I just also want to concur, I've had 3 Mandarin juveniles jump, 2 male, 1 female, only one at any one time in a 2 year established Fluval 13.5 as a species-specific tank, outside of snails and LPS corals. The 2 males 1" inch length is from biota and the 1.5" inch female was unknown. In both sexes, when the light shuts off immediately at night, they have a darting behavior, not as if being chased, but as if they are looking for some escape. It was very apparent, one minute they are slow moving hunt and peck observers, and when the light is out, they are in beast mode scanning and mapping out the light from the dark stock hood. The first male exhibited the behavior, and I was rookie and had a Finnex hang on back refugium in the back that cracked the lid open the lid a smidge. I watched this odd behavior for about an hour, and went to bed only to find the fish chip in the morning. The second male, I had left the lid closed, including covering the feeding hole. I did however removed the filter covering off, leaving a little thin exit close the first chamber entrance. As the light went off, the first week, I noticed the same darting behavior, and turned on the light and entice it to spot feed live baby brine, but there was no way of getting his attention. Eventually, by the end of two weeks, he found the hole by the first chamber even though I laid a filter pad to cover it. Next day, filter floss was flipped into the first chamber and fish chip on the floor. The third female also exhibited the same behavior, and I closed all holes. I did hear at night, an odd knocking from the room, which later on I pieced together was probably the failed jumps. But during cleaning one day, lights out, I had the hood propped open, I went to get a drink. When I came back I couldn't find her, and there was no water droplet trace that she jumped. It wasn't until next day I found her in one of my cleaning buckets that I swear was 3-4 feet away from the tank, dried up.
I was guessing it was the angle of incidence that specifically this tank was exhibiting when the stock light was off. They all behaved aggressively to their mirror images from the glass, and it might be because of the many incidences that it would occur in this small tank, the higher likely triggering the behavior. Overall, I'm glad someone else can speak about the same observation, because I thought it was just me. It couldn't have been just bad luck.