Home / Research / Statistics

Dated populations · explicit denominators · no silent addition

The state of dead tokens.

Crypto mortality is measured three different ways below: last active trade, first observed website failure and human-reviewed project evidence. Each series keeps its own definition, source date and missingness.

TRADING FAILURES · 2021–2513.41M

GeckoTerminal aggregate

FAILED IN 202511.56M

year of last active trade

PUBLIC DOSSIERS37,294

explicit review tier on every record

EVIDENCE-REVIEWED41

deep DeadTokens case files

01 / ON-CHAIN

Trading failures by year

CoinGecko source ↗

Tokens once listed on GeckoTerminal, with at least one trade, grouped by the year they last traded. Pump.fun tokens were counted only after graduation.

2021
2,584
2022
213,075
2023
245,049
2024
1,382,010
2025
11,564,909
53.2%reported failed share across the study populationStopped trading does not establish that the website, team or code also stopped.

02 / LAUNCHPAD

Pump.fun lifespan

Study ↗
68.67%final Pump.fun bonding-curve trade on launch day
Observed
18.68M
Same day
12.83M
One day
2.18M
90+ days
850.18K

Through 2026-06-18. Excludes tokens with no trades and any later post-graduation trading on Raydium, Meteora or PumpSwap.

03 / REVIEW

Evidence coverage

Research desk →
13explicitly ceased
11possible silence leads
9unresolved records
3shortcut falsification

This curated cohort tests the evidence model. It cannot estimate market-wide mortality.

04 / WEBSITE CHECKS

Offline detections by month and year

Blockspot’s published benchmark records when its checker flagged a site—not necessarily the historic day a project died. Scanner and backfill batches can dominate a month.

Annual detections

2026*
6,919220% vs new
2025
10,38175% vs new
2024
12,488149% vs new
2022
10% vs new
2021
10% vs new

Trailing 13 months

Jul '2527532% vs new
Aug '258612% vs new
Sep '2520319% vs new
Oct '253,518265% vs new
Nov '2528721% vs new
Dec '251,321309% vs new
Jan '263812% vs new
Feb '266516% vs new
Mar '26407% vs new
Apr '26669% vs new
May '2616826% vs new
Jun '265,5401865% vs new
Jul '26*1,002593% vs new
Do not call “deaths ÷ new listings” a mortality rate.

It compares two same-period flows, not the share of a launch cohort that failed. June 2026’s 1,865% is more consistent with a large detection/backfill batch than a cohort survival estimate.

Inspect the source page ↗

05 / CHAINS

Website-offline records by blockchain

External Blockspot buckets at the dated capture, shown for comparison. The captured chain total exceeds the captured annual total by 56 records; the mismatch is retained rather than hidden.

Ethereum7,568
Binance Smart Chain7,472
Solana4,104
Base711
Polygon509
Avalanche325
Arbitrum294
Tron263
Fantom209
TON122
Other tokens2,065
Other cryptocurrencies6,204

READ BEFORE QUOTING

Three numbers, three populations.

13,407,627 is a last-trade aggregate from CoinGecko/GeckoTerminal, not our searchable records.

Provider directory withheld. Row-level aggregator data is not published without documented redistribution rights.

37,294 is the publication-safe dossier corpus: 37,253 automated evidence scans and 41 deep evidence-reviewed files.

Read definitions, failure modes and correction rules →