It so much depends on the subject and the lighting...
Mainly, I cover swimming events (70%), with other sports being a major part of the remainder (20%). Over a two-day event I will take 5-8,000 shots. For outdoor sports, I expect that about 70% of these will be technically acceptable shots, (both the subject in focus and correctly exposed), but after review only about 10% are acceptable for display/sharing (composition, expression/action etc). This may be due to inevitable duplication, or that someone just looks 'odd', because of their exertion.
When reviewing the typical (indoor) swimming events, that 70% reduces to about 30% (and lots of very sharp, perfectly focussed splash!). I'm very happy if I get 2-300 quality shots of different swimmers. Typically, some of the swimmers are easy to capture, with a very predictable motion, others I just can't quite capture.
Another consideration is the edit/review time: after a day's good shooting I can generally find my 2-300 'real' keepers in just a few hours.
In contrast, I had a very poor and unevenly lit swimming gala - the light varied between lanes, as well as up and down the pool: it took me over a dozen editing/review session (over several weeks) to produce an acceptable collection.
There's no doubt that Don sums it up perfectly...
Don Haines said:
If your standards are low enough, every shot can be a keeper.....
Most of us here on this site are hoping to learn and improve - your 'worst' could be better than my 'best', but I'll never get better without critical review (self or peers).
Anyway, back to the statistics:
This shot was "1 of 900" for that race,
- "1 of 5,000" for the gala
- "1 of 150" for that particular length of the race
- "1 of 20,000" over similar galas
or "1 of the first burst of 8", because the swimmer I wanted was in the first heat after the break...
...I got lucky!