[Cryptography] MAybe he's Adam Back, maybe not

[Jon Callas]

> On Apr 16, 2026, at 19:08, Peter Gutmann via cryptography <cryptography at metzdowd.com> wrote:
> 
> iang <iang at iang.org> writes:
> 
>> iang, also tagged twice as SN on writing style & British spelling.
> 
> Maybe a bunch of us could all say we're Satoshi and, like a CDO, dilute the
> risk by spreading it around?
> 
> The thing that most of the graphology-analysers keep missing is coding style.
> Some of the folks on the cypherpunks list did actually write code, so there's
> an extensive corpus of code out there that doesn't require cloud-picture-
> analysis, you just look at the BTC code and other code the person has written
> and you can immediately see it's not them.

I forget who it was -- one of the malware researchers guys, whom I follow on Twitter -- criticized the NYT article precisely on stylometry grounds, saying that his analysis of Adam Back's code is that Adam writes C++ like a C programmer who came to it late, and that the bitcoin code looks like an MFC programmer. This is precisely why I don't think it was Hal Finney, myself. To me, it looks nothing like Hal's usual code. And yet there are other knowledgeable people who have other opinions.

I think that discussions about stylometry either in code or English break down if you assume the writer was trying not to be caught. The whole idea of stylometry is that there are unconscious signals in someone's writing, but if someone is putting in explicit misdirections, or using some simple tools like word frequency counts to develop countermeasures.

I think Satoshi tried to hide style both in the paper and code, myself, and as fun as it is to debate it, it's really easy to forget that we're looking through a lens that intentionally had stuff smeared on it.

	Jon