Does delay cause more close approaches, or deadlier ones?

The question

Losses rise with delay while time in the bubble barely moves. What changes about a close approach when the positions craft hear are older?

What was tested

We re-flew 96 runs with a recorder that logged, for each close approach, how old and how wrong the heard position was, and how fast the two craft closed on each other. A timing fault in the recorder was fixed and all runs re-flown.

What was found then

Delay does not add close approaches (about 10–11 per craft-hour at every delay); it makes them deadlier. Comparing approaches of similar closing speed, the chance of a collision rose steadily with delay, reaching 4.8× that of perfect information at 2 s. Faster approaches add up to 1.9× on top. How wrong the heard position was did not tell colliding pairs from survivors. An earlier reading, 'closing speed decides', was withdrawn.

What we learned later

What changed

Nothing.