This article turns a folklore concept into a measurable one and reviews history against the measurement. It predicts no season and recommends no positioning. Nothing here is investment advice. Decisions are yours.
"Alt season" is crypto's most yearned-for phrase and one of its least defined: the period when alternative assets broadly and dramatically outperform BTC. Yearning is not a measurement, so this article builds one, checks history against it, and lists what has actually been true whenever the season arrived. The short version: real alt seasons are rare, shorter than memory suggests, and preceded by measurable conditions, none of which is a tweet.
A definition you can compute
Folklore defines alt season by vibes; measurement defines it by breadth of relative performance: the share of the liquid alt universe outperforming BTC over a trailing window. The convention popularized by public season-trackers uses 75% of the top assets beating BTC over 90 days as the threshold, and the exact numbers matter less than the structure: a breadth measure, over a window, against the reference asset. Note what the definition excludes on purpose. A handful of memes multiplying while the median alt bleeds is not a season, it is a lottery with good marketing. And alts falling slower than a crashing BTC technically "outperform" while everyone loses money, which is why season readings only mean anything alongside the absolute regime check the regime dial publishes. Our board methodology carries both ingredients daily, breadth and BTC-relative context, so the season question is answerable from published formulas rather than from whoever is loudest.
History against the ruler
Run the measurement backward and the episodes shrink to a short list. Late 2017 into January 2018: the canonical season, breadth extreme, nearly everything multiplying against BTC, and the finale of a mania that ended in an 80% market-wide winter. Early 2021, roughly January through May: the last broad, unambiguous season by the breadth ruler, ETH and the large-alt complex dramatically outrunning a strong BTC. Late 2021 was already narrower, sectors rotating while the long tail lagged its own folklore. And since then, stretches of days-to-weeks where breadth spiked without sustaining, but no 90-day broad season through the entire 2024-2025 stretch of BTC dominance. Two honest observations follow. Real seasons are rare, a couple per decade, not per year. And they are late-cycle phenomena: both canonical episodes bloomed in the final phase of their bulls, which is romantic in memory and dangerous in real time.
The three preconditions
Whenever the measured seasons arrived, three conditions were already true. A calm or grinding-up BTC: not necessarily ripping, but stable enough that the reference asset's risk stopped dominating every book, because nobody walks out the risk curve while the anchor is dragging, the one-factor structure documented across this library. Fresh external capital: seasons coincide with new-participant inflows, the cohorts who arrive after BTC's run and reach for higher beta, rather than with internal reshuffling among incumbents. And a narrative substrate: 2017's token boom, 2021's DeFi and NFT complex, some story machine converting attention into allocations across many assets at once, breadth being exactly what distinguishes a season from a lottery. The 2024-2025 absence is explained by the same list read in reverse: institutional flows expressed through vehicles that never touch alts, no broad narrative substrate, and a market whose new capital had, for the first time, a way to buy the asset class without buying the casino.
The costs of waiting for one
The season's rarity has a portfolio arithmetic that folklore omits. Holding a broad alt basket between seasons has meant holding assets that bleed against BTC for quarters at a time, which our own transplant experiment paid to document across 2024-2025, and the bleed compounds: an alt book down 40% against BTC needs a monster season merely to reach the benchmark it could have held passively. The strategy of "positioning early and waiting" is therefore a standing short of the reference asset with undefined duration, and the measured seasons' late-cycle timing adds the final insult: those who wait longest are rewarded closest to the cliff. None of which says seasons are untradeable. It says the trade is a regime response, act when breadth actually turns, not a faith position held against the reference clock.
Watching without narrating
The watchlist is three published numbers, no prophecy required. Breadth: the share of the universe above trend, and, sharper, the share outperforming BTC over the window, turning up and sustaining, days do not count, the definition is a window for a reason. The regime dial: risk-on, because every measured season lived inside one. And relative strength at the asset level for anything you actually follow, a ranking step the derivation journal still has queued. When all three agree, whatever is happening no longer needs the name; when they do not, the name is doing the work the numbers refused to. Alt season is real, rare, and measurable. Insist on the third property before believing anyone about the first two. Decisions are yours.