If you’ve ever completed a Sustained Attention (SART) session and noticed that your Minute-on-Minute (MoM) accuracy steadily improves — while your Grand Mean (GM) accuracy stays unusually low — you’re not alone. It’s one of the most common questions we get from users, athletes, and coaches.

And it makes sense. When your performance clearly improves during the session, a low GM score can feel like a bug or miscalculation.

The truth? It’s working exactly as intended.

The reason comes down to how this task is designed — to isolate and test two different brain systems:

This article explains how the scoring logic works — and why it’s critical that these two metrics are tracked separately.

The Sustained Attention task (SART) presents a rapid-fire stream of single-digit numbers, shown every 750–1050 milliseconds.

Your instructions:

This creates two competing demands:

That’s what makes this task neurologically powerful. You build up a tapping rhythm — then have to break it suddenly. To measure these two systems properly, we track them separately: with MoM and GM.

Let’s break the session into two behavior streams:

This tracks how well you tap when you’re supposed to — i.e., for non-3s. Since non-3s appear most of the time, MoM provides a clean minute-by-minute view of your focus, rhythm, and consistency.

It answers: “Are you staying focused and tapping correctly under pressure?”

GM tracks your ability to stop tapping when you see a 3. Because 3s are rare and appear randomly, GM is measured across the whole task.

It answers: “Did you successfully inhibit your response when it mattered?”

Combining all responses into one score would be misleading. Here's why:

By splitting them:

This gives you much more meaningful insight — and ensures high scores are earned through real performance, not inactivity.

This athlete tapped consistently on non-3s and avoided tapping on 3s. Both scores were high.

Result: MoM and GM aligned — elite-level execution and restraint.

Mean Data

MoM Data

This athlete struggled early, missing taps and failing to inhibit on 3s. MoM improved later, but GM remained low due to early inhibition errors.

Result: MoM shows recovery, but GM reflects full-session inhibition failures.

Mean Data

MoM Data

The reason GM and MoM can diverge — and must be tracked separately — is because this task challenges two very different parts of your brain: focused attention, and impulse inhibition. A single accuracy score wouldn’t be able to capture both systems reliably.

MoM tells you how well you’re managing attention under pressure, minute by minute. GM tells you whether your brain is overriding automatic impulses when it needs to. This is especially important in environments like high-performance sport, combat, or aviation, where a single mistimed action (or failure to act) can be the difference between success and failure.

By separating these scores, we ensure the task gives honest, reliable, and specific feedback — even when the answers aren’t pretty.

GM stands for Grand Mean — and it’s the overall average of your performance across every trial in a session.

Definition:

The Grand Mean is the total average of all correct responses across the full task, calculated from individual trials (not grouped by time).

What it tells you:

MoM stands for Minute on Minute — it’s the average of each minute’s performance, treated equally regardless of how many events occurred per minute.

Definition:

The MoM is the unweighted average of the accuracy values calculated for each minute block during the task.

What it tells you:

Because “3”s appear randomly and infrequently, and each one is critical.
Missing even a few has a big impact. This is by design - GM reflects your ability to override automatic tapping and shows how well your brain handles high-stakes inhibition moments.

GM is a cumulative score over the entire session.

Even if your MoM accuracy rises (indicating improvement), early failed inhibitions on “3”s will still count against your GM. It's a realistic picture of your inhibition control over time, not just at the end.

SART is designed to simulate real-world challenges where:

By splitting tapping (MoM) and inhibition (GM), we get precise data on two critical mental systems — which is why this approach is the most accurate and informative way to score this task.

Minute-on-Minute (MoM) accuracy tracks how accurately you’re tapping on non-3 numbers during each minute of the task.

Because most of the numbers you see are not 3s, the majority of your taps are on valid targets. If you're tapping correctly on most non-3s, your MoM accuracy will stay high — even if you occasionally tap on a 3 by mistake.

That said, tapping on a 3 is still considered incorrect and will reduce your MoM accuracy. But since there are fewer 3s than non-3s in each minute, a few errors won’t drop the percentage significantly unless they happen frequently.

This is completely normal — and it's one of the most important distinctions in how the Sustained Attention (SART) task is scored.

Your MoM accuracy reflects how well you're tapping on non-3s — the majority of stimuli. It shows your ability to stay focused, maintain rhythm, and respond quickly over time. That’s why MoM can trend upward minute by minute, especially if you’re finding your flow mid-task.

Your GM accuracy, on the other hand, is much more strict. It tracks how well you’re inhibiting your response when a “3” appears. Because 3s appear rarely and randomly, every incorrect tap on a 3 carries a heavy weight. Just a few failed inhibitions can pull your GM score down dramatically — even if your tapping performance is otherwise flawless.

So while MoM rewards sustained attention and tapping accuracy, GM exposes lapses in impulse control — and it doesn’t forget early mistakes.

Think of it like this:

Together, they show you how your brain is balancing execution speed with restraint, which is crucial in environments like sport, medicine, aviation, and military decision-making.