Boinker, Rolls, Notcoin: clicker mechanics evolution on TON

Clicker mechanics went through a characteristic three-stage evolution in the TON ecosystem over two and a half years: from the pure tap of Notcoin (January 2024), through Hamster Kombat’s combo cards (mid-2024), to meta-games like Boinker and social formats like Rolls (2025-2026). Each step tried to solve the problems of the previous generation. This article is a design analysis of what worked, what burned out, and what survived as a useful technical pattern.

The tone is critical. Most tap-to-earn projects have no positive LTV for the median player. Precisely for that reason it is interesting to look at the design decisions that move the industry forward, even when the commercial outcome of each individual project is debatable.

Stage 1 (January 2024 — May 2024): Notcoin and pure-tap

Notcoin launched in January 2024 as a minimal game: one screen, a tap button, a coin counter, referral links. Game design was almost absent — it was a social experiment about whether the simplest mechanic could go viral in Telegram.

The experiment worked: by May 2024 Notcoin had 35M+ active users and ran a successful airdrop of the NOT token. This created the template that later games followed.

What worked in Notcoin:

1. Low entry barrier. One tap, no tutorial, no balance to top up.
2. Virality via referrals. Each invitee gave a permanent passive boost — a strong social incentive to share the link.
3. Clear promise. “Tap now — get a token later.” A simple message, understandable to anyone.
4. Telegram-native UX. No download, opened from a chat link — playing immediately.

What did not work:

1. No long-term loop. After day one the mechanic became monotonous. Players left tabs open with auto-tap scripts.
2. No meta-goal. Why more coins? Only for the airdrop. After the token drop the reason to play vanished.
3. Sybil vulnerability. One person registered dozens of accounts. Post-airdrop filtering cut a significant share of farm networks.

Stage 1 lesson: pure-tap works as a marketing funnel for airdrop but not as retentive gameplay.

Stage 2 (mid-2024): Hamster Kombat and the combo-card era

Hamster Kombat shipped in summer 2024 with the ambition of fixing two main Notcoin problems: give a meta-goal and hold the player longer. The solution was a combo-card system.

Hamster Kombat design:

• The player taps and earns base coins.
• Spends coins on “combo cards” representing virtual “crypto exchange departments” (legal, devs, marketing).
• Each card generates passive coin income that continues offline.
• Card levels go up, exponentially increasing cost and yield.
• Daily ciphers add a return-rhythm.

What worked:

1. Idle mechanic. The player no longer had to tap eight hours — after card levelling, passive income piles up offline. Opening Telegram becomes “collect what accumulated.”
2. Decision-making. Which cards to level up in what order is a real choice with math. That holds attention longer than pure-tap.
3. Daily cipher. The player opens the app each day to avoid missing the daily code — a proven casual mechanic.

What broke:

1. Airdrop disappointment. With 300M+ registered there was nothing to divide: the computed share per user turned out tiny. The sybil filter removed even more people. Active channel comments: “played 4 months — got $5.”
2. No real product after airdrop. The “crypto exchange” remained purely notional — there is no real exchange inside Hamster Kombat, it was a visual metaphor.
3. Power creep. Cards grew so expensive that only active daily-combo grinders could maintain competitive income. The casual audience left.

Stage 2 lesson: idle mechanics add days of retention but do not solve the fundamental problem — there is no reason to stay after the token. And when the token undershoots expectations, trust in the category drops.

Stage 3 (2025-2026): second generation, Boinker and meta-games

By early 2025 it became clear: tap-to-earn in pure form was exhausted as a format. Teams continuing in TON gaming started adding deeper meta-games aimed at retention without promising financial outcome.

Boinker as case study:

Boinker is one of the most discussed projects of 2025-2026. A team with Notcoin-era experience openly built a game with a retention-first approach. Structure:

• Base layer: tap and base currency (like Notcoin).
• Episodes: game content split into episodes (like a season pass) with different mechanics each time. One episode is cooperative resource gathering, another PvP, a third narrative with choices.
• In-game resources. Multiple currency types with distinct functions — this creates a mini-economy inside the game.
• NFT achievements. Completing episodes gives NFT items with value outside the game (also on-chain assets).
• Social layer. Clans, chats, leaderboards.

What works in Boinker:

1. Continuous content updates. A new episode every few weeks prevents the player from burning out on one mechanic.
2. Decoupling from airdrop. The game is not framed only as “farm the token.” If a token comes — bonus. If not — the player still got gameplay.
3. Experienced team. Unlike the anonymous teams of 2024, Boinker is built by people with public reputation.

What is not yet proven:

1. Long-term retention. Nobody has played Boinker for a year — too early.
2. Token model. If/when an airdrop comes, will it avoid the Hamster Kombat sybil problem — open question.
3. Content load. Maintaining new episodes every few weeks is a heavy team workload. Episode fatigue is real.

Stage 3b: social clicker, Rolls and PvP

In parallel with meta-games, social formats emerged. Rolls is a dice game between users. This is no longer a clicker in the strict sense, but genetically related: the same path of adding retention on top of a simple mechanic.

Rolls design:

• Base mechanic — dice roll on a stake.
• Matches against other players (PvP), not against the house.
• Telegram chats serve as match rooms.
• Social layer — leaderboards, clans.
• Transparent RNG — on-chain commitment-reveal scheme, both sides can verify fairness.

Why it worked:

1. Social engagement. A game with a live human holds attention more strongly than playing against the system.
2. Transparent RNG. The main problem of traditional online casinos is distrust of RNG. On-chain commitment-reveal solves it mathematically.
3. Casino mechanic is time-tested. Variable reward is the strongest engagement pattern known in psychology. It works on any audience.

What is bad:

1. House edge re-framed as PvP rake. The platform takes a fee on every match. For most players the sum of stakes loses to rake over the long run.
2. Regulators. PvP dice with stakes is a game of chance in most jurisdictions. That does not change because the transactions are on-chain.
3. Addiction and health. Casino mechanics are dangerous to vulnerable players. In traditional jurisdictions this is balanced by regulation and self-exclusion tools; in TWA — barely at all.

Comparison of the three generations

Retention figures are approximate estimates from public metrics. Teams rarely disclose exact data.

What stuck from clicker evolution

A few technical and design patterns that moved from stage to stage and became TON gaming standards:

1. Telegram-native UX. Launch from bot, no login page.
2. Referral trees. Standard distribution tool — every second- and third-generation game keeps the referral mechanism.
3. Daily activities. Ciphers, missions, login bonus — psychological anchor for return.
4. TON Connect for payments. Standardised canon since 2024, no one builds their own wallet integration.
5. Stars + TON dual payment. Micropayments in Stars, large ones in TON.
6. NFT achievements. A way to give the player a sense of “accumulation” without promising a token.
7. Sybil-aware design. Whoever does not account for sybil filtering in token distribution after Hamster Kombat loses trust in round one.

What burned out and does not return

Patterns that were tried, failed, and are no longer used by serious teams:

1. Pure-tap without meta. Notcoin was first, but the formula no longer goes viral — users know it wastes time.
2. Promise of guaranteed airdrop. After Hamster Kombat such promises are met with suspicion.
3. Anonymous teams. Trust fell sharply; new projects with a public team get an advantage.
4. Monotone combo cards. That mechanic is monotonous; episode-based content replaces it.

Lesson for the TON game developer in 2026

If you plan to build a new game:

1. Do not repeat Notcoin/Hamster. That happened, the audience knows the end is disappointment.
2. Think retention from day zero. Not “how to attract a million” but “how to hold those who came.”
3. Decoupling from airdrop. The game must be playable without the promise of a token.
4. Content plan. Episode-based requires a serious content pipeline. That is investment, not one-shot.
5. Transparency. Team is not anonymous, tokenomics published before launch, RNG verifiable.
6. Regulatory awareness. If the mechanic resembles gambling — evaluate local risk in advance.

Conclusion

The evolution of clicker mechanics on TON is a miniature history of all Web3 gaming. First a simple, viral, but non-retentive formula. Then complication trying to add retention through idle. Then acceptance that idle alone is not enough, and a shift to real gameplay content with meta loops and sociality.

Each stage was useful — even the burned Hamster Kombat gave the industry an understanding of sybil. The next fourth generation of games likely arrives via AI-generated content (dynamic episodes built on LLM) or integration with real content (sports betting, prediction markets with regulatory licences).

Tap-to-earn as a standalone genre died. But the techniques born in it (Telegram-native UX, TON Connect, referral trees) live on and form the foundation of TON gaming in 2026.