How we actually test a game before we score it

People sometimes ask if we just play a level and slap on a number. We don't. Every game gets multiple sessions across at least two different phones, usually a newer one and something a couple of years old, because performance matters and not everyone upgrades yearly.

We keep notes as we go — first impressions, where it got annoying, whether the controls held up under pressure. Those raw notes usually contradict the trailer, and that gap is where the useful part of a review lives.

Only after all that does anyone talk about a rating. We argue about scores more than we'd like to admit, which is probably a good sign. A number nobody questioned is a lazy number.

The point is repeatability. If two of us play the same game we should land in the same neighborhood, and if we don't, we go back and figure out why.