Processing speed is how quickly you take something in, work out what it is, and act on it. It is one of the most clearly measurable cognitive abilities, one of the first to slow with age, and one of the few that responds visibly to short daily practice. It is also one of the easiest things to oversell, so this covers what the drills actually load and where the evidence stops.
What people call processing speed is really three separate stages, and a slow overall time can come from any of them.
Perceptual speed — how briefly something can appear and still be identified. The Blink Test isolates this: a character flashes for milliseconds and the only question is whether you registered it.
Decision speed — how long it takes to choose among options once you have perceived them. Lightning Choice measures it, and it is where most of the time goes.
Motor speed — how fast you can produce the response. Reaction Time with a single signal is close to a pure motor measure, because there is nothing to decide.
Comparing your simple reaction time against your choice reaction time tells you how much of your speed is decision rather than motor. The gap between them is the interesting number, and it is usually larger than people expect — Hick described the relationship in the early 1950s: response time rises roughly with the logarithm of the number of alternatives.
Speed is not just about being quick. When a step is slow, it occupies attention that would otherwise be available for whatever the step was part of. This is why arithmetic fluency matters: if 7×8 arrives rather than being computed, the attention you would have spent computing it is free for the problem the multiplication belongs to.
It is also why processing speed shows up as a common factor behind so many other measures. A slow perceptual stage degrades performance on everything downstream of it.
The age story, honestly
Processing speed declines gradually and consistently with age across every large study that has looked. That is one of the more robust findings in cognitive ageing, and it is not controversial.
What is contested is whether training it slows the decline. Speed-of-processing training has been studied in older adults with mixed and much-debated results. Some trials report durable effects on the trained measures and on self-reported everyday function; independent replication has been patchy. Treat the strong versions of that claim with caution.
Blink Test — perceptual threshold, shortens as you succeed.
Lightning Choice — choice reaction with a growing number of options.
Number Storm — rapid filtering under high visual load.
Speed Sorting — categorical decisions at a rising rate until you break.
Math Drills — thirty seconds of arithmetic, which is the retrieval-speed version.
Number Grid — a Schulte table, which trains search efficiency and peripheral awareness.
The speed games hub collects them, and the reaction games hub covers the motor end.
Learn the mapping before you start
Any hesitation about which key does what is pure added time on every single trial. This is the single largest avoidable cost in every choice-reaction drill.
Filter on one feature and commit
In high-load drills like Number Storm, pick one distinguishing feature and use it for the whole round. Changing criteria mid-round is the most expensive mistake available.
Do not trade accuracy for speed
Every one of these drills penalises a wrong response by more than the time a correct one would have cost. Going faster than your accuracy allows produces a worse score, not a better one.
Short sessions, same device
Ten minutes is plenty; accuracy degrades quickly after that. And because display and input latency move timed results by tens of milliseconds, compare only runs made on the same hardware — the same caveat that applies to every cognitive benchmark.
How quickly you take something in, work out what it is and act on it. It has three stages — perceptual, decision and motor — and a slow overall time can come from any of them.
You will get faster at the drills, quickly and reliably. Whether that spreads to everyday function is contested: speed-of-processing training has produced mixed results and patchy independent replication. Be cautious of strong claims.
Yes, gradually and consistently, in every large study that has looked. It is one of the more robust findings in cognitive ageing.
Lightning Choice, because the gap between it and a simple reaction test tells you how much of your speed is decision rather than motor.
Yes, all of them, with no account and nothing to install.
Measure the three stages separately, learn the mapping before you start, and never trade accuracy for speed. Then compare against your own history on one device, because everything else is hardware. Start on the [speed games hub](/speed-games).