I have a 50w MV bulb that I thought was EOL due to a fried starting electrode after a ballast failure but recently, out of curiosity to see what happened, I put it in a fixture that I had never tested it in. I plugged the light in expecting nothing to happen as I had assumed the bulb was dead but to my surprise, it worked. I then decided to unplug the light before the bulb became too hot to handle. I unscrewed it, put it in the fixture I originally wanted to use it in and again applied power. It worked but this time I let it warm up fully and amazingly, it was still outputting an amazing amount of light for a bulb that I thought was EOL.
I'm trying to figure out what may have caused the bulb to start working again after a ballast failure and multiple failed tests on a known working ballast.
Any ideas as to why this happened would be great

I had an MV lamp where the mercury condensed and bridged the starting electrode with the main electrode and the lamp wouldn't strike. I was able to light it on another ballast and then it worked normally after that. Could also be a marginal connection somewhere in the lamp so current may not be getting to an electrode, or maybe the starting resistor is damaged and not working correctly, or conditions were just such that the ballast OCV wasn't enough to light it before but are now. Could even be a socket issue, or many other possibilities.

I have seen so many electrical and mechanical things over the years develop a problem for one day and not work right or work at all and then work fine the next time...and never could explain why.

I have this well used GE 100w A23 that has trouble starting on most 100w MV ballasts. It would strike on a 70w MH ballast either. It did start on a 100w MH ballast with the ignitor disabled.

I'd reckon it's worn out the point where it won't start on a correct ballast. The 100w MH ballast probably gives it a bit of a extra kick to start it.

After my first time getting it to strike on the ballast I will be using it on, I did a cold start the next morning and it worked again! I have one more MV lamp I was trying to revive but I believe it's EOL because the ballast died and allowed the lamp to draw too much current, frying the starting electrode/resistor in the process.