Is the sunk cost fallacy always a fallacy?

[woweedd]

Eh... I don't think you have to go that far. The comic hasn't really used the unreliable narrator trope very much.

The bigger problem- as I've pointed out- is that people just assume way too much in situations where we aren't given very many hard facts in the first place. "Don't screw it up" is very ambiguous, and a lot of people take ambiguity as confirmation of their priors. You don't have to think that Redcloak is lying, you just have to strip away the heavily biased narrative he filters his worldview through and you can see that the facts, on their face, don't really support the conclusions he's coming to.

What the comic has done on a very regular basis is provide new information that invalidates people's interpretations of old information.

To be fair, it has used unreliable narration. The entire backstory Lord Shojo provided had some...Rather crucial omissions.

[BloodSquirrel]

To be fair, it has used unreliable narration. The entire backstory Lord Shojo provided had some...Rather crucial omissions.

Omissions are different than false information- and what, really, did Shojo omit that he knew or would be relevant at the time? As far as we know, the story he told was basically accurate. There's a lot more to it, but there's always more to every story, and none of it was deceptive.

[woweedd]

Omissions are different than false information- and what, really, did Shojo omit that he knew or would be relevant at the time? As far as we know, the story he told was basically accurate. There's a lot more to it, but there's always more to every story, and none of it was deceptive.

I mean, the story explictly says the second world is our world, when it just isn't. That's not omission, that's just unreliable narration. Shojo wasn't lying, but he didn't have all the info.

[BloodSquirrel]

I mean, the story explictly says the second world is our world, when it just isn't. That's not omission, that's just unreliable narration. Shojo wasn't lying, but he didn't have all the info.

It doesn't, actually. There are a couple of gaps where "It broke out and they tried again" would fit in without directly contradicting anything.

But that's beside the point- the story was still basically true. If Shojo had told the full story, it wouldn't have changed much, other than upping the stakes a little bit. It was Thor's color explanation that really added a new dimension to things. This is very materially different than the kind of "Jirix was lying about what The Dark One said" speculation that tomandtish was talking about.

[Peelee]

It doesn't, actually.

It does. It doesn't use the specific phrasinf "this is the second world and there were nine between it and the first" but it tells the story of gods creating a world, it being destroyed, gods creating a second world, and mortals finding the rifts in that world. It is clearly saying this is the second world, which is what Shojo believed, and which was incorrect.

But, as you say, that's beside the point. We do know that when crayons are broken out it is unreliable narrator. The author has told us that.

[Kish]

I think two crayon scenes in 1271+ strips (two, because that's counting the Start of Darkness one) does not contradict "hasn't really used the unreliable narrator trope very much."

I don't think that rules out the possibility that Jirix was lying to Redcloak, since the Dark One's words to Jirix were only stated by him and not shown on-panel, but I also don't think that's very likely.

(A better example than the crayon here would be Redcloak, in Start of Darkness, describing Rat as an evil god, which Rich explicitly spelled out is an inaccuracy that is not a lie and is significant for conveying that Redcloak assumes any ally of the Dark One must be evil; and it still falls under Not Very Much.)

[Peelee]

I think two crayon scenes in 1271+ strips (two, because that's counting the Start of Darkness one) does not contradict "hasn't really used the unreliable narrator trope very much."

I apparently missed the "very much" part of the claim.

[Gurgeh]

(A better example than the crayon here would be Redcloak, in Start of Darkness, describing Rat as an evil god, which Rich explicitly spelled out is an inaccuracy that is not a lie and is significant for conveying that Redcloak assumes any ally of the Dark One must be evil; and it still falls under Not Very Much.)

Rat isn't referred to as an evil god in Start of Darkness, even by Redcloak - only Loki and Tiamat are shown to take TDO's side in the initial confrontation where Redcloak's narration refers to "older evil gods". Rat turns up later (alongside Tiamat) when TDO is asking about the Snarl; in this scene, Redcloak characterises it as TDO "confront[ing] his few allies among the elder gods".

So while a reader could certainly be forgiven for assuming Rat was evil based on SoD's flashbacks (it'd be a natural inference to draw that Rat, associating with the explicitly evil gods who've sided with TDO, would share their alignment), neither the book or Redcloak said as much.

[Rad]

Anyway what I wonder is if sunk cost doesn't have a value if you do have limited time? A scenario where you invest resources for some gain, and based on what you get, you make your next decision. The alternative of ignoring whatever you got seems unlikely to be better?

Here is an example. In a game of chess you have 2.5 hours to finish a game, so does your opponent, so in total of 5 hours. The game may be 50 moves long, that means each player makes 50 moves for a total of 100 moves... 5 hours for a 100 moves may sound like a lot, but you actually only have 3 minutes per move in average.
With sound time investment, a large part of the moves does not require 3 minutes, but even if we eliminate half the moves it still leaves out 6 minutes per move.
Now I won't go into chess terminology, but imagine you have a position that requires you to think a lot.. So far you have invested your time well, so you can afford at least 30 minutes to consider your options, but then you get absorbed in a specific line of moves for the majority of the duration and now you really cannot spend any more time considering the position. Yet all you have for the time spend is a very deep understanding of one move and the variations which follows, and a very shallow understanding of perhaps two other moves.
You have realized that the move you spend the most time on will give you a position from which you have to fight for a draw if your opponent plays correctly, and you also know there are many ways your opponent can go wrong (perhaps the very reason you were lured to spend so much time on this particular move). However, presently, you see nothing wrong with one of the other moves you considered, and the third you decided to disregard completely, as it doesn't seem to be any good. Also you just discovered a fourth move that you haven't really looked at, but at first glance it does look like a forced win for you, or close to at least.

So what do you do? Please have in mind all moves are supposed to be very complicated, so you cannot be very confident in any other move than the one move you don't think is very good if your opponent plays well against it.

In my opinion the resources spend, is what should determine the outcome. In the chess example, if you know little to almost nothing of other candidate moves than the one you spend so much time on, isn't it a much larger risk to go for any other move then?
Sometimes people play dubious openings in chess knowing they can be punished, but have so much preparation in hand, that it is unlikely to happen based on their current opponent. If this is a valid strategy, is it not also better to go for where your resources went, than to invest in avenues you haven't explored properly, even though at first sight they look so much more promising?

(I'll get to your chess example, I promise)

The sunk cost fallacy is a fallacy as a process to come to a conclusion. It can happen that the "best" conclusion is the one that has been invested into, but the decision on which conclusion to take should be based on what the situation is at decision time.

It is noteworthy that the situation at the decision time will include everything that came from the previous investment. The key issue here is that the "sunk cost fallacy" evaluates based on the cost of the investment rather than its outcome.

In the chess example, you invested some time into exploring a branch that would promise you a draw but not a win. Now you have to evaluate on what to do based on the information you acquired and the time you have left (NOT the time you have spent). Taking the draw might be a good or great option depending on how you estimate a number of factors but it can also be bad.
The point is that you should choose based on what you have (information and time left), not on what you spent to get it.
Maybe you had a material advantage and the information is useful in the sense that you now know not to follow that sub-par branch. Maybe you're left with too little time to play anything else and trying to make a different move is a sure loss (in this last case you painted yourself in a corner and made a blunder, which is hardly a situation where you can brag you've been right all along).

[brian 333]

Sunk cost fallacy: Give me liberty or give me death!

Not sunk cost fallacy: Give me liberty or life so I can take another shot at liberty later on.

Sunk Cost, the amount invested already, is gone and usually cannot be retrieved. Using the amount spent to justify future actions, ("I've killed so many to get here, killing a few more won't matter!") Is the fallacy. There may not be better options than the one you are pursuing, but the decision should be based on what it will take moving forward, not based on what it took to get to the decision point.

For example, having killed so many to get this far, now I evaluate the situation and plan to move forward with what has been learned. The cost going forward will be only what is needed to attain my optimum outcome. Killing may still be required to succeed, but the point is, I weigh projected future cost against likely future gain.

The same decision based in Sunk Cost would be, "I've killed to get this far, so if I just keep killing, eventually I will accomplish my goal." In this case, future cost is not important. Only the cost already paid matters.

[gbaji]

Sunk cost fallacy: Give me liberty or give me death!

Not sunk cost fallacy: Give me liberty or life so I can take another shot at liberty later on.

To be fair, that's actually a false dilemma fallacy.

Sunk Cost, the amount invested already, is gone and usually cannot be retrieved. Using the amount spent to justify future actions, ("I've killed so many to get here, killing a few more won't matter!") Is the fallacy. There may not be better options than the one you are pursuing, but the decision should be based on what it will take moving forward, not based on what it took to get to the decision point.

For example, having killed so many to get this far, now I evaluate the situation and plan to move forward with what has been learned. The cost going forward will be only what is needed to attain my optimum outcome. Killing may still be required to succeed, but the point is, I weigh projected future cost against likely future gain.

The same decision based in Sunk Cost would be, "I've killed to get this far, so if I just keep killing, eventually I will accomplish my goal." In this case, future cost is not important. Only the cost already paid matters.

That's still not technically correct. The fallacy is that "I've killed this many, so they will have died for nothing if I don't finish what I started." It's not about whether killing more people is needed or desired (or should become some sort of habit, I guess), but that this cost has already been paid, and somehow should factor into the equation.

"Many Bothans died to get us this information" (from Return of the Jedi) is a sunk cost fallacy. It implies that the reason they must succeed isn't because of the value of success relative to failure, nor that it's worth the additional risk for the reward of stoping the second death star (althouh it clearly is), but that if they fail, those lives will have been lost for nothing. Of course, we're not explicitely told that the reason to succeed is because of that cost, so it's not 100% certain. I'm reasonably sure they would have gone forward with the plan even if no one had died along the way, so we can probably just chalk it up to dramatic language. But yeah, it's still there, hanging over their heads, presumably pressuring them in some way to "make those lives worth the cost". Maybe...

It's worth noting that there's a reason why logical fallacies are commonly used in debates. They are actually quite effective at motivating people and inspiring emotions.

[brian 333]

Spoiler

Show

To be fair, that's actually a false dilemma fallacy.



That's still not technically correct. The fallacy is that "I've killed this many, so they will have died for nothing if I don't finish what I started." It's not about whether killing more people is needed or desired (or should become some sort of habit, I guess), but that this cost has already been paid, and somehow should factor into the equation.

"Many Bothans died to get us this information" (from Return of the Jedi) is a sunk cost fallacy. It implies that the reason they must succeed isn't because of the value of success relative to failure, nor that it's worth the additional risk for the reward of stoping the second death star (althouh it clearly is), but that if they fail, those lives will have been lost for nothing. Of course, we're not explicitely told that the reason to succeed is because of that cost, so it's not 100% certain. I'm reasonably sure they would have gone forward with the plan even if no one had died along the way, so we can probably just chalk it up to dramatic language. But yeah, it's still there, hanging over their heads, presumably pressuring them in some way to "make those lives worth the cost". Maybe...

It's worth noting that there's a reason why logical fallacies are commonly used in debates. They are actually quite effective at motivating people and inspiring emotions.

A fair criticism. Though the first bit was a shot at humor. I keep forgetting that to be internet-funny I need to include some cat videos, or perhaps a guy getting hit in the 'nads.

[BloodSquirrel]

To be fair, that's actually a false dilemma fallacy.

It's neither. It's a statement of personal value, phrased as an imperative. A false dilemma is when you state that there are only two options. Saying that you would prefer death to all but one of the options is just establishing a rank order of your preferences.

It's worth noting that there's a reason why logical fallacies are commonly used in debates. They are actually quite effective at motivating people and inspiring emotions.

As I've pointed out before, people who are engaging in the sunk cost fallacy usually aren't doing so explicitly. What usually happens is that it biases them toward a conclusion, which they then rationalize as the correct conclusion using optimistic assumptions, pessimistic assumptions regarding the alternatives, and potentially some other logical fallacies thrown into the mix.

"If I just keep killing, I'll win eventually" is the kind of conclusion that someone might come to as a result of the sunk cost fallacy, even if it isn't an example of the fallacy itself.

[gbaji]

A fair criticism. Though the first bit was a shot at humor. I keep forgetting that to be internet-funny I need to include some cat videos, or perhaps a guy getting hit in the 'nads.

Hah! Fair enough.

I think I've been somewhat programmed over time to make sure even when responding to a joke to make some statement clarifying any incorrect assumption contained within that joke (although I do usually try to say something like "I'm sure you're joking, but..."). I do try to restrict this to jokes resting on an incorrect assumption about the subject of the topic itself though (like what a sunk cost fallacy is in this case).

I've been burned in far too many online discussions where such jokes are used to prime a positive response (to the joke), and later used to dismiss disagreement (to the incorrect assumption contained within the joke) based on that initial positive rsponse ("well, you didn't think it was wrong before", "now you're backpedalling so you clearly don't know what you're talking about", etc...). Yeah.

Not at all saying that's what you were doing, of course. Just clarifying why I responded that way. I'd love it if jokes were always just jokes and we could all just laugh and move on. But, unfortunately, on the interwebs, it's sometimes impossible to tell if someone is making a clearly incorrect statement as a joke, or it's something they actually think is true and by presenting it in joke form, they hope to get people to appear to agree with them (or something). Dunno. There are weird people on the internet. And I'm certainly one of them.

It's neither. It's a statement of personal value, phrased as an imperative. A false dilemma is when you state that there are only two options. Saying that you would prefer death to all but one of the options is just establishing a rank order of your preferences.

Oh absolutely. But if it *were* a fallacy, it would be a false dilemma. Like if Patrick Henry actually believed there were only two options or something.

[pearl jam]

It's not that, either, though. The issue isn't that he believed those were the only two options available. It's that he was rejecting all the other alternatives. If he could not have liberty, then the only other option that was acceptable to him was death.

[Peelee]

The Mod on the Silver Mountain: Let's change examples.

[brian 333]

The Mod on the Silver Mountain: Let's change examples.

I'm playing chess and I performed a castle early and dedicated my strategy to defense. My opponent has foolishly left himself vulnerable to a queen-bishop trap. Instead of going for the trap I continue to work my defense because it was the plan I put my effort in, and all that effort would be wasted if I changed strategies.

Whether my defense will work or not is unknown, and beside the point. Whether the trap would have worked is likewise beside the point. What makes it a sunk cost fallacy is that I based my decision on what I have already spent rather than what I must spend going forward, and the likelihood of either method succeeding.

So long as Redcloak's decisions are based on making sure the sacrifices were worth it, he is stuck in sunk cost.

[BaronOfHell]

I'm playing chess and I performed a castle early and dedicated my strategy to defense. My opponent has foolishly left himself vulnerable to a queen-bishop trap. Instead of going for the trap I continue to work my defense because it was the plan I put my effort in, and all that effort would be wasted if I changed strategies.

Ah! But haven't you heard, offense is the best defense.. :p

[gbaji]

I'm playing chess and I performed a castle early and dedicated my strategy to defense. My opponent has foolishly left himself vulnerable to a queen-bishop trap. Instead of going for the trap I continue to work my defense because it was the plan I put my effort in, and all that effort would be wasted if I changed strategies.

Whether my defense will work or not is unknown, and beside the point. Whether the trap would have worked is likewise beside the point. What makes it a sunk cost fallacy is that I based my decision on what I have already spent rather than what I must spend going forward, and the likelihood of either method succeeding.

So long as Redcloak's decisions are based on making sure the sacrifices were worth it, he is stuck in sunk cost.

If Redcloak were to declare that his choices were to continue with the Plan or accept Durkon's offer to seal the rifts (each being equally viable and acceptable in his mind), would anyone label that as a sunk cost fallacy? No, we wouldn't.

It's the futher decision to choose A over B because of a sunk cost that makes it a sunk cost fallacy. The mere existence of a choice does not.


And to be fair, the two fallacies really exist in different modes. Typically, the sunk cost fallacy is internal. It's a fallacy one engages in when making decisions. It's one tricking oneself into making a poor decision. The false dilemma fallacy is usually external. It's about tricking an opponent into making a poor decision by presenting them with only poor decisions to make. If Brian's opponent were to table talk and say "Well Brian, you have to choose between taking my Queen or continuing your defensive strategy. What's it gonna be?", it's possible said opponent has realized that there's some other move Brian could make that would demolish their own position, so they make it seem as though Brian only has those two choices to make. If Brian accepts the "false dilemma" and spends all his time considering the pros and cons of just those two options, maybe he misses a much better move.

The point being that declaring that there are only two choices isn't about what the person making the declaration is choosing for themselves, but what he want's to trick his opponent into thinking are the only two choices available. Which is why I labeled it a false dilemma in the first place.

[WanderingMist]

"Many Bothans died to get us this information" (from Return of the Jedi) is a sunk cost fallacy. It implies that the reason they must succeed isn't because of the value of success relative to failure, nor that it's worth the additional risk for the reward of stoping the second death star (althouh it clearly is), but that if they fail, those lives will have been lost for nothing. Of course, we're not explicitely told that the reason to succeed is because of that cost, so it's not 100% certain. I'm reasonably sure they would have gone forward with the plan even if no one had died along the way, so we can probably just chalk it up to dramatic language. But yeah, it's still there, hanging over their heads, presumably pressuring them in some way to "make those lives worth the cost". Maybe...

This would only be "sunk cost fallacy" if they kept going with only the information the Bothans brought them when new, better things came up, e.g., if they later learned that Palpatine was not, in fact, on board the Death Star and was in fact on Coruscant, yet continued with their plan as though he were still aboard it.

[gbaji]

This would only be "sunk cost fallacy" if they kept going with only the information the Bothans brought them when new, better things came up, e.g., if they later learned that Palpatine was not, in fact, on board the Death Star and was in fact on Coruscant, yet continued with their plan as though he were still aboard it.

The sunk cost is the dead Bothans, not whether the information they obtained is accurate. The dead Bothans should not have had any impact on their decision to go forward regardless of other conditions or changing information later. If it did, it was a logical fallacy, again, regardless of whether new information came to light or not.

I'll also point out (think I may have touched on this before) the difference between argument and debate. Argument is what you use to make decisions. Logical fallacies should be avoided when making arguments (and decisions based on them) because arguments are based on logic, and logical fallacies are, well, not logical (duh). Debate, on the other hand, is what you use to influence other people to join you in some way (perhaps to get them to act on a decision you've made). Logical fallacies are common and to be honest, quite effective, in debate.

The statement about the dead Bothans is a logical fallacy (sunk cost). But in the case it was used, it's not inappropriate to make it. Mon Motha is not (hopefully) actually using their deaths to make her own decision, but to inspire her people to go forward with the plan. Logical fallacies are terrible methods to use to make decisions, but great methods to get other people to act on them.

Win one for the Gipper, right? Almost every rallying cry ever noted historically includes some form of sunk cost reference. People wouldn't use them (and we wouldn't still remember them often centuries or more later) if they weren't effective.

[Peelee]

The sunk cost is the dead Bothans, not whether the information they obtained is accurate. The dead Bothans should not have had any impact on their decision to go forward regardless of other conditions or changing information later. If it did, it was a logical fallacy, again, regardless of whether new information came to light or not.

I'll also point out (think I may have touched on this before) the difference between argument and debate. Argument is what you use to make decisions. Logical fallacies should be avoided when making arguments (and decisions based on them) because arguments are based on logic, and logical fallacies are, well, not logical (duh). Debate, on the other hand, is what you use to influence other people to join you in some way (perhaps to get them to act on a decision you've made). Logical fallacies are common and to be honest, quite effective, in debate.

The statement about the dead Bothans is a logical fallacy (sunk cost). But in the case it was used, it's not inappropriate to make it. Mon Motha is not (hopefully) actually using their deaths to make her own decision, but to inspire her people to go forward with the plan. Logical fallacies are terrible methods to use to make decisions, but great methods to get other people to act on them.

Win one for the Gipper, right? Almost every rallying cry ever noted historically includes some form of sunk cost reference. People wouldn't use them (and we wouldn't still remember them often centuries or more later) if they weren't effective.

Those aren't sunk costs. They're motivational tools. They were going to try to win regardless of if Gipper died, the coach just wants them to honor him. The Rebels would have tried to attack the Death Star even if the Bothans who got the information all lived, she was just stating how much it cost them. In either case, the desired result would have still been tried for either way, so it can't be a sunk cost.

[gbaji]

Those aren't sunk costs. They're motivational tools.

They are both. Sunk costs tend to act as motivators, which is why they are used to motivate people. It's also why the fallacy exists though. It's because we humans are strongly wired to be motivated by those costs, but in reality they should not actually be used when making decisions.

In a purely logical reasoned calculation, those decisions should all be ones the folks involved should want to do anyway. We both agree on that. The baseball team wants to win whether Gipp existed or not. The Rebels want to destroy the Death Star regardless of whether or not Bothans died to get the info. Numerous military forces throughout the ages all want to win the battle/war regardless of whether they "Remember <insert name of person/place/ship/whatever lost in battle previously>!".

So why do leaders mention these things? It should not make a difference at all, as you said. But it clearly does. And the reason it does is because people react emotionaly (and irrationally) to sunk costs. Most people are susceptible to the fallacy and will be influenced by it and allow it to affect their decision making processes. If it didn't, leaders wouldn't continue to use it.

But yeah, from the point of view of the people being motivated? To the degree that it actually motivates them? That's a sunk cost fallacy on their part. It should not matter what previous costs were incurred when making decisions, but if by being reminded of that cost, you play harder, or fight harder, or are willing to take greater risks and make greater sacrifices, then *you* are committing the sunk cost fallacy. And again, it clearly works, so yeah, clearly a lot of people fall into it. It's human nature. If it wasn't, it wouldn't work, and leaders wouldn't use it.

[Peelee]

They are both.

Again, they did not affect the choice to pursue the actions. Whether Gipper died or not, Notre Dame would still have tried to win. If the Bothans got the info with no casualties, the rebels would have still assaulted the second Death Star. There was no sunk cost that influenced anything.

[Snails]

If a person in a leadership position makes a sober logical decision, are they somehow duty bound to abstain from using emotional and arguably irrational arguments to motivate the people implementing the plan?

Of course, not. Peelee is correct that the Sunk Cost Fallacy is about the decision, specifically an incorrectly framed context for a decision, not the details of the implementation.

[WanderingMist]

This is from the Wikipedia article about the fallacy:

"Or, if they hold private information about the undesirability of abandoning a project, it is fully rational to persist with a project that outsiders think displays the fallacy of sunk cost."

Redcloak is in the opposite situation, he has private information about the undesirability of continuing the project, yet continues to act as though the project must be completed.

[Reach Weapon]

Redcloak is in the opposite situation, he has private information about the undesirability of continuing the project, yet continues to act as though the project must be completed.

In so far as Redcloak's belief that information provided by those dwarves is untrustworthy and misleading is accurate he's not actually in that situation. As such, I don't think this is good evidence he's engaging in the sunk cost fallacy, even should that belief prove to be wrong.

[Kish]

In so far as Redcloak's belief that information provided by those dwarves is untrustworthy and misleading is accurate he's not actually in that situation.

Similarly, insofar as Redcloak's belief in the dangers of dihydrogen monoxide is accurate, he's being very logical in avoiding drinking water.

[brian 333]

In so far as Redcloak's belief that information provided by those dwarves is untrustworthy and misleading is accurate he's not actually in that situation. As such, I don't think this is good evidence he's engaging in the sunk cost fallacy, even should that belief prove to be wrong.

I agree that Redcloak does not believe Durkon. Durkon is not the only source of information Redcloak has. He has been at four gates so far, and every time so far the result has been a destroyed gate. Even if we assume Durkon lied or was deceived by those lying gods of Good, Redcloak has ample evidence that his attempt to secure this gate at this time is most likely to result in a destroyed gate.

He is literally pursuing a goal that he knows is unlikely to be achieved, but he refuses to believe that because he is trying to make his sacrifices 'worth it.'

[Reach Weapon]

[...] the dangers of dihydrogen monoxide [...]

Unless I am watery-gravely mistaken, sunk costs are amoungst those dangers.

He is literally pursuing a goal that he knows is unlikely to be achieved [...]

On the other hand, Redcloak was explicitly directed to not screw it up, not to give up. Then he added, "No pressure, though," which I thought was nice.