% Massachusetts · a number, not a pick
LastW 37–10 vs Sacred Heart · Sat Sep 12 · William Watson III 152 pass yds, 1 TD, 40 rush yds
Nextvs Stonehill · Sat Sep 19 · 3:30 pm ET · Warren McGuirk Alumni Stadium · Game preview
Updated Tue Sep 15 · through week 2
Cautious optimism is building around UMass.
Joe Harasymiak's year 2 (2026): 2-0, ahead of Shane Montgomery's 0-12 in the same spot.
belief cooled ↓ 3.0vs prior week
Fanbase belief
Leaning up
55/100 · 33 mentions · top 17% of FBS
~ our readThe model
#137
CFB Zeitgeist · -8.3
✓ measured4The Pulse on UMass
55
Δ -3.0 vs last wk
33
mentions this week · conversation velocity
What moved it — last 14 days
—
Awaiting next signal window
"Rise as one."
— UMass fanbase · recurring line
Fanbase Signals
Cohort Divergence3 pctSpread between sub-fanbase cohorts in the Pulse window.
The Fever Year
Position = mood · width = volume · color = the kind of feeling
Inside the fanbase — all 7 corners
Cr/CFB
GGoogle News
HHustle Belt
CCampus paper
BBluesky
YYouTube
The split, staged
Warmest corner+0.00
Casuals
most bought-in cohort our read
vs
Coldest corner−0.00
Bettors
coolest cohort our read
Loudest in the room this week real · by engagement
All 7 corners of the fanbase
skeptical
buying in
Casuals+0.24
Bettors+0.16
Cohort sentiment is a source-weighted model read — never a quote pinned to a person.
What this fanbase keeps coming back to our read · topic model
Topics are model-labeled from Reddit — a rough signal of what this fanbase is circling right now.
250 mentions · 6 sources · week 2026-02. Counts, sources and the divergence rank are measured (fact). Cohort labels and sentiment are a source-weighted model read (our read) — never a quote attributed to a person.
Belief checklight on evidence
~ our readMassachusetts fans are split right down the middle, with a bit less hard evidence than most fanbases.
backed by facts
the mood
✓ counted We read 95 posts from Massachusetts fans this week.
Press Pulse
8 quotes collected this season
What the beat said, in their own words.
We are finding ways to be really effective in the second half of games, and we are starting games really well, but we aren’t connecting the two
And now we really got to figure out over this next week who’s going to go in the game if they have to with the twos.
So now is like the grind, like the dog days of camp. So our drills have taken over into that
Five Lenses
What people are saying right now
The broader fanbase8 posts
News & campus coverage12 posts
CFB analyst takes2 posts
Bluesky · @rodger.bsky.social
1w ago
Fanbase health · UMass
Fanbase health · UMassUMass's fanbase is growing — 54 of 100.
Loud (664 weighted voices in the latest week), losing on the field (0-12 last season, 2-0 so far), and not arguing about much.
54
On the field — 0-12 last season, 2-0 so far40
How loud — 664 weighted voices in the latest week86
How much they argue — not arguing about much35
3 of 3 signals · on-field reading uses last season until 4 games are in · 0–100, bands at 26 / 51 / 76
More in this tab
Arguments
Fan mood
Fans are feeling good about Massachusetts
Good morning Happy Saturday!
Will someone who’s in Southwest or central switch dorms with me? I am in Dickinson (orchard hill). Please!
Boston Massachusetts Aggie!
Fan mood · 5 fans, 7 posts scanned · every quote verified verbatim
This page may include content derived from public news sources, social media, and fan communities. We relay these sources as-is — we are not reporting them. Sources may be summarized or quoted imperfectly; for anything that matters, go look at the original source directly. The original source may also be wrong. AI-generated content is labeled and may contain factual errors. All analysis is editorial opinion, not professional reporting. Data sources · About this site