Yeah there's plenty of ways statistics can be skewed one way or another. For example, perhaps the "accidents on descent" wisdom refers to "mountain" objectives (where there is an actual descent to deal with, and it might be different from the ascent route) and the stats in this post might include all climbing, where indoor climbing (including top ropr and auto-belays) and sport climbing might not have so much of a descent, you're just sitting down in the rope and that's it.

And this is why I'm not a big fan of infographics like the one we see here. It looks nice, but it's actually quite hard to understand what it's actually saying.

I would suspect part of that is how they assign "main causes". If you fall, and a piece pulls out when the rope comes taught, is that a "fall" accident or a "piece" accident?

I can see how someone might get hit by lightning, but I can't really see how an accident could happen from gear just failing spontaneously.