[tomcatfish]

This reply is smug but dreadfully silly.

The giveaway is the handling of information and curiosity. You argue for throwing both away, and it's not clear why. When the author takes away more decimals than they should, the article becomes useless. When the author leaves in more decimals than they should, I round "with my eyes" to my desired precision. As a bonus, I can take their numbers and spot-check them easily.

The author put up a fun piece on a board game review website and summarized that the dice are fine. You ask what the threshold is, I say use your brain and eyes to pick one. We only need to read this once, not grade 200, so we don't need to invent an arbitrary cutoff.

If you treat students or coworkers in this way, I hope it is clear to them that you respect rubrics more than the actual "Ask a question, gather data, answer it in public" scientific process and that they do not mistake stodgy rules for must-follow procedures. It would be a shame to scare people off from rolling dice on the internet because someone may say there are too few p-values or too many decimal places.

[exmadscientist]

I'm actually the anti-rubric guy! What matters to me is whether your arguments are appropriate to support your conclusions. Not whether you jump through the right hoops.

This guy played it straight: these measurements, this result, that conclusion. But the evidence chain was bad: results couldn't be derived from those measurements, and conclusion couldn't be derived from those results. So I called it out.

This is literally my day job, so I don't really like seeing poorly reasoned research, and maybe I'm more sensitive than most. But if you're going to play it straight, I think you should get it right. If that means you use a lot more weasel words, so be it -- if something is your opinion I can't argue with it. But when you state it as a fact, you better be able to back it up.