Talk:Appeal to probability/Archives/2012

This article is misleading

It is not a good idea to invent new names for logical moves and/or fallacies simply for use on Wikipedia. Insofar as there is a fallacy discussed here (and some of the examples in the article are indeed fallacious), "Appeal to probability" is not a helpful name for it. If the intention was to state that the inference from "P is probable" to "P" is not an entailment, then the point is trivial. But calling this a fallacy strongly suggests that it is epistemically illicit (see the cross-link in the article on "Moroni," for example), whereas accepting highly probable propositions is a paradigmatically reasonable thing to do. (Please, no one bring up the lottery paradox here until you've read Kyburg's "Conjunctivitis.") The argument used elsewhere in the discussion of this article regarding the probability of a person's dying and the number of people who do die daily misses the point that if we are talking about one (randomly-selected) individual about whom we know nothing else pertinent to his living out the day, survival is indeed the way to bet. Of course non-monotonic inference runs risks, but that fact alone does not qualify such inferences as fallacies.

In any event, the further discussion of modal fallacies is not germane since it does not involve an appeal to probability, only to possibility.

66.227.161.249 00:15, 23 October 2007 (UTC)

I'm dubious that this is a real fallacy and the examples aren't appropriate. For example, claiming there's a fallacy because there is a microscopic chance of the claim being false? Isn't that in itself something like the nirvana fallacy? That is, the argument isn't perfect in that it allows a low probability alternative, hence it is a fallacy (the true nirvana argument would go a bit further and claim the argument is false)? In the firewall example, there's no time limit given in the claim, so if there's a finite probability per unit time of being hacked, then a hack will eventually occur since the probability is 1. But the problem is that no one is going to run their server forever.

At best, you can say it's a fallacy due to oversimplification. Something like "Don't run with those scissors, you'll hurt yourself!"

The third example is a "correlation means causation" fallacy combined perhaps with a little something else (the statement implies soccer is correlated with hooliganism, which sounds to me a bald assertion or maybe proof by media coverage). There's no implication of probability there.

The second example seems to be the only true example of the fallacy given in this article. You have a combination of an extremely unlikely event even over a long life span combined with a negative payout. A similar example might be "Don't wear seatbelts because you could be pinned in the vehicle." That ignores that as far as we know, even if automobile accidents were the sole cause of death for us (and assuming we kept driving rather than more sensible act of completely avoiding the only way we could die), then you're far more likely to die of something other than pinned by a seatbelt, if you always fasten your seatbelt. And you're more likely to die, period, if you don't wear seatbelts. In other words, the probability isn't inevitable despite having both a positive probability and an indefinite span of time. -- KarlHallowell (talk) 20:47, 10 November 2009 (UTC)

From the other side?

That is, "if something that has happened did not happen, it would never have happened in the past of future", bad example but hopefully you get the idea: "if the shoe was not invented, we would not have the sock". Is there a name for this fallacy? Ljcrabs (talk) 16:57, 1 October 2009 (UTC)

Notability

I removed the sources tag, as it stated that there was a list of references (which there wasn't). The article incorrectly appeals to modal logic: modal-possibility isn't statistical-probability. Unless there are some academic references, I will recommend this article for deletion. 203.129.44.15 (talk) 11:09, 19 October 2009 (UTC)

I just looked for sources, and couldn't find any, other than copies of Wikipedia; this doesn't actually seem to be a notable fallacy. I don't think it's even a common fallacy: when have you ever heard someone claim that because something could happen, it must happen? This seems like something someone made up, and I agree, it should probably be nominated for deletion. (For the record, I'm not the same person as the IP above.) Robofish (talk) 23:17, 22 October 2009 (UTC)

I've checked as well and none of the fallacies used for appeal to probability seem to be unique to the premise of the fallacy; that because something can happen, it will. It is also poorly written. The second sentence of the introduction, beginning with "Although a fallacy when used in absolute terms...", has little to do with the fallacy. I don't think it's necessary to propose that probability is a valid school of thought considering it's widespread acceptance in science. Likewise the examples can all be attributed to other fallacies, the first one being gambler's fallacy, the second is correlation does not imply causation, and the third is affirming the consequent (not correlation does not imply causation as noted in the page). The links to infinite monkey theorem and Murphy's law are tangentially related but I do not feel warrant an entire poorly-written page. Until some sort of academic source can be found linking these ideas into a specific fallacy I propose this page be deleted. HunterIV4 (talk) 22:36, 5 February 2012 (UTC)

Vanishingly small

As far as I can tell, the core idea of this article is as follows: "Proving that something is vanishingly small does not prove that it is zero." But the article appears to ignore the numerous cases where proving that something is vanishingly small produces useful results by itself, such as the Heisenberg uncertainty of an object with a large enough mass that the classical mechanics apply to within measurement error. To put it another way:

Possibly the sun won't explode tomorrow.

Therefore, the sun won't explode tomorrow.

This is strictly a fallacy, but because the chance of the sun dying is vanishingly small, the fallacy is a useful approximation. --Damian Yerrick (talk | stalk) 20:06, 20 November 2009 (UTC)

Long Winded

The first example is unnecessarily lengthy going into details about gambling behaviors. They need to be short and sweet to narrow down what the precise characteristics of the fallacy are. --99.48.61.207 (talk) 15:57, 4 September 2011 (UTC)

My bad. I was amused at the irony of the counter-argument's appeal to probability. I've actually got quite a few edits I never finished where I pair that back down. Pithy takes me awhile.—Machine Elf ¹⁷³⁵ 01:08, 5 September 2011 (UTC)

If you say "My bad" then either fix it or let someone else fix it. The behavior has nothing to do with this article. — Preceding unsigned comment added by 87.22.38.235 (talk) 05:31, 26 January 2012 (UTC)

You're bad? It's incredibly naive to claim anyone confuses probability with certainty that way, least of all anyone who's ever rolled a pair of dice. I'd call it misguided WP:OR, not a simpler “example”. As I explained in my edit summary, the description of what constitutes problem gambling makes numerous appeals to probability, much more so than the gambler, who would appeal to anything but… As I also explained in my edit summary, the IP is just flat-out wrong that “it is theoretically possible to play the lottery an infinite number of times and never win”. Furthermore, unless we're talking nonsensically about a “lottery” that's impossible to win, it doesn't follow that it's somehow “invalid” to characterize winning as an optimistic or best-case scenario.

I'm not preventing anyone from editing. What I said above, is that I agree it should be paired down, not that it should be removed. Would the “behavior” you mention be ′gambling’ or ‘describing problem gambling’? My comment was regarding the later, but the term “behavior” implies the former. Are you suggesting it's somehow self-evident that statistical analyses aren't vulnerable to the supposed fallacy?—Machine Elf ¹⁷³⁵ 09:01, 26 January 2012 (UTC)

Van Gogh Example

I'd say it's not immediately clear at first why this 'Van Gogh fallacy' is an appeal to probability. I think this is at least partially because it's not laid out well with the Ps and Qs. Read the example and decide what P means, it seems like it means two different things in two different places. If P is penniless artist and Q is artistic genius, then more appropriately it should say something like Van Gogh was both P and Q, and I am P so I should be Q, which I think is a stretch for this fallacy anyway, but I see where it's trying to go. And more importantly this is definitely not affirming the consequent, unless we're to also assume that Van Gogh was penniless because he was a genius. If we don't have that Q implies P (which I don't see why we would), then assuming P implies Q isn't affirming the consequent, it's more like confusing correlation with causation. — Preceding unsigned comment added by 174.102.96.164 (talk) 06:44, 30 September 2012 (UTC)

---

I didn't see the Murphy's Law thing before it was removed. But as a lay person, Murphy was the first thing that sprung to my own mind in reading the explanation of the fallacy as currently provided. I humbly suggest that there must be some way to treat the fallacy as a misapplication of Murphy. Murphy is helpful in risk planning because when it is applied properly, it frames planning as a matter of prioritizing all possible risks rather than as a question of whether something is practically impossible, and thus to be ignored even when there's nothing else on the agenda for someone on retainer to prevent losses. In my limited understanding, the article's treated fallacy, when applied to possible positive outcomes assumes that one is ignoring other more important things that are more likely to happen first; such as in the debt/lottery example, the possibility that debt can eventually stop one from being able to play the lottery is much more likely than being able to play enough to win a jackpot. The way this fallacy is described, "appeal to probability", it actually seems more like "neglect of comparative improbability".

Whatever the case, I like things to be framed concisely at the tops of articles, but I also like a certain amount of in-depth discussion, even if it is necessary to mark something as disputed or otherwise potentially problematic, and I can't help but suspect that there is some significant proportion of users who would also prefer that. I don't see deleting practically everything and posting it for argument on the talk page as necessarily being the best solution.

Can't we do any better? Please?

Joshua Clement Broyles — Preceding unsigned comment added by 186.29.237.199 (talk) 22:10, 5 December 2012 (UTC)

Would this be a part of this article?

I had a little debate with someone recently, and I thought their argument was a logical fallacy, but I couldn't find anything that was exactly what they were using. Perhaps someone could point me in the right direction? Their argument was that, given a 50% probability of X, as you increase the sample size, it is more likely for X to occur on the next trial. Basically they said that the more times you flip a coin and it lands on tails, that increases the chance for the next flip to land on heads. Is this a formal fallacy of any kind? This article was as close as I could come. (I know that the argument isn't true, I just want to know if it has a name :) ). 76.185.17.196 (talk) 04:48, 20 May 2008 (UTC)

to put it mathematically, you both have valid arguments. If you keep the probability at 50%, then each individual toss has the same chances of landing on tails. However, if you increase the sample size, the chances of the coin landing on tails exponentially decrease. For instance, if I were to flip a coin once, the probability of it landing on tails would be 50%. If I flip it twice, this goes up. This is demonstrated in the chart below:

	Trial 1 (1 flip)	Trial 2 (2 flips)	Trial 3 (3 flips)
Possibility 1	H	HH	HHH
Possibility 2	T	HT	HHT
Possibility 3		TH	HTH
Possibility 4		TT	THH
Possibility 5			HTT
Possibility 6			THT
Possibility 7			TTH
Possibility 8			TTT
Percent chance that bold is outcome	50%	25%	12.5%

... and so on. I hope this helps. —

Supuhstar * § 22:34, 18 August 2009 (UTC)

"to put it mathematically, you both have valid arguments" -- To put it factually, you are wrong. As the number of flips increases, the probability that the coin will land tails on some flip of course increases, but the claim was that it increases for the next flip, which is fallacious -- the gambler's fallacy, precisely. Aside from that, this discussion is off-topic -- WP talk pages aren't a help bureau, they are for improving the associated article. -- 96.248.226.133 (talk) 19:39, 21 December 2012 (UTC)

It's the gambler's fallacy which covers any situation where one claims that the probability of a random event can be determined (wholly or in part) by independent events. A more specific label is the "we're due for one" argument for claiming the likelihood of an event (often used when speaking of disasters) is increasing because it hasn't happened in a while. For something like an earthquake or volcano where there is a buildup and release mechanism, this makes some sense. You'd expect the likelihood per unit time to increase as the time since the last event increases. For things like hurricanes (and typhoons/cyclones) where the odds of a hurricane in a season are pretty much independent of any other season, there should be no expectation of a change in frequency. -- KarlHallowell (talk) 20:58, 10 November 2009 (UTC)

Biased

This article is very biased to the idea that the Appeal is always false. There should be a tag on it, but I have no time to look up which one to use. Alternatively, someone could unbias it, but, again, I haven't the time. — Supuhstar * § 22:19, 18 August 2009 (UTC)

This is gross nonsense, like complaining that an article about World War II is biased toward it having happened. Either "appeal to probability" is a real fallacy, supported by reliable sources, in which case it is always fallacious (not "false"; that's the wrong concept), or there shouldn't be an article about it. And no one cares about your lack of knowledge of WP tags or your other personal problems. -- 96.248.226.133 (talk) 19:57, 21 December 2012 (UTC)

The lottery example is absurd

The problem with playing the lottery repeatedly until you win it so as to pay off your debts is not that you might never win but that a) your payoff is negative because of the cost of lottery tickets and b) you're not likely to win before you need the money ... if debts could be held off indefinitely, you wouldn't even need to play the lottery.

And this article can't be redeemed with better examples because a) "probable" and "possible" are not interchangeable and b) appeal to probability is rational ... in fact it's the basis of rational behavior in the real world. That probability doesn't logically entail inevitability is a trivial matter that doesn't bear mention, let alone a Wikipedia article. Finally, the sole citation for this article does not meet the requirements of a reliable source. -- 96.248.226.133 (talk) 19:30, 21 December 2012 (UTC)