It was a verdict that reflected the often painful complexities and contradictions of blood testing company Theranos.

Four guilty verdicts, four acquittals and three charges on which the jury couldn't agree.

For many who have followed the Theranos saga, the podcasts, the documentaries, the books, you might have thought that Elizabeth Holmes's conviction was nailed down.

After all, she had claimed her diagnostics machines could test hundreds of diseases, when they couldn't. And she was the founder and chief executive of Theranos, so surely the buck would stop with her in court?

But underestimate Holmes at your peril. This is a woman who created a $9bn (£6.6bn) company she set up when she was 19. A woman who at one point had the world at her feet - who Bill Clinton and Joe Biden both praised.

There was another reason to think Holmes might be acquitted. These fraud cases are extremely difficult to prosecute. Jurors were asked to consider hundreds of technical documents and sit through evidence from dozens of witnesses.

Holmes has just had a baby too, and some commentators believed she would strike a sympathetic character.

Holmes also personally gave evidence in court, an unusual thing to do in a fraud trial. She described a relationship she had with Theranos's chief operating officer, Ramesh "Sunny" Balwani, who she claimed exercised coercive control over her. She also said she had been sexually abused. Mr Balwani categorically denies the allegations.

Some onlookers have remarked that this part of Holmes's evidence was particularly effective. The jury was listening.

Holmes's defence also believe they had one killer argument - that Holmes never sold her shares. Even when the company was worth nearly $10bn (and Holmes nearly $5bn) all wrapped up in shares, she didn't cash out. Those shares are now worthless.

Her defence argued that if she was a genuine fraud she would have taken the money and run. Instead, they said, she believed in what she was doing.

Yet both things can be true. You can have a vision, a goal, a mission as Holmes would call it, and still commit fraud.

At times Holmes was shown evidence that she herself admitted didn't look good. She said on more than one occasion that there were things she would have done differently in hindsight.

One particular example that sticks out is when the logos of Pfizer and GlaxoSmithKline were used to suggest the two pharmaceutical companies had endorsed Theranos. They had done no such thing. The prosecution made this a major plank of their argument - as close as they could get to a smoking gun.

Holmes's defence also had a major hole in it. Almost everyone you speak to about Theranos says that Holmes ran the company like an obsessive autocrat. She knew everything about what was going on.

Yet part of her defence was that she didn't know what was happening in her company - or the major problems with the tech. All too often she said she wasn't aware of information put to her by the prosecution, or that she didn't remember key events. It didn't ring true.

Holmes always wanted to be in control. And some have speculated that's why she decided to testify, to be in the driving seat of her own defence. It didn't work.

Silicon Valley is full of crooks, frauds and charlatans. "Fake it until you make it" is the brash term often used. Holmes certainly talked a good game. Yet the confidence, the certainty of success she conveyed was bluster.

However, some have noted it's telling that one of the very few high-profile female tech chief executives has been found guilty of fraud - when other high-profile (men) have not.

The verdict will, in theory, make founders sit up and take note - that there are consequences to not telling investors the truth.

But others wonder whether it will change anything much. In Silicon Valley there are still major rewards for selling dreams - for telling investors what they want to hear - rather than the truth.