Correlation, not causation.

Thanks for your thoughts. I am wondering if we're too hard on ourselves. One could argue that underpowered studies also contribute to getting closer to the truth without being able to make definitive claims. So could it be that the problem is not about underpowered studies but how we discuss them?

When talking about professional sports, the smallest effect size of interest is mostly very small. At the same time, it's nearly impossible to get large sample sizes. It's also unlikely that there will be enough similar studies so that a useful meta-analysis can be done in the future. What now?

So, we should give up then? What's your take on how to deal with this?

Do you also have absolute numbers for runs in behind?