Farmer Detection Framework
⚠️ Data confidence: LOW–MEDIUM — Farmer classifications (Cented, Domy) are based on a single day’s data and one month’s aggregate. The framework logic is sound; individual verdicts should be re-evaluated monthly. See cented for updated Cented profile.
Summary
Traders who farm copy traders have a distinct pattern: unrealistic win rates combined with high visibility and extreme short holds. Real skilled traders maintain realistic win rates (50-60%) even with fast execution. This framework helps distinguish between sustainable trading edges and exploitative farming strategies.
What is a “Farmer”?
A farmer is a trader who:
- Builds a public profile (leaderboard ranking, high visibility)
- Attracts copy traders to follow their trades
- Executes trades very quickly (7-10 seconds)
- Exits BEFORE copy traders’ orders fill
- Profits from the spread created by copy traders, not from actual trading edge
The Farmer Mechanism
T=0s: Farmer buys 100 SOL of token X
T=1s: Leaderboard updates (farmer shows +5% profit)
T=2s: Copy traders see the signal
T=3s: Copy traders start buying (placing orders)
T=5s: Token price rises due to copy volume
T=7s: FARMER EXITS (sells at peak)
T=10s: Copy traders' buys finally fill (at high price)
T=20s: Token dumps (initial buying pressure exhausted)
T=30s: Copy traders are underwater
Result: Farmer made +5%, copy traders made -2%. Farmer profited FROM copy traders, not WITH them.
Red Flags: How to Spot a Farmer
🚩 RED FLAG #1: Unrealistic Win Rate
| Win Rate | Hold Time | Interpretation |
|---|---|---|
| 90%+ | Any | 🚨 FARMER (statistically impossible) |
| 75%+ | < 30s | 🚨 FARMER (too high for scalping) |
| 60-70% | < 30s | ⚠️ SUSPICIOUS (verify other flags) |
| 50-60% | < 30s | ✅ REALISTIC (possible with real edge) |
| 40-50% | Any | ✅ REALISTIC (normal trading range) |
Why? At extreme speeds (7-10 seconds), even the best traders face execution risk. A 94.6% win rate is statistically implausible unless the edge comes from copy traders (farming), not from market movements.
Real Data Evidence:
- theo: 55.9% win rate = REALISTIC (hold time disputed — see note below)
- radiance: 54.45% win rate = REALISTIC
- Cented: 94.6% win rate = UNREALISTIC 🚨
⚠️ Data Conflict — theo’s hold time: This page originally cited “7-second holds” for theo (sourced from Theo_Live_Trade_Analysis, a 99-minute on-chain snapshot on Apr 22). The operational strategy pages (Theo_Exit_Rules, Theo_Step_by_Step_Guide) describe 30–60 second holds. The Theo_Front_Run_Strategy analysis shows 30–60 sec as the average. The 7-second figure may reflect a subset of very fast trades or a measurement artifact. Treat theo’s hold time as 7–60 seconds depending on trade type; the 55.9% win rate is consistent across sources.
🚩 RED FLAG #2: High Visibility + Leaderboard Dominance
| Metric | Farmer Signal | Skill Signal |
|---|---|---|
| Daily Rank | #1, #2 (very high) | #9, #20, #25 (lower) |
| Broadcast Hold Time | YES (brags about speed) | NO (keeps it private) |
| Social Media Presence | High (recruits copy traders) | Low (doesn’t need them) |
| Mentions of “Copy Me” | Explicit | Never mentioned |
Why? Farmers NEED copy traders to farm. They attract visibility. Skilled traders don’t need farmers — their edge works independently.
Real Data Evidence:
- Cented: #1 daily + #1 monthly = FARMING SIGNAL (10-second hold time asserted; no direct on-chain verification in raw data)
- theo: #9 daily → #2 monthly, hold time not broadcast = SKILL SIGNAL
- radiance: #8 daily → #20 monthly, quiet profile = SKILL SIGNAL
🚩 RED FLAG #3: Daily Spike → Monthly Fade
| Pattern | Interpretation |
|---|---|
| #1 daily, #1 monthly | ✅ Consistent skill OR farming both |
| #1 daily, #50 monthly | 🚨 FARMER (daily edge doesn’t compound) |
| #2 daily, #27 monthly | 🚨 FARMER (ranking collapsed) |
| #9 daily, #2 monthly | ✅ SKILL (slow-building edge) |
| #8 daily, #20 monthly | ✅ SKILL (sustainable) |
Why? Real edges compound monthly. Farming edges collapse monthly (copy traders adapt, find new farmers).
Real Data Evidence:
- Domy: Daily #2 → Monthly #27 (25-position drop) = FARMER
- theo: Daily #9 → Monthly #2 (rank improved) = SKILL
🚩 RED FLAG #4: Win Rate Doesn’t Improve with Scale
| Trader | Daily Win % | Monthly Win % | Trend | Signal |
|---|---|---|---|---|
| Cented | 47.4% | 94.6% | DOUBLED | ⚠️ Suspicious (improved too much) |
| theo | ~56% | 55.9% | Flat | ✅ Realistic (consistent) |
| radiance | 43.5% | 54.45% | Slight improvement | ✅ Realistic (normal learning) |
Why? Real traders improve 2-10% with scale (better execution, learning). Farmers show extreme swings (47% → 95%) = they’re exploiting copy traders differently daily/monthly.
🚩 RED FLAG #5: Extreme Short Holds ONLY
| Trader | Hold Times | Pattern | Signal |
|---|---|---|---|
| Farmer | All 5-15s | Never varies | 🚨 FARMING (can’t adapt) |
| Skilled | Mix (5s, 30s, 2m) | Adapts to market | ✅ SKILL (flexibility) |
Why? Real traders adapt to market conditions. Farmers are locked into the 5-15s window (where copy traders take time to fill). If they held longer, copy traders would catch up.
Real Data Evidence:
- theo: 7s hold, but portfolio shows variation (different tokens, different timing) = SKILL
- Cented: Strict 10s holds, always exits at that window = FARMING PATTERN
The Farmer Detection Checklist
Score 1 point for each red flag:
Trader Analysis Template
Trader Name: _______________
[ ] 1. Win Rate: Is it 75%+ on fast holds (< 30s)?
(90%+ = automatic farmer flag)
[ ] 2. Visibility: Are they #1-3 on leaderboards?
(Do they broadcast hold times to attract copy traders?)
[ ] 3. Daily vs Monthly: Did daily rank drop >10 spots monthly?
(e.g., #2 daily → #27 monthly)
[ ] 4. Consistency: Did win rate jump >20% from daily to monthly?
(47% → 95% = farming signal)
[ ] 5. Flexibility: Do they only hold 5-15 seconds?
(No variation = farming lock)
SCORE: ___ / 5
0-1 points: ✅ REAL SKILL (safe to copy)
2-3 points: ⚠️ SUSPICIOUS (investigate more)
4-5 points: 🚨 FARMER (do not copy)
Case Studies: Real Examples
✅ CASE STUDY #1: THEO (Real Skill)
Win Rate: 55.9% monthly ✅ REALISTIC
Visibility: #9 daily, #2 monthly ✅ NOT CHASING #1
Hold Time: 7–60 seconds (source-dependent) ✅ WITHIN REALISTIC RANGE
Daily → Monthly: Improved ✅ SKILL
Flexibility: Portfolio varies ✅ ADAPTABLE
Score: 0/5 → REAL SKILL ✅
Conclusion: theo’s edge is real execution, not farming. Safe to copy. Hold time varies by data source (7s from on-chain snapshot vs 30–60s from strategy docs); win rate is consistent at 55–59% across all sources.
✅ CASE STUDY #2: RADIANCE (Real Skill)
Win Rate: 54.45% monthly ✅ REALISTIC
Visibility: #8 daily, #20 monthly ✅ LOW PROFILE
Hold Time: 8 seconds ✅ CONSISTENT
Daily → Monthly: Stable ✅ SUSTAINABLE
Flexibility: Portfolio shows variety ✅ ADAPTABLE
Score: 0/5 → REAL SKILL ✅
Conclusion: radiance is a steady professional. Edge is real. Safe to copy.
🚨 CASE STUDY #3: CENTED (Likely Farmer)
Win Rate: 94.6% monthly 🚨 UNREALISTIC
Visibility: #1 daily, #1 monthly 🚨 MAXIMUM VISIBILITY
Hold Time: 10 seconds (published) 🚨 BROADCASTS FOR FARMERS
Daily → Monthly: Both #1 🚨 MAINTAINS #1 (farming advantage)
Flexibility: All 10-second holds 🚨 LOCKED PATTERN
Score: 5/5 → LIKELY FARMER 🚨
Conclusion: Cented’s 94.6% win rate is too high to be real skill. Likely exploiting copy traders. Risky to copy.
🚨 CASE STUDY #4: LUKEY (Likely Farmer)
From monthly leaderboard (rank #30):
Win Rate: 96.4% (27 wins / 1 loss) 🚨 STATISTICALLY IMPOSSIBLE
Monthly PnL: +94.12 SOL (low absolute PnL despite perfect record)
Visibility: Ranked monthly #30 ⚠️ MODERATE (building profile)
Hold Time: Unknown — insufficient data
Daily Appearance: Not in daily top 30 ⚠️ INCONSISTENT
Score: 3/5 → LIKELY FARMER 🚨
Conclusion: 27/1 win rate over a 30-day period is mathematically implausible for market-based trading. The low absolute PnL (+94 SOL despite 96% win rate) suggests small position sizes — consistent with early-stage farming while building a follower base. Monitor for ranking increase and win rate decline as farming saturates.
⚠️ CASE STUDY #5: WUGI (Suspicious)
From monthly leaderboard (rank #17):
Win Rate: 73.1% (19 wins / 7 losses) ⚠️ ELEVATED
Monthly PnL: +246.77 SOL
Visibility: Ranked monthly #17 ⚠️ MODERATE VISIBILITY
Hold Time: Unknown — insufficient data
Daily Appearance: Not in daily top 30 — no cross-platform check available
Score: 2/5 → SUSPICIOUS ⚠️
Conclusion: 73% monthly win rate clears the farming suspicion threshold (75%+) by a narrow margin. Insufficient data to confirm. Needs hold time data and daily/monthly cross-check before copying. Treat as suspicious until more data available.
Why Farmers Eventually Fail
The Farmer Lifespan
PHASE 1: Growth (Weeks 1-4)
- Trader builds profile
- Attracts first copy traders
- Farming edge works (spreads are juicy)
- Win rate: High (94%+)
- Copy traders: Happy (early followers do well)
PHASE 2: Saturation (Weeks 5-12)
- More copy traders follow
- Orders fill faster (less spread to farm)
- Edge narrows
- Farmer has to trade faster/more aggressively
- Win rate: Starts declining
PHASE 3: Collapse (Weeks 13+)
- Copy traders figure out they're being farmed
- Exit trades early (front-run the farmer)
- Farmer's edge disappears
- Win rate: Crashes to 20-30%
- Monthly ranking: Drops to #50+
- Copy traders: Abandon (lost money)
Result: Farmer’s returns unsustainable. Real skill (theo, radiance) compounds. Farmers fade.
The Skill vs Farmer Win Rate Gap
Why This Matters
Real skilled trader at 8-second holds:
- Win rate: 55%
- Loses 45% of trades (real market risk)
- Edge: Comes from better timing, entry accuracy, risk management
- Sustainable: Yes (works regardless of copy traders)
Farmer at 8-second holds:
- Win rate: 94%
- Loses only 6% of trades (impossible without farming)
- Edge: Comes from exploiting copy trader fills
- Sustainable: No (collapses when copy traders leave)
The Math:
- 55% win rate at 8s = realistic (market noise, execution risk)
- 94% win rate at 8s = farming (copy traders create the spread)
How to Verify Your Assessment
Tier 1: Score the Checklist (0-5)
Use the template above. Score 4-5 = likely farmer.
Tier 2: Analyze Win Rate Math
- Calculate expected win rate for the hold time
- 8-second holds should be 45-60% max
- 90%+ = automatic red flag
Tier 3: Track Over Time
- Watch the trader for 2-4 weeks
- Do rankings change?
- Does win rate decline?
- Farmers will show:
- Win rate declining
- Monthly rank dropping
- Copy traders leaving
Tier 4: Check Community Feedback
- Are copy traders complaining about fills?
- Are they talking about “farming”?
- Do experienced traders avoid this trader?
Actionable Rules
✅ SAFE TO COPY (Real Skill)
- Win rate 40-60% on any hold time
- Win rate 50-65% on short holds (< 30s)
- Daily rank ≠ Monthly rank (skill improves with scale)
- Low visibility (not trying to recruit copy traders)
- Hold times vary (adaptable to market)
🚨 DO NOT COPY (Likely Farmer)
- Win rate 75%+ on any hold time
- Win rate 90%+ on short holds (statistically impossible)
- 1-2 daily AND 1-2 monthly (farming advantage)
- Broadcasts hold times (recruits copy traders)
- All trades in 5-15 second window (locked pattern)
⚠️ INVESTIGATE FURTHER (Suspicious)
- Win rate 65-75% on short holds
- Daily #1 OR monthly #1, but not both
- Medium visibility (some recruitment)
- Recent entrant to leaderboard (unproven)
- Portfolio shows zero losses (impossible)
The Key Insight
Real traders show realistic statistics. Farmers show impossible statistics.
The difference between theo (55.9% win rate) and Cented (94.6% win rate) on similar hold times is not experience — it’s that theo trades against the market, Cented trades against copy traders.
Final Checklist: Before You Copy Anyone
[ ] Win rate makes mathematical sense for hold time?
[ ] Daily/monthly rankings are consistent or improving?
[ ] Trader doesn't broadcast speed or attract copy traders?
[ ] Win rate is stable or improving (not declining)?
[ ] Portfolio shows adaptation (not locked pattern)?
[ ] Been profitable for 30+ days?
[ ] Monthly ranking is top 50 (proven over time)?
If all YES → Copy this trader
If any NO → Investigate further
Sources
- axiom_trades_apr22.txt
- kol_scan_daily_apr22.txt
- kol_scan_monthly_apr22.txt
- Cielo trader profiles (hold time data)
Related Pages
- theo — Real skill example (55.9% win rate, sustainable)
- cented — Farmer example (94.6% win rate, suspicious pattern)
- winner-checklist — Complementary signal validation framework
- theo-style — How skilled traders execute