On Tuesday, within a few hours of each other, Wired, and then Gizmodo, published articles contending that Satoshi Nakamoto, the name under which bitcoin was introduced to the world in November, 2008, is likely the pseudonym of an Australian information-security specialist named Craig Steven Wright. Both publications were working from anonymously sourced documents provided to them in November, and although both stopped short of expressing certainty Wired risked a headline proclaiming “Bitcoin’s Creator Satoshi Nakamoto Is Probably This Unknown Australian Genius.”

Soon after, I asked Gavin Andresen, who has been bitcoin’s de-facto lead developer since 2011, whether he thought the story was legitimate. He replied, “I have no idea if it’s true. I’d never heard of Craig Wright until skimming the Wired article just now.”

Attempt after attempt has been made to identify Nakamoto (Joshua Davis had a go, in 2011). The more journalists fail, the sweeter and more tempting the prize has become, like a burgeoning lottery jackpot. But past failures have led bitcoin watchers to become skeptical of such claims, and soon enough there were growing indications that key evidence linking Wright to Nakamoto was suspect.

There are two kinds of evidence that would verify the identity of the real Nakamoto beyond reasonable doubt. One is a digitally encrypted signature originating from private keys known to belong to him. The blockchain, bitcoin’s distributed ledger, contains a record of every bitcoin transaction ever made, and Nakamoto himself conducted the first transactions on the so-called genesis block, which left unfalsifiable records of his personal bitcoin addresses. He could send bitcoins from one of these addresses in order to prove his identity or use the address to “sign” an encrypted message. Nakamoto also published an ordinary P.G.P. fingerprint, of the kind commonly used for encrypting email, in his early writings; an encrypted signature originating from that address would provide similarly strong proof. But because it would be possible to steal or otherwise acquire this information from the real Nakamoto, even the appearance of a digital signature would still require careful investigation.

The second kind of evidence would be knowledge-based. The real Nakamoto would be able to describe in detail the work that he did before launching bitcoin, in January, 2009; he would be able to produce, or at least convincingly describe, private conversations and projects that he shared with key bitcoin figures such as Andresen, who has been a core developer of the cryptocurrency’s software from its earliest days, and the late Hal Finney, a pioneering developer of P.G.P. public-key cryptography. It goes without saying that this second kind of evidence would require that the candidate be willing submit to an interview.

Neither Wired nor Gizmodo was able to produce ironclad, cryptographically signed, or interview-based evidence of this kind. Instead, their cases rely on e-mails and documents purportedly taken from Wright’s computer. These include a draft trust agreement for the disposition of a cache of 1.1 million bitcoins, to be held by a company in the Seychelles whose ownership is linked, in the agreement, to an e-mail address and public keys known to have once been controlled by Nakamoto. The stories also include accounts of the strange, sad history of the late Dave Kleiman, Wright’s former colleague, who is alleged in the leaked documents to have been involved in helping conceal Wright’s identity as bitcoin’s inventor, and who died in April, 2013. Gizmodo also quoted two of Kleiman’s business partners, who claimed that Wright had told them that he and Kleiman were involved in the invention of bitcoin.

The convoluted, circumstantial nature of this evidence makes the stories look unusual, at best. To their credit, both publications acknowledged the holes; the authors of the Wired story, Andy Greenberg and Gwern Branwen, were particularly careful to allow the possibility of an elaborate hoax. As for Gizmodo, the site’s editor-in-chief, Katie Drummond, told me in an e-mail that her team had performed due diligence. “We didn’t think the hack alone was enough to go with, as compelling as some of these documents were,” she said. To try to verify the documents’ authenticity, Gizmodo hired a stringer in Sydney; sent the reporter Andy Cush, who co-wrote the story with Sam Biddle, to Florida; and relied on various experts.

But many questions remain. Gizmodo’s version of events makes much of a 2014 transcript of a meeting between Wright, his lawyer, and a representative of the Australian Taxation Office, in which Wright said, “I did my best to try and hide the fact that I’ve been running bitcoin since 2009.… By the end of this I think half the world is going to bloody know.” The same remark appears as a pull quote in Wired’s piece. Even if the transcript is genuine, though, it’s unlikely that “running bitcoin” was meant in the sense of “running a business.” Nobody “runs” open-source software like bitcoin; not even Andresen would ever have referred to himself in this way, because such software is designed specifically to give users the chance of having some hand in its development. It’s more likely that the phrase meant running the bitcoin software used to mine the currency. (The program is a sort of computational competition for a bitcoin reward that is generated roughly every ten minutes.) It would make sense, on one level, for Wright to profess a desire to conceal that he had been mining for so long, if it were true. Anybody who’d been mining steadily since 2009—when the network consisted of a few ordinary computers, rather than thousands of massively powerful purpose-built ones, as it does today—would have amassed a ton of bitcoin; you could have bought five thousand bitcoins for around twenty-seven dollars in cash in 2009 and they would now be worth around two million dollars.

It’s also possible that the Wright didn’t really have a huge cache of bitcoin, but that he, or someone else, wanted people to believe that he did. But why?