Bertrand Russell on Miracle Patches

Supposing you got a crate of oranges that you opened, and you found all the top layer of oranges bad, you would not argue, ‘The underneath ones must be good, so as to redress the balance’; You would say, ‘Probably the whole lot is a bad consignment’

There is nothing wrong with wanting to be pleasantly surprised by the next update that promises to solve all the problems, but you must be surprised. Can you flip a fair coin and get heads 10 times in a row? Sure, that is only a 1-in-1024 chance, it must happen all the time in a world with many coin flips. But if someone is taking your money based on that coin, after 10 flips, you should be looking for a two-headed coin.

“They have learned their lesson” is rarely a safe assumption. If someone did a lousy job last time, you must raise your probability that he will do a lousy job next time. Otherwise, you are taking bad work as evidence that good work must be coming. Do you take good work as evidence that bad work must be coming? If you take both good and bad work as reasons to believe that good things are coming, you are shaky on concepts like “evidence” and “reason.”

To end on a concrete example, before turning it back to Mr. Russell, Age of Conan had what was by all accounts a miracle patch at the end of beta, and perhaps several post-launch. At it still had that quality we have come to expect from Funcom.

I wish to propose for the reader’s favourable consideration a doctrine which may, I fear, appear wildly paradoxical and subversive. The doctrine in question is this: that it is undesirable to believe a proposition when there is no ground whatever for supposing it true.

: Zubon

8 thoughts on “Bertrand Russell on Miracle Patches”

  1. Yeah, you’re exactly who he’s talking to. Poor epistemic hygiene will not make you happy in the long run, Bryan Caplan’s “rational irrationality” aside. Building up false hope so you can be happy until it is dashed is just getting further up the mountain before you jump, because you enjoy the view.

  2. I’ve often wondered: if an event is inevitable, the more you wait, the more likely it is to happen. If an even is not inevitable, the more you wait, the less likely it is to happen. A true ‘miracle’ patch, not being inevitable (in addition to the evidence that suggests bad will follow bad), makes it even more likely a miracle patch will never come.

    AKA you can’t polish a turd, inevitable or no.

  3. If an event is not inevitable, then waiting makes it neither more nor less likely to happen. Roll a die. The odds of it being 1 are 1/6. Roll a die. The odds of it being 1 are 1/6…

  4. I still remember that Age of Conan “Miracle Patch” and how it made the game a “Must Buy”…as it did for myself and so many others..

    And then how every patch after that started to bring the game to a worse position than it was in beta…lol.

    Good thing this is not the case anymore.

    But, Miracle Patch is a set of words that should not ever exist together!

  5. You’re looking at it backwards, warcabbit. Imagine that someone tells you they will roll a die and let you know if it rolls a 1 or not. How many rolls will it take before you decide that it’s not a 6-sided die? How many times can they roll before you think that the die is either weighted or doesn’t actually have a 1?

    In other words, we don’t know the probability of a miracle patch. But for each patch that is released that isn’t a miracle patch, we can lower our estimate of the probability. Eventually we reach a point where it is either too low to be considered or just outright impossible.

  6. Yeah – it’s not that the probability of X happening increases over time. it’s the probability of X /having happened/ goes up. For the purposes of the maxim, you don’t stop checking just because it’s happened already.

    And the inverse is true too – it’s “never happened” instead of “won’t happen”.

Comments are closed.