CoinSkeptic
1,900+ crypto ideas tested · 305 coins tracked · 11+ years of prices · every result shown, pass or fail
Free: the 9 rules that passed, on one page. Plus every new test by email. No coin picks, no hype. Unsubscribe anytime.Sign up →

← The Evidence

FAILED Nov 7, 2026

Should you buy the crypto that looks cheapest each week?

Short answer: No. Putting each week's money into the coin that looked cheapest ended with more money than an even split in only 45% of 1,900 groups, worse than picking at random (49%).

What we tested: Should you put each week's money into the coin that looks cheapest? Every coin Coinbase has listed, dead ones included, with a year of prices and $500k+ a day of trading: 1,900 random groups of 7 coins at 19 start dates from 2020 to 2024. Each week's $1 went into the coin lowest between its support and resistance, furthest below its record and its 200-day average and after the biggest 30-day fall, or into the opposite coin, instead of being split evenly; 104 weeks of buying, 1% cost each way

What counted as a pass (decided before we looked): Each pick had to end with more money than the even split in at least 60% of groups, with 5% or more extra money in the middle group, end ahead in at least 55% of groups with a middle result above even for start dates in 2020-2022 and in 2023-2024, and pass a luck check (fewer than 2.5% of random weekly picks as good)
Result: Neither pick worked. The cheapest-looking coin ended with more money than the even split in only 45% of 1,900 groups, 1.6% less in the middle group, worse than picking a coin at random (49%). It was ahead for start dates in 2020-2022 (54%) but behind for 2023-2024 (33%); the opposite pick did the reverse (40%, then 61%) and was ahead in 49% overall, 0.4% less in the middle. Alone, the coin furthest below its record was ahead in only 37% (9.4% less in the middle) and the coin lowest between its support and resistance in 51%, a coin flip

What it means for you

Buying plans and portfolios: what this family of tests supports, in plain moves. Information only, not financial advice.

Related tests

FAILEDNov 7, 2026
Do coins with a strong past keep winning for long-term buyers? Every coin Coinbase has listed, dead ones included, with a year of prices and $500k+ a day of trading, at 20 start dates from 2020 to 2024: the third whose weekly buying did best over the past year vs the worst third, then buying weekly for the next 2 years, 1% cost each way
No. Coins whose weekly buying did best over the past year ended ahead over the next 2 years in only 21% of cases, vs 25% for the worst ones. Only 7% beat Bitcoin. Full result →
FAILEDNov 7, 2026
Buy double when Bitcoin is 50%+ below its record, or the same amount every week? Both plans set aside $100 a week; the double plan buys $75 in normal weeks and saves $25, then spends $100 plus up to $100 of the savings in deep-fall weeks. Every start week since 2011, 1% cost on each buy and the sale
Not by saving up for it. Keeping $25 a week back to buy double in deep falls was ahead in only 34% of four-year runs and 0% of ten-year runs: the waiting cash missed Bitcoin's rises. Full result →
MIXEDOct 27, 2026
Lump sum vs $100 a week ($5,200 a year, value after 365 days)
It depends on the coin. Buying $100 a week beat a lump sum on 76% of start dates across all coins (most kept falling), but on Bitcoin and Ether a lump sum usually won. Full result →
MIXEDOct 27, 2026
Daily vs weekly vs monthly buying (the same $5,200 a year)
It barely matters on the big coins. Daily beat monthly by 3.9% across all coins, but on the four test coins it lost on 3 of 4 over all history. Full result →
MIXEDOct 29, 2026
All coins: $100 a week only in the BUY ZONE vs every week
Mixed. Buying $100 a week only in the BUY ZONE helped on smaller coins but not on the biggest ones ($500M+, $1B+), and with more coins added it mostly failed. Full result →

All 11 tests on DCA and buying plans →

Free: the 9 rules that passed, on one page. Plus every new test by email. No coin picks, no hype. Unsubscribe anytime.
Send me the 9 rulesWatch the tests on YouTube

Tested on past prices with no fees unless stated; the past is no promise. Information only, not financial advice. Every other test →