Interface - Designing Games - page 244

The right frame rate depends on the needs of a design. Some old
text adventure games would run at less than 1 fps, since they took time
to type descriptions on the screen. Shooters like Quake 3 can be run at
200 fps or more. Games running at 1 fps are viable if they are turn-based
and controlled through key presses. The 200 fps rate is the upper level
of frame rates pursued by competitive PC shooter players. These players
run 10-year-old games on hardware 30 times more powerful than the best
available at the game's release because they want every ounce of competi-
tive advantage they can get.
The vast majority of games, though, run at either 30 fps or 60 fps.
Few players can feel an improvement higher than 60 fps, and frame rates
lower than 30 fps are so low as to be visually choppy. So usually, you'll face
a choice between these two alternatives. At 30 fps, with the optimal four
frames of latency, there is 100 to 133 ms of latency plus screen lag. At 60
fps, the delay drops to 50 to 67 ms plus screen lag.
In a world without limits on computing power, we'd always choose 60
fps. In real life, though, 60 fps has serious costs. John Carmack, legendary
programmer of the Doom and Quake games, said that a 60 fps game has
about one-third the processing power available per frame as a 30 fps game.
That's a costly trade-off in graphics and processing complexity.
The choice between 30 fps and 60 fps depends on the expected skill
of players, the type of control interface, and the game's specific emotional
triggers. A game based on art or narrative is often best left at 30 fps be-
cause this allows for richer visuals. So is a skill-based game that doesn't
depend on rapid input, like a turn-based strategy game. Where 60 fps be-
comes essential is when high-skill players interact with systems demanding
rapid-response input. Shooters, racing games, and fighting games are the
most common candidates for 60 fps.
Often, there's pressure from other development stakeholders to
back down from the 60 fps standard, even when it's the right choice.
This happens because the costs of 60 fps directly affect many individual
stakeholders in the development process while the benefits are hard to
see. Artists want to push limits of scale, detail, and beauty. Programmers
want more CPU cycles to add more complexity to their AI and physics
systems. Marketers and publishers want the most impressive screenshots
and tech blurbs. These groups are often justifiably dubious about sacrific-
ing two-thirds of their per-frame processing power for a seemingly invis-
ible benefit. So there's a natural political bias against 60 fps. That's why