Alert Outcome Analysis — April 29, 2026
Period: Apr 29, 2026 (00:59–09:11 UTC) Total alerts: 26 2x definition: ATH market cap > $70K post-migration Data sources: GeckoTerminal OHLCV, DexScreener, bot alert_cache.db (VPS) Bot status: First full session on DigitalOcean VPS (deployed Apr 28)
Results by Tier
| Tier | 2x | Total | Win Rate |
|---|---|---|---|
| EXTREME | 4 | 7 | 57% |
| VERY_FAST | 1 | 1 | 100% |
| FAST | 0 | 3 | 0% |
| MODERATE | 2 | 4 | 50% |
| SLOW | 2 | 6 | 33% |
| STALLED | 1 | 5 | 20% |
| TOTAL | 10 | 26 | 38% |
Winners (confirmed ATH > $70K)
| Tier | Name | Peak MC | Traders in BC | Total SOL |
|---|---|---|---|---|
| EXTREME | Actually Useless | $224K | 1 (parsiix) | 2.38 |
| SLOW | chadhouse | $110K | 2 (chester, dv) | 6.70 |
| EXTREME | Mr. Nice Guy | $102K | 1 (kevnszn) | 4.39 |
| MODERATE | sam moonman | $98K | 2 (dv, parsiix) | 3.93 |
| VERY_FAST | OpenAI Coin | $92K | 2 (dv, west) | 13.33 |
| EXTREME | Broke Retards On Solana | $86K | 3 (theo, chester, dv) | 8.95 |
| STALLED | Frok | $83K | 2 (theo, parsiix) | 5.49 |
| EXTREME | Garlic Dog | $76K | 4 (dv, chester, jijo, parsiix) | 9.01 |
| SLOW | TerminAItor | $73K | 1 (dv) | 3.45 |
| MODERATE | Ewun Mogman | $70K | 1 (trenchman) | 4.89 |
Near-Misses (ATH $45–69K)
| Tier | Name | ATH MC | Gap to 2x | Traders |
|---|---|---|---|---|
| SLOW | The Human | $64K | -$6K | trenchman, errol, bandit |
| STALLED | Chudjak | $56K | -$14K | chester |
| FAST | Ketalon | $49K | -$21K | nosa1x, parsiix |
Hard Failures (ATH ≤ $40K — never meaningfully left migration MC)
| # | Token | Tier | Est. ATH | Traders |
|---|---|---|---|---|
| 1 | VAMPCOIN | EXTREME | $36K | jijo |
| 2 | codemaxxing | EXTREME | ~$35K | chester |
| 3 | Psyop (“You have to buy this”) | FAST | ~$35K | decu, dv |
| 4 | AGI (“Abeyant Goblins Inside”) | FAST | ~$35K | chester, dv |
| 5 | the farm | SLOW | ~$35K | trenchman, errol |
| 6 | GremlinCoin | STALLED | ~$35K | dv, trenchman |
| 7 | PumperNickel | MODERATE | ~$35K | dv |
| 8 | LIL FCKIN GOBLIN (LFG) | SLOW | $34K | dv, errol, trenchman |
| 9 | That Fucking Bird (LARRY) | SLOW | ~$35K | west |
| 10 | Agent Tusk | MODERATE | ~$35K | kevnszn, leck |
| 11 | Just-Dice | STALLED | ~$35K | errol |
| 12 | agent tusk (copy) | STALLED | ~$35K | kevnszn, errol |
| 13 | Goolag | EXTREME | ~$35K | nosa1x |
Pattern Analysis
Pattern 1 — EXTREME tier improved dramatically (57% vs 30% on Apr 27–28)
7 EXTREME alerts, 4 winners. This is a major shift from the previous session where EXTREME was 30% (3/10). The winners all had identifiable hooks: Actually Useless (viral self-deprecating meme), Mr. Nice Guy (Trump Truth Social post linked on DexScreener), BROS (degen community identity), Garlic Dog (4-trader consensus — highest trader count in dataset).
The 3 EXTREME losers: VAMPCOIN (vampire naming = negative signal), codemaxxing (generic), Goolag (derivative of Google).
Key difference from Apr 27–28: the EXTREME winners this session had more known-trader SOL committed. Garlic Dog had 9.01 SOL across 4 traders. BROS had 8.95 across 3. The losers had 1-2 SOL from single traders. Total BC SOL is a better predictor than tier speed for EXTREME alerts.
Pattern 2 — VERY_FAST is a new tier that produced a winner
OpenAI Coin migrated with VERY_FAST speed and hit $92K. This is the first appearance of this tier in the dataset. Two traders (dv + west) committed 13.33 SOL total — the highest SOL commitment of any alert this session. The combination of extreme speed + high SOL + recognizable brand name (OpenAI) created a strong setup.
Note: west committed 9.88 SOL — the single largest position in the entire dataset. This trader was new to the tracker but showed immediate high-conviction behavior.
Pattern 3 — Copy token pattern reconfirmed: Agent Tusk appeared twice
Agent Tusk (2zT8K…, MODERATE, alerted at 07:45) and agent tusk (3niS4…, STALLED, alerted at 09:11). Both failed. The first had kevnszn + leck. The copy had kevnszn + errol. Same pattern as Dog In Vest / Vesting in Apr 27–28. The original didn’t even win — and the copy still attracted known traders (kevnszn appeared in both). The duplicate name detector would have caught this.
Pattern 4 — dv appeared on 6 winners but also 5 losers
In the Apr 27–28 session, dv was described as “0 confirmed hard failures.” That no longer holds. This session dv appeared on: Psyop ❌, AGI ❌, GremlinCoin ❌, PumperNickel ❌, LFG ❌. All hard failures.
However, dv’s win rate is still 55% (6/11) — significantly above the session average of 38%. And critically, dv’s winners include every tier: EXTREME (BROS, Garlic Dog), VERY_FAST (OpenAI), MODERATE (sam moonman), SLOW (chadhouse, TerminAItor). dv remains the best signal but is not infallible. dv alone is insufficient — the narrative filter still matters.
Pattern 5 — parsiix is the most efficient trader (4/5 = 80% hit rate)
parsiix appeared on 5 alerts: Actually Useless ✅, Garlic Dog ✅, sam moonman ✅, Frok ✅, Ketalon (near-miss). That’s 4 winners out of 5, an 80% hit rate. parsiix also had the session’s biggest winner (Actually Useless, $224K).
Combined with the Apr 27–28 data where parsiix had 2 winners, the cumulative record is 6 winners across 2 sessions. parsiix picks winners at a higher rate than any other tracked trader.
Pattern 6 — 3+ trader consensus produced the highest peaks again
| Token | Traders | Peak MC |
|---|---|---|
| BROS | 3 (theo, chester, dv) | $86K |
| Garlic Dog | 4 (dv, chester, jijo, parsiix) | $76K |
This matches the Apr 27–28 pattern where both 3-trader alerts hit 64K) — 3 traders + weak narrative ≠ winner.
Refinement: 3+ traders is a positive signal, but narrative quality determines if it breaks $70K. The Human had 3 traders but no narrative hook. Garlic Dog had 4 traders AND a memeable name.
Pattern 7 — High SOL commitment ≠ guaranteed win
LARRY had west committing 8.89 SOL (single largest individual bet) and still failed. Agent Tusk had kevnszn at 10.13 SOL — also failed. Individual high SOL is not predictive by itself. What matters is SOL from multiple traders (consensus) or SOL from the most accurate traders (dv, parsiix).
Compare: OpenAI Coin had 13.33 SOL from 2 traders (dv + west) and won $92K. But LARRY had 8.89 SOL from west alone and failed. Multi-trader SOL > single-trader SOL.
Pattern 8 — STALLED win rate dropped (20% vs 40% on Apr 27–28)
Only 1 STALLED winner (Frok, $83K) out of 5 alerts. The Apr 27–28 session’s STALLED thesis was “a trader accumulating for 30 minutes has conviction.” This session’s STALLED losers included GremlinCoin (dv + trenchman) and agent tusk #2 (kevnszn + errol) — multi-trader STALLED alerts that still failed. STALLED is not inherently high-quality. The slow-BC conviction thesis holds only when the traders involved have high hit rates (parsiix, dv on good narratives).
Pattern 9 — New traders: west, errol, jijo, nosa1x, leck, bandit
6 traders appeared for the first time. Quick assessment:
| Trader | Alerts | Winners | Win Rate | Notes |
|---|---|---|---|---|
| west | 2 | 1 (OpenAI) | 50% | Huge SOL size (9.88 avg). Mixed — big win, big loss |
| errol | 5 | 0 | 0% | Appeared on 5 alerts, 0 winners. Noise trader |
| jijo | 2 | 1 (Garlic Dog) | 50% | Early data — need more |
| nosa1x | 2 | 0 | 0% | Goolag ❌, Ketalon near-miss |
| leck | 1 | 0 | 0% | Only appeared once (Agent Tusk) |
| bandit | 1 | 0 | 0% | Only appeared once (The Human near-miss) |
errol is the session’s noisiest trader — 5 appearances, 0 winners. If errol is in BC without a higher-quality trader present, the alert is likely noise.
Synopsis
The Apr 29 session produced a 38% win rate on 26 alerts, slightly better than the Apr 27–28 session (32% on 31 alerts). The detector’s first full session on the VPS ran without interruption and captured a higher-quality batch of alerts.
Key shifts from the previous session:
-
EXTREME improved from 30% to 57% — driven by higher SOL commitment and better narratives in winning tokens. EXTREME + high total SOL + identifiable narrative = real signal.
-
dv is still the best signal but now has confirmed failures. 6/11 win rate (55%) vs 5/10+ in the previous session. The difference: dv’s failures this session all had weak/no narratives (Psyop, PumperNickel, GremlinCoin). dv in BC + strong narrative = high conviction. dv in BC + no narrative = noise.
-
parsiix is the most accurate tracker — 4/5 (80%) this session. Combined across both sessions: the trader to watch.
-
Copy token pattern reconfirmed — Agent Tusk appeared twice, both failed. Duplicate detector would have caught the second.
-
New trader errol is a negative signal — 0/5 wins. If errol is the only known trader in BC, consider skipping.
P&L Simulation — 0.1 SOL per Alert, Sell at ATH
Assumptions
- Entry MC: $33K (pump.fun migration price)
- Exit: peak ATH market cap — theoretical sell at exact top
- Position: 0.1 SOL per trade, no fees or slippage
- Hard failures estimated at $33–36K ATH
Total trades: 26 Total deployed: 2.6 SOL
Winners
| Token | Tier | Peak MC | Mult | Return | P&L |
|---|---|---|---|---|---|
| Actually Useless | EXTREME | $224K | 6.78x | 0.678 SOL | +0.578 SOL |
| chadhouse | SLOW | $110K | 3.33x | 0.333 SOL | +0.233 SOL |
| Mr. Nice Guy | EXTREME | $102K | 3.09x | 0.309 SOL | +0.209 SOL |
| sam moonman | MODERATE | $98K | 2.97x | 0.297 SOL | +0.197 SOL |
| OpenAI Coin | VERY_FAST | $92K | 2.78x | 0.278 SOL | +0.178 SOL |
| BROS | EXTREME | $86K | 2.61x | 0.261 SOL | +0.161 SOL |
| Frok | STALLED | $83K | 2.52x | 0.252 SOL | +0.152 SOL |
| Garlic Dog | EXTREME | $76K | 2.31x | 0.231 SOL | +0.131 SOL |
| TerminAItor | SLOW | $73K | 2.22x | 0.222 SOL | +0.122 SOL |
| Ewun Mogman | MODERATE | $70K | 2.13x | 0.213 SOL | +0.113 SOL |
| Subtotal | 3.074 SOL | +2.074 SOL |
Near-Misses
| Token | Tier | Peak MC | Mult | Return | P&L |
|---|---|---|---|---|---|
| The Human | SLOW | $64K | 1.95x | 0.195 SOL | +0.095 SOL |
| Chudjak | STALLED | $56K | 1.70x | 0.170 SOL | +0.070 SOL |
| Ketalon | FAST | $49K | 1.49x | 0.149 SOL | +0.049 SOL |
| Subtotal | 0.514 SOL | +0.214 SOL |
Hard Failures
| Token | Tier | Est. Peak | Mult | Return | P&L |
|---|---|---|---|---|---|
| VAMPCOIN | EXTREME | $36K | 1.09x | 0.109 SOL | +0.009 SOL |
| 11 tokens @ ~$35K | Various | $35K | 1.06x | 1.166 SOL | +0.066 SOL |
| LFG | SLOW | $34K | 1.03x | 0.103 SOL | +0.003 SOL |
| Subtotal | 1.378 SOL | +0.078 SOL |
Full Summary
| Category | Trades | Deployed | Returned | P&L |
|---|---|---|---|---|
| Winners | 10 | 1.0 SOL | 3.074 SOL | +2.074 SOL |
| Near-Misses | 3 | 0.3 SOL | 0.514 SOL | +0.214 SOL |
| Hard Failures | 13 | 1.3 SOL | 1.378 SOL | +0.078 SOL |
| TOTAL | 26 | 2.6 SOL | 4.966 SOL | +2.366 SOL |
| Deployed | 2.6 SOL |
| Returned | 4.966 SOL |
| Net P&L | +2.366 SOL |
| ROI | +91% |
P&L Simulation — 2x Take-Profit
Rules: Tokens hitting 66K → sell at ATH (same as base).
All 10 winners hit 2x → each returns 0.200 SOL → winners subtotal: 2.000 SOL Near-misses unchanged: 0.514 SOL Hard failures unchanged: 1.378 SOL
| Deployed | 2.6 SOL |
| Returned | 3.892 SOL |
| Net P&L | +1.292 SOL |
| ROI | +50% |
Left on table: 1.074 SOL from winners (Actually Useless alone lost 0.478 SOL of upside).
P&L Simulation — Tiered Exit
Rules:
- TP1: sell 50% at 2x ($66K MC)
- TP2: sell 30% at 3x ($99K MC)
- TP3: sell remaining 20% at trailing stop −20% from peak
- No TP fires → trailing stop on full position at peak × 0.80
Winners — Hits 3x+ (3 tokens)
| Token | Peak | TP1 (50%@2x) | TP2 (30%@3x) | TP3 (20%@peak×0.8) | Total | P&L |
|---|---|---|---|---|---|---|
| Actually Useless | 6.78x | 0.100 | 0.090 | 0.109 | 0.299 | +0.199 |
| chadhouse | 3.33x | 0.100 | 0.090 | 0.053 | 0.243 | +0.143 |
| Mr. Nice Guy | 3.09x | 0.100 | 0.090 | 0.049 | 0.239 | +0.139 |
Winners — Hits 2x but not 3x (7 tokens)
| Token | Peak | TP1 (50%@2x) | Remaining 50% @trailing | Total | P&L |
|---|---|---|---|---|---|
| sam moonman | 2.97x | 0.100 | 0.119 | 0.219 | +0.119 |
| OpenAI Coin | 2.78x | 0.100 | 0.111 | 0.211 | +0.111 |
| BROS | 2.61x | 0.100 | 0.104 | 0.204 | +0.104 |
| Frok | 2.52x | 0.100 | 0.101 | 0.201 | +0.101 |
| Garlic Dog | 2.31x | 0.100 | 0.092 | 0.192 | +0.092 |
| TerminAItor | 2.22x | 0.100 | 0.089 | 0.189 | +0.089 |
| Ewun Mogman | 2.13x | 0.100 | 0.085 | 0.185 | +0.085 |
Winners subtotal: 2.182 SOL returned on 1.0 SOL — P&L: +1.182 SOL
Near-Misses — Trailing stop on full position
| Token | Peak | Trailing Exit | Return | P&L |
|---|---|---|---|---|
| The Human | 1.95x | 1.56x | 0.156 | +0.056 |
| Chudjak | 1.70x | 1.36x | 0.136 | +0.036 |
| Ketalon | 1.49x | 1.19x | 0.119 | +0.019 |
Near-misses subtotal: 0.411 SOL returned on 0.3 SOL — P&L: +0.111 SOL
Hard Failures — Trailing stop on full position
Most at 1.06x → trailing at 0.85x → 0.085 SOL each 11 × 0.085 = 0.935 | VAMPCOIN (1.09x) = 0.087 | LFG (1.03x) = 0.082
Hard failures subtotal: 1.104 SOL returned on 1.3 SOL — P&L: −0.196 SOL
All Strategies Comparison
| Strategy | Deployed | Returned | P&L | ROI |
|---|---|---|---|---|
| ATH exit | 2.6 SOL | 4.966 SOL | +2.366 SOL | +91% |
| 2x take-profit | 2.6 SOL | 3.892 SOL | +1.292 SOL | +50% |
| Tiered exit | 2.6 SOL | 3.697 SOL | +1.097 SOL | +42% |
What the Apr 29 Math Tells Us
2x TP outperforms tiered this session (+50% vs +42%). This is the opposite of the Apr 27–28 result. The reason: no massive outlier. The biggest winner was Actually Useless at 6.78x — meaningful but not extreme. In the previous session, Vesting (9.3x) and Justice For Luca (6.97x) made the trailing slice on tiered worthwhile. Here, only 3 tokens hit 3x+, and the trailing slice on a 6.78x is only 0.109 SOL.
The tiered strategy’s value proposition is outlier capture. When the session contains no 10x+ runners, the trailing slice adds complexity without proportional reward. The 2x TP is simpler and more profitable in a “normal” session.
However, you don’t know in advance which session has the outlier. The Apr 27–28 session had SCAM (485x) and Vesting (9.3x) which made tiered dramatically better. The right approach is still tiered — the rare sessions with outliers more than compensate for the normal sessions where 2x TP edges ahead.
P&L Simulation — Worst Case (Hard Failures + Near-Misses → $0)
| Strategy | Winners Return | Losses (→ $0) | Total | P&L | ROI |
|---|---|---|---|---|---|
| ATH exit | 3.074 SOL | 0 SOL | 3.074 SOL | +0.474 SOL | +18% |
| 2x take-profit | 2.000 SOL | 0 SOL | 2.000 SOL | −0.600 SOL | −23% |
| Tiered exit | 2.182 SOL | 0 SOL | 2.182 SOL | −0.418 SOL | −16% |
Even in worst case, ATH exit stays green (+18%) because the 10 winners average 3.07x. Both 2x TP and tiered go negative. The stop loss is still the most important variable — in the realistic simulations above, losers returned ~0.85–1.06x instead of zero, which is the difference between a profitable and losing book.
Trader Leaderboard — April 29, 2026
Scoring: 1 point per coin bought in bonding curve that hit $70K+ ATH post-migration.
| Rank | Trader | Points | Winning Coins | Hit Rate |
|---|---|---|---|---|
| 🥇 1 | dv | 6 | BROS, Garlic Dog, chadhouse, sam moonman, OpenAI Coin, TerminAItor | 6/11 (55%) |
| 🥈 2 | parsiix | 4 | Actually Useless, Garlic Dog, sam moonman, Frok | 4/5 (80%) |
| 🥉 3 | chester | 3 | BROS, Garlic Dog, chadhouse | 3/6 (50%) |
| 4 | theo | 2 | BROS, Frok | 2/2 (100%) |
| 5 | kevnszn | 1 | Mr. Nice Guy | 1/3 (33%) |
| 5 | west | 1 | OpenAI Coin | 1/2 (50%) |
| 5 | trenchman | 1 | Ewun Mogman | 1/5 (20%) |
| 5 | jijo | 1 | Garlic Dog | 1/2 (50%) |
| — | errol | 0 | — | 0/5 (0%) |
| — | nosa1x | 0 | — | 0/2 (0%) |
| — | decu | 0 | Psyop (❌) | 0/1 (0%) |
| — | leck | 0 | — | 0/1 (0%) |
| — | bandit | 0 | — | 0/1 (0%) |
Combined Leaderboard — Apr 27–29, 2026
| Rank | Trader | Apr 27–28 | Apr 29 | Combined | Combined Rate |
|---|---|---|---|---|---|
| 🥇 1 | dv | 5 | 6 | 11 | ~57% |
| 🥈 2 | parsiix | 2 | 4 | 6 | ~67% |
| 🥉 3 | chester | 2 | 3 | 5 | ~45% |
| 4 | theo | 2 | 2 | 4 | ~67% |
| 5 | kevnszn | 2 | 1 | 3 | ~38% |
| 6 | trenchman | 1 | 1 | 2 | ~25% |
| 7 | decu | 1 | 0 | 1 | — |
| 7 | west | 0 | 1 | 1 | — |
| 7 | jijo | 0 | 1 | 1 | — |
Tier 1 traders (most reliable): dv (11 wins), parsiix (6 wins, 67% rate), theo (4 wins, 67% rate) Tier 2 traders (consistent): chester (5 wins, 45% rate), kevnszn (3 wins) Tier 3 traders (emerging): trenchman, west, jijo — too few data points Noise traders: errol (0/5), nosa1x (0/2), leck (0/1), bandit (0/1)
Cross-Session Comparison
| Metric | Apr 27–28 | Apr 29 | Trend |
|---|---|---|---|
| Total alerts | 31 | 26 | ↓ fewer alerts |
| Win rate | 32% (10/31) | 38% (10/26) | ↑ improved |
| EXTREME win rate | 30% (3/10) | 57% (4/7) | ↑ significantly improved |
| STALLED win rate | 40% (2/5) | 20% (1/5) | ↓ regressed to mean |
| Avg winner peak MC | $140K | $102K | ↓ smaller winners |
| Biggest winner | Vesting $307K | Actually Useless $224K | ↓ no 9x+ runner |
| Unique traders | 8 | 13 | ↑ wider tracker net |
| ROI (ATH exit) | +110% | +91% | ↓ slightly lower |
| Copy token failures | Dog In Vest | agent tusk #2 | Same pattern |
The detector is getting more accurate (38% vs 32%) while seeing fewer total alerts. Quality over quantity. The main weakness this session: no breakout 5x+ runner except Actually Useless, which caps the upside on every strategy.
Suggestions for Improving Alert Quality (Updated)
From Apr 27–28 (validated by Apr 29):
- ✅ Name blacklist — VAMPCOIN would have been caught by a “vamp” check (vampire ≈ scam-adjacent). Consider expanding:
["scam", "rug", "ponzi", "exploit", "drain", "vamp"] - ✅ Duplicate name detector — Agent Tusk appeared twice. Second was a STALLED copy that failed.
- ✅ Surface total BC SOL — This session confirms: high total SOL (9+ across 2+ traders) correlated with winners.
New from Apr 29:
-
Add trader-level weighting to alerts. parsiix (80%), theo (100% this session), and dv (55%) are categorically more reliable than errol (0%) or nosa1x (0%). Weight the alert priority by trader quality:
- Tier 1 trader (dv, parsiix, theo) → ELEVATED priority
- Tier 2 (chester, kevnszn) → standard priority
- Unknown/new trader → reduced priority until track record established
-
Flag errol as a noise trader. 0/5 this session. If errol is the only known trader in BC → deprioritize or suppress the alert.
-
Track VERY_FAST as a distinct tier. OpenAI Coin was the only VERY_FAST alert and it won. The tier appears between EXTREME and FAST. Need more data to assess whether VERY_FAST is reliably high quality or this was a one-off.
Key Takeaways
- 38% overall win rate — 10/26 alerts resulted in 2x+, up from 32%
- dv is still #1 — 6 winners, but now has confirmed failures when narrative is weak
- parsiix is the most accurate — 80% hit rate this session, 67% combined
- EXTREME improved — 57% win rate when supported by high total SOL and narrative
- Copy tokens confirmed — Agent Tusk duplicate failed, same as Dog In Vest pattern
- errol is noise — 0/5, consider filtering when solo
- Stop loss remains critical — worst-case simulations go negative without it
- No session-defining outlier — biggest winner was 6.78x vs previous session’s 9.3x (or 485x SCAM)
Related
- alert-outcomes-apr27-28 — Apr 27–28 analysis (31 alerts)
- hard-failure-postmortem-apr2026 — loser case study validating Winner Checklist
- winner-checklist — go/no-go decision framework
- narrative-triggers — narrative tier list and scoring
- migration-speed-signal — speed tier definitions