Every game here is timed on the same machine, from double-click to first input, and then again on every load the game asks for afterwards. The verdicts are about design, but the numbers are about your evening.
Reviews cover the whole game, but every one of them carries the timings: cold boot, warm boot, fast travel, respawn after death, and the wait between pressing continue and regaining control.
Review
8.6
A large open world that loads in four seconds and hides no seams. The fast travel is close to instant, the death loop is under two seconds, and the result is a game that lets you be careless without punishing you with a wait.
Review
6.4
The combat is genuinely excellent and you will see roughly a third less of it than you expect. We timed the respawn cycle across two hundred deaths and the arithmetic of a two-hour session is not kind.
Review
9.0
The best-behaved game in the file. One load at launch, none afterwards, no transition screens between the workshop and the village, and an autosave that never once stalled the frame. Small game, large lesson.
Review
7.2
Eleven seconds of elevator for every deck change, which is craft on the first day and tax on the twentieth. We counted 340 lift rides across a single campaign and worked out what that costs in hours.
Review
7.8
A superb detective story delivered in three-second instalments. Individually trivial, collectively the reason we started timing dialogue transitions at all, and the clearest case in the file for streaming ahead of the player.
Note
A short note on why tactics games now take longer to reach a first turn than open worlds take to reach a first step, and what the two seconds between selecting a unit and seeing its range actually costs a long campaign.
Every figure printed on this site was taken the same way, in the same room, on the same desktop, with the same drive. That consistency is the whole value of the exercise: it does not tell you how fast a game will be on your machine, it tells you how these games compare with each other.
Loading is the only part of a game that is pure cost. Nothing about it is play, nobody enjoys it, and it is the one area where an improvement has no design trade-off attached.
A nineteen-second respawn teaches caution more effectively than any difficulty setting, and it teaches it for the wrong reason. Players stop experimenting when failure is expensive in time rather than in stakes.
A three-second pause is beneath complaint and above nothing. Multiply it by four hundred conversations and you have an hour of a twenty-hour game spent watching a fade, which is precisely the arithmetic reviews never do.
A ninety-second wait with a percentage is tolerable. The same wait with an animated shape and no information produces the belief that something has broken, and several games in this file have taught players to force-quit a working process.
The number is a review score for the whole game. Timings sit alongside it as recorded fact, and where a game is excellent in spite of its waits the review says so before it says anything else.
We do not run a correspondence page, but readers repeat their measurements and post them publicly, and the ones that changed how we work are worth recording.
I timed the same lift rides on a slower drive and got fourteen seconds rather than eleven. Your relative ranking held exactly, which is the useful part.
Nobody had put a number on the respawn loop before. I stopped playing that game at the point where I realised how much of my session was the fade-in.
The point about silence rather than slowness is the one I keep repeating at work. A progress figure costs nothing and changes everything about how long a wait feels.
