League Playoff Prosperity Baseline (2021-22 through 2025-26)

Generated 2026-05-11. Public PBP-V3 baseline for the prior Celtics prosperity analyses. Compares Boston's lead-protection behavior against every other NBA playoff team across five postseasons.

Coverage: 399 unique playoff games (87 / 84 / 82 / 84 / 62 per season). 352 of 399 games succeeded on first pass; 47 read-timeout failures were retried at 180s timeout and all succeeded. 114,345 tagged events across 28 distinct teams, 80 team-seasons (16 playoff teams per season).

Sources:

• data/league_playoff_games_discovered_2021_22_to_2025_26.csv (399 games)
• data/league_playoff_tagged_events.csv (114,345 events)
• data/league_playoff_tagged_events_enriched.csv (same + usage groups)
• data/league_playoff_fetch_log.csv
• data/league_playoff_context_summary_by_team.csv
• data/league_playoff_context_summary_by_team_group.csv
• data/league_playoff_delta_vs_baseline.csv
• data/celtics_vs_league_prosperity_baseline.csv
• data/celtics_vs_league_per_season.csv
• data/league_top_two_vs_other_gap.csv

---

Definitions

Identical to the Celtics analyses:

• prosperity_lead_8plus: period >= 3, team's margin >= 8 entering event.

• smaller_lead_5plus: period >= 3, margin >= 5.

• non_prosperity_pressure: period >= 3, margin between -5 and 0.

• second_half_all: period >= 3, any margin.

• full_series: every team event in the team's playoff run.

• ender / creator roles. ender_only view is used for headline comparisons (no creator double-counting).

• Force event proxy: rim attempt OR free throw event OR assisted three created.

For each team-season, top_usage_player is the player with the most ender_only events in that team's playoff run. second_usage_player is the next. top_two_usage_combined is both. other_teammates is everyone else.

---

Sample sizes (ender_only, league-wide)

Boston's prosperity ender sample: 477 / 428 / 556 / 307 / 130 across the 5 seasons = 1,898 events.

Champions detected (heuristic: most W in round 4)

• 2021-22: GSW
• 2022-23: DEN
• 2023-24: BOS
• 2024-25: OKC
• 2025-26: still in progress (R2 ongoing as of data pull)

---

A. League baseline. What is normal playoff prosperity behavior?

Across 80 team-seasons of playoff data, the average team-season shows these prosperity-vs-full-series deltas in ender_only view:

A1. Does the average playoff team lose rim pressure with an 8+ second-half lead? Yes. 80% of team-seasons see their rim attempt rate fall in prosperity. Mean drop is -3.54 pp. This is the most consistent prosperity behavior in the league.

A2. Does the average playoff team take more threes? No, not really. The mean three-rate delta is -1.82 pp, the median is +0.01 pp, and half of teams go up while half go down. The "teams jack threes when up" intuition does not survive 80-team-season inspection. (This is one of two findings that updates a Celtics-specific claim from prior reports.)

A3. Does the average playoff team turn it over less? No. Mean turnover-rate delta is essentially flat at -0.19 pp; median is positive at +0.11. Roughly half of teams turn it over more in prosperity and half turn it over less.

A4. Is prosperity usually more efficient, less efficient, or neutral? Slightly less efficient. Mean PPP delta is -0.09 (median -0.02). The standard deviation is 0.25, so single-season swings are large; the center of the distribution is just below zero.

A5. What goes up in prosperity? Free throw events (+1.51 pp mean, 66% of teams rise). This is consistent with the late-clock-with-lead pattern of fouls and bonus free throws.

Plain-English summary of the league baseline:

When a playoff team has an 8+ second-half lead, four out of five teams attack the rim less, three out of five teams lose force-event rate, and the median team loses a few hundredths of a point per possession. Three rate does not systematically rise; turnover rate does not systematically fall. The one thing that reliably rises is free-throw trips.

---

B. Boston comparison

Boston ender_only prosperity PPP vs league each season:

Boston ranks top 8 in prosperity PPP in every season and is #1 in the league in 2025-26 (1.038 vs league mean 0.775).

Boston ender_only prosperity-vs-full-series deltas vs league:

B1. Is Boston worse than league average in prosperity PPP? No. Boston averages +0.03 PPP delta in prosperity; the league averages -0.09. Boston is in the 75th percentile on this metric across team-seasons. Boston's prosperity PPP is above the league mean in every one of the five seasons studied. Boston ranks 8 / 6 / 5 / 3 / 1 in the league across the period.

B2. Is Boston worse than league average in rim-rate drop? No, slightly better. League rim rate falls -3.54 pp on average; Boston falls only -2.37 pp on average. Boston is at the 54th percentile (neutral).

B3. Is Boston worse than league average in force-rate drop? Slightly worse but well within range. Boston averages -2.43 force-rate delta vs league -2.03. Boston is at the 36th percentile (more force loss than league median, but only marginally).

B4. Is Boston abnormal in turnover behavior? Mildly favorable. Boston's team-wide turnover rate falls -0.67 pp in prosperity vs league average -0.19 pp. Boston is at the 40th percentile. The Tatum-specific TO surge documented in prior reports does not drag the team's overall TO behavior below league norm; the rest of the roster overcompensates.

B5. Is Boston abnormal in three-rate behavior? Yes, unusually high. Boston's three-rate delta in prosperity is +2.89 pp vs league average -1.82 pp. Boston is at the 81st percentile. Boston shoots more threes with leads than most playoff teams. That is a Boston-specific pattern (or specifically a Boston era pattern, Mazzulla-era specifically) and not a league norm.

B6. Is Boston abnormal in free-throw behavior? Yes, opposite direction. Boston gets fewer FT events in prosperity (-0.05 pp) while league gets more (+1.51 pp). Boston is at the 34th percentile. This is consistent with the three-heavy diet: a team taking more threes with leads is taking fewer rim/contact attempts and drawing fewer late-clock fouls.

---

C. Star structure comparison

C1. Across the league, are top-two usage players generally less efficient than other teammates in prosperity?

Yes, slightly. Across 74 team-seasons with both groups present:

So the league norm is: top-two stars score about 0.05 PPP LESS than the rest of the roster in prosperity windows. The 25th-percentile team has a gap of -0.21. A quarter of teams have positive gaps (top two outscore others).

C2. Is Boston's Jays vs Other Celtics gap unusually large?

Boston by season:

The 2022-23, 2023-24, and 2024-25 seasons show BOS in the bottom 3-5 of the league on this gap. The 2023-24 season is especially interesting: Boston had the second-largest top_two-vs-other prosperity deficit in the NBA, and Boston won the championship that year. A "Jays underperform the role players in prosperity" gap was not disqualifying.

In 2021-22 (Finals run) and 2025-26 (first-round loss to PHI), the gap was positive: the top two outscored the rest. So the Jays-vs-others gap is volatile season to season and not predictive of postseason outcome.

C3. How do Boston's top-two deltas rank among teams?

In 2 of 5 seasons (2021-22, 2025-26): top-two slightly better than other teammates, around the 12th rank (out of 15-16). Bottom of the league on star-deficit.

In 3 of 5 seasons (2022-23, 2023-24, 2024-25): top-two clearly less efficient than other teammates, ranking 2nd to 5th most negative. In those years Boston's stars are taking the harder shots and the role players are converting easier looks. That is consistent with how the Mazzulla/Stevens-era Celtics are constructed: heavy outside-shooting roster around two heliocentric scorers.

The 2024-25 -0.251 gap (3rd most negative in the league that season) and the 2023-24 -0.178 gap (2nd most negative) are the two seasons where the Jays-vs-other structural gap was most pronounced. One of those seasons was a championship.

---

D. Verdict on Boston's prosperity issue

Mapped to the six options from the brief:

The cleanest characterization: Boston is a top-quartile prosperity team that has had the same Jays-vs-others structural gap for three straight seasons including a championship year. Within that structure, Tatum is the only player on the team whose turnover rate goes up when Boston has the lead, but his team has been efficient in prosperity anyway.

---

E. Message-board paragraph

The league baseline reframes the entire prior conclusion. Across five postseasons of public play-by-play, Boston ranks 8th, 6th, 5th, 3rd, and 1st in the NBA in prosperity PPP. Boston's prosperity efficiency delta is +0.03 PPP (75th percentile); the league loses -0.09 PPP. Most playoff teams get worse with leads; Boston gets slightly better. Boston shoots more threes with leads than the league norm (+2.89 vs -1.82 league average), which is a Boston-specific era pattern, not the "settle under pressure" story Stevens' transcript hinted at. The "Jays underperform the role players in prosperity" gap is real (-0.111 average vs -0.075 league average) but exists in the same shape on most top teams, and Boston rode that exact gap to a championship in 2023-24. Within Boston, the Tatum-specific TO surge in prosperity from earlier reports survives this baseline (he is the only Celtic whose TO rate goes up with leads), but it does not turn the team into a below-average prosperity outfit. The Stevens "prosperity collapse" narrative does not match the multi-year league-relative data: the Celtics have been one of the best prosperity teams in the NBA for five years running.

---

Caveats

• Public PBP only. True possession initiator is not identifiable; events are labeled by possession_ended_by and possession_created_by.
• Free-throw events are attempts not trips. A two-shot trip is two events.
• No shot-clock or defender-distance data.
• "Force event" is a public-data proxy, not Second Spectrum.
• Top/second usage players are determined by ender-event count, not by league-tracked usage rate.
• 47 of 399 games initially failed with 30s read timeouts; all were recovered on retry at 180s timeout and are included in the final totals. No games are excluded.
• 2025-26 playoffs are still in progress (62 games discovered to date, only R1 + early R2). Boston's 2025-26 prosperity sample is small (130 ender events) and the BOS-PHI series is the only series included.
• "Champion" classification is a heuristic (most R4 wins per season) and may need manual override in edge cases.
• This is a 5-season window. League baseline is era-specific (era characterized by high three-point rate, more spacing).