Farcaster

walked & thought (i’m not recording myself…) but here’s my head on killing the masters, and driving sustainable retention & “interesting content” to the platform by fixing the leaderboard 

reality is that nuking the OGs is not a binary toggle—it’s a systems-design question with second- and third-order effects; here’s how i see the moving parts:

1/ status & incentives

early adopters earned outsized reach by just being here first. that’s fine until the network’s marginal user can no longer break through the legacy moat. if a day-one video post autowins 100 likes while an actually useful deep dive & product update gets 20, you’re watching path-dependence strangle discovery

in practice, it means:

- new builders ship less because nobody sees them
- ct lurkers visit, scroll, decide “same clique, pass,” and bounce
- rewards funnels up to incumbents who push out tips, compounding the gap

→ the core unease isn’t envy; it’s a mispriced signal. the algo keeps saying “this is the good stuff,” while the builders & natives screams otherwise

2/ the three cohorts

- tribe 1 ~ signal ogs: ex/builders, polymaths, tech weirdos. crypto-adjacent but maybe not crypto-maxi. they add context, nuance, real reach outside the bubble. lose them & you forfeit narrative diversity

- tribe 2 ~ squatter ogs: first-mover clout, little net-new value. their likes are reflexive, their posts rarely travel. they’re not malicious, just noise-amplifiers

- tribe 3 ~ crypto devs / ct raiders: care about composability, capital, memetics. they build tooling, pump charts & bring liquidity—but also volatility

→ healthy farcaster = tribe 1 anchoring discourse & tribe 3 stoking velocity

→ tribe 2 is ballast; useful only until it turns to drag (today)… time to kill 

3/ re-weighting without collapse

full de-weighting of all OGs flips one monoculture for another… better:

- time-decay any early-fid bonus: followers earned 18 months ago lose some power each quarter. still rewards history, but the ladder pulls up slowly (probably inline)

→ better: where do the net-new users on the platform go? $tipn/ $noice? video content? micro-blogging? the data tells the story 

- multi-metric karma: marry onchain contribution (deploy a mini-app, tip from your own wallet, ship a hook) with social metrics (unique commenters, repost ratio, outbound clicks)

→ forces tribe 3 to do things, rewards tribe 1 for context, and leaves tribe 2 exposed

- topic entropy boost: posts that pull fresh tags or cross-graph links get a temporary surfacing bump. incentives exploration over echo (probably inline)

- weight across the week: monday post at 11am is not weighted the same as thursday post at 7pm or saturday post at 2pm (i’ve posted about this before). this is a countermeasure to protect the aggregate signal of group 1 against groups 2 & 3 (farm the leaderboard with food or thirst posts in last six hours, get nuked into oblivion; seriously, rate limit this garbage & mark em spam, it’s zero net-new value to the network)

→ none of that kills masters overnight; it just makes the throne contestable again, or at least directional pushes us that way 

4/ why ct still matters

every network graduates from intimacy to economy. we’re at the hinge: either become a crypto super-app (wallet layer, mini-apps, liquidity) or remain a nice niche forum. ct is the only funnel with enough bodies & capital to fund that leap. the irony: ct won’t stay unless discovery feels meritocratic. fixing the leaderboard is table stakes for courting them

→ killing tribe 2 & losing 20% of tribe 1 might be the only way; either way, pivoting ahead of a bull market when builders here are finally drawing ct attention is BAD for builders & their projects

we integrated crypto but never tried to make it work, now we’re pivoting. if crypto doesn’t work here, nothing likely does. it’s inherent & fundamental. i don’t make the rules. no one is coming here for science or micro blogging (they will churn when the tribe 2 & tribe 1 leaderboard fabrication ceases & their engagement & rewards go to zero as we have seen before)

- tam + wallets on tap: fc’s core audience is maybe 2k daily users; ct has 100× that. you need their wallets to even sniff meaningful scale—tips, mints, gas, all flow from that crowd

- money memes = growth loops: every jump in fc activity—$degen, cashtags & charts, $noice, $bnkr—was bolstered by ct migrations. the price or airdrop lit the beacon, the beacon pulled new users, new users → more builders → more apps. rinse, repeat. no ct, no meme ignition

- builder economics: mini-apps survive on throughput. if $10-50k/day of tips moves consistently, someone will ship the next $degen. if it’s $500/day, they ship nothing & leave. ct’s speculation engine bankrolls the option set

- escape-velocity & critical mass: wallet-native super-app = thousands of experiments, most failures. ct supplies the risk capital & dopamine to fund those failures. without it, fc tops out as a pleasant but tiny forum

5/ the video pivot™ 

→ mostly orthogonal, and sometimes harmful, imo

shorts diversify content but solve none of the above. they actually magnify clout concentration because video creation has higher friction; incumbents with built-in audiences capture outsized algorithmic placement. good for dopamine, bad for surface-area innovation. pursue it, sure—but only after the incentive layer is patched. how does this play here? 

→ 100% of tribe 2 & a good chunk of tribe 1 are the king makers here. they drive discovery for new incumbents with less established existing audience — this, at the expense of tribe 3 & a portion of crypto-focused tribe 1 — meaning tribe 2 is largely shit in & shit out & should NOT have anywhere near this amount of power in terms of driving leaderboard winners & content highlights (if tribe 2 drives the trending page, we are cooked, but why?) →

- intent mismatch: people historically open fc to talk markets, tokens, charts, governance, scaling, direction, building—words + wallet. 60-sec walk-and-talk vlogs & grwm shorts scratch a totally different itch. when they dominate the leaderboard, crypto folks bounce immediately: “wrong feed”

- winner-take-all dynamics: video reach scales with audience you already have. the early squatter ogs keep winning; new builders stay invisible. we try to drain the moat, and short creators backed by tribe 2 fill it right back up

- product focus tax: hosting, transcoding, moderation—video soaks up engineering cycles that could ship hooks, sdk docs, or wallet ux. opportunity cost is real

→ they can coexist, but siloed: if fc has to test video, give it a “clips” tab & its own karma bucket. keep the main feed merit-based onchain contribution + discourse. otherwise the brand message—“crypto builder hub”—gets fuzzy

6/ acceptable casualties imo

- we probably lose 10-20% of tribe 1 who equate any deeper crypto move with “wasteland”

- we probably see 60-70% of tribe 2 churn out once their engagement & rewards fall off, or more likely, they’re forced into creating more “interesting content”

7/ endgame

→ if new builders can’t acquire 500 organic followers in 90 days, the algo is clogged

simple. 

8/ bottom line

we don’t need to torch every OG; we need to make status renewable. decaying privilege, rewarding provable contribution, and diversifying surface area gets us there… 

peace follows not from silencing masters, but from making mastery a moving target 

9/ tldr 

- you need ct because it’s the only funnel with both bodies & capital

- video can coexist, but only siloed; in the main feed it reinforces squatter moats & blurs the crypto thesis

- double down on low-friction onchain toys, leaderboard re-weighting, and reward-back loops to maximize the odds of another $degen-style shockwave

- protect tribe 1, trim tribe 2, and give tribe 3 room to build—the network compounds when status is earned, not timestamped

in short:

→ “interesting content” is contextual: a chart breakdown of liquidity funnels is interesting to ct; a day-in-the-life vlog rarely is. flood the top slots with the wrong genre & you repel the very cohort that can bankroll growth 

→ allow tribe 2 to remain in power & the reward loops degrades over time, ending with content that’s not interesting to any single niche. tribe 2 produces uninteresting content & then pushes more of it to the top through first-movers. this alienates builders at its most fundamental level. moreover, all of these new kings will churn as novelty fades & rewards collapse, leaving us stuck with a pot of slop that builders don’t care about. so builder churn. rewards dry up. new creators churn. then its gg. 

→ interesting shorts on any niche can be found on almost any social media platform today. interesting crypto content can be found on maybe 2 platforms in total & the first is dying. now is our time to capitalize, not alienate.

Composable feeds.
Allow me to create my own feed and set it as the default

Allow mini apps to install feeds

Would love to see neynar, quotient,.. feeds I can select from

$hamster /hamster 
martinvr.eth 
deploy.degen.eth

Building /recaster, a third-party Farcaster client

Yeah, Recaster app has this feature, can import starter pack as a feed, but not in miniapp.

@haole I'm probably not original and you are already doing this?

❤️ @lovejoy
🪿 /honk 
📸 /365photo 
🌌 /fargallery 
📼 /tape

@franmarengo3 this could be a good project for you, scraping all of the starter packs

https://docs.farcaster.xyz/reference/warpcast/api#get-starter-pack-members

if we can find where the ids are

Systems Engineering student 🇦🇷 | Data Analyst @quotient | Learning & building

sounds like a solid idea, gonna dig into how starter packs are structured. If we can extract them cleanly, Quotient can turn them into usable feeds