Your Brain Score is your aggregate best performance across all 18 Corflex modes, normalized to a 0–100 scale. It is the average of your personal best on each mode, expressed as a percentage of the theoretical maximum. Get a perfect 50 out of 50 on every mode, and your Brain Score is 100. Average 25 out of 50 across the board, and your Brain Score is 50.
That is the headline number. The interesting story is everything underneath it: which scores count, how recent they have to be, why we split easy and hard into separate scores, and what the metric deliberately ignores. If you are going to use your Brain Score as a benchmark for your own progress (or compare it to other players on the global leaderboard), it helps to understand exactly what it is measuring.
For each of the 18 modes, we look at your highest score ever submitted on a given difficulty. We add those 18 personal bests together. The maximum possible total is 18 modes times 50 points each, which equals 900. Your Brain Score is your total divided by 900, multiplied by 100, rounded to the nearest whole number.
The denominator (900) does not change as you play more games. It is fixed by the structure of the platform. Every mode caps out at the same theoretical maximum, so each mode contributes equally to your Brain Score. Sequence and Reflex, despite being completely different cognitive tasks, count the same toward your aggregate rating. We do not weight one cognitive domain over another, because the whole point of the metric is to capture cross-domain performance.
Your personal best on a mode is just your highest score, period. We do not require recency. A great Sequence run from three months ago counts the same as one from yesterday. This is intentional: the Brain Score is meant to reflect your peak demonstrated capability, not your current form on any given day. If you want a more time-sensitive view of your performance, the daily and weekly leaderboards are better tools for that.
For a long time, Brain Score was a single number. We calculated it by taking your highest score per mode regardless of which difficulty you played on. In practice, this meant easy scores almost always won, because easy mode is naturally tuned to allow more room near the maximum. Hard scores, despite representing meaningfully more skill, almost never affected the metric. Players who grinded hard mode for the challenge saw their effort vanish into a metric dominated by their easy bests.
So we split it. There is now a separate Brain Score for each difficulty, with its own leaderboard and its own unlock threshold. Easy Brain Score requires playing all 18 modes on easy and combines your easy and daily bests (since dailies are always played on easy). Hard Brain Score requires playing all 18 modes on hard and uses only your hard solo bests. The two are computed independently and displayed side by side everywhere they appear: on your profile, on the home page, on the dedicated leaderboard.
The practical consequence is that hard mode is now meaningful for the prestige metric. If you want to compete at the top of the Hard Brain Score leaderboard, you have to actually play hard mode across every category. There is no shortcut through easy. Conversely, casual players who stick to easy and dailies still get a complete and competitive Easy Brain Score that rewards the play patterns they actually engage in.
This is where the most common confusion lives. Three game contexts contribute scores that do not feed into your Brain Score:
The only contexts that count are solo (which uses a fresh randomized seed each time you play) and daily (where everyone plays the same daily seed but each player is limited to one attempt). These are the contexts where score comparisons are statistically meaningful, so these are the ones we use.
Brain Score is most useful as a breadth measure. A high Brain Score means you have demonstrated competence across a wide range of cognitive domains: working memory, spatial reasoning, pattern recognition, reaction time, color and pitch perception, symbolic learning, and more. You cannot game it by being elite in one area. To score 90 on Brain Score, you need to be solidly above-average in essentially everything we test.
It also captures consistency over time in a soft way. Because we use your personal best on each mode, a single great run on each will eventually unlock a high score, but reaching the absolute top of the leaderboard typically requires repeatedly improving your best on multiple modes. Players who push the upper bounds tend to be ones who keep coming back and chipping away at their weakest mode scores.
And finally it captures specialization tradeoffs. If you compare your scores mode-by-mode against the leaderboard, you can see exactly where your cognitive strengths and weaknesses are. Players often discover surprising patterns: someone with great working memory might score poorly on color perception, or a fast reactor might struggle with spatial tasks. The aggregated number tells you your overall standing, but the per-mode breakdown is where the actual self-knowledge lives.
Brain Score is not a measure of intelligence in any rigorous sense. The modes test specific cognitive skills under specific conditions. They do not measure crystallized knowledge, verbal reasoning beyond simple word patterns, complex problem-solving, creativity, social cognition, or any of the other capabilities that real intelligence research considers important. A high Brain Score correlates with quick processing of perceptual and pattern-based information. It does not say anything reliable about your IQ, your professional ability, or your decision-making in real-world contexts.
It is also not a clinical measure. The modes are designed for entertainment and friendly competition, not diagnostic assessment. If you have concerns about cognitive function, online brain games are not the right tool. They lack the controlled administration, normative data, and validation that clinical assessments require.
And it does not adjust for device or context. Playing reaction-time and visual search games on a phone with high-latency input will produce different scores than playing on a desktop with a wired mouse. We do not normalize for this. Your Brain Score reflects your performance under whatever conditions you happened to play in, which is the honest answer but means cross-player comparisons should account for hardware differences when they get specific.
The fastest path to a higher Brain Score is to find your worst mode and focus there. Because the metric is an average, lifting your weakest scores has more leverage than improving your already-strong ones. Going from a 30 to a 40 on a mode you are bad at adds 10 to your sum (a bigger gain than going from 45 to 48 on a mode you are great at).
The slower but more rewarding path is to play across every mode regularly. Brain training research is consistent on one point: practice generally improves performance on the practiced task, but the improvements only generalize when the task itself is varied. By rotating through 18 different cognitive domains, you give yourself a much better shot at the kind of broad cognitive engagement that the research suggests is more meaningful than narrow grinding.
And do not skip the dailies. They are designed to give every player one fair attempt at the same puzzle each day, which makes them an unusually clean way to track your true peak performance over time. If your Brain Score is plateauing on solo, your dailies are often where breakthrough scores happen.