Cupping Benchfive roasts a season

How we cup

Four marks, one score out of 100

Every game on the bench gets a fresh account and at least four weeks of daily play, usually more, in Chrome and Firefox on an ordinary laptop. We don't spend during that window, so the balance mark reflects the non-paying experience. Pricing gets checked separately afterwards.

35%Body

Depth. Are the decisions still interesting in week six, or is it timers all the way down?

30%Balance

The shop. How a player who never pays is treated, and how loudly the store asks.

20%Brightness

Pace. How soon a fresh account reaches its first real decision, and whether the middle keeps moving.

15%Finish

Polish. Interface, stability and load times in an ordinary browser over long sessions.

The cupping score

Each mark is out of ten. The weighted total, multiplied by ten, is the cupping score out of 100. Nothing under 70 goes on the bench: a game that scores lower gets no label rather than a bad one, because a shelf of things we don't recommend isn't useful to anyone.

Roast level

Roast isn't a mark. It describes how hot the competition runs, from light (nobody can attack you) to dark (you can be raided while offline), so you can pick the cup that suits the evening you want.

Who pays for the beans

Affiliate commissions, and nothing else. "Brew a cup" buttons lead to each publisher's official site through a tracking link, and if you create an account afterwards the publisher may pay us. That doesn't change the price you pay, and it has never changed a score. Publishers don't see notes before they're published and can't buy a place on the bench.

Re-cupping

Every roast is cupped again each season. When a mark moves, the tasting notes say why and the season on the label changes.