Seven of Eight Crypto Whitepapers Are Harder to Read Than the Tax Code
Eight crypto whitepapers scored with the readability formulas schools use. Seven are harder than IRS Publication 17. The one exception is the document explicitly written to teach rather than to establish priority.
It scores 47.5 on Flesch Reading Ease and grade 12.2 on Flesch-Kincaid — roughly a high-school-senior reading level. That is harder than IRS Publication 17, which scores 49.7.
Get new posts in your inbox
No spam, unsubscribe anytime.
About the Author
Sujit KarkiFinance Researcher & Market Analyst
Independent finance researcher and market analyst with expertise in macroeconomics, equity markets, and personal finance. I help regular investors make better-informed decisions through rigorous, data-driven analysis.
Eight crypto documents measured with standard readability formulas. Seven are harder to read than IRS Publication 17.
The one exception is Monero's Zero to Monero, at 52.1 — a 55,000-word textbook, not a whitepaper.
Bitcoin's whitepaper is the most readable of the papers proper, at grade 12.2 — a high-school-senior reading level.
Cardano's Ouroboros summary is the hardest at 27.3 reading ease, with 27.4% complex words. Ethereum's demands the highest grade level, 15.9.
The formulas measure sentence and word length, not conceptual difficulty, so these are a floor on how hard these documents actually are.
"Do your own research" is the standing advice in crypto, and the whitepaper is
where you are told to start. Every project has one. It is presented as the
document that lets an ordinary person evaluate the thing before putting money in.
So: can an ordinary person read it? That is a measurable question, and the tools
for measuring it are older than the internet.
The measurement
Flesch Reading Ease and Flesch-Kincaid Grade Level are the formulas schools,
insurers and government plain-language rules have used for decades. They are
arithmetic over three counts — sentences, words, syllables:
Reading Ease = 206.835 − 1.015 × (words per sentence) − 84.6 × (syllables per word)
Higher is easier. 60–70 is considered plain English.
No judgement, no vocabulary list, no model deciding what counts as clear. Anyone
with the same text gets the same number to the decimal.
I ran thirteen crypto documents through it — eight resolved — alongside five
documents whose difficulty you already have an intuition for, so the scores land
on a scale rather than floating free.
Harder than the tax code
Full results — crypto documents and references for scale
Document
Words
Reading ease
Grade level
US Constitution
4,707
23.8
21.5
Cardano (Ouroboros)
888
27.3
14.2
Wikipedia: Blockchain
12,923
31.8
11.8
Ethereum
14,030
34.8
15.9
Chainlink
54,415
36.0
13.0
Ripple
4,721
39.6
13.8
Solana
7,507
45.1
12.4
Bitcoin
3,450
47.5
12.2
Tezos
5,874
48.7
10.9
IRS Publication 17
144,954
49.7
10.3
Monero (Zero to Monero)
55,692
52.1
10.7
FTC crypto scam guide
2,183
53.4
9.8
Alice in Wonderland
30,325
73.4
7.7
Seven of the eight crypto documents are harder to read than IRS Publication
17.
That is worth sitting with. Publication 17 is the general guide to filing a
federal tax return — the document people invoke as shorthand for unreadable
government writing. It scores 49.7. Bitcoin's whitepaper scores 47.5, Ethereum's
34.8, Cardano's Ouroboros summary 27.3.
Bitcoin's asks for a high-school-senior reading level. Ethereum's asks for a
college junior. Cardano's has 27.4% complex words — more than one word in
four with three or more syllables, denser than the Constitution.
The exception proves the point rather than undermining it. The single
document easier than the tax code is Monero's Zero to Monero, at 52.1 — and it
is not a whitepaper. It is a 55,000-word textbook written to teach the system to
people who want to understand it, as opposed to a paper written to establish a
technical claim. Given a document whose purpose is explanation rather than
priority, crypto can produce prose at a tenth-grade reading level. Almost nobody
chooses to.
Warning
Correction, 7 September 2026. This post was first published with five
documents and the headline claim that every whitepaper measured is harder than
the tax code. Widening the corpus to thirteen attempts, of which eight resolved,
falsified that: Monero's Zero to Monero scores 52.1, above IRS Publication
17's 49.7. The title, the takeaways and the table have been revised, and the
finding is now stated as seven of eight. The original five scores are unchanged
— the claim was too strong for the sample it rested on, which is exactly what a
wider sample is for.
One reference complicates this, and it belongs in the table. Wikipedia's
article on blockchain — written by volunteers for a general audience, with no
incentive to sound impressive — scores 31.8, harder than seven of the
eight whitepapers — every one of them except Cardano's. Writing about this
subject is genuinely difficult, and some of what the whitepaper scores capture
is the topic rather than the authors.
That is a real confounder and it caps how strong the claim can be. What it does
not explain is the spread. Wikipedia and Ethereum are within three points of
each other; Monero sits seventeen to twenty points above both, on the same
subject matter, because someone set out to teach it. The topic sets a floor on
difficulty. It does not decide where in the twenty-point range above that floor
a document lands.
Pro Tip
Notice which document is easiest after the children's book: the FTC's guide to
crypto scams, at 53.4. The clearest writing about crypto in this sample comes
from the agency warning people about it, not from the projects asking for their
money. That is not an accident — plain-language standards are a legal
requirement for federal consumer material.
Why this matters more than it sounds
The stock answer is that whitepapers are technical documents and technical
documents are hard. Fair enough for a cryptography paper aimed at cryptographers.
But that is not the role the whitepaper plays. It is cited to retail buyers as
the thing that makes an investment legible — read the whitepaper is offered as
the answer to how do I know this is real. A document requiring a college
reading level cannot serve that function for most people, and roughly half of US
adults read below the level Ethereum's whitepaper demands.
The practical consequence is that "read the whitepaper" functions less as advice
than as a filter. It sounds like due diligence and mostly produces the appearance
of it — people skim, recognise a few words, and conclude the project is
substantial because the document was hard. Difficulty reads as rigour when you
cannot evaluate the content.
If you cannot get through the whitepaper, that is not a personal failing and it
is not a reason to buy anyway. It is a reason to treat the "do your own research"
framing with suspicion, because the research it points at is not written for you.
What the formulas cannot see
Four limits, and the last is the one that matters most.
Sample size. Eight documents resolved, not eight hundred. These are among the
best-known projects with stable, publicly fetchable documents. A broader sample
would likely be worse, since obscure projects have less editorial attention, but
that is a guess and this does not test it.
Five documents failed. Polkadot, Filecoin and Algorand returned 404 at the
URLs their own sites publish; Avalanche's and Litecoin's pages did not yield
enough continuous prose to score. They are reported as failures in the data
rather than quietly dropped, because a missing row that goes unmentioned is how
samples get flattering. That three of thirteen official whitepaper links are dead
is its own small finding about the genre.
Not every document is the same kind of document. Two are book-length — Zero
to Monero at 55,692 words and Chainlink's at 54,415 — and are technical guides
rather than papers. At the other end, Cardano's row is an 888-word summary of
Ouroboros, not the full academic paper. Reading ease is length-independent, so
the scores remain comparable, but the documents are not equivalent artefacts and
the table names what each one is.
The formulas measure length, not difficulty. This is the real limitation.
Flesch counts syllables and sentence length; it has no idea what an elliptic
curve is. A paper of short sentences about hard mathematics scores as easy, and a
paper of long sentences about nothing scores as hard. The US Constitution's 23.8
is largely an artefact of 18th-century sentence construction, not conceptual
density — which is exactly why it is in the table, as a visible reminder of what
these numbers can and cannot tell you.
So read the results as a floor. The whitepapers are at least this hard. On
conceptual difficulty — the thing a reader actually runs into — they are
certainly harder, and no formula in this post can tell you by how much.
Methodology
The full results are published with this post as
readability.json, including the five documents that
failed to resolve. The script is
research/readability.py; it re-runs
against live documents, so the numbers here are what it returned on the date
below rather than a transcription.
Documents. Thirteen crypto documents attempted, fetched from each project's
own domain; eight resolved (Bitcoin, Ethereum, Solana, Ripple, Cardano,
Chainlink, Monero, Tezos) and five failed. Plus five references: the US
Constitution, IRS Publication 17, the FTC's crypto scam guide, Wikipedia's
blockchain article, and Alice in Wonderland.
Extraction. PDFs parsed with pypdf; HTML stripped of script, style and
navigation before counting. URLs, citation markers and non-ASCII artefacts are
removed, since they distort word length without being words anyone reads.
Formulas. Flesch Reading Ease and Flesch-Kincaid Grade Level as stated above.
Syllables counted by the standard vowel-group heuristic with a silent-final-e
adjustment — approximate, applied identically to every document, so comparisons
between them hold even where an absolute count is off by a little.
Why formulas rather than a model.The previous experiment on this
site tried to score qualitative
text by keyword matching and the metric was wrong in both directions. Readability
indices are deterministic arithmetic with no vocabulary list and no judgement,
which is the right instrument for a question like this one.
Disclosure. The commentary was drafted with AI assistance; the document
selection, the method and the conclusions are mine.
Sujit Karki||8 min read