JasoninHalifax wrote:
It's kinda like in swimming terms, who's faster, phelps or leclos. But you put each of them in pools in the ocean subject to currents and waves, you let the salinity and temperature vary considerably without measuring it, you change the height of the starting blocks, the pool is "roughly" 50m long, but you aren't quite sure because each wall is free floating.
How many trials do you need to run to get a meaningful number?
And I'd say to that - again, your best bet would be to use tests that mirror your racing conditions.
You want to find out who is the fastest at 100m free in a pool? Use a 50m POOL test.
You want to find out who is fastest in a 3-mile OWS? Test in OWS.
If the best swimmer has a much bigger gap than all the others, they'll likely win in all test trials, in all conditions, and in all races, which makes it easy to find the winner, no surprise.
But let's say you have 2 twin Michael Phelps's, both who train similar intensity, but one who does a lot of focused training as an open water 3 mile specialist, and the other specializes in pool racing from 100 to 1500m. THe difference will be a lot less, and your best bet in assessing their performance in a 3mile OWS vs 100m free in the pool - is to test it in that race condition. (Or in this case, just race them in that condition.)
In terms of the required sample size (n), that depends on the magnitude of the difference. It's the equation for standard error, you can wiki it.