Over the past several tests Morgan has used different solutions to the same problem, even though we've been controlling every possible variable. Room A, for example. In Test 9 he/she exercised precise control to move the boxes one at a time. In Test 10 he/she simultaneously levitated them all straight up a few inches, technically completing the test. Then, in 14 he/she incinerated them. If not for the safety glass, you'd be dead.
What is happening after reset that's leading to this variance?