Altcoin Season Index: How It Is Calculated

An altcoin season index is a count: the share of the largest coins whose price rose by more than bitcoin's over the last 90 days. When 75 percent or more beat bitcoin, the two best-known versions call it altcoin season, and when 25 percent or fewer do, bitcoin season. Rebuilt from public prices for the 90 days ending October 22, 2026, 58 of the 99 other coins in the top 100 beat bitcoin (58.6 percent), and 27 of the 49 others in the top 50 did (55.1 percent). Both readings sit in the middle band, and neither is close to 75.

The more useful result is how far the answer moves when any one rule is changed. The same day gives 38 percent on a 30-day window and 59 percent on 90 days. It gives 59 percent with stablecoins and tokenized funds removed and 45 percent with them left in. The count is a description of past price changes. It does not say that a season is under way or about to start.

Two published rules, side by side

The two indexes most often quoted are run by CoinMarketCap and by Blockchain Center. Their own pages, read on October 23, 2026, state the rules this way.

Three things the rules leave open matter for the numbers that follow. Bitcoin is itself one of the top coins and cannot outperform itself, so a "75 percent of the top 100" test is either 75 of 99 or 75 of 100, and neither page says which. The CoinMarketCap page does not define the band between 25 and 75. And neither page publishes the list of coins it counted or the formula that turns the count into an index number, so an outsider can reproduce the rule but not the exact figure.

Rebuilding the count from public prices

The coin ranking is the CoinGecko list of the 200 largest coins by market cap, pulled at about 11:27 UTC on October 23, 2026. Prices come from the Coin Metrics community API for the 82 of the 100 coins it carries under matching names, and from CoinGecko daily prices for the other 18. Both give a 00:00 UTC price, and the dates are aligned so that every coin is measured at the same instant: the ending point is Coin Metrics' October 22 fixing, which is midnight UTC at the end of that day, and the starting point is the one 90 days earlier, July 24. Bitcoin went from 64,100.11 to 81,257.02 dollars over those days, a rise of 26.8 percent.

The ranking contains items that are not crypto assets in the sense either index intends. Of the 100 largest entries by market cap, 24 are dollar-pegged tokens (18), tokenized funds (4) or tokenized metals (2). The rule used here removes them and counts the next 100 coins down:

The classification here was made by reading each entry's name, and the CoinGecko list used here contains no wrapped bitcoin or staked ether tokens, so none needed removing. After the removals the 100th coin is Tezos, at rank 134 on the raw list.

The medians are taken over the other coins in each set, and bitcoin's own change was +26.8 percent. The share of the full set, with bitcoin in the denominator, is 58.0 percent for the top 100 and 54.0 percent for the top 50. Blockchain Center showed 53 on the same day. The two are close, but its coin list is not published, so the difference cannot be traced to a rule.

Eight ways to move the answer

Changing one rule at a time on the same day shows how much of an index number is a rule and how much is the market.

The window changes the answer more than any other rule: 38.4 percent at 30 days and 58.6 percent at 90 days from the same coins on the same day. The coin set matters least in this snapshot and the treatment of ether matters least of all, because ether's 28.1 percent return was within 1.3 points of bitcoin's and left the count nearly unchanged.

A count that ignores how much

The index records whether each coin beat bitcoin and discards the margin. In this snapshot that hides a lot. Thirty-five of the 99 coins beat bitcoin by more than 20 percentage points, and thirteen finished within 5 points of it on either side, so the count is sensitive to small differences in bitcoin's own number:

A five-point shift in bitcoin's return moves the count by six or seven coins. Reaching 75 percent would have needed 75 of the 99 coins to beat bitcoin, 17 more than did. The coins that count as winners are also very unequal: the six best returns over the window were Bitway (+2,800.6 percent), Akedo (+1,042.2), Pearl (+583.7), Quack AI (+515.4), Derive (+233.6) and Quant (+194.3), and the six worst were Rain (-39.2), Stable (-32.9), MemeCore (-28.0), Provenance Blockchain (-25.2), BinanceLife (-21.0) and Canton (-14.0). Eighty-two of the 99 rose in dollar terms and only two fell by more than 30 percent. The index treats Bitway's 2,800 percent and Tezos beating bitcoin by a tenth of a point as the same one vote.

When bitcoin falls, the dollar tokens start to win

Whether stablecoins are left in is not a cosmetic choice, because their dollar price barely moves. When bitcoin rises, they cannot beat it. When bitcoin falls, every one of them beats it. The table uses the same group of coins, the top 100 by raw market-cap rank, and shows the count with and without the 24 tokens, for several end dates and with bitcoin's 90-day return for each.

When bitcoin rose, leaving the tokens in lowered the share by 10 to 14 points; when it fell by 17 to 20 percent, leaving them in raised it by about 5 points. The two August and July rows also show the other limit of a backward look: 29 of the coins on today's list had no price 90 days earlier, so the sets compared are different sets. Seven of the 24 tokens in these rows were assumed flat.

Moving the start date by a few days

An index quoted to one decimal place suggests a precision the rule does not have. The window is fixed at 90 days, but its start day is a single price for each coin, and a few days either way changes both bitcoin's return and the count. Holding the end at October 22 and the top 100 list fixed:

Across these nine start dates the count ranges from 54 to 62 coins, 54.5 to 62.6 percent, and every one sits in the neutral band. That spread of about eight points from the start date alone is a fair measure of the noise in a single published reading. A move from 53 to 58 between two days on a chart can be the window sliding past one day's price, not a change in anything about the market.

Position by position

The counts above can be checked line by line. This table lists the 49 coins that make up the top 50 after bitcoin, in market-cap order, with the 90-day change to October 22, 2026 and whether it beat bitcoin's +26.8 percent. Litecoin missed by a tenth of a point (+26.7 percent) and Cardano cleared by 2.4 points (+29.2 percent), which shows how thin the line is for coins near bitcoin's own return.

Grouping the top 100 by position shows that no band carries the result. Among the nine coins after bitcoin in positions 1 to 10, six beat it (66.7 percent, median change +29.4). Of the 15 in positions 11 to 25, eight did (53.3 percent, median +29.2); of the 25 in positions 26 to 50, thirteen did (52.0 percent, median +29.2). Positions 51 to 75 had 16 of 25 (64.0 percent, median +50.8) and positions 76 to 100 had 15 of 25 (60.0 percent, median +31.8). Medians are taken within each band. The smaller coins did slightly better on this count, which is the pattern an altcoin-season story would predict, but the margin between the bands is a few coins and the lowest band reading is still 52 percent. The 75 percent line would have required about three of every four coins in every band to beat bitcoin, and no band came close.

The same coins, weighted by size

A count gives Bitway and Ethereum one vote each, though Ethereum is worth far more. Weighting by market value gives a different picture of the same 99 coins. Their combined market cap in the October 23 snapshot was about 799 billion dollars, of which Ethereum was 36.6 percent. The coins that beat bitcoin held about 601 billion dollars of it, or 75.2 percent, so by value three quarters of the altcoin market belonged to coins that beat bitcoin, while by number only 58.6 percent did. The cap-weighted 90-day return of the 99 coins was +44.0 percent, the unweighted mean +90.4 percent (pulled up by Bitway's +2,800.6 percent) and the median +32.8 percent, against bitcoin's +26.8 percent.

The weighted figures are biased in a way that the count is not. The weights are end-of-window caps, which already include each coin's gain over the window, so winners carry more weight than they did on July 24. A weighting by starting caps would give a smaller share, and it cannot be computed from the data pulled here. The point is that the three summaries of the same coins on the same day, 58.6 percent by count, 75.2 percent by end-of-window value and +90.4 percent for the mean change, answer three different questions, and only the first is what the published indexes report.

What the count would have read in recent weeks

Because the ranking changes daily, a historical index uses a different list on each date. The table instead fixes today's list and recomputes the 90-day count at weekly dates, which shows the swing but builds in a bias: the coins on today's list are today's survivors, and they are more likely than the full list at the time to include recent winners. The denominators fall in the earlier rows because coins listed during the window have no starting price.

Over these 13 weeks the top 100 reading ranged from 34.6 to 61.6 percent and never came near either threshold. Blockchain Center's page lists a flip on September 2, 2026 from Bitcoin Season to Neutral, which means its index was at or below 25 before that date. This recomputation does not reproduce a stretch that low: its lowest top 50 reading is 34.7 percent. A different coin list, the survivor effect or a rule not stated on the page could explain the gap, and this page cannot say which.

What a published index number can be checked against

A reader who wants to test a published reading has four things to compare. The first is the coin list, which neither site publishes, so the check is whether a reading falls inside the plausible range of rebuilt versions: Blockchain Center's 53 on October 23 compares with 54.0 percent (27 of 50) or 55.1 percent (27 of 49) here. The second is the window, where a mismatch of a few days moves the count by up to eight points, as the start-date table shows. The third is the treatment of dollar tokens and tokenized funds, which moves the count by 10 to 14 points on this data. The fourth is bitcoin's own return over the window, because a count of coins above a moving bar depends on the bar. A reading that falls within a few points of a rebuilt value tells you the rule was followed; it does not tell you which of the many reasonable lists was used.

What the number can and cannot say

A reading of 58.6 percent says that in a particular 90 days, a majority of a particular list of large coins did better than bitcoin. It does not say whether money is moving from bitcoin into other coins, because a coin can beat bitcoin by falling less. It does not distinguish a broad rise from a rise concentrated in a few coins that the count weights equally. And it is silent about the next 90 days. Market cap weights are a different lens on the same question: the bitcoin dominance guide explains the share of total market value that bitcoin holds, and the market cap versus FDV guide explains why the rankings used here depend on which supply figure a provider counts. For the cycle history that these indexes are usually read against, see the halving cycles page.

Frequently Asked Questions

What is the altcoin season index?

It is the share of the largest coins that outperformed bitcoin over a rolling window, 90 days in the two best-known versions. CoinMarketCap counts the top 100 and Blockchain Center the top 50, both excluding stablecoins and asset-backed tokens. Altcoin season is called at 75 percent or more, and bitcoin season at 25 percent or fewer.

Is it altcoin season right now?

By the rule both published versions state, no, on the data used here. For the 90 days ending October 22, 2026, 58.6 percent of the top 100 and 55.1 percent of the top 50 beat bitcoin, below the 75 percent line. Blockchain Center showed 53 and "It is not Altcoin Season!" on October 23. This is a statement about past price changes and not a forecast.

Why do different altcoin season charts show different numbers?

Each chooses a coin set (50 or 100), a window (30, 90 or 365 days), a treatment of stablecoins and tokenized funds, and a formula for the number. On the same day this page found 38.4 percent for a 30-day window and 58.6 percent for 90 days, and 44.9 percent with dollar tokens left in against 58.7 percent with them removed on a comparable set.

Does a high reading mean altcoins will keep outperforming?

No. The index counts which coins beat bitcoin in a past window, and the count moves by six or seven coins for every five-point change in bitcoin's own return. Nothing in the rule links one window to the next, and this page makes no prediction.

Can stablecoins affect the index?

Yes, if they are not excluded. A dollar token cannot beat bitcoin when bitcoin rises and always beats it when bitcoin falls. In the top 100 by raw rank, 24 entries were dollar tokens, tokenized funds or tokenized metals, and leaving them in lowered the share by 10 to 14 points when bitcoin was rising.

Sources

CryptosEyes publishes general educational research, not investment, legal, or tax advice. This article counts past price changes under stated rules. It does not identify a market phase, predict any coin's price, or recommend buying or selling anything.