This is the argument:

If someone makes consistently bad decisions, then it must be intentional, because if you were to make decisions at random, some of them would be good.

At first, I found this argument compelling, but the more I think about it, the more flawed it seems. Here’s my take:

  1. A “good” or “bad” outcome to a decision is completely relative. What’s a good outcome for the decision maker, may not be for anyone else.
  2. The consequences of a decision cannot be fully seen ahead of time. What the decision maker may believe to be a good outcome for them, may not come to pass or even turn out to be a bad outcome for them.
  3. The “no better than random” argument doesn’t apply here. This is used in science where the outcome (hypothesis) is tested against the null hypothesis to ascertain a statistically significant difference (or otherwise). The hypotheses are deduced by observation. In the decision maker’s case, it’s the other way around in terms of sequencing.

As an aside: If it’s the decision maker’s job to make the best decision for some party, and they consistently fail to do that, then that’s incompetence, not malice.

Is the original argument a logical fallacy? Does my analysis stand up; particularly my third point, which I only have a layperson’s insight on? Are there any other flaws in the original argument?

  • 1
    This looks like some argument from ignorance fallacy, obviously one doesn't make decisions completely at random and there's other possibilities as you rightly inferred such as incompetence... Oct 10, 2022 at 3:01

4 Answers 4


Is the implication that making consistently bad decisions must be intentional a logical fallacy?

Yes. It is a basic non-sequitur, in that the conclusion does not necessarily follow from the initial observation.

In this case, the initial observation is of a person making 'consistently bad decisions'.

As the OP and answers to date have noted, there are many reasons why a person might make consistently bad decisions, not one.

Therefore, it is fallacious to conclude that any one reason (ie. intentionality) must be the reason. It is a conclusion that does not rest on evidence provided by the initial claim, but upon an unsubstantiated assumption. In other words, the conclusion does not logically follow from the initial premise.

It's worth noting that a conclusion arrived at by means of non-sequitur can nonetheless be correct. Fallacies typically identify the means via which a conclusion is reached, not the accuracy of the conclusion.


Is the implication that making consistently bad decisions is intentional a logical fallacy?

This question is a hard one. Sometimes it seems yes, but other times no.

The IRS notes that most arithmetic errors on tax returns would result in a lower tax bill, and thus favor the taxpayer. Intentional? Probably. Here, there is an obvious self-interest and that factor helps guide the conclusion that yes, the nominal errors are indeed intentional.

At the other end of the spectrum are the bad decisions made by the perpetrators on the true crime television shows. Notwithstanding an intense self-interest in not getting caught, the perp will often make a series of lousy decisions which results in their identification and capture, often within a few days. The errors cannot be intentional. Here, the anxiety and confusion after a murder override the process of rational decision-making.

After all that, my answer is yes, it is fallacious, because both of my examples require additional factors to explain the outcome. Without those factors, nothing follows from the single premise of consistently bad decisions. If I had to pick a fallacy, I would call this one Hasty Generalization.

  • Errors that lead to a perp getting caught are not random, they are heavily filtered. A crime usually requires multiple errors before the cops have even a mild chance of catching the perp. It is a form of reverse survivor selection. Only those who get caught make big stacks of errors. The ones who don't make so many errors tend to get away with it.
    – BillOnne
    Oct 11, 2022 at 1:58

It is not automatically a fallacy. It is not something that can have a hard-and-fast answer. The estimate of the chance that the errors happened in the way they were observed is complicated in the extreme and subject to many errors, biases, and filters.

You would need to examine the types of errors and their expected range, the possible effects of errors in the expected range, survivor bias, and a bunch of other things I'm forgetting just now.

Just as an example, consider my comment on @Mark Andrews's answer. Perps that make very few mistakes tend not to get caught. So those who get caught will tend to have made lots of mistakes that increase their chance of getting caught.

A lovely example of survivor bias comes from data about damage to aircraft in WW2. But it is a long side-bar. Please read at the link if you are interested.

So data about errors needs to be extremely carefully examined. Maybe there is some "innocent" reason for the errors to be biased. Innocent in that the person is not doing them deliberately.

One trivial example: Consider a guy who has a lot of car crashes. When driving, any error steering is going to have a high chance of creating a crash. You would need to examine these quite carefully to conclude they were deliberate.

So, if the person had some neurological disorder that causes improper hand-eye coordination, the errors might not be deliberate. The subject might be unable to do any better. (Though, if he knows he has the condition, we might still hold him culpable for continuing to drive.)

If, however, he was constantly racing, driving much too fast, driving on the wrong side of the road, passing on the shoulder, following too close, etc., then these sorts of errors would be hard to interpret any other way than deliberate. That is, you must characterize the type of error and have some information about the origin of the error.

There are some ways that the data can be characterized fairly convincingly. Consider that, at some future point, the POTUS is from the Stupid Party. And the media is constantly making mistakes that make him look bad. Dozens of mistakes, day after day, of all different types. Mis quoting, mis attributing, getting dates wrong, reporting false facts, etc. etc. Then there is an election and the Stupid Party is out, and the the Evil Party is elected for their turn. Now the same media continues to make about the same number of mistakes per week. But this time, they are reporting things in a way that makes the POTUS look good.

Because it is the same exact people making similar mistakes but with the opposite polarity, it is difficult to believe that they could be doing it accidentally. When they report "the chocolate ration fell" during the Stupid Party administration (and it actually stayed the same), but report "the chocolate ration rose" during the Evil Party admin (and it still stayed the same) it is difficult not to turn to malice for an explanation. The only factor that changed was which party was in power. It was similar data in a similar situation about a similar subject. Bias seems inescapable.

There is one more caveat to keep on your dashboard. There is probability involved here. There is no hard threshold past which you conclude it is deliberate. You still cannot mind read. You must examine the data on the basis of the probability you conclude from the data and stay as open minded as the probability justifies. If it's 5 coin tosses that all came up "heads" it is still only 1 in 32. You begin to be mighty sus. But it is not logical proof. Not good enough for criminal court, but likely good enough for civil law suit.

  • "It is not automatically a fallacy. It is not something that can have a hard-and-fast answer. " THAT is a fallacy. Like in deductive reasoning you're meant to produce hard-and-fast rules, you have a set of premises and the conclusion MUST follow from them and every failure to do so is a fallacy. Of course in real world scenarios you encounter inductive arguments where it's just a weak or strong claim but not a certain one, but in that case you're not speaking of fallacies as it's about statistics not logic.
    – haxor789
    Oct 11, 2022 at 10:27
  • @haxor789 claims it must have a hard and fast answer. Ignores the many pages of explanation of why it cannot have a hard and fast answer. Mouths platitudes about statistics that are already in the answer. Gets ignored.
    – BillOnne
    Oct 11, 2022 at 15:10
  • It has a hard-and-fast answer and that answer is yes. As you've enumerated yourself cases in which it doesn't have to be intentional but can have other reasons, it should be quite obvious that the implication of "must it" can be answered with no. Meaning it's a logical fallacy. Now is it always wrong because of that? No that question does not have a hard-and-fast answer as you've stated, but that is a different question.
    – haxor789
    Oct 11, 2022 at 15:34

Yes it obviously is a fallacy see Futilitarian's answer.

Just to give a few more examples where it breaks down (though mind you ANY of these and the other answers is proving it's a fallacy as you just need one counter example where it goes wrong).

First of all "decision making" is not a generalizable task. Like it could be multiple choice, no-choice, infinite choice, aso and it could be about decisions that matter, that matter to you, that matter to others, that don't matter and on topics that you have low, high or medium knowledge and that's all before you argue whether the outcome is good or bad. So the idea of a "good" or "bad" decision maker is already kind of a flawed.

The more you are experienced the more it's not a decision as much as following an algorithm and the less experience you have the more it is about confidently choosing between more or less arbitrary options, where the outcome is likely equally arbitrary.

But also the more generalized version of this argument: "Having consistent results implies that you're following a plan" is not necessarily correct either. Like you could employ the same strategy of risk taking or risk avoidance in different scenarios and despite doing the same get vastly different outcomes or the other way around you could do something else every time and get the same outcome. Also it's not clear whether the pattern that you've found is on your side or on the side of the external scenario. Like if you play the lottery every Sunday and always lose that's not your intention the game is just statistically rigged against you. So in a sense it IS intentional but not on your end. And in terms of forces of nature, it might not even be "intentional" concerning any known agent.

But even the "if things happen often, you've found a pattern" generalization doesn't really work because that streak might just have happened to be a fluke. Like just because a fair coin shows a 50:50 pattern after a lot of tries doesn't mean you can't have streaks of 20 times the same side in a row. It's statistically rarer than shorter streaks but it's not impossible. And you might just happen to live in that edge case.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .