Findings
Fourth-quarter rate shifts and their limits
Descriptive splits, not a skill. Per-player fourth-quarter rate shifts are shown with their confounds stated. They are not a late-game skill and not a prediction.
quoted on 2 pages
How it was tested
- Gate: descriptive: Q4 versus Q1-Q3 per-36 rates with an appearance floor
- Corpora: local per-quarter boxes, 2024-25, 2025-26 present in this cache
- Folds: none
Replication
single corpus
Inspector
No inspector page for this finding.
Descriptive Q4-vs-Q1-Q3 per-36 scoring, rebounding, and assist rate shift per player, from local per-quarter boxes -- not a late-game-skill or predictive claim.
The site's clutch_context module honestly REFUSED a clutch number because the grain was wrong. This is what it took to turn that refusal into an honest number: not 'clutch', just the measured Q4-vs-earlier per-36 shift. Published alongside the refusal.
Method. Per-player Q4 vs Q1-Q3 per-36 rates from local per-quarter player boxes; a floor of >=25 Q4 appearances. Descriptive historical splits, not a clutch metric and not a prediction.
Window: 1,231 games (2024-25, 2025-26 present in this cache) — a fixed set of games in this cache, not the current season.
Points shift
floor: q4_games>=25 and q13_min>=100 and q4_min>=25 · n_qualified 409 of n_considered 626
Top risers
| Player | Q4 /36 | Q1-3 /36 | Shift | Q4 games |
|---|---|---|---|---|
| Sandro Mamukelashvili | 29.26 | 15.74 | +13.52 | 52 |
| Doug McDermott | 20.33 | 11.5 | +8.83 | 31 |
| Anfernee Simons | 27.92 | 19.42 | +8.5 | 68 |
| Zion Williamson | 36.59 | 28.84 | +7.75 | 32 |
| Norman Powell | 30.74 | 23.03 | +7.72 | 53 |
| Collin Gillespie | 21.19 | 13.73 | +7.46 | 30 |
| Jamal Cain | 17.94 | 11 | +6.94 | 35 |
| Tristan Thompson | 10.72 | 4.15 | +6.57 | 35 |
| CJ McCollum | 28.24 | 21.78 | +6.45 | 53 |
| Donovan Mitchell | 32.95 | 26.63 | +6.32 | 71 |
| Luka Garza | 24.32 | 18.16 | +6.16 | 30 |
| Olivier-Maxence Prosper | 16.33 | 10.18 | +6.15 | 43 |
| Zach LaVine | 28.45 | 22.34 | +6.11 | 70 |
| Thomas Bryant | 20.68 | 14.85 | +5.83 | 49 |
| Neemias Queta | 16.93 | 11.15 | +5.78 | 56 |
Scroll horizontally for all columns.
Top fallers
| Player | Q4 /36 | Q1-3 /36 | Shift | Q4 games |
|---|---|---|---|---|
| Nikola Jović | 9.87 | 18.49 | -8.61 | 36 |
| Domantas Sabonis | 14.39 | 21.93 | -7.54 | 57 |
| Branden Carlson | 15.27 | 22.04 | -6.77 | 31 |
| Myles Turner | 13.63 | 20.32 | -6.69 | 62 |
| Micah Potter | 3.44 | 10.12 | -6.68 | 27 |
| Gary Harris | 1.88 | 8.43 | -6.55 | 31 |
| Andre Drummond | 8.92 | 15.41 | -6.49 | 25 |
| Steven Adams | 4.87 | 11.27 | -6.4 | 39 |
| Tim Hardaway Jr. | 9.25 | 15.62 | -6.37 | 67 |
| Josh Okogie | 10.97 | 17.11 | -6.14 | 29 |
| Vít Krejčí | 8.63 | 14.73 | -6.1 | 53 |
| Lonzo Ball | 6.73 | 12.7 | -5.97 | 34 |
| Kristaps Porziņģis | 19.49 | 25.41 | -5.92 | 37 |
| Jaylin Williams | 8.98 | 14.78 | -5.8 | 36 |
| Michael Porter Jr. | 15.68 | 21.32 | -5.64 | 66 |
Scroll horizontally for all columns.
Rebound shift
Top risers
| Player | Q4 /36 | Q1-3 /36 | Shift | Q4 games |
|---|---|---|---|---|
| Dalen Terry | 7.49 | 3.66 | +3.82 | 51 |
| Trey Jemison III | 11.81 | 8.46 | +3.34 | 34 |
| Jae'Sean Tate | 9.03 | 5.88 | +3.15 | 46 |
| Ariel Hukporti | 9.88 | 6.75 | +3.13 | 25 |
| Kelly Olynyk | 10.7 | 7.6 | +3.1 | 34 |
| Andre Jackson Jr. | 9.31 | 6.45 | +2.86 | 35 |
| Tari Eason | 11.34 | 8.57 | +2.77 | 48 |
| Charles Bassey | 16.39 | 13.71 | +2.68 | 26 |
| Deandre Ayton | 13.66 | 10.98 | +2.68 | 36 |
| Jett Howard | 5.04 | 2.38 | +2.65 | 53 |
| Johnny Furphy | 7.83 | 5.25 | +2.58 | 41 |
| Christian Koloko | 11.15 | 8.58 | +2.57 | 33 |
| Max Strus | 8.14 | 5.63 | +2.51 | 40 |
| Baylor Scheierman | 7.29 | 4.92 | +2.37 | 31 |
| Tre Jones | 6.55 | 4.18 | +2.36 | 39 |
Scroll horizontally for all columns.
Assist shift
Top risers
| Player | Q4 /36 | Q1-3 /36 | Shift | Q4 games |
|---|---|---|---|---|
| Blake Wesley | 8.36 | 4.76 | +3.6 | 49 |
| Pat Spencer | 8.48 | 5.23 | +3.26 | 31 |
| Maxi Kleber | 4.68 | 1.72 | +2.95 | 30 |
| Rob Dillingham | 8.53 | 5.6 | +2.93 | 38 |
| Dejounte Murray | 10.19 | 7.32 | +2.88 | 26 |
| Jalen Pickett | 7.62 | 4.88 | +2.74 | 41 |
| Shake Milton | 6.08 | 3.63 | +2.45 | 48 |
| Reed Sheppard | 5.55 | 3.38 | +2.17 | 44 |
| Aaron Holiday | 4.86 | 2.87 | +1.99 | 58 |
| Jordan Goodwin | 4.38 | 2.49 | +1.89 | 27 |
| Craig Porter Jr. | 5.84 | 3.99 | +1.84 | 44 |
| Micah Potter | 3.01 | 1.24 | +1.77 | 27 |
| Ousmane Dieng | 3.92 | 2.3 | +1.62 | 33 |
| Rayan Rupert | 2.58 | 1.01 | +1.57 | 46 |
| Orlando Robinson | 4.67 | 3.11 | +1.56 | 32 |
Scroll horizontally for all columns.
Confounds
- Q4 minutes are heavily lineup- and blowout-dependent: stars sit in blowouts, benches play garbage time -- this is NOT playing-time-neutral.
- per-36 on small Q4 minutes is noisy even at the floor; treat as descriptive, not a ranking of ability.
- NOT a clutch metric (clutch requires score/time context this does not have) and NOT predictive.
- fixed historical game set; not current-season.
Descriptive Q4-vs-Q1-3 per-36 splits only, from a fixed observation window — not a clutch metric, and not a forecast-quality measurement.
Limitations
- Sample and independence: quarters repeat players and games; per-36 rates from few Q4 minutes are noisy.
- Scope: the seasons present in the local cache, players above the floor only.
- Does not establish clutch skill or predict future fourth quarters.
Verdict history
- : current: no packet verdict (descriptive exhibit)