Played memory match for a week. Did my working memory actually improve?

I ran this experiment because I've been skeptical about "brain training" apps since Lumosity's 2014 FTC settlement. Whether cognitive training actually transfers to daily life is a long-running argument in psychology. So this time I put myself in. One week of Mind Match, then look at what actually changed.

The setup

I should be upfront: this is not a scientific experiment. Sample size of one, no control, no blinding. I just wanted numbers beyond "I feel sharper."

What I did:

• 8 AM every day after waking, 30 minutes of Mind Match Hard mode (24 cards)
• 10 PM every day, a digit-span test (read a random number sequence aloud, then reverse-recite, see how many digits I could hold)
• Logged for Mind Match: average moves, average time, first-pair-success rate
• Logged for digit span: 5 trials each night, averaged

Baseline was the prior week, no training, same nightly digit-span test.

In-game data

The improvement here was unmistakable:

• Day 1: 42 moves average, 92s average
• Day 3: 35 moves, 78s
• Day 5: 30 moves, 68s
• Day 7: 28 moves, 65s

Moves down 33%, time down 29%. Looks great. I was ready to write the clickbait headline: "30 minutes of memory match per day, 30% memory improvement."

Then —

The digit-span data

Baseline: reverse recall, 6.2 digits average.

Post-experiment:

• Day 1: 6.4
• Day 3: 6.6
• Day 5: 6.4
• Day 7: 6.6

Variation within ±0.4 of baseline. Statistically indistinguishable from "no change."

Memory match made me 30% better at memory match. My ability to remember digits didn't move.

This is what cognitive psychologists have called the "near-transfer problem" for decades. Task training makes you better at the trained task. Transferring to a structurally different task is much harder.

So what did I actually practice?

I analyzed my Mind Match strategy. Day 1 I flipped cards essentially at random — see a symbol, try to remember it. Day 7 I was using a "chunked scan": flip the top-left 2x3 region first, roughly memorize those 6 positions, then start matching systematically, then scan the top-right.

See it? What I learned wasn't "bigger working memory." It was a strategy specific to memory match. Method improvement, not cognitive improvement.

Useful but easily mythologized distinction. Every "brain training" app sells the second thing (cognitive improvement). What they actually deliver is the first thing (method improvement).

Does that mean memory match is useless?

Of course not. It gave me three real benefits — none of which have anything to do with "training your brain":

1. A low-stakes morning focus warm-up. I used to start the day by checking email — quickly hijacked by whatever's unread. Memory match for 30 minutes gave me a calmer morning rhythm. That's a ritual effect, unrelated to memory.

2. A concrete sense of progress. Most things at work aren't measurable. Move count and seconds are. Seeing daily improvement gave me small motivation. That's self-efficacy, unrelated to memory.

3. It's genuinely relaxing. Failure cost is low (just replay). Successful pair feedback is immediate and clear. For anxious-type personalities (like me), this "low-risk high-feedback" loop is a real calming agent.

So: is daily memory matching worth 30 minutes? Yes. But don't play it because "I'm training my brain." Play it because "I enjoy it." The two reasons have very different staying power.

About commercial brain-training apps

Now the potentially uncomfortable part.

Over the past decade, Lumosity, Elevate, Peak, CogniFit and others have been examined by independent researchers. The 2014 joint statement from Stanford and the Max Planck Institute provides a useful baseline; the 2016 FTC ruling on Lumosity is another. The summary: training the specific task works, but evidence for transfer to general cognitive improvement is limited (research keeps evolving; readers should check current literature).

This isn't to call these apps scams. It's to say their core marketing pitch — "smarter," "anti-aging brain" — is not currently supported by high-quality experimental evidence.

Closing

I played memory match for a week. I became a better memory match player. I did not become "someone with better working memory."

That finding sharpened BverGame's own positioning. We don't sell "brain training." We sell "two subway stops, no regrets." That's the honest pitch.

If you ever hear any casual game (ours included) claiming it'll boost your cognition, please ask: is there transfer data?

Max is BverGame's co-operator. This is personal experimentation, not medical or psychological advice. For cognitive-training research, see review papers on "cognitive training transfer" or coverage from independent science outlets.