DroneX · Radio

Does the channel do anything? Drones can transmit four numbers nothing names, and the ground can hear them. Whether either matters is answered by silencing one and flying the same genome against the same opponents again.

Does the radio matter

silencing a channel · + means the swarm got worse without it
island air ground all runs volume entropy
beam00 8ddd
n22
n19
n25
2831 75788379 1683
beam01 a6b1
n27
n26
n25
2981 79398848 2052
beam02 60ac
n27
n23
n24
3001 50864190 1874
beam03 e649
n26
n22
n26
2987 97574243 1927
msi00 d44f
n21
n19
n20
1478 100311834 883
How this is measured, and why ±25 is not a result

An island re-runs the same engagements with one channel silenced and reports the difference in the raider's score. Same genome, same opponents, one thing changed — which makes it the only number here immune to the hardware differences between these machines, and the only one that can answer why rather than what. ⚠ It is measured over a handful of engagements, so it moves in steps of about 25 points and one fight changing hands is a whole step. Each dot is one exercise, not one sample. The wire republishes a result until the next is run and the board samples every 30 seconds, so a single exercise used to be drawn up to eighty times: a cloud whose weight came from the sampling rate rather than from evidence. Consecutive repeats are collapsed and the count that survives is printed as n. The run count in the table is how many exercises the ISLAND has run, which is far larger than how many distinct results have reached this page. Treat a single ±25 as noise. What would settle it is the same sign holding over many, which is why nothing is claimed below eight. Volume is how much was transmitted and entropy is in millibits over sixteen buckets: a channel driven with a constant is silence wearing a signal's clothes, and it would pass a volume test. Talking, saying something, and it mattering are three different facts.

Each island's history

One exercise cannot settle a sign. What settles it is the same number drifting off zero across many, so these are the trajectories rather than the latest reading.

beam00 8ddd

Does the radio matter, over time

The change in the raider's score when a channel is silenced, sampled over time. One reading is noise — the measure moves in steps of about 25, so one engagement changing hands is a whole step. What settles it is a line sitting off zero and staying there. Drawn as steps, because the wire republishes one exercise until the next is run. Written down, so it survives a deploy.

beam01 a6b1

Does the radio matter, over time

The change in the raider's score when a channel is silenced, sampled over time. One reading is noise — the measure moves in steps of about 25, so one engagement changing hands is a whole step. What settles it is a line sitting off zero and staying there. Drawn as steps, because the wire republishes one exercise until the next is run. Written down, so it survives a deploy.

beam02 60ac

Does the radio matter, over time

The change in the raider's score when a channel is silenced, sampled over time. One reading is noise — the measure moves in steps of about 25, so one engagement changing hands is a whole step. What settles it is a line sitting off zero and staying there. Drawn as steps, because the wire republishes one exercise until the next is run. Written down, so it survives a deploy.

beam03 e649

Does the radio matter, over time

The change in the raider's score when a channel is silenced, sampled over time. One reading is noise — the measure moves in steps of about 25, so one engagement changing hands is a whole step. What settles it is a line sitting off zero and staying there. Drawn as steps, because the wire republishes one exercise until the next is run. Written down, so it survives a deploy.

msi00 d44f

Does the radio matter, over time

The change in the raider's score when a channel is silenced, sampled over time. One reading is noise — the measure moves in steps of about 25, so one engagement changing hands is a whole step. What settles it is a line sitting off zero and staying there. Drawn as steps, because the wire republishes one exercise until the next is run. Written down, so it survives a deploy.