HOOPSIM
Sign in Get the app Play Now
← All columns State of the Game

Halfway through RSB, the numbers keep leaning toward the real league

Vol. 17 · Oct 22, 2026. RSB has reached the halfway mark, and across 12222 new games most of the league's key lines have kept moving toward real real-league numbers.

Play Now →

HoopSim is a free, browser-based basketball GM game, alone or in leagues with friends. Every game your club plays counts toward the next column.

The gameA free basketball GM sim in your browser. You run one of 30 pro clubs.
The leagueThe other 29 are AI GMs, your friends, or both.
The priceFree to play. AI conversations are Pro, with a 7-day trial. Free vs Pro →
The appNo app store. Get it on your phone →
Reading · Vol. 17Scroll to read
HoopSim
HOOPSIM · OFFICIAL COLUMN VOL. 17 · OCT 22, 2026
State of
the Game
25281 GAMES · ALL LEAGUES · 22 real-league CHECKS
Dr. PossessionAri Lindqvist · @DrPossession · Engine & realism

Halfway through RSB, the numbers keep leaning toward the real league

RSB has reached the halfway mark, and across 12222 new games most of the league's key lines have kept moving toward real real-league numbers.

Halfway is a good place to stop and read the whole notebook. Since Vol. 16 we have added 12222 new games, bringing the sample to 25281 regular-season games across 203 HoopSim leagues.

Our Realism index now has 22 checks, with 19 on target (86%) and all 22 within striking distance. Nothing sits in the lab as a miss, and the remaining work is about refining the last few.

REALISM INDEX · 22 CHECKS VS 2025-26 real-league
100%NEAR-real-league
22 / 22
within striking distance
19 / 22
dead on target
Every regular-season game across all HoopSim leagues, measured against the 2025-26 real-league.

Pace and scoring both drifted down

Pace moved from 106.6 to 103.8, while the real league sits at 99.4. The gap is closing, but it is still the clearest open question in the lab, and I want to know whether this is a true tempo preference or a question of how possessions are counted before I call it.

WHAT MOVED SINCE VOL. 16
TOWARD THE real-leagueAWAY / FLAT
  • ▼Rebounds46.0 → 44.8real-league 43.8
  • ▼Pace106.6 → 103.8real-league 99.4
  • ▼Points / game119.4 → 116.4real-league 115.6
  • ▼Assists28.0 → 27.3real-league 26.7
  • ▼Blocks5.2 → 5.1real-league 4.8
  • ▼Turnovers14.3 → 14.0real-league 14.5
Before → after, with the real league mark on the right.

Points per game moved from 119.4 to 116.4, against an real-league figure of 115.6. Scoring is now sitting close to the league's own level, which is exactly where a sim should land after a stretch of drift.

“Halfway isn't a finish line. It's the point where the numbers start telling you what to measure next.”

Ball security holds steady, passing a touch lighter

Turnovers moved from 14.3 to 14.0, a small step away from the real league's 14.5, and that is a shift I'm watching rather than dismissing. Turnover rate sits at 12.2% against the real league's 12.7%, and both lines remain on target.

Steals are at 7.6 against an real-league figure of 8.4, which is the gap I'd most like to narrow next. Assists eased from 28.0 to 27.3, still a hair above the real league's 26.7, so ball movement is staying in a healthy range while the lab keeps testing it.

Rebounds and blocks keep closing in

Rebounds moved from 46.0 to 44.8, now just above the real league's 43.8. That is the kind of slow, honest convergence we want to see as the sample grows.

Offensive rebound rate sits at 24.4% against the real league's 26.0%, which the Engine Lab marks as close rather than on target. Blocks moved from 5.2 to 5.1 against an real-league figure of 4.8. Whether the offensive rebound gap traces back to rim and paint share is the next thing on the bench, and I'll report it when the test is complete.

Contenders, rebuilds and a crowded middle

Across 1889 teams, net rating runs from +32.1 at the top to -32.1 at the bottom, with a full spread of outcomes in between. The top buckets are populated: 96 teams sit between +10 and +15, and 78 clear +15.

PARITY · 1889 TEAMS BY NET RATING
≤−10: 172−10…−5: 274−5…0: 4780…5: 5395…10: 25210…15: 9615+: 781722744785392529678≤−10−10…−5−5…00…55…1010…1515+NET RATING (PTS PER 100 POSSESSIONS)

The middle is where the league spends most of its time. The buckets from -5 to +5 hold more than half of the 1889 teams, while 172 teams land at -10 or worse. That shape is what a competitive league looks like, with room for a rebuild and a contender in the same season.

The three-point line settles near the real league

Three-point attempts make up 40.5% of shots, against an real-league share of 41.5%, and the weekly line has spent most of its recent stretch hovering around that same mark. Shot-making has landed exactly on the real league's 36.0%, which tells me the volume is settling for the right reasons.

3PA SHARE · BY WEEK▲ +9.4
60%50%40%30%20%10%real-league 41.5%WK OF 10/13: 32.7%10/20: 32.0%10/27: 34.0%11/03: 32.7%11/10: 33.1%11/17: 33.1%11/24: 33.6%12/01: 33.4%12/08: 33.9%12/15: 32.2%12/22: 33.9%12/29: 34.0%01/05: 32.9%01/12: 33.5%01/19: 33.5%01/26: 33.7%02/02: 32.8%02/09: 32.4%02/23: 33.2%03/02: 34.0%03/09: 33.5%03/16: 31.9%03/23: 32.2%10/16: 28.5%10/23: 27.4%10/30: 26.7%11/06: 25.9%11/13: 27.3%11/20: 26.7%11/27: 25.8%12/04: 27.1%12/11: 26.3%12/18: 26.3%12/25: 27.8%01/01: 31.3%01/08: 34.6%01/15: 34.0%01/22: 34.5%10/20: 41.5%10/27: 42.8%11/03: 42.8%11/10: 42.8%11/17: 42.5%10/19: 40.6%10/26: 40.0%11/02: 40.3%11/09: 40.2%11/16: 40.3%11/23: 40.5%11/30: 40.3%12/07: 40.5%12/14: 40.9%12/21: 41.9%12/28: 41.8%01/04: 41.1%01/11: 41.2%01/18: 41.1%01/25: 41.4%02/01: 41.5%02/08: 41.5%02/22: 41.5%03/01: 42.1%03/08: 42.6%03/15: 41.6%03/22: 41.9%03/29: 42.2%04/05: 42.5%10/18: 43.0%10/25: 42.3%11/01: 42.5%11/08: 43.1%11/15: 43.9%11/22: 42.8%11/29: 43.6%12/06: 43.3%12/13: 43.2%12/20: 43.1%12/27: 43.2%01/03: 42.5%01/10: 43.1%01/17: 42.8%01/24: 43.0%01/31: 43.4%02/07: 44.9%02/21: 44.1%02/28: 44.9%03/06: 45.7%03/13: 44.7%03/20: 45.1%03/27: 45.5%04/03: 45.5%10/16: 45.0%10/23: 41.8%10/30: 42.0%11/06: 41.1%11/13: 42.4%11/20: 42.2%11/27: 41.7%12/04: 42.4%12/11: 43.5%12/18: 41.7%12/25: 42.8%01/01: 42.1%01/08: 43.5%01/15: 43.7%01/22: 42.6%01/29: 43.5%02/05: 43.1%02/12: 43.5%02/19: 41.5%02/26: 43.9%03/05: 43.5%03/12: 43.3%03/19: 42.8%03/26: 42.7%04/02: 42.1%32.7%42.1%WK OF 10/1311/0911/0110/2304/02

I still want to know if three-point volume explains the offensive output gap or if something else is doing the work. That answer belongs in the next column, once I've tested it against the weekly data.

NEXT WEEK IN THE LAB
1Pace question. Pace sits at 103.8 against the real league's 99.4, and I'll decide whether it is tempo or counting.
2Rim and paint. Offensive rebound rate is at 24.4% against the real league's 26.0%, and I'll test whether rim and paint share explains it.
3Scoring drivers. Points per game is at 116.4 against the real league's 115.6, and I'll check whether three-point volume or something else drives the remaining gap.

Thank you for reading along for half a season of measuring. The best part of the lab is still ahead of us.

— Dr. P

SOURCE: Engine Lab · 25281 regular-season games across all HoopSim leagues · benchmarks: real-league 2025-26