Facing 110 years in prison, Sam Bankman-Fried "can't recall" what he did at FTX

That's both, in the general case, technically true (the processes isn't adding up all the statutory maxima at one point) and substantively false (as for each charge, the sentence is clamped to the statutory range, and thus the sum of the statutory ranges is the actual legal maximum that can be applied, even though at no point in the sentencing calculation are the statutory maxima summed to get to the actual sentence; the actual sentences that are scheduled to run consecutively, which are already clamped to the statutory range, are summed.) Moreover, its particularly wrong, IIRC, and if the various changes to the charges haven't changed this, in SBF's case, because working through the facts in the indictment at least originally against the guidelines calculations for each charge very plausibly reached a guidelines range at or beyond the statutory maximum.

> That's not how federal sentencing guidelines work,

Federal sentencing guidelines are not mandatory, and federal courts can make departures within the range between the statutory minimum and statutory maximum above or below the guideline sentence. They are, to borrow from Pirates of the Carribean, more what you'd call guidelines than actual rules. Moreover, clamping sentence for each charge to the statutory range is exactly how the guidelines work.

Yes, it would be very useful (though highly subjective; even when departures aren't ultimately involved, whether conditions in the guidelines are applicable based on the jury verdict and the evidence at trial is the domain of often extensive post-verdict litigation that can amount to a mini-bench-trial on the guidelines; jounralists predicting the outcome of such would be getting in to highly speculative and error-prone rather than fact-based reporting) for coverage of federal criminal trials to present useful context around how the guidelines might apply based on the facts alleged in the indictment.

Popehat's repeated banging on the point that the reporting by the media of statutory ranges is generally useless for any real purpose is accurate and something that adds to the knowledge of someone who encounters it who has no pre-existing knowledge of how federal sentencing works. And for people who get the oversimplified idea from it that "adding up the sum of the statutory maxima never tells you what the realistic outer limit is for sentence in federal criminal cases, it is always significantly lower" will probably, for most cases they see reported in the media, mostly not have the wrong idea, even if they are working from a rule that is almost as oversimplified as "the sum of the federal statutory maxima is a realistic, if guaranteed to not be low, sentence estimate in federal criminal cases in general".

But several facts remain true:

1. The sum of the statutory maxima is the actual legal maximum total sentence, and

2. There are cases where that will ultimately set the maximum sentence under the guidelines, because the pre-clamping guidelines range for each conviction will exceed the statutory maximum,

3. There are cases where the actual sentence will exceed the guidelines sentence, and the maximum there will still be that set by the statutory limits on each offense.

To actually assess the likely range of federal sentences for a given indictment, you would need to review the indictment, the guidelines, and try to assess each of the potential guidelines factors against the allegations in the indictment (both the full set of facts alleged and the set of alleged facts that would allow conviction on each charge but produce the minimum guidelines range), and then compute what the actual sentence would be for the various possibilities under the guidelines, and then be familiar with the history of similar federal cases (especially in the same trial court and judge) and whether and in what circumstances departures from the guidelines ranges have been given, and assess that. All of which often gets into highly subjective, highly contentious, opinion, even when experts are involved. OTOH, you can not purport to report a likely sentence, and then report a maximum sentence, and you will be reporting a verifiable, clear legal fact, even if arguably one of often limited usefulness in predicting what people might like to predict.

Yes, he was.

https://www.justice.gov/media/785641/dl?inline

    I look then to Section 5G1.2{d) of the guidelines, which
    tells us that where there are multiple counts, and the
    guideline range exceeds the statutory maximum for the most
    serious count, the court must impose consecutive terms of
    imprisonment to the extent necessary to achieve the total
    punishment.
    
    There is a little bit of ambiguity, however, as to
    what is meant by "total punishment" where the guideline
    calculation calls for ~ife imprisonment, but Second Circuit
    case law makes clear that in such a situation, the district
    court is to stack or add up the maximum sentences for all
    the counts.
    
    In United States v. Evans, for example, 352 F.3d 65,
    where the guideline calculation called for life
    imprisonment but no count carried a life sentence, the
    court held that the guideline range is 240 years, the
    maximum sentences for all the counts added together.
    
    Accordingly, here the guideline range is not life
    imprisonment, but 150 years, the maximum sentences for each
    of the 11 counts added together.

None of this happened but a pardon isn’t the only way.

Doing what he's done means that there's nearly zero chance of leniency. I'd take even a small chance of leniency over a tiny chance of it.

It's also possible, as delusional as it would be, that he really thinks he's in the right here and is standing on principle. If that's the case, then he's doing what he should do.

A jury does not generally review each piece of evidence to determine whether the evidence is "valid" or not.

This evidence was already introduced at trial (in support of the conspiracy charges, since the campaign financing charge was dropped back in August). This means that the judge already reviewed it and deemed it to be valid, admissible evidence. (That happens before a trial, not after it, though in some cases may happen during a trial with the jury excused from the courtroom.)

It's weird to blindly believe OpenSecrets, which also does not link to supporting documents, while disbelieving widely-reported news reporting that has been covered for the past year...without any refutation by the Republican beneficiaries of SBF's donations.