How a verdict is made
Updated 22 September 2026
Every verdict on this site is made the same way. It is slow on purpose.
Who rules
Two of us govern each realm on fresh accounts on a normal live server, without press access or free premium currency. The third reads both sets of notes and asks the awkward questions. Minimum three weeks; longer where a world has a round structure worth seeing through.
The four criteria
| Criterion | Weight | The question |
|---|---|---|
| Time | 35% | Can you make real progress in a short session? Are timers pacing or pressure? Does it punish a day off? |
| Fairness | 30% | How far does a free ruler get, and does paying buy convenience or victory? Are prices shown plainly? |
| Depth | 20% | Is there still something to learn in week three? Do alliances, guilds or fellowships create real social play? |
| Polish | 15% | Does it load fast, run in the four major browsers and stay readable on a laptop screen? |
Each criterion is scored out of ten by both rulers independently. Where they differ by more than two points we play another week and settle it. The verdict is the weighted average rounded to one decimal place.
Example: 9.0, 8.5, 8.5 and 8.5 gives (9 × 0.35) + (8.5 × 0.30) + (8.5 × 0.20) + (8.5 × 0.15) = 8.675, published as 8.7. That is Forge of Empires this sitting.
The paid week
After the free period, one of us spends a fixed amount, roughly a cinema ticket where we live, the way a typical player would and records what it bought and how far it moved the account. That week feeds the fairness score.
Re-tests and lost seats
Realms are re-tested when they change their shop, release a major update or when readers report something material. A realm loses its seat if it breaks an entry rule, scores below 7.0 or its community shrinks to the point where multiplayer stops working.
Independence
Affiliate links carry the sponsored attribute. Verdicts are locked before links are added. Publishers have no editorial input. If you think a verdict has something wrong, tell us and we will look again.