Team Subjective Highlights

Executive summary

How a video editor interprets the data

When we win, the video tends to

Breathe more.

Fewer total PBPs, but each moment stays on screen longer, and the whole video is a bit longer (~4%). It feels less frantic, more “let this moment land.”
Lean into joy.

After our goals, we spend a lot more time on celebrations (especially in games where both teams scored). The story is: we scored, and we’re feeling it — linger on hugs, runs, crowd, bench.
Show our misses a bit less.

We include fewer “we almost scored” moments overall — the video doesn’t need to prove we could have; the result already did.
Let the rival sweat (a little).

When they fluff a chance, we give more of their frustrated reactions — it reads as “they had their moments but couldn’t take them,” which flatters our win.

When we lose, the video tends to

Cut faster and pack more in.

More PBPs, shorter average PBP length — snappier, more “here’s everything that happened,” less time to sit in any single beat.
Spend less time celebrating our goals (where both teams scored).

We still show we scored, but we don’t milk the happy bits as long — the mood of the day pulls the cut forward.
Show more of our near-goals.

More own-team non-goal chances — often reads as “we were in it,” “look how close it was,” or simply “we didn’t get the breaks.” It’s the classic “we fought / could’ve been different” layer.
Give the opponent a bit more spotlight.

More editorial emphasis on their threatening moments — visually, the opponent’s story gets a bit more air without leaning on small-sample replay counts in this report.

One-line takeaway for the desk

Wins sell emotion and clarity: longer holds, bigger celebrations, fewer “almost” PBPs, a touch of rival pain. Losses sell effort and narrative: quicker pace, more “our chances” and “their moments” to frame the game, with celebrations trimmed so the piece doesn’t pretend it was a party.

Highlights from the data

Bottom line from paired fixtures; exact n varies by column completeness in the sheet — see each figure. % convention: where winner videos are higher on average, we report 100×(W−L)/L (“vs loser baseline”). Where loser videos are higher (e.g. PBP count), we report 100×(high−low)/low using the lower mean as baseline.

Figure 1 — Rival non-goal chances (optional exposure)

Optional non-goal attacking PBPs for the opponent in the fixture (counts).

Figure 2 — Self non-goal chances (optional own-team exposure)

Non-goal attacking PBPs for your own team — shots and attacks you chose to show beyond mandatory goals. Compares to optional rival chances in Figure 1.

Figure 3 — Pacing (PBP structure)

Average length of a play and how many plays make up the highlight (excluding total runtime — see Figure 4).

Figure 4 — Total runtime

Figure 5 — Persona & emotion (score-adjusted)

Raw celebration seconds mostly track how many goals each side scored, which biases winner videos upward. Here we control for score and exposure volume: self celebration ÷ goals scored by that channel’s team (from the match score columns), self reaction ÷ self non-goal chances (only rows where that side had at least one self non-goal chance), and rival reaction ÷ rival non-goal chances (cordial rival face-time per optional rival chance shown). Celebration and reaction are not opposites — both are persona dwell; these ratios make videos comparable across different scorelines.

Self celebration per goal scored

Self reaction per self non-goal chance

Rival reaction per rival non-goal chance

Deep dive context for future use

Caveat

Sample sizes are small (n in the teens–30s for some metrics). Treat this as directional editorial instinct, not a rigid formula — brand, opponent, and how ugly the loss was still override averages.

Cordiality

Any extra exposure to the rival is “generous” in this framework — not only optional chances. Rival reaction time is not the opposite of self celebration; it is another form of screen time you can give the opponent. High rival exposure (reactions, rival celebrations you still show, etc.) reads cordial in the same direction as showing rival non-goal chances — you are ceding space in the video.

What this study does not try to explain

We did not analyze what happened in the match in terms of type of action (e.g. shot shape, set pieces, “quality” of save or goal). The play-by-play content of the match is given: goals and obviously strong plays will appear. There is no meaningful “editorial voice” in which PBP moments get in at that level. The editorial voice shows up in pacing, how things are split (how many of each kind of beat, not which specific plays), exposure to people (how much time on self vs rival personas), emotional emphasis, and how much room is left to the opponent in the cut.

What each metric means

Short definitions aligned with the tagging sheet and docs/methodology.md.

Rival non-goal replay counts remain in the workbook and stats.json; this dashboard does not chart them (small n).

Later & exploratory (not the headline claim)

Per-club views need larger n per club — optional follow-up.
Rival identity (e.g. big-six vs rest): future scope for whether pacing or generosity shifts by opponent draw.

Before any per-team view

Per-team charts are exploratory only. The aggregate window spans only matchweeks; slice-by-club sample sizes are tiny — use as a conversation starter unless tagging expands.

References

docs/results.md — paired tables & synthesis.
docs/methodology.md — definitions, PBP accounting.
scripts/build_stats.py — regenerate stats.json.