So I was playing around with the spreadsheet. From what I can tell, this option doesn't work. One of the Constitution-Bs needs to be delayed to Q2 or else there's a one quarter 75sr deficit.
Also, we can gamble on a 5sr deficit (depends on Captain's Log rewards and no repair costs) in 2310Q2 if we scheduled another Constitution-B in Q2. So a risky version of this is:
I'm opposed to gambling on a deficit. So is Briefvoice. Part of the point here is to maintain at least a minimal reserve capacity for repairs and emergencies, because while having more cruisers is important it's not
that important.
That goes double with some of our ships actively pursuing the Syndicate, and with the Sydraxians having been eating their Wheaties lately. The odds of one or more of our ships ending up in the hospital some time this year are actually rather high, probably higher than normal- and in a normal year the odds are at least 50/50.
I don't suppose there's way to make conditional orders? Something like:
[] 2 Explorers, 2-3 Connie-Bs, Snakepit
Start 2 Excelsiors in Q1 in SF berth and 40 Eridani berth
Start 1 Constitution-B in Q1 in 40 Eridani berth
Start 1 Constitution-B in Q2 in 40 Eridani berth
Start 1 Constitution-B in Q2 in UP berth if there's currently at least 100sr available in Q2
Request Excelsior's worth of resources in Snakepit
Sounds nice.
@OneirosTheWriter, is that a thing that we could do?
There are good reasons why sunk cost fallacies can be okay, and you've already stated them before (basically psychology), but you're conflating them with the raw rational economic implications with these examples.
The thing is, the psychology and the raw rational economics are tied together, because they both relate to the real underlying issue I'm talking about-
decision-making.
Decision-making strategies that are seemingly illogical, but which consistently lead to good results,
are not wrong. They may not even be illogical. There is often an underlying higher-order logic at work. The higher logic may have to do with psychology or it may not- it varies.
Sticking to a long-term plan even in the face of sticker shock and setbacks is one of those things that pays off. It's not a rule to follow
all the time, but it's a rule that works quite a lot more often than not. And it is even more reliable when coupled with a bit of judgment and basic common sense.
The problem with [the scenarios Simon cited] is that they're incomplete - fully open to what-ifs. The whole point of the sunk cost fallacy is that a) there are scarce resources (typically time and money), and b) there's an opportunity cost for those scarce resources.
To give some crazy counter-examples:
30 yr mortgage: if there was an awesome deal for a better house and I can't afford to both keep both my current mortgaged house while buying the better one, I may be inclined to sell my current house to mortgage the better house
Moon rocket: if the current design is so incredibly inefficient that a newer design would actually be finished more quickly and at lower cost, I would favor the new design
College degree: if there's an economic recession and the degree I'm working towards is now suddenly worthless, you'd bet your ass I'd be inclined to switch degrees (assuming I'm aiming to get a job out of it)
See, the thing is, all those counter-examples
ARE crazy, in that they are unlikely to occur. They're all remotely plausible, but you are unlikely to see situations where:
1) The better house is so much better
and cheaper that it's worth writing off the sunk costs in your current home, and accepting that you won't fully own your new home for thirty years instead of fifteen for the old one.
2) The current moon rocket is so lousy that it is
obviously going to be faster and cheaper to design a new rocket from scratch. As opposed to basing this on over-optimistic estimates of how
little the new one will cost compared to realistic (fact-based) estimates of how much the old one cost. I could go into more detail and provide considerable evidence for this in my opinion, but it would veer into the politics of the past fifteen years in the United States.
3) The economic rewards of seeking and hopefully finding work for two years (in a recession) exceed the
expected lifetime value of a bachelor's degree. Or even the expected lifetime value of the bachelor's degree for the few years before I could get around to finishing my degree "whenever."
Basically, while dismissing sunk-cost reasoning as a fallacy works
very well when resources are nigh-infinite and logistics is not being kept track of, it works much less well when resources are scarce and logistics is critical.
There will always be
exceptions, times when it is best to write off a sunk cost, usually when the entire plan was a bad idea in the first place, or when something totally unforeseen happens that renders a good plan irrelevant.
However, the existence of exceptional cases that make reasoning in a certain way
incorrect does not mean that this category of reasoning is fallacious.
I'm not saying any of these are realistic. But addressing sunk cost fallacies requires dealing with the exact circumstances surrounding it, and an economic decision among alternatives, AND separately considering the psychological implications of cancelling a project with sunk costs. (Though technically, it's theoretically possible to put a price on those psychological implications - analysts do something like this all the time with market research.)
Which I did. My argument is that no good alternatives exist, we need the
Constitution-Bs badly,
AND that it is bad policy to abandon our up-front investment at a critical moment when we were planning to capitalize on it by adding a little more investment now- and then no more.
Every case where one is dealing with sunk costs has to be addressed on its own merits. And most cases with
realistic sunk costs, when addressed on their merits, point to the conclusion that you should not abandon your initial investment, UNLESS:
1) You have strong reason to think your original plan was a bad one because the odds were against you all along (e.g. gambling)
2) Your initial investment did not in any way actually increase your likelihood of a future payoff (e.g. gambling)
3) The rules just changed so drastically that there is an easy way to win far more than you had once expected (your lucky day).
4) The rules just changed so drastically that you are in far more danger of losing a great deal than you expected (e.g. it's time to completely abandon your home and business and flee the country because a massive civil war just broke out)
Here, the rules haven't changed, there is no reason to think our original plan is a failure, and the initial investment did in fact make the payoff we anticipate starting in 2312 possible.
...
Basically, the real role of the sunk cost fallacy is as a way of deterring people from pouring resources uselessly down a hole or becoming addicted to gambling. When you try to apply it to the middle stages of a
rational long range plan, the best that can happen is that you throw a monkey wrench into the works.
Now, in the context of the Connie-B, there's currently no good alternatives to building at least 6 Connie-Bs (pessimistically) or 8-10 Connie-Bs (more likely). But what if we suddenly got some admiral that could cut down the Renaissance prototype time to half (like Patricia Chen with explorers) so that it finishes just before 2312Q1. I would totally be willing to cancel some Connie-Bs and save up for a Renaissance construction surge in 2312. (Though a Renaissance this early would seriously mess up with our careful crew balancing.)
Not in a bad way, since it would cost more redshirts (of which we have many) and fewer blueshirts (of which we have few). It'd be awkward in the context of our just having reconfigured the Academy to provide a boost to our supply of techs over the next four years, which we anticipated needing to crew ConnieBees, but we would manage.
Thing is... that's not going to actually happen. The rules haven't changed that dramatically; all that is happening is sticker shock.
Well, sunk cost reasoning, aka escalation of commitment, isn't the only justification for staying the course. Lots of it is simply risk aversion - status quo may be suboptimal, but it could be worse when risking a change. It's a pretty optimal general strategy, but can be trumped by detailed analysis, or more commonly, peer pressure.
EXACTLY. Which is why I am advocating for escalation of commitment, in the face of what I perceive to be peer pressure. People see the "we're spending 20pp on getting more ships" and have sticker shock. This creates pressure to economize, but that isn't actually based on recalculating the
reasons why we originally committed to building ConnieBees. Those reasons still exists, and those needs will not be satisfied until 2020 or later even if we are able to build
Renaissance-class cruisers in large numbers.
