Review

24 games in a year: Max and Leo's honest retrospective

By Max & LeoJun 25, 2025~1450 words · 6 min
24 games 6 played repeatedly 8 played occasionally 10 almost no plays 25% hit rate, matches our expectation This is indie gaming's true distribution
BverGame's 12-month live data, in raw form

This is the last article in the BverGame series. Leo and I released Flux 2048 a year ago, today released the 24th game Letter Hunt. The set is complete. While memory is fresh, we want to honestly retrospective this year. What worked, what didn't, what the data says.

Background: why 24

The 24 number was decided for romantic reasons: 2 games per month, finish in a year. Leo thought this was an "executable commitment." Initially we considered 50, 12, 6, various numbers — finally 24 was the sweet spot. Enough to feel like a "game library"; not so many that quality drops.

But in practice there were many deviations. Detailed data below.

Real game distribution (by user dwell time)

After AdSense approval, we could see each game page's "average dwell time." Very intuitive:

Tier 1 (avg 8+ minutes):

These 6 account for ~65% of total dwell time. Common trait: all are classic mechanics with strategic depth.

Tier 2 (3-6 min): Neon Snake, Block Drop, Mind Match, Number Stack, Bubble Pop, Pixel Path, Pong Echo, Word Drift. 8 games, ~22% of dwell time.

Tier 3 (< 2 min): Tap Rhythm, Color Merge, Loop Trail, Flip Tiles, Reaction Dot, Bird Bounce, Box Crash, Push Box, Tile Slide, Letter Hunt. 10 games, combined ~13%.

Honestly, the 10 tier-3 games — if it weren't for the "24 commitment" — we could have skipped. But we made them. About 30% of Leo's work hours and energy went into products almost no one plays.

Why this distribution is normal

The 6 : 8 : 10 distribution reminds me of the venture capital "power law." VCs invest in 10 startups — 8 fail, 1 small return, 1 explodes — the one covers the other nine. Indie games are similar.

Specifically, our "top 6" supported the "tail 10." If we'd only made the top 6, we might earn similar money (traffic concentrates there anyway), but site "breadth" would shrink dramatically, and SEO would suffer. So the "opportunity cost" of making tail products is actually SEO value + brand completeness — even with no plays per individual game, existence itself has value.

This insight made us re-examine "indie game success" standards. Not "every product is a hit," but "is the portfolio reasonable?"

What we did wrong

Honest list:

1. The rhythm game Tap Rhythm was completely the wrong direction. Two months in, Tap Rhythm averaged 1.2 min dwell — players play one round and leave. Cause: Web Audio synthesized drums don't sound like real music; players expect "music" enjoyment from rhythm games; we delivered "geometric beats." Cognitive misjudgment. Unfixable; will remember.

2. Color Merge was a "seems-fun" trap. Before launch I played it and thought it was interesting; Leo also said OK. Result: second-to-last dwell time. Cause: looks simple (drag colors to merge), but actual operation is clunky, goal unclear. Lesson: designer's own "this is fun" is the least reliable feedback source. For later games we let 10 outsiders try before launch.

3. Article publishing schedule was irregular. 24 articles — first 3 months we shipped 12, then a 2-month gap, then sprint at the end. Not great for SEO. Google prefers "consistent updates." We should have done 1-2 weekly, not bursts.

4. No "daily" or "weekly" content type. Later I observed high-traffic game sites (e.g., Crazy Games) have daily picks, weekly leaderboards, monthly reports. We had none. If we had, might have built "weekly visit" habits. Note for next project.

5. We applied for AdSense too early. Applied 2 weeks after launch, rejected. Reapplied after 60 content pages, approved. Waiting from the start would have saved rework.

What we did right

Can't only self-flagellate; also note what worked:

1. Bilingual strategy was right. 50% of visits come from non-Chinese users. Chinese-only would have halved traffic. Bilingual lets one piece of content cover the world; marginal cost is negligible (each English version is just ~30 extra minutes of translation).

2. Fully static + no signup was right. No user system, no database, no server, ops cost approaches zero. Even if BverGame earned $0/day, our loss is just domain renewal (~$11/year). This "low-floor failure" model let us work calmly, not warped by ops pressure.

3. Content + games combination was right. A pure game site struggles to build "authority"; Google doesn't rank it high. Our 24 deep articles brought lots of "long-tail search" traffic — queries like "why is 2048 still classic," "is Sokoban NP-hard," etc., land on our articles. Then users see game links in articles and naturally play. The "content → games" funnel effect is visible.

4. Choosing 6 game genres was right. Our 24 games cover Logic / Arcade / Memory / Strategy / Puzzle / Rhythm — 6 categories. Makes "game library" not look like a pile of one style, but a "small but complete" curated set. If we'd done 24 2048 variants, SEO trust would suffer greatly.

5. No IAP, no off-AdSense. Just AdSense banners — no video interstitials, no purchases, no paywalls. Keeps user experience clean; word-of-mouth feedback is real.

Honest numbers

On money, we believe in transparency. Annual totals:

At $1/hour effective wage, far below any full-time job. But this was never a "make money" project. BverGame's real value isn't money:

1. Real portfolio. Leo can show 24 hand-coded games to any employer; Max can show 24 deep articles to any media company. Market value far exceeds $720.

2. Real product experience. We completed 0-to-1 for the first time: planning → development → ops → content → SEO → monetization. Hit pitfalls and solved them at every step. Next product, our efficiency will be 3x.

3. Real relationships. Leo and I worked together a year without a fight, growing more in sync. The rarest asset in any commercial project.

The biggest return on indie products isn't money. It's the existential certainty of "I completed something."

Future plans

BverGame won't stop updates but won't expand. Specifically:

1. 24 games maintained as a static library. Bug fixes or browser compatibility — we'll patch, but no new games actively added.

2. Articles every 2-3 months, on current game industry observations. No more chasing volume.

3. Leo starts next project (still in stealth). Max considering writing a book on "two-person indie products," based on this year's experience.

4. BverGame domain we'll keep renewing, even if traffic drops to zero. It's our "portfolio museum."

For readers who made it here

If you're considering indie games or indie products, our advice:

1. Don't wait for the "perfect idea." Our first Flux 2048 was a clone, not "original genius." But it got us moving. Those waiting for perfect ideas stay waiting 5 years without starting.

2. Find a partner. Solo product work is isolating to the point of breakdown. Two people can encourage each other, reflect together, slap each other awake. Leo pulled me out of "I quit" mode many times.

3. Set executable commitments. 24 games + 24 articles + 1 year. This numeric commitment leaves no room for escape. Whenever we wanted to slack, a glance at the calendar — we'd grit teeth and continue.

4. Process matters more than result, but don't lie to yourself. We didn't earn much, but learned priceless things. At the same time, we don't pretend "$720/year is huge success." Honestly mark costs and returns; only then can you retrospect.

Closing

This is article 24, the last in the BverGame series. If you followed from #1 to here, thanks for the company. If you only read this one, also welcome — take a look at those 24 games whenever.

Wish you do your own thing. A year goes fast.

— Max and Leo, June 2025

Authors Max and Leo are BverGame's operator and engineer. Data is real-as-of-publishing site stats, partly anonymized to protect business details but distribution and ratios are real. This is the finale of BverGame's 24-article series.