wuwiandClaude Opus 5 19c13986c5 Ball 30% faster, all four pace numbers together
baseShotSpeed 16 -> 20.8, qualityShotSpeed 12 -> 15.6, serveSpeed 19 -> 24.7,
serveQualitySpeed 22 -> 28.6. All exactly 1.3x, so nothing about the relative
feel moves: rally and serve, tennis and splatter, his ball and the CPU's, which
share baseShotSpeed.

Swept the full-pull serve down the court: 53.3 m/s against the old 41 at every
range, flight times ~23% shorter (30% more pace is 23% less time in the air).

One exception, and it is geometry rather than tuning: a shambler standing right
on the tape is loft-limited. That shot has to come almost straight down, so the
net-clearance loop trades the extra pace straight back and it ends up only ~14%
quicker. Nothing to fix unless it reads as sluggish next to everything else, in
which case serveSpin is the knob — a harder dip buys pace over the tape.

Tennis serves 8/8 into the box at the new pace; 900 frames clean.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-02 18:53:33 +02:00
2026-07-11 09:23:01 +02:00
S
Description
No description provided
149 MiB
0 Stars 1 Watchers 0 Forks
Languages
JavaScript 75.2%
Python 15.8%
HTML 7.3%
Shell 1.4%
Swift 0.2%