TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Bertrand's Paradox

82 pointsby jcralmost 10 years ago

15 comments

andrew-luckeralmost 10 years ago
There should be infinitely many &quot;solutions&quot; to this if you are allowed to define the sampling method arbitrarily.<p>Intuitively, one would think that random points are uniformly distributed, but that is not maintained in two of these solutions. If you randomly generate two points in [0,2pi) with a uniform distribution then you will quickly get the 1&#x2F;3 probability empirically or provably by the limit.<p>However, if you are allowed to derive the second point from the first (such as the cross section example) or create some other dependency, then you are given great liberty as to what the resulting &quot;probability&quot; will be. Sure you can even guarantee that every two points on the circle will be present &quot;eventually&quot; but infinity is very large and how you approach it will determine your eventual outcome. Here we could subsample preferred points first and actually derive any specific probability (0,1) and maybe even make it sound reasonable. This behaviour will even be maintained into the limit, but it is certainly not intuitive, at least to me.
评论 #10002153 未加载
评论 #10017019 未加载
ihmalmost 10 years ago
I got 1&#x2F;3, but with a &quot;different&quot; distribution. (I took symmetrized pairs rather than pairs. But I realized after I wrote this comment that folding a square in half preserves relative area so symmetrizing makes no difference.)<p>I started off with a flat torus (two ordered points on the circle) and then quotiented out by (x, y) ~ (y, x) (since the chord corresponding to both is the same) to get a triangle as in the diagram [0].<p>By geometric intuition&#x2F;trigonometry&#x2F;etc it&#x27;s clear that a chord has length &lt; sqrt(3) iff the difference between the angles x and y is less than 2pi&#x2F;3. The region corresponding to pairs with distance distance less than 2pi&#x2F;3 is shaded in [0]. One adds it up and sees that the unshaded region is 1&#x2F;3 of the whole.<p>[0]: <a href="http:&#x2F;&#x2F;i.imgur.com&#x2F;ow8Dkgd.png" rel="nofollow">http:&#x2F;&#x2F;i.imgur.com&#x2F;ow8Dkgd.png</a>
jcralmost 10 years ago
This will most likely be helpful as well:<p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Bertrand_paradox_%28probability%29" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Bertrand_paradox_%28probabilit...</a>
评论 #10000997 未加载
评论 #10001129 未加载
jamesromalmost 10 years ago
It appears each approach depends on the way random chords are selected. Each approach yields different probabilities of selecting a chord with a certain property.<p>But my intuition suggests that constructing a proof for <i>all</i> chords should be relatively simple, and indeed 1&#x2F;3 of them hold the property that they are longer than √3.
评论 #10001183 未加载
评论 #10001224 未加载
评论 #10001158 未加载
评论 #10001144 未加载
amalconalmost 10 years ago
One could just as easily make an argument for a ratio of sqrt(3)&#x2F;2. The chord is selected by randomly picking a length out of the range [0, 2], placing it arbitrarily on the circle, and then rotating it randomly between zero and one revolution[1].<p>Before we can select a random chord, we must first map the set of chords onto the set of real numbers (or, equivalently, tuples of real numbers). This is required in order to meaningfully apply a distribution to them. The problem is that there are multiple ways to do this. Which mapping we select has as much effect on the outcome as the probability distribution.<p>Likewise, if we want to &quot;count&quot; all random chords, we need to do the same thing (map them onto the set of real numbers). We will get different &quot;counts&quot; depending on which mapping we select.<p>[1]- Edit: This is not a very satisfying approach, but it does demonstrate the problem well. Being satisfying isn&#x27;t a requirement for a mathematical construct.
sopooneoalmost 10 years ago
Years ago I posted a question on CraigsList (<a href="https:&#x2F;&#x2F;forums.craigslist.org&#x2F;?ID=20211237" rel="nofollow">https:&#x2F;&#x2F;forums.craigslist.org&#x2F;?ID=20211237</a>) Science and Math discussion board, that I think may be nearly equivalent, and equivalently ambiguous in it&#x27;s notion of random, to this post. It went on for quite some time, and was even reprised later (<a href="https:&#x2F;&#x2F;forums.craigslist.org&#x2F;?areaID=4&amp;ID=20432604" rel="nofollow">https:&#x2F;&#x2F;forums.craigslist.org&#x2F;?areaID=4&amp;ID=20432604</a>) by someone else. Helpful people went back and forth with beautiful 3d renders to illustrate their solutions, though those links are sadly broken now.<p>It is very gratifying to read this post, as I think it explains why we were never able to come to a satisfying answer.
dstyrbalmost 10 years ago
I don&#x27;t agree with any of these solutions. In my opinion the circle should be positioned in an infinite plane, then random vectors should be drawn on the plane and ignored unless they intersect the circle. Then after making some large amount of intersecting vectors you calculate. I&#x27;m guessing that this essentially reduces to the random endpoints method.<p>The other two methods don&#x27;t make sense at all to me. Random chords is just obviously not a random sampling of possible vectors-- it&#x27;s essentially the same as sampling a gaussian(x) distribution using an equal sampling in x and ignoring the fact that a gaussian is a PDF and you can&#x27;t sample that way, you need to invert it into a CDF and take an equal sampling in _y_. Very basic I think...<p>And random midpoint doesn&#x27;t strike me as random either as it forces the orientations of the vectors to be non-random and systematically smaller than random because of the pi*r^2 area-dependence of concentric circles...<p>But, what do I know.
评论 #10001344 未加载
评论 #10001670 未加载
评论 #10001315 未加载
评论 #10001353 未加载
transfirealmost 10 years ago
I came up with a third -- it seemed fairly obvious after some thought. I do not quite get the 1&#x2F;2 and 1&#x2F;4 explanations. (Unfortunately these gif analogies do not do such a great job of clarifying.) So I decided to ask good old &quot;Reality&quot;.<p><pre><code> N = 1000 SQRT3 = Math.sqrt(3).to_r def point_on_circle x = (rand().to_r * 2) - 1 y = Math.sqrt(1 - x**2) y = -y if rand(2) == 0 return [x,y] end tally = 0 N.times do x1, y1 = point_on_circle x2, y2 = point_on_circle d = Math.sqrt( (x2 - x1)**2 + (y2 - y1)**2 ) tally += 1 if d &gt; SQRT3 end puts &quot;#{tally}&#x2F;#{N}&quot; </code></pre> Oddly, it gives an answer close to a third (yeah me!). But... it is always a bit high at approx 36%.
评论 #10001266 未加载
评论 #10002116 未加载
arexsuttonalmost 10 years ago
The 1&#x2F;4th answer has a typo.<p>&quot;The probability of being hit is the probability of being inside the area of fire divided by the total area&quot;<p>No.<p>The probability of being hit is the probability of being inside the area of fire.<p>The probability of being inside the area of fire is the area of fire divided by the entire area.
aaron695almost 10 years ago
You and another incredibly smart human have to chose an answer (of the three) in secret.<p>You differ you die.<p>Which one do you chose?<p>To me this is the correct answer. I don&#x27;t think the other two would even be in consideration.<p>I also think aliens would agree, although I&#x27;m not sure that would be fair since the original question (on this website) is being asked to me, a human.<p>[edit] Interesting this is also not the exact Bertrand paradox since they perhaps have added units to it to get around Edwin Jaynes solution *See Wiki
lordnachoalmost 10 years ago
IMO the most natural interpretation of &quot;random&quot; with no qualifiers is the one that gives you 1&#x2F;3. Pick a point, pick another point, a third of the time the second point in within the far arc of the triangle.<p>I&#x27;m not sure the ninja method gives you what I&#x27;d call a representative sample of possible chords. Looks like he only cuts vertically, so he can never give you a skew-slice.<p>Similar with the dragon.
评论 #10005425 未加载
flowerjenialmost 10 years ago
It seems every methodology relies on upon the way arbitrary harmonies are chosen. Every methodology yields diverse probabilities of selecting a harmony with a certain property. <a href="http:&#x2F;&#x2F;www.trainingintambaram.in&#x2F;ccna-training-in-chennai.html#" rel="nofollow">http:&#x2F;&#x2F;www.trainingintambaram.in&#x2F;ccna-training-in-chennai.ht...</a>
reilly3000almost 10 years ago
The average of the provable probabilities is the resultant probability. About 36%.
agnes49almost 10 years ago
The Bertrand paradox is a problem within the classical interpretation of probability theory. <a href="http:&#x2F;&#x2F;www.trainingintambaram.in&#x2F;web-designing-training-in-chennai.html" rel="nofollow">http:&#x2F;&#x2F;www.trainingintambaram.in&#x2F;web-designing-training-in-c...</a>
murbard2almost 10 years ago
For extra credit compute the probability under Jeffrey&#x27;s prior.