We’ve already published the factual recap of Thought Farm’s first Yakkamon dev stream. This is the other half: what we think of it.
Nine decisions, graded. Some of this is critical — that’s the point. This is a team with five years of live-game experience shipping a second title, and they’ve earned a serious read rather than a hype post.
The scorecard
Nine decisions, graded. Eight came out of the stream itself; the ninth predates it but underpins everything else.
| Decision | Grade |
|---|---|
| Specialization through monsters, not items | A |
| New X account, made fast | A |
| Reusing $FLOWER instead of launching a token (predates the stream) | A |
| Referral rule change | A− |
| Chapter-based rollout, PvP after launch | A− |
| Decaying deposit multiplier | B |
| Black-box genetics | B |
| Hiding the leaderboard | B− |
| The doubling size bonus | C |
What they got right
Specialization comes from your monsters
The designer said the primary way you specialize will be the monsters you have — your farming team and your combat team — rather than separate boost items. That’s the best decision of the stream, and it’s the one that makes everything else cohere.
It means the creatures are the game. In a lot of web3 games the creatures are skins and the real power sits in equipment, which quietly makes the collection meaningless. Putting power in the monsters means hunting matters, breeding matters and trading matters — three pillars reinforcing each other instead of competing. Their stated lesson from Sunflower Land was blunt: players only collect things with utility, and nobody wants cosmetic PFPs.
Abandoning the compromised X account
Their official account was compromised and a phishing link went out. They chased malicious domains through the night, then stopped waiting for X support and started a new account.
Correct, and the speed is the impressive part. Every hour that account stayed compromised it was a weapon pointed at their own community, with their verification badge on it.
The recovery move was smart too. Rather than just asking people to follow the new handle, they wired it into the dashboard: reconnect X and previous connectors keep their points and earn more, for 20 in total. That converts an administrative chore into a reward, so the migration actually happens.
No new token
Not from this stream, but underpinning all of it. In a space where a new project almost always means a new token to sell, choosing to reuse $FLOWER is the strongest signal of intent they’ve given.
Shipping narrow
Early access in Q4 ships the fundamentals only: gathering, monsters, hunting and basic combat. PvP and advanced breeding come after launch, in chapter releases.
Right call, and they’ve run this playbook for five years on Sunflower Land. But there’s a marketing tension worth naming: the airdrop is sold on Legendary monsters described as top of the power curve and built for combat. If combat is “basic” at launch, the headline reward is partly a promise about a system that doesn’t exist yet. Not dishonest — but someone chasing a Legendary should know they may hold it for months before the system that showcases it arrives.
Where we’d push back
The doubling size bonus — our only C
The size bonus doubles at every step, up to +80% at 50,000 $FLOWER. Stack that on the opening week’s 3× multiplier and a single 50,000 deposit produces 190,000 points.
For comparison: grind every free action perfectly for two months, with no referrals, and you land near 900.
What we’d do instead: taper the size bonus rather than double it. Reward larger deposits with diminishing returns rather than accelerating ones, so the curve flattens at the top. You keep the incentive to deposit meaningfully while leaving the upper reaches reachable by effort.
To be fair: deposits are withdrawable and spendable in-game, which is genuinely unusual and far better than the lockups this space normally runs. And the team has never claimed the top 100 was reachable free. The early access cutoff is the top 100,000, and that is reachable free.
The wallet-splitting hole
This one came out of a reader conversation after the stream, and it’s the sharpest criticism of the deposit design we’ve seen.
Because the size bonus is rated per deposit, splitting a large sum across several wallets costs surprisingly little — while multiplying your leaderboard slots.
| Same 50,000 $FLOWER, opening week | Points | Reward slots |
|---|---|---|
| One wallet | 190,000 | 1 egg + 1 NFT |
| Ten wallets of 5,000 | 17,000 each — 170,000 total | 10 eggs + 10 NFTs |
Splitting costs 10.5% of your points and turns one reward slot into ten. A 17,000-point account should land comfortably inside the top 5,000, so the trade is: give up one top-100 finish, gain roughly ten top-1,000 finishes, for about a tenth of your points. If the NFTs are tradable, that’s obviously worth doing.
Three fixes, in order of how well they’d work:
- Apply the referral fix to the airdrop. They already require the friend to link a wallet and deposit 100 $FLOWER before a referral counts. Make the same thing a condition of airdrop eligibility — an account qualifies only with a minimum of points earned from non-deposit sources. Ten split wallets would each need a distinct aged Discord, a distinct X account and a week of daily play. Expensive at scale, and it costs honest players nothing.
- Announce wallet-clustering checks in advance. Wallets funded from a common source are trivially detectable on-chain, and referrals are already double-verified before the leaderboard is finalized. Saying so beforehand deters more than catching people later.
- Cap reward slots per verified trainer. The exploit pays because ten Legendary Egg Cs are worth more than one Egg A — and since the team has since confirmed A, B and C are all the same power level, differing only in species and scarcity, steepening the curve isn’t available as a fix without breaking that design. Capping slots per verified identity is.
The blunt fix — making top rewards non-tradable — would end it instantly but breaks the tradability promise. Not worth it.
Hiding the leaderboard
As launch approaches they’ll hide the leaderboard so people can’t game it on the final day. The reasoning is sound — a visible board near a deadline becomes a bidding war.
But the cost lands on the wrong people. A visible leaderboard is how a free player knows whether their effort is working. Take it away and you’re asking people to grind daily toward a threshold they can’t see. Some will simply stop.
Black-box genetics
Asked whether there would be formal design proposals, the designer said he wants to hide base-level systems — especially genetics — so the community back-solves how they work.
We love the instinct. Games that publish every formula get solved in a week and then you’re playing a spreadsheet. Hidden systems create the discovery culture that keeps games alive for years.
The nervousness is about the market. Monsters are tradable on-chain, so when the underlying stats are hidden and the assets are tradable, information becomes an asset too — captured by whoever can datamine hardest or buy enough monsters to infer the distributions.
Our suggestion: publish the ranges, hide the rolls. Tell people a species has an attack range and that training shifts it. Don’t tell them the exact numbers on their monster. You keep the discovery and remove the scraping advantage.
The thing nobody mentioned
This is inference rather than fact, so treat it as such.
By the team’s own description, the Yakkamon core is roughly one developer at ninety percent, one at fifty-fifty, and a twelve-week game design intern working on combat. The rest of the studio is still mainly on Sunflower Land.
That is a small team for a Q4 early access. It isn’t a criticism of competence — this is a studio that has run a live game for five years with over a hundred open source contributors, and they were open about the allocation rather than pretending otherwise.
But it means the Q4 window is tight, and “fundamentals only” starts to look less like philosophy and more like necessity. It also means a bus-factor risk they demonstrated on the stream itself: one developer got ill and the deposit launch moved five days. A minor slip, handled well — and a preview of what a bigger disruption looks like.
Four questions for the next stream
- What’s the rough chapter order after launch? Not dates — just whether PvP is early or late.
- What’s the sink for common monsters, and when does it ship? Contracts were mentioned; details weren’t.
- Will the hidden leaderboard show bands, or nothing at all?
- Will stat ranges be published even if individual rolls aren’t?
Our overall read
The design instincts on this team are genuinely good. Power in the creatures rather than in items. No forging, for tonal reasons. Sped-up clocks so real life isn’t punished. Narrow launch, expand by chapter. Those are decisions made by people who have watched games fail and understand why.
The economy is where we’d push back, and specifically the acceleration of the size bonus — which turns the top of the leaderboard into something you buy rather than something you contest, and opens the wallet-splitting hole above.
And the biggest risk isn’t anything they decided. It’s bandwidth.
They’re being unusually honest about all of it — the bot numbers, the illness, the things they don’t know yet. That doesn’t guarantee anything. But it’s a better starting position than most projects in this space ever manage.