This isn't a guessing game, really. You have some prior knowledge of the expected effect size - that allows you to calculate the sample size required for your s

This isn't a guessing game, really. You have some prior knowledge of the expected effect size - that allows you to calculate the sample size required for your statistical test to have a certain power. That is an essential and very simple part of a good experimental design which, unfortunately, is very often ignored.