# Possibly apocryphal worst estimate of probability?

**URL:** <https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833>\
**Category:** Factual Questions\
**Created:** [January 23, 2020, 3:23am UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833 "2020-01-23T03:23:49Z")\
**Posts on this page:** 18\
**Page:** 2

<div class="post-metadata">

**Author:** ![Andy\_L](https://avatars.discourse-cdn.com/v4/letter/a/c67d28/32.png) [@Andy\_L](https://boards.straightdope.com/u/Andy_L)\
**Post date:** [January 23, 2020, 2:42pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/21 "2020-01-23T14:42:50Z")

</div>

> [@ftg](#):
>
> The weird claim regarding male-male breeding cited by the OP’s source is at odds with what I’ve been seeing in recent years. Namely, that it is the men who will die out (fairly soon) and women will be reproducing without them.
> 
> The earliest (serious?) cite I see regarding this is a book called [Adam’s Curse: A Future Without Men](https://en.wikipedia.org/wiki/Adam's_Curse) from 2003. It says that shrinking y chromosomes mean human males only have about 5000 generations (or ~125,000) years before they go bye-bye.
> 
> Some variation of this has been touted regularly. Often with the number of years given as drastically less.
> 
> Of course the fact that hominid males, never mind mammalian males, have been around far longer than that, which suggests that dying out is not a likely occurrence, doesn’t seem to matter.
> 
> .

Some mammals have developed sex-determination processes that don’t involve a Y-chromosome, possibly because an ancestral species lost its Y chromosome via shrinkage [Tokunoshima spiny rat - Wikipedia](https://en.wikipedia.org/wiki/Tokunoshima_spiny_rat)

---

<div class="post-metadata">

**Author:** ![brossa](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/brossa/32/3040_2.png) [@brossa](https://boards.straightdope.com/u/brossa)\
**Post date:** [January 23, 2020, 3:02pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/22 "2020-01-23T15:02:19Z")

</div>

OP here -  
My thanks to Senegoid for the hint that I might not be imagining the whole thing, and particular thanks to Dr. Strangelove for posting a link to an example of this fallacy. The Daily Show bit is really close to my hazy memory, but it’s from ten years ago, and the non-uterus haver pregnancy setup came from Oliver, not the nutcase. It’s quite possible that my memory is bad and this is the actual clip I was misremembering, but I’m holding out hope that there’s a better fit out there somewhere.

I appreciate the ancillary discussion of probability and such, and long may it continue, but can anybody come up with additional concrete examples of people committing this sort of blunder publicly in interviews or speeches or such? It wouldn’t need to specifically be about non-uterus haver pregnancy, although that is the gold standard…

---

<div class="post-metadata">

**Author:** ![Buck\_Godot](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/buck_godot/32/6573_2.png) [@Buck\_Godot](https://boards.straightdope.com/u/Buck_Godot)\
**Post date:** [January 23, 2020, 3:25pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/23 "2020-01-23T15:25:19Z")

</div>

> [@DPRK](#):
>
> 50:50 is a prior probability estimate where you have _absolutely no idea_ which alternative is more likely, or when you really believe they are equally likely. This is discussed a little bit as the “[principle of indifference](https://en.wikipedia.org/wiki/Principle_of_indifference)” and [uninformative priors](https://en.wikipedia.org/wiki/Uninformative_prior) which cite how it is centuries old.
> 
> ETA there is an interesting discussion there of other probability distributions representing uncertainty such as the Haldane prior and Jeffreys prior.

Certainly a flat 50/50 prior is useful as a starting point in Bayesian statistics, but it is only that a starting point, which you modify in the presence of actual data.

One could argue that without data a 50/50 is a valid point estimate, but it has rather large error bounds, on the order of +/-50% 😉

---

<div class="post-metadata">

**Author:** ![RedSwinglineOne](https://avatars.discourse-cdn.com/v4/letter/r/a9adbd/32.png) [@RedSwinglineOne](https://boards.straightdope.com/u/RedSwinglineOne)\
**Post date:** [January 23, 2020, 7:41pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/24 "2020-01-23T19:41:31Z")

</div>

> [@scudsucker](#):
>
> Choosing two tickets with the same numbers is actually a winning strategy, one of the few when playing the lottery.
> 
> If you win, against enormous odds, and someone else also - again, against enormous odds has the same numbers, you will get 2/3 of the prize, and they will get 1/3.

How do you figure that this is a winning strategy? :dubious:

---

<div class="post-metadata">

**Author:** ![Bill\_Door](https://avatars.discourse-cdn.com/v4/letter/b/50afbb/32.png) [@Bill\_Door](https://boards.straightdope.com/u/Bill_Door)\
**Post date:** [January 23, 2020, 7:51pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/25 "2020-01-23T19:51:34Z")

</div>

> [@RedSwinglineOne](#):
>
> How do you figure that this is a winning strategy? :dubious:

I think the best strategy would be to keep all of your picks higher than the number 31. People who pick their own numbers tend to concentrate on “birthday numbers”, so choosing numbers that don’t match up with calendar dates should help your chances of an unshared win.

---

<div class="post-metadata">

**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\
**Post date:** [January 23, 2020, 8:25pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/26 "2020-01-23T20:25:15Z")

</div>

> [@Wendell\_Wagner](#):
>
> Anyone who says that something could happen or it could not happen means that the probability of it happening is 50/50 doesn’t remotely understand probability.

I’ve normally heard that in a facetious/jocular context, i.e. not to be taken seriously as a probability estimate. It’s a bit of a joke. Now, are there some people dumb enough to believe it? I’m sure there are, but when I’ve heard it, it’s not been meant to be taken seriously.

---

<div class="post-metadata">

**Author:** ![kaylasdad99](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/kaylasdad99/32/3398_2.png) [@kaylasdad99](https://boards.straightdope.com/u/kaylasdad99)\
**Post date:** [January 23, 2020, 9:26pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/27 "2020-01-23T21:26:50Z")

</div>

> [@Bill\_Door](#):
>
> I think the best strategy would be to keep all of your picks higher than the number 31. People who pick their own numbers tend to concentrate on “birthday numbers”, so choosing numbers that don’t match up with calendar dates should help your chances of an unshared win.

What about if I call dibs on the whole thing? I don’t see why THAT should be deprived of legal force.

---

<div class="post-metadata">

**Author:** ![septimus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/septimus/32/410_2.png) [@septimus](https://boards.straightdope.com/u/septimus)\
**Post date:** [January 24, 2020, 6:15am UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/28 "2020-01-24T06:15:45Z")

</div>

> [@scudsucker](#):
>
> Choosing two tickets with the same numbers is actually a winning strategy, one of the few when playing the lottery.
> 
> If you win, against enormous odds, and someone else also - again, against enormous odds has the same numbers, you will get 2/3 of the prize, and they will get 1/3.

No. It is better to play two numbers (call them #1 and #2) rather than to bet twice on #1.

If nobody else bets the winning number, you are twice as likely to win the whole thing. If one other person bets the winning number, you have two chances at one-half the jackpot (one-half plus one-half sums to one net chance) rather than one chance at two-thirds the jackpot. If two other people bet the winning number, you have two chances at one-third (summing to two-thirds) rather than one chance of two-fourths. And so on.

```auto
I knew a guy who stated the "Either yes or no, so odds are one-half" as a demonstration that he couldn't think in probabilities. I'm not sure how serious he was. But the guy was smart, rising from technician to Engineering V.P. of a fair-sized company. He invented the "Zero Propagation-Delay Latch." I'm sure this invention wasn't novel, but I'd never heard of it before or since. (Any digital circuit designers reading this: Do you know what a Zero Propagation-Delay Latch is?)
```

---

<div class="post-metadata">

**Author:** ![don\_t\_ask](https://avatars.discourse-cdn.com/v4/letter/d/e68b1a/32.png) [@don\_t\_ask](https://boards.straightdope.com/u/don_t_ask)\
**Post date:** [January 24, 2020, 6:54am UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/29 "2020-01-24T06:54:45Z")

</div>

In Richard Feynman’s [What Do You Care What Other People Think?](https://www.amazon.com/What-Care-Other-People-Think/dp/0393320928) he discusses the difference between management and engineers in estimating the likelihood of failure of a Challenger space shuttle flight. Management were confident that the chance of shuttle failure was 1 in 100,000 but the engineers estimated it to be closer to 1 in 100 or 1 in 200, and were crossing their fingers during every flight.

---

<div class="post-metadata">

**Author:** ![Mangetout](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mangetout/32/19_2.png) [@Mangetout](https://boards.straightdope.com/u/Mangetout)\
**Post date:** [January 24, 2020, 8:20am UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/30 "2020-01-24T08:20:11Z")

</div>

> [@bob\_2](#):
>
> On that basis, I should start buying lottery tickets because, since I will either win or not win, it will be a 50/50 chance.

I have encountered this one in the wild - a guy I used to work with bought lottery tickets saying exactly: “you’re either going to win it or you’re not; it’s 50/50” (and he was not intentionally being funny or anything).

---

<div class="post-metadata">

**Author:** ![Wendell\_Wagner](https://avatars.discourse-cdn.com/v4/letter/w/8491ac/32.png) [@Wendell\_Wagner](https://boards.straightdope.com/u/Wendell_Wagner)\
**Post date:** [January 24, 2020, 8:33am UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/31 "2020-01-24T08:33:14Z")

</div>

pulykamell, I’m not sure if I’ve ever heard anyone in conversation say that the probability of something happening is 50/50 when all they meant is that it may or may not happen but they had no idea what its probability was. They certainly wouldn’t say it just as a joke. That doesn’t mean that you might hear people say such things, but that doesn’t mean that your experience or my experience is typical of how most people talk. I think we need to be careful not to take our experience in conversations as being typical of everyone else’s experience in conversations. In any case, anything I’ve written in this thread is about what the best way to make predictions works, not about how it’s done in random conversations.

---

<div class="post-metadata">

**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\
**Post date:** [January 24, 2020, 12:42pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/32 "2020-01-24T12:42:46Z")

</div>

> [@Mangetout](#):
>
> I have encountered this one in the wild - a guy I used to work with bought lottery tickets saying exactly: “you’re either going to win it or you’re not; it’s 50/50” (and he was not intentionally being funny or anything).

Yes, that’s exactly the kind of usage I encounter it in. I also don’t believe that that person really believes they have a 50% chance of winning.

---

<div class="post-metadata">

**Author:** ![Hari\_Seldon](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/hari_seldon/32/5173_2.png) [@Hari\_Seldon](https://boards.straightdope.com/u/Hari_Seldon)\
**Post date:** [January 24, 2020, 1:35pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/33 "2020-01-24T13:35:41Z")

</div>

> [@Buck\_Godot](#):
>
> Certainly a flat 50/50 prior is useful as a starting point in Bayesian statistics, but it is only that a starting point, which you modify in the presence of actual data.
> 
> One could argue that without data a 50/50 is a valid point estimate, but it has rather large error bounds, on the order of +/-50% 😉

Sure but if you Bayesian statistics you have to modify your results by statistics and taken over the entire history of humanity, the probability asymptotically tends towards 0 and very quickly.

---

<div class="post-metadata">

**Author:** ![Mangetout](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mangetout/32/19_2.png) [@Mangetout](https://boards.straightdope.com/u/Mangetout)\
**Post date:** [January 24, 2020, 2:16pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/34 "2020-01-24T14:16:41Z")

</div>

> [@pulykamell](#):
>
> Yes, that’s exactly the kind of usage I encounter it in. I also don’t believe that that person really believes they have a 50% chance of winning.

No, although they probably did vastly overestimate their chances of winning (evidenced by the fact that they participated in a lottery), just not all the way to 50%.

---

<div class="post-metadata">

**Author:** ![gnoitall](https://avatars.discourse-cdn.com/v4/letter/g/bb73d2/32.png) [@gnoitall](https://boards.straightdope.com/u/gnoitall)\
**Post date:** [January 24, 2020, 6:03pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/35 "2020-01-24T18:03:33Z")

</div>

To be contrarian, I don’t thing the naïve (and probably intellectually dishonest) “two alternatives, must be 50% odds” estimate is the most egregious.

We’ve all experienced the “That will never happen, you can’t prove to me it’ll ever happen, I am just going to assume it’ll never happen so I can avoid thinking about it or expending resources on contingency measures or insurance or any of that.”

“We’ll assume it can’t happen, because if it does there’s nothing we can do about it anyway” is a very close second.

---

<div class="post-metadata">

**Author:** ![Fotheringay-Phipps](https://avatars.discourse-cdn.com/v4/letter/f/d9b06d/32.png) [@Fotheringay-Phipps](https://boards.straightdope.com/u/Fotheringay-Phipps)\
**Post date:** [January 24, 2020, 6:34pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/36 "2020-01-24T18:34:26Z")

</div>

> [@pulykamell](#):
>
> I’ve normally heard that in a facetious/jocular context, i.e. not to be taken seriously as a probability estimate. It’s a bit of a joke. Now, are there some people dumb enough to believe it? I’m sure there are, but when I’ve heard it, it’s not been meant to be taken seriously.

I’ve encountered people who used the term in this context and meant it seriously. (Generally it’s people who have bought a lottery ticket on a rare occasion and are trying to feel good about it.) But it wasn’t that they misunderstand probability so egregiously. They were simply misunderstanding and misusing the term “50/50”.

None of these people genuinely thought it was as likely that they win as that they don’t win. They just thought that “50/50” simply meant that one of two possibilities was going to happen. To the extent that misunderstanding probability factored into it, it was in thinking the fact that there were only two possibilities was of any significance at all in terms of the overall likelihood.

I think the way it worked is like this. These people heard “50/50 chance” used by other people to mean something that has a high likelihood of happening. And they thought that anything with only two possibilities is a “50/50 chance”. So they thought in their own case they had a high likelihood of winning, since it’s a “50/50 chance”. They did not appreciate that “50/50 chance” means that one possibility is as likely to happen as not, which would not apply in their case.

---

<div class="post-metadata">

**Author:** ![dauerbach](https://avatars.discourse-cdn.com/v4/letter/d/c2a13f/32.png) [@dauerbach](https://boards.straightdope.com/u/dauerbach)\
**Post date:** [January 24, 2020, 8:52pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/37 "2020-01-24T20:52:53Z")

</div>

I think it is more useful, rather than 50/50, which at least my brain translates as a coin toss, is to realize the personal probability is actually 0% or 100%. And you will likely decide that one outcome is much more likely than the other.

---

<div class="post-metadata">

**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\
**Post date:** [January 24, 2020, 9:43pm UTC](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833/38 "2020-01-24T21:43:05Z")

</div>

> [@dauerbach](#):
>
> I think it is more useful, rather than 50/50, which at least my brain translates as a coin toss, is to realize the personal probability is actually 0% or 100%. And you will likely decide that one outcome is much more likely than the other.

It is typical to interpret the probability of an event as its [relative frequency](https://en.wikipedia.org/wiki/Frequentist_probability) in a sequence of repeated trials, but it is also legitimate to express a degree of belief as a probability, as in [Bayesian probability](https://en.wikipedia.org/wiki/Bayesian_probability), even if it is just a specific event which will either occur or not occur.

[Previous page](https://boards.straightdope.com/t/possibly-apocryphal-worst-estimate-of-probability/846833.md?page=1)
