FlyTris.

A tiny player at the controls.

Falling-rules checkpoint · evaluated

256 neurons · 9089 connections
8 learned output weights

READY TO PLAYDrag to look around
01 /THE WIRED PLAYERSix legs. Six connections. One game.
CURRENT PLAYER / FALLING RULES

A new player at the controls.

190.8Before · test lines/game
280.1Saved checkpoint · test lines/game
+46.8%Mean change on this test set

3.19 hours trained · 63 generations.

Mixed evidence: 24 paired games across eight independent sequences. Mean gain 89.3 lines; 95% interval -45.0 to 205.7. Some sequences regress, and the interval includes zero.

The replay and live opponent now use falling ticks, midair rotations and legal slides. Replay: first test seed at medium gravity.

HUMANS & ONE VERY SMALL OPPONENT

The scorebook.

Two-minute, server-verified ranked matches. Score = lines cleared; pieces placed break ties.

Loading the scorebook…

Human vs. fly

Humans 0 lines

Fly 0 lines

No completed matches yet.

Difficulty defaults control the fly pace. Only games played with those defaults enter the current scoreboard.

EasyDefault difficulty settings

Player / dateHumanFly

MediumDefault difficulty settings

Player / dateHumanFly

HardDefault difficulty settings

Player / dateHumanFly

Top 10 human scores per difficulty, paired with the fly’s score from the same match. Dates use your local time. Names are player-provided.

THE MEASUREMENTS

Original placement experiments.

RunBeforeAfterGain [95% CI]Training
110.1240.71+40.58 [33.33, 48.62]1.1 min
227.46103.83+96.38 [86.75, 103.88]1.4 min
330.0868.58+68.50 [58.33, 78.50]1.3 min

Mean lines cleared on identical, unseen piece sequences. These historical trials used placement rules; their results are separate from the current falling-rules checkpoint.

42.25Trained raw-feature control
0.04Random placements
UNDER THE SHELL

What actually learns?

Board features enter a fixed fly-derived spiking circuit. Eight learned output weights choose a placement from its activity.

The fly animation illustrates recorded decisions. It is not a biomechanical simulation. Removing recurrent connections did not consistently hurt performance in these trials, so fly wiring has not been shown to be essential.

Source & GPL license