1
3 Comments

I built an RPG character generator for Ethereum wallets in 7 days — here's what happened when I launched

I had a simple idea: what if your Ethereum wallet was an RPG character?

Paste any ETH address, and in 10 seconds you get a full character card — class, stats, AI-generated lore, and a Combat Power score. You can even battle other wallets in turn-based PvP and compete on a global leaderboard. No wallet connect, no signing, completely read-only.

Try it: https://ethrpg.app

What it does

The app reads your on-chain transaction history and maps it to RPG mechanics:

  • 8 classes based on behavior patterns (NFT collectors become Hunters, DEX traders become Rogues, diamond hands become Guardians, etc.)
  • 7 stats calculated from real data using log-scale normalization (so both 10-tx wallets and 50,000-tx whales get meaningful scores)
  • PvP battles — pit any two wallets against each other in deterministic turn-based combat
  • Global ranking — leaderboards by Combat Power, battle record, and explorer score
  • AI lore that references actual crypto events (The Merge becomes "The Great Convergence," DeFi Summer becomes "The Alchemy Renaissance")
  • Shareable card image rendered server-side — optimized for social feeds

The 7-day build

  • Day 1-2: Blockchain data pipeline + transaction classifier + stat formulas
  • Day 3: AI lore generation (with 4-step fallback) + card image renderer
  • Day 4: Frontend (landing, result page, share flow)
  • Day 5: PvP battle system + global ranking
  • Day 6: Polish, error handling, deploy to Vercel
  • Day 7: Launch day

Solo dev. Tech stack: Next.js, TypeScript, Alchemy SDK (on-chain data), Claude API (lore), @vercel/og (image rendering), Vercel hosting.

48-hour launch results (honest numbers)

After 2 days:

  • 30-34 external visitors (not counting my own testing)
  • 2 card generations from real users (both on Day 1 — Day 2 had zero new ones)
  • 2 users completed the full funnel (card -> battle -> ranking)
  • 2 return visitors came back 24 hours later
  • 1 organic Twitter share by a user (first viral loop trigger)
  • $0 revenue (it's free)

Not exactly a rocketship. But the details tell a more nuanced story.

What I learned about launching with zero audience

I prepared a "multi-channel simultaneous launch" across 6 platforms. Here's what actually happened:

  • Product Hunt — 6 upvotes, traffic halved by Day 2. New maker account = low algorithmic visibility.
  • Hacker News — Shadow banned. New account + low karma = auto-filtered.
  • Farcaster — 0 engagement. No Power Badge = can't post in channels.
  • Reddit — Couldn't post. Subreddit karma requirement not met.
  • Twitter — 3 clicks total, but 1 organic user share. 0 followers = shouting into void.
  • Dev.to — 14 views. Only channel with zero barrier.
  • Indie Hackers — 4 visitors, 0 conversions. Product page live, community post up.

4 out of 7 channels were effectively blocked. Every major platform has invisible gatekeeping for new accounts. I didn't know this going in.

The good news buried in the data

The product works. Distribution doesn't.

Here's what the small sample actually showed:

  1. Users who find it go deep. The one HN user who saw my post before it got killed? They hit an error (empty wallet), tried a different address, generated a card, fought a battle, and checked the ranking. Full funnel in 2 minutes.

  2. Retention is real. Two Day 1 users came back 24 hours later — same time of day, same behavior. The ranking system creates a reason to return.

  3. First organic share on Day 2. An IH visitor ran 3 battles in a row, then shared the result on Twitter. The viral loop triggered for the first time — even if it didn't cascade yet.

  4. Card-to-battle conversion is high (~60%). Once someone has a card, they want to fight. The RPG hook works.

The bottleneck is the top of funnel. 30+ people visited the landing page. Only 2-3 actually typed in a wallet address. That's a ~15% conversion rate on the first step. The rest of the funnel is fine — the problem is getting people to try it.

What I changed after Day 1

Based on the data, I redesigned the landing page:

  • Added clickable wallet chips (vitalik.eth, stani.eth, etc.) so visitors can try instantly without typing
  • Improved the empty wallet error to suggest famous addresses instead of a dead end
  • Stripped redundant CTAs — one clear action, not three

The original landing page asked users to paste an Ethereum address. That's a huge ask for a cold visitor. The chips reduced friction to a single click.

If I did it again

  1. Build platform presence 4+ weeks before launch — karma, followers, badges. Every major platform gates new accounts.
  2. Personal network first, platforms second — DMs convert 10x better than cold posts. I should have started here.
  3. Zero-barrier channels only on Day 1Dev.to and Indie Hackers are the only places where a new account can post immediately.
  4. Ship the "try it instantly" UX before launch — asking cold visitors to paste a wallet address is too much friction. Should have had the clickable demo wallets from Day 0.

The fun part

I tested ~60 known wallets. Some results:

  • vitalik.eth — Hunter, 62,065 Power. Ethereum founder — hunts relics across chains.
  • stani.eth — Priest, 62,005 Power. Aave founder — channels sacred protocol mana.
  • pranksy.eth — Hunter, 70,140 Power. NFT legend — highest power score yet.
  • sassal.eth — Summoner, 57,995 Power. ETH educator — bridges between realms.

All 8 classes appeared in the wild. The rarest? Summoner — heavy bridge users are a tiny minority.

What's next

  • Building karma on HN and Reddit for a relaunch attempt
  • Daily Twitter posts tagging crypto figures with their character cards
  • Direct outreach to crypto/dev friends (should have started here on Day 0)
  • Considering open-sourcing the classification engine as a standalone library
  • Watching if the landing page redesign moves the 15% conversion needle

Would love to hear:

  • Has anyone else launched to crickets and turned it around? What was the inflection point?
  • What's your experience with platform gatekeeping as a new account?

https://ethrpg.app — paste any ETH address (or click a famous wallet), see what hero you are.

posted to Icon for group Artificial Intelligence
Artificial Intelligence
on February 27, 2026
  1. 2

    This is one of the most honest launch post-mortems I've read. The platform gatekeeping section hit hard. I launched on HN today for the first time and hit the exact same invisible wall.
    Your observation that '4 out of 7 channels were effectively blocked' should be required reading for anyone launching a new product. The advice to build platform presence 4 weeks before launch is something I wish I'd known earlier.
    The data point that actually stood out to me: users who found the product went deep immediately . That's a strong signal the product works.
    How are you thinking about the relaunch timing? Are you targeting a specific karma threshold on HN before you try again?

    1. 1

      Thanks! The HN invisible wall is brutal — especially when you spend hours crafting a Show HN post and it just... disappears.

      For the relaunch, I'm not targeting a specific karma number but rather watching for two signals:

      1. Account age > 30 days — from what I've read, this matters more than raw karma for Show HN visibility
      2. Consistent comment history — I'm doing 2-3 genuine technical comments per day. Currently at ~15 karma after a few days. Aiming for 30+ before I retry.

      The tricky part is that HN's anti-spam heuristics are completely opaque. There's no documentation, no error message, no "your post was flagged." You just post and it vanishes from /new. I only figured out mine was shadow-killed by checking /newest in an incognito window.

      One thing I'd do differently: don't launch on HN as your first-ever post. Build a comment history first so the system recognizes you as a real person, not a drive-by spammer.

      What did you launch? Did your post show up on /new at all, or did it get the same silent treatment?

  2. 1

    Happy to answer any technical questions! The stat formulas and battle system are probably the most interesting parts from an engineering perspective.

Trending on Indie Hackers
Why Indie Founders Fail: The Uncomfortable Truths Beyond "Build in Public" User Avatar 136 comments Your AI Product Is Not A Real Business User Avatar 85 comments I got tired of "opaque" flight pricing →built anonymous group demand →1,000+ users User Avatar 48 comments The Clarity Trap: Why “Pretty” Pages Kill Profits (And What To Do Instead) User Avatar 32 comments I built an enterprise AI chatbot platform solo — 6 microservices, 7 channels, and Claude Code as my co-developer User Avatar 28 comments I went from 40 support tickets/month to 8 — by stopping the question before it was asked User Avatar 16 comments