The Explainer: Standards of evidence and shifting burdens of proof

A reader wants to know how much proof is required to convict an athlete on a doping charge.

Dear Explainer,
As a follow-up to your recent articles on the Contador controversy, I have to ask what burden of proof is required to substantiate doping claims and lead to a suspension?
Thanks much!
John Marshall
Columbia, Missouri

Hello John,
That question is an interesting one when it comes to doping cases, but let’s start with a minor distinction. “Burden of proof,” roughly means who has to present a case and “standard of evidence” is the measure used to determine how strong of a case that has to be.

Which standard will these men need to meet to prove to 'reasonable' people that the accused is a witch?
Which standard will these men need to meet to prove to 'reasonable' people that the accused is a witch?

In keeping with the legal profession’s inexplicable penchant for Latin, we can begin the discussion with the phrase “semper necessitas probandi incumbit ei qui agit,” which basically means if you level a charge be prepared to prove it. The burden of proof initially rests on the charging party.

But just what and how much do they actually need to “prove?” By what standard is their evidence judged? There is the old non-judicial – and nearly impossible to meet – standard of “beyond a shadow of a doubt,” but you won’t find that applied in any court. If you think of the old “scales of justice” model, everything would be on the prosecution’s side of the scale. The burden would be so onerous that even DNA evidence wouldn’t pass, because a one-in-four-billion chance that the profile would belong to someone else might qualify as that “shadow.”

Instead, we apply a somewhat looser series of standards of evidence, depending on the case at issue.

There’s proof and then there’s proof

The classic standard of evidence with which we are most familiar is the “beyond a reasonable doubt” applied in criminal cases in the U.S. and several other – but by no means all – established democracies. It is regarded as the highest burden of proof one side must meet in presenting a case.

In order to convict a defendant in criminal trials in the U.S., you need to present sufficient evidence so that there exists “no reasonable doubt in the mind of a reasonable person that the defendant is guilty.”

In most civil cases, the standard is much, much lower. The finder of fact – a judge or a jury – need only find that the charging party has presented a “preponderance of evidence” to show that a balance of probabilities would suggest that his case is more likely to be true than not.

There is also the intermediate “clear and convincing” standard, which requires that the truth of the assertion is substantially more likely than not. That’s used in some civil cases and criminal appeals.

But how do you quantify those? We can start by pointing to the “preponderance” standard, because it’s probably the simplest. Since it uses “more likely than not” as a guide, the best way to describe that is if your case is more than 50 percent believable, you win. The scale need only be ever-so-slightly off-balance.

There is a great deal of debate, however, about what percentage of proof might apply to the others. Is “beyond a reasonable doubt” indicative of a 90-percent degree of certainty? Is it 80 percent? Maybe 70 percent? Whatever that is, the “clear and convincing” standard falls somewhere between the other two.

It’s a subjective standard and we’re trying to put a number to it. As with most unanswerable questions, there is always someone willing to try and answer it. There have been a number of attempts to place those burdens on the scale of probability, with judges noting what those standards are in their rulings and academics wrestling with the question.

I think one of the better recent efforts was described a few years ago in the Valparaiso University Law Review. If you’re into that sort of thing, check out “Distributions of Interest for Quantifying Reasonable Doubt and Their Applications.” (I like to read it to my kids at bed time.)

Looking for Mr. (and Ms.) Reasonable

The problem, though, is that application of those standards depend on human beings, whose biases, foibles and personal histories all come into play. Legal minds try to address that issue by assuming that most of these decisions are made by the ever-elusive “reasonable person.”

That “Reasonable Person” is someone who pops up in most areas of modern common law. Mr. Reasonable decides guilt or innocence. Ms. Reasonable’s behavior is used to establish the standard of care in negligence cases. In contract law, the existence of an agreement is often judged by the standard of a “reasonable observer.”

I’ve actually never met a completely Reasonable Person, but if you do, show them proper deference. Mr. and Ms. Reasonable are apparently the most powerful figures in our legal system.

Proof in doping cases isn’t just reasonable, it’s comfortable

So what happens if your favorite – or least favorite – rider trips the old Dope-O-Meter®?

Reading the World Anti-Doping Code gives us a whole new standard. Article 3 of The Code notes that “the standard of proof shall be whether the Anti-Doping Organization has established an anti-doping rule violation to the comfortable satisfaction of the hearing panel.”

Okay, but what does it take to make a panel – or at least two of its three members – comfortably satisfied? Fortunately, The Code goes on to explain that “this standard of proof in all cases is greater than a mere balance of probability but less than proof beyond a reasonable doubt.”

If you try to pin that down on the standards we discussed earlier, we’re looking at something akin to the American “clear and convincing” standard.

In a doping case, though, there are some things that the prosecution need not prove. For example, if an A sample comes back positive and that finding is confirmed by a test of the B sample, The Code allows the prosecution to move forward on the presumption that the test was conducted properly. The mere introduction of the test result shifts the burden of proof to the defense.

Shifting burdens

As you might recall, the bulk of the defense mounted by Floyd Landis was based on the laboratory’s alleged violation of the International Standards for Laboratories. The Code allows for such a defense to be raised and imposes only a “preponderance of the evidence” burden upon a defendant making that charge. Indeed, in 2005 Iñigo Landaluze successfully appealed his conviction on charges that he tested positive for testosterone at the Dauphiné Libéré that year by doing just that.

At the time, that was all Landaluze needed to show. According to a Court of Arbitration for Sport decision, Landaluze demonstrated by a preponderance of the evidence that the testing lab had violated its own procedures. At that point, the burden shifted back to the prosecution (in this case, it was the UCI) to show that those procedural errors had not affected the outcome of the test (that it wasn’t a prejudicial error). All the UCI needed to do then was to satisfy the lower “preponderance” standard and show that the error was not prejudicial.

In reviewing the transcript of the hearing, however, the CAS panel noted that the prosecutors for the UCI merely stood up and declared that the error wasn’t prejudicial.

CAS found that the UCI had not met that burden by simply stating that assertion; they should have provided evidence to show that the lab’s departure from acceptable practices did not ultimately mean the results themselves were not flawed.

The CAS panel then concluded that even though Landaluze was quite likely guilty, the UCI’s failure to meet its burden in the original hearing meant that he had to be acquitted.

“The arbitrators have also specified that even though Iñigo Landaluze benefited from this flawed procedure to be acquitted, the CAS decision does not constitute a declaration of his innocence.”

In other words, the CAS panel concluded that it was better to let a guilty man go free than to put the integrity of the entire system at risk. Score one for the sanctity of law!

Well, the decision sent shockwaves through the system and WADA took the decision to heart. Yup, they sure did. At the next World Conference on Doping in Sport, delegates approved a change to The Code, leaving the burden of proof on the defendant to show that not only was there an error, but that it was prejudicial. Normally a fan of WADA and its mission, this one underscores my big separation-of-powers gripe about WADA. The organization has essentially granted prosecutors the legislative authority to change the rules after losing a case.

There are two things to keep in mind when it comes to the lab defense raised in USADA v. Landis. First, that case was prosecuted under the old rule, meaning that the burden shifted back to USADA to show that the error didn’t affect the outcome. Second, in reviewing the case, CAS found that USADA had met its burden in showing that none of the departures from standard procedures affected the actual outcome of the test.


So, what does this mean in the case of Alberto Contador? For one thing, he isn’t challenging the results of the test. He’s acknowledging that there were, indeed, low levels of clenbuterol in his system.

If it were to stop there, that would be tantamount to a guilty plea. The prosecution wouldn’t have to do much beyond that. But Contador will – assuming the case actually goes before a panel – be raising an affirmative defense that the clenbuterol came from tainted beef. He bears the burden of proof, but need only meet the lower standard of proof to make that claim and shift the burden back to the prosecution to refute.

Of course, even if he does succeed, The Code – as we discussed in earlier columns – would still allow him to be penalized. Much of that will be left to the discretion of the panel.

“The Explainer” is a regular feature on If you have a question related to the sport of cycling that our editors might be able to answer, feel free to send your query to and we’ll take a stab at answering. Not all letters will be published and some questions may be combined with those of other readers. Please include your full name and hometown.