THIS IS F**KING ABSURD
I WENT TO SLEEP WITH $100. GROK COPIED TRADERS ALL NIGHT. I WOKE UP TO $4,256.
At 20:00 I connected Phantom. Grok could see the balance and copy trades. It couldn't move a cent on its own.
Then I split the $100 into six tiny bets. Under $20 each, so no single wallet could hurt the night.
Grok watched 1,284 wallets on FOMO, flagged the ones with real history and picked six to follow. Every buy they made, it made. Every sell, it sold.
At 00:33 the account was only 4x. Two of the six were deep red. One had turned $20 into 27 cents.
I'd have panicked and closed them. I was asleep.
The bot did nothing dramatic. No revenge trade, no new wallet, no bigger size, no panic at 1am. It just kept mirroring the six.
By 07:32 the same wallets had sold into coins worth millions in market cap. The one that started at $15 finished at 87x.
Even the wallet that was at 27 cents came back to 20x.
It also knew when not to copy. At 23:34 a whale exited a coin. Ignored. At 06:37 an early wallet sold. Not copied.
5 years of trading taught me small size is what lets you sit through the ugly middle of the night.
8 buys. 58 sells. $100 to $4,256 by 08:00.
The full copy setup is below. Free, and almost nobody starts this small. Save it before bed.
if half of a https://t.co/xRyHOvuIy6 coin's first 100 buyers came in the first 10 seconds, buying it at $100K won 25% of the time
calmer starts: 58%
i took every coin launched by 202 https://t.co/xRyHOvuIy6 devs from nov 14-21 (4,935 launches). for the 285 that reached $30K i pulled the first 100 buyers and counted how many bought in the first 10 seconds
the trade: buy when it crosses $100K (or $30K), sell at 2x, stop at -50%, 4% costs
buy at $100K:
> under 50 of the first 100 in 10 sec: 43 trades, 58% won, +24% per trade
> 50 or more: 24 trades, 25% won, -25% per trade
buy at $30K:
> under 50: 68 trades, 38% won, -2% per trade
> 50 or more: 76 trades, 16% won, -37% per trade
at $30K, 35 of 144 coins had all 100 first buyers inside 10 seconds
same split in both halves of the week: 63% vs 25%, then 54% vs 25% at $100K
sniper tags agree: 20+ snipers in the first 100 → 16% won at $30K, 35% with fewer
and you can see all of it before you buy
how to check on gmgn (20 sec):
1.) token page → Early Buyers. the first 100 wallets in order. most rows showing the launch time = the crowd got there before you
2.) Holders → Snipers count. 20 or more: skip it at $30K
3.) info panel → Bundler %. a high number backs up the no
4.) calm start and it reaches $100K → that's the group that won 58%
one week, 285 coins. the crowd count is known at entry, nothing here looks into the future
$10,524.66 → $1,242,743.31 118x
6 fomo wallets. each put in $1-3K. each took out $120K+. all UTC:
> jul 14: sockzt buys PONS 4 times. $1,011.90 at $788K-912K mc
> aug 6: degenone buys CETS. $1,065.14 from $56.8K
> aug 11-13: degenone sells 125 times. $139,384.90
> aug 20: kyle buys MOO once. $1,990.45 at $136.2K
> aug 29: Pote_korea buys GG. $1,474.09 from $56K
> sep 5: Pote_korea is out. $131,852.92
> sep 8, 02:54: KnownGraySmelt and Je3x buy 4STOCK in the same minute
> sep 8, 11:54: Je3x is out. $154,328.67
> sep 9: KnownGraySmelt is out after 55 sells. $433,057.46
> oct 3: kyle is out after 51 sells. $123,176.70
> nov 10: sockzt's last sell. all 6 are out
in → out:
> KnownGraySmelt: $2,989.92 → $433,057.46 (145x)
> sockzt: $1,011.90 → $260,942.66 (258x)
> Je3x: $1,993.16 → $154,328.67 (77x)
> degenone: $1,065.14 → $139,384.90 (131x)
> Pote_korea: $1,474.09 → $131,852.92 (89x)
> kyle: $1,990.45 → $123,176.70 (62x)
19 buys, 308 sells. transfers excluded, every dollar out is a sale
together: 118x
the 4Stock pair alone: $4,983.08 → $587,386.13
if half of a https://t.co/xRyHOvuIy6 coin's first 100 buyers came in the first 10 seconds, buying it at $100K won 25% of the time
calmer starts: 58%
i took every coin launched by 202 https://t.co/xRyHOvuIy6 devs from nov 14-21 (4,935 launches). for the 285 that reached $30K i pulled the first 100 buyers and counted how many bought in the first 10 seconds
the trade: buy when it crosses $100K (or $30K), sell at 2x, stop at -50%, 4% costs
buy at $100K:
> under 50 of the first 100 in 10 sec: 43 trades, 58% won, +24% per trade
> 50 or more: 24 trades, 25% won, -25% per trade
buy at $30K:
> under 50: 68 trades, 38% won, -2% per trade
> 50 or more: 76 trades, 16% won, -37% per trade
at $30K, 35 of 144 coins had all 100 first buyers inside 10 seconds
same split in both halves of the week: 63% vs 25%, then 54% vs 25% at $100K
sniper tags agree: 20+ snipers in the first 100 → 16% won at $30K, 35% with fewer
and you can see all of it before you buy
how to check on gmgn (20 sec):
1.) token page → Early Buyers. the first 100 wallets in order. most rows showing the launch time = the crowd got there before you
2.) Holders → Snipers count. 20 or more: skip it at $30K
3.) info panel → Bundler %. a high number backs up the no
4.) calm start and it reaches $100K → that's the group that won 58%
one week, 285 coins. the crowd count is known at entry, nothing here looks into the future
the half-life of a https://t.co/xRyHOvuIy6 pump: buy at $100K and the median coin is worth half your entry in 52 minutes
after 24 hours it's worth 8%
i took every coin launched by 231 https://t.co/xRyHOvuIy6 devs from nov 13-20 (5,482 launches), bought each one when it crossed $30K, $50K, $100K or $250K, and sold on a timer
how long until the median coin is worth half your entry:
> bought at $30K: 20 min
> bought at $50K: 22 min
> bought at $100K: 52 min
> bought at $250K: ~2 hours
bought at $100K, 63 trades, 4% costs, sold after:
> 15 min: +19% avg, 44% won
> 1 hour: +6% avg, 35% won
> 3 hours: -28% avg, 19% won
> 24 hours: -69% avg, 13% won
at $250K, held 24h: 0 of 39 trades in profit
the average is pulled up by a few runners. the median trade is red at every timer. what changes is how red: -14% at 15 min, -60% at 1 hour, -96% at 24 hours
third sample in a row with the same shape. buy at $100K and hold:
> coins traders bought, nov 13-19: 1h +5%, 6h -29%, 24h -60%
> launches by 209 devs: 1h +10%, 6h -29%, 24h -36%
> launches by 231 devs: 1h +6%, 6h -44%, 24h -69%
at $30K and $50K every timer lost on average. the clock only helps once a coin has already proved itself
how to use it:
1.) pick your exit time before you buy. still in after an hour = you're holding the median coin
2.) right after you buy, set a sell at 2x in gmgn and a 60 min timer on your phone
3.) "i'll hold it overnight" was the most expensive trade in this test. 24h was the worst timer at every entry
4.) still running after an hour? sell most of it, keep a small bag
yesterday my backtest said: buy coins that cross $100K in minute one
i tried to break it on 2 new samples. one held. one didn't
the rule: buy when a https://t.co/xRyHOvuIy6 coin crosses $100K, sell at 2x, stop at -50%, 4% costs per trade
yesterday (nov 13-19): crossed in minute 1 → 75% of trades won. crossed later → 24%
test 1, new days. 1,019 launches from nov 19-20 the old test never saw. coins picked the same way (what real traders bought):
> minute 1: 18 trades, 50% won, +19% per trade
> later: 87 trades, 23% won, -28% per trade
held. smaller edge
test 2, every launch. all 5,570 coins launched by 209 devs in 7 days, no filter:
> minute 1: only 6 trades, 33% won, -3% per trade
> later: 69 trades, 32% won, -10% per trade
gone. same rule at $30K: minute 1 won 24% (17 trades), later 32% (184)
why they disagree: test 1 only has coins someone bought. a coin that ripped in minute 1 and died in minute 3 barely gets buyers, so it barely makes the sample. test 2 has all of them
across all launches: 4.6% ever reach $30K, 1.7% reach $100K, 5 of 5,570 reached $1M
what i'm keeping:
> the rule as a filter on coins already trending on gmgn. on raw launches it did nothing
> a second sample for every backtest before i call it a rule
> small size. 18-20 trades in the winning group is still a small number
how to check any memecoin backtest, 3 questions:
1.) which coins? "trending" and "coins people bought" have already dropped the instant rugs
2.) out of sample? it has to hold on days the test never saw
3.) how many trades in the winning group? under 30 is a hint at best
the half-life of a https://t.co/xRyHOvuIy6 pump: buy at $100K and the median coin is worth half your entry in 52 minutes
after 24 hours it's worth 8%
i took every coin launched by 231 https://t.co/xRyHOvuIy6 devs from nov 13-20 (5,482 launches), bought each one when it crossed $30K, $50K, $100K or $250K, and sold on a timer
how long until the median coin is worth half your entry:
> bought at $30K: 20 min
> bought at $50K: 22 min
> bought at $100K: 52 min
> bought at $250K: ~2 hours
bought at $100K, 63 trades, 4% costs, sold after:
> 15 min: +19% avg, 44% won
> 1 hour: +6% avg, 35% won
> 3 hours: -28% avg, 19% won
> 24 hours: -69% avg, 13% won
at $250K, held 24h: 0 of 39 trades in profit
the average is pulled up by a few runners. the median trade is red at every timer. what changes is how red: -14% at 15 min, -60% at 1 hour, -96% at 24 hours
third sample in a row with the same shape. buy at $100K and hold:
> coins traders bought, nov 13-19: 1h +5%, 6h -29%, 24h -60%
> launches by 209 devs: 1h +10%, 6h -29%, 24h -36%
> launches by 231 devs: 1h +6%, 6h -44%, 24h -69%
at $30K and $50K every timer lost on average. the clock only helps once a coin has already proved itself
how to use it:
1.) pick your exit time before you buy. still in after an hour = you're holding the median coin
2.) right after you buy, set a sell at 2x in gmgn and a 60 min timer on your phone
3.) "i'll hold it overnight" was the most expensive trade in this test. 24h was the worst timer at every entry
4.) still running after an hour? sell most of it, keep a small bag