Probability and the Kennedy Assassination

How Big a Conspiracy?

One of the worst problems of conspiracy theories is their tendency to involve dozens, scores, or hundreds of conspirators.

It's easy to see why this might happen. Given the sheer amount of evidence against Oswald, one has to either believe a lot of it or posit quite a lot of people faking evidence. Was there a clear paper trail linking Oswald to the Mannlicher-Carcano rifle? It must have been forged, and FBI investigators had to allow it to be introduced into evidence as though it was genuine.

Do three backyard photos link Oswald to the rifle, the pistol used to shoot Tippit, and to radical leftist periodicals? They must have been forged in a sophisticated government photo lab. And of course all the Dallas police officers who claimed to have discovered the photos were lying. Then, of course, an FBI expert told the Warren Commission that the photos were genuine. He must have been lying too. And the House Select Committee put a whole crew of top photo experts on the issue, and they said the photos were authentic. So they must have been lying too.

Repeat this analysis for other pieces of evidence, and for all the witnesses who gave testimony damaging to Oswald, and soon you have the proverbial "cast of thousands."

And this is just the conspiracy to cover up the fact that a conspiracy killed Kennedy, not the conspiracy to actually shoot the president.

How Large a Conspiracy?

A basic principle of conspiracies is that the larger the conspiracy, the less likely it is to be able to cohere and maintain secrecy.

Let's assume, for the sake of argument, that three people are involved in a conspiracy. They successfully kill JFK and then move on to other pursuits. Assume, further, that the probability that any single conspirator will "defect" and "blow the whistle" on his fellow conspirators is one in twenty (.05). The probability that he won't defect is therefore .95.

What is the probability then, that none of the three conspirators will defect? It's:

.95 x .95 x .95 = .857

Such a conspiracy would likely be successful. There are a bit under three chances in 20 that one or more of the members would defect.

Now let's suppose that the conspiracy has 20 plotters, each of whom must stay quiet to avoid "blowing the plot." The probability that none of them will do so is:

.9520 = 0.358

In other words, the odds are almost two to one against such a conspiracy "hanging together." If there are 30 members, the probability is:

.9530 = 0.215

A Conspiracy Too Big?
Fred Litwin's classic article, "A Conspiracy Too Big? Intellectual Dishonesty in the JFK Assassination" does what amounts to an inventory of the people who would have to have been implicated in a plot to kill JFK and then cover up the evidence.


Of course, the above analysis holds only if each conspirator's decision to defect is independent of each of the others. This is probably a servicable assumption. But it wouldn't hold if, for example, a "head conspirator" imprisoned the other 29 conspirators. In that case, if security at the prison is maintained, none of the 29 will talk, but if it breaks down all or most are likely to escape and, assuming they resent the imprisonment, likely to talk.

But this scenario begs the question of just who is going to imprison the other 29 conspirators. The "prison" must have "guards," and of course they know of the conspiracy.

It might seem more sensible to simply kill off, rather than imprison, members of the assassination conspiracy.

But the same logical problem holds for the notion of a "cleanup squad" of people going around killing off people who might "spill the beans." The members of the cleanup squad themselves become members of the conspiracy who have to be imprisoned, killed off, or allowed to run around with very dangerous knowledge.

Mysterious Deaths

Another problem for the "cleanup squad" is the need to kill people off without getting caught and leaving a trail that leads back to the chief plotters. Presumably, conspirators can hire first-class hit men, but even they are not infallable. So this leads us back to the same sort of logic that we started with. If the probability that the "cleanup squad" can successfully kill any individual and get clean away is .95, the probability that they can kill 20 people all with equal success is:

.9520 = 0.358

". . . exactly how big was the conspiracy to kill JFK? Are we talking about one assassin with an accomplice or are we talking about something larger? If one were to believe the current literature, we are faced with not just 'something larger' but a monster conspiracy that consists of several assassins, several accomplices, and the destruction and forgery of vital evidence. The critics have constructed a conspiracy so massive that it ultimately falls of its own weight." — Fred Litwin

Of course, one can posit that the conspirators are able to suborn local law enforcement officials. If this is so, this might drive up the probability of a "success" as even sloppy murders are covered up by cops. But of course this comes at the expense of even more people who know that something conspiratorial is going on. For example, assume that bribery, threats, blackmail or such applied against five law enforcement officials will drive up to .99 the probability that any given murder can be done successfully, then twenty murders can be done with the odds of success in the total enterprise being:

.9920 = 0.818

That sounds like pretty good odds for the conspiracy. But then there are 100 law enforcement officials running around who might decide to "blow the whistle." If each of them has a .98 probability of remaining silent, the probability that all of them will remain silent is:

.98100 = 0.1326

Those aren't particularly good odds. What has happened is that plotters have purchased short-term success at the expense of a high liklihood that someone will eventually speak out.

In fact, when one looks at the details of the "mysterious deaths" they usually don't seem so mysterious. If a conspiracy killed Kennedy, it's not boneheaded enough to go around killing dozens of additional people.

How Perfect a Conspiracy?

If it's implausible that a large conspiracy could keep everybody "in the know" silent for 40 years, it's equally implausible that any complicated assassination plot could work perfectly. And the conspiracists do posit very complicated plots. Most, for example, believe in shooters from two or three locations in Dealey Plaza. Most believe that shadowy agents of the conspiracy were in Dealey Plaza after the shooting. Most believe that Commission Exhibit 399 (the "Single Bullet") was planted at Parkland Hospital. Most believe there was a "coverup" involving dozens or hundreds of Dallas cops, Federal investigators, media people, and so on. What if somebody slipped up? What happens if a bullet from a rifle other than Oswald's is discovered in Kennedy's body? What happens if some of the dozens of people with cameras in Dealey Plaza that day happen to get one of the shooters in a photo? What happens if Oswald has an iron-clad alibi at the time the shots ring out? What happens if the person assigned to plant CE 399 at Parkland Hospital is seen doing it?

The probabilities involved here are obvious by now. If there is a .95 probability that each individual conspiracy task will be properly executed, the probability that every one of a dozen conspiracy tasks will be properly executed is .9512 which equals .358. As Gary Sumner has observed in his essay "Logic and the Killing Of John Kennedy":

According to well-known Murphy’s Law, if anything can go wrong, it will. Imagine all the things that could have gone wrong in attempting a difficult, dangerous operation such as killing the President of the United States. Considering the idiotic nature of the plan, the slightest mishap, the tiniest unforeseen circumstance, could have brought the operation to ruin. But nothing went wrong. The killers achieved perfection.

And since then they have successfully resisted the urge to talk about it. Various authors have postulated anywhere from a couple of dozen conspirators to several hundred. At this writing the assassination took place 40 years ago, yet no conspirator has talked. Not one has gotten drunk and revealed the murder to his wife or mistress, who has then gone to the authorities or the media. Not one has made a death-bed confession. Not one has left behind a letter of explanation in his lawyer’s safe to be opened after his death.

Think about it. These mysterious men, many of whom must not even have known one another before the plot was hatched, got together, planned and carried out the crime of the ages, in public and on television, then vanished ghostlike into history. Nobody saw them and they didn’t make any mistakes. None of them ever talked. They committed the perfect crime, using the stupidest plan imaginable, and got away with it.

This sort of logic, of course, doesn't apply if one has a very simple conspiracy theory. Why not, for example, posit that someone persuaded Lee Oswald to take his gun to work and shoot Kennedy? Why not posit that Oswald, rather than being a manipulated "patsy" who might turn on his manipulators and talk, was a witting member of the conspiracy who fully understood and agreed with its purposes? Why not posit that only a very few people, or even one person, was a part of the conspiracy with Oswald?

The reason is simple. If you do this you have to give up the notion of a Grassy Knoll shooter. You have to accept the Single Bullet Theory. You have to discard notions about faked and forged evidence. You have to let LBJ, the CIA, the FBI, anti-Castro Cubans, the Mafia and all the rest of the usual suspects off the hook. You have to quit claiming that the Warren Commission was a "coverup." Whatever the logical virtues of such an approach, it's grossly unsatisifying.


One can quibble with the exact numbers here. Should it be a 30 person conspiracy or a 75 person conspiracy? Is the probability that any given conspirator will remain loyal .90 or .95 or .99? But the general outlines of the issue are clear. Does this analysis rule out all conspiracy theories? No, but it pretty much does rule out the ones that are any fun.
Return to Kennedy Assassination Home Page