5 ms·
>Oh my lord you guys that is not how probabilities work. You can’t look at what actually happened and then say “Well that was the chance.” If I flip a coin 20 t
by joshuamorton 9y ago
>Oh my lord you guys that is not how probabilities work. You can’t look at what actually happened and then say “Well that was the chance.” If I flip a coin 20 times and happen to end up with a 15:5 split that doesn’t mean the chance of heads is 3:1.
Erm, over a large sample size, yes this is exactly how things work. Sure 15/5 may be too small, but 150/50 isn't. That's well after the point where its reasonable to believe that your coin is rigged.
As the other user mentioned, if you don't have the prior that the coin is fair, a 15/5 outcome may (and does) indeed imply that heads are more likely than tails. In fact, 15/5 would imply a less than 1% chance of a fair coin, given uniform priors. (Beta[15, 5] -> 1st percentile is 50.175)
- pps43 9y agoBut in this case the samples are small. If you flip a coin once and got tails, what is the probability of heads?
- joshuamorton 9y agoThat's not quite the same question as what I'm answering. I'm saying "given you flip a coin once and it comes up heads, what is the probability that its a fair coin" (or previously, given you flipped a coin 20 times, and it came up 15/5, what is the likelyhood it is a fair coin). And with a single flip, you can't actually answer that question (BetaDist[1, 0] is undefined). Now, once you've answered that question, you can integrate to get a probability that the next coin is heads, but with a single flip you can't even answer the first question. You need to have gotten heads and tails each at least once.
- isostatic 9y agoFlip a coin 50 times and get 50 heads, it's almost impossible it's a fair coin. Does your BetaDist function work in that case?
- joshuamorton 9y agoNo, the beta distribution BetaDist[a, b] is only defined for a, b strictly > 0. When a or b is very large and the other is zero, you can treat it as 1 and get a decent approximation, but it'll be a an estimate that biases towards the center.
- joshuamorton 9y agoI should also mention that if you're willing to make assumptions like a uniform prior, this totally works, but it can lead to some surprising results for small a, b. Notably a uniform prior is Beta[1, 1]. Beta[2,1] implies something like a 75% chance of the true mean being above .5, which seems a tad overzealous.
- pps43 9y agoExactly, that's not the question you're answering, but it is the question that is being asked in this context: what is the probability of rapid unplanned disassembly of SpaceX rockets.
- joshuamorton 9y agoHence the "but you can integrate to answer that". If your rocket has a 50% chance of being explosion proof and a 50% chance of blowing up half the time (you could imagine this as any number of situations, like two otherwise identical rockets which you use), any given launch has a 25% chance of rapid unplanned disassembly. You can do the same thing over a continuous random variable.
- pps43 9y agoBut you know neither the chance of the new rocket being explosion proof, nor the chance of the rocket blowing up half the time. You built the rocket, launched it once, and it did not blow up. What should be integrated to get the probability of rapid unplanned disassembly?
- joshuamorton 9y agoEnter the Beta(variate) distribution. The beta distribution[1] is a cool statistical distribution defined by BetaDist{a,b} (or alpha, beta, but that's too much work), where a is the number of successes and b is the number of failures you've sampled. It has a number of cool properties, chief among them that given X = BetaDist{a,b}, then cdf(X, x) = the probability that the mean of the distribution you are approximating is less than x. It has a bunch of other nice properties too (like E[X] = a / (a + b), which should be obvious), but those aren't as relevant here. So let's say that you assume a uniform prior. This is defined as BetaDist{1,1} [2]. this is probably the wrong prior. So you might have a better idea. If, for example, you believe there is a 10% chance of your rocket exploding based on some calculations you've done, you might use a differently tuned beta distribution, like BetaDist{9,1} or BetaDist{4.5,.5} if you were feeling uncertain (but in general it would likely be better to use {8,2} in that situation iirc). But let's assume {1,1} for now. So you launch your rocket. Everything goes great. You update your distribution. Its a success. So you get BetaDist{2,1} [3]. So what is the chance your rocket explodes? Well the cdf of your beta distribution is the probability that the mean is less than x. So the pdf of the beta distribution is the probability that the mean is exactly x. So then The integral from 0 -> 1 of `(1 - x) * pdf(X, x) dx` is the estimated probability that your rocket explodes on its next launch, since that's "for every value x, the likelyhood of the distribution being that one multiplied by the chance your rocket explodes given that distribution". For the one rocket case, this happens to be equal to E[X] = a / (a + b), so it's 1/3. For the two rocket case, you apply reinforcement learning/k-armed bandit strategies like UBC1[4] or Thompson Sampling[5]. These are algorithms that will result in you picking the best rocket with as few unnecessary explosions as possible, provably. You can see some related discussion I've had on HN about these algorithms [6]. [1]: https://en.wikipedia.org/wiki/Beta_distribution https://en.wikipedia.org/wiki/Beta_distribution [2]: http://www.wolframalpha.com/input/?i=beta+distribution+(1,1) http://www.wolframalpha.com/input/?i=beta+distribution+(1,1) [3]: http://www.wolframalpha.com/input/?i=beta+distribution+(2,+1) http://www.wolframalpha.com/input/?i=beta+distribution+(2,+1... [4]: http://banditalgs.com/2016/09/18/the-upper-confidence-bound-algorithm/ http://banditalgs.com/2016/09/18/the-upper-confidence-bound-... [5]: https://en.wikipedia.org/wiki/Thompson_sampling https://en.wikipedia.org/wiki/Thompson_sampling [6]: https://news.ycombinator.com/item?id=17014232 https://news.ycombinator.com/item?id=17014232
- ebbv 9y agoYes if you actually run an experiment with real sample sizes you can begin to form hypothesis around and unknown probability. But my point was you cannot look at a historical event and say "Well this happened, so that was the chance." That's the basis for the silly internet joke "The chance is 50/50, either it happens or it doesn't." Which is what the people I replied to were doing.
- joshuamorton 9y ago>But my point was you cannot look at a historical event and say "Well this happened, so that was the chance." If the event happens repeatedly, you absolutely can! If someone attempts something 20 times, and it works 10 of them, you can conclude that there's an approximately 50% chance of success. You have a sample size! The exact same thing is true for a statement like "the chance of an astronaut dying is approximately 1 in 25". We have the sample size to show that. >That's the basis for the silly internet joke "The chance is 50/50, either it happens or it doesn't." No, that's totally different. That's a misunderstanding of priors. What you're doing is more like forgetting that the law of large numbers is a thing.
- ebbv 9y agoI’m not forgetting anything. The actual shuttle missions were not a statistical experiment or a consistent action where you can say “well this is what happened so that was the chance of death on an individual shuttle flight.” It only means that’s your chance if you have the technology from quantum leap (the tv show) and you randomly land in the body of one of the participants. as I said already the shuttle and everything around it evolved. The crews changed. Trying to say “the actual chance of death was this based on how many people died.” is silly.
- joshuamorton 9y ago>It only means that’s your chance if you have the technology from quantum leap (the tv show) and you randomly land in the body of one of the participants No, then the event already happened, so you know the outcome with certainty. > The actual shuttle missions were not a statistical experiment or a consistent action where you can say “well this is what happened so that was the chance of death on an individual shuttle flight.” Indeed there are confounding factors that make the error nontrivial. This doesn't invalidate the entire experiment. >Trying to say “the actual chance of death was this based on how many people died.” is silly. Its exactly as silly as saying "the actual chance of getting heads is based on how many heads you get". The problem with your argument is that that isn't silly. To calculate how likely getting heads is, you flip the coin a bunch of times, see how many heads you get, and then you have your answer (and a confidence level). Its the opposite of silly. In other words, there's no big difference between "We'll flip a bunch of coins to see how likely we are to pull heads" and "we'll launch a bunch of people into space to see how likely they are to end up dead". The second isn't as rigorously controlled as the first, but that's fine as long as you account for it.