Oh. I don't get it, then.
Announcement
Collapse
No announcement yet.
I’d rather ask you all than google it
Collapse
X
-
David Spiegelhalter can give you a few minutes on the general idea. A key difference from classical probability work is that this is not about calculating a theoretical probability for an uncertain event (lotteries, radioactive decay), it is about finding the likelihood that a thing did happen or did not happen.
Comment
-
David Spiegelhalter can give you a few minutes on the general idea. A key difference from classical probability work is that this is not about calculating a theoretical probability for an uncertain event (lotteries, radioactive decay), it is about finding the likelihood that a thing did happen or did not happen.
I get the Bayesian basic idea, but I don't know how the math works.
Comment
-
The underlying concept of frequentist statistics is (cutting several corners) that the probability of an event is roughly equivalent to the frequency of the event happening in a series of observations. The idea is that you can scratch "roughly" from that sentence if you stretch the series of observations to infinity.
So if I want to compute the probability of a coin coming up heads, I flip the coin ten times, and observe the number of heads as variable X. Then the probability of the coin coming up heads is X/10. If, instead of repeating coin flipping ten times, I repeat it an infinite number of times, this probability will necessarily converge to the real probability.
The good thing about frequentist statistics is that it is objective: everybody will agree on the resulting probabilities. The bad thing about frequentist statistics is that you need an infinite number of observations, and who has the time for that?
Comment
-
Conversely, the core concept of Bayesian statistics is that everyone has an opinion beforehand on the probability that a given coin comes up heads. So either you think that the coin is fair, and you assign probability 1/2 to the event "coming up heads", or you have reason to believe that the coin is unfair, and assign a different probability to that event. Now if we have a series of observations, these might either match your pre-assigned probability, or they don't. In the former case, your prior belief is fortified. In the latter case, your prior belief needs to be adjusted with this new data, which will lead you to assign a new posterior probability to the event "coming up heads".
The good thing about Bayesian statistics is that it allows you to say something about the probability of an event, way before you have seen an infinite number of observations. The bad thing about Bayesian statistics is that everyone is allowed to have a different prior belief, and hence, everyone may end up with a different posterior belief: even after observing a sequence of outcomes, people may still disagree on what the probabilities are.
Comment
-
To illustrate the difference between the two schools, the German tank problem may be instructive.
Comment
-
Originally posted by Sits View PostWhy do male and female shirts, jackets etc. button on opposite sides?
Edit: I was half-right, but it's the other way round. Men dressed themselves and women had maids to do it.Last edited by Stumpy Pepys; 07-03-2018, 12:33.
Comment
-
A long time ago I read a kid's book about a kid in the dark ages who survived by poaching game from a small boat that was round - like a big wok without a handle. I think it was propelled by polling rather than paddling. Like punting, but the boat was small and round, not long and flat-bottomed. And I feel like I read somewhere that there are still people in Wales especially who use such boats for recreation and fishing now.
Did I just dream that? Or is that really a thing and if so, what's it called? Because the wiki page on "punting" doesn't seem to have it.
Comment
-
ad hoc - OTF's 'coracle oracle'.
The (British) coracle was traditionally used widely in Wales and Shropshire: there was one guy, the late Fred Davies, who earned fame/infamy during the seventies for watching Shrewsbury Town matches from his coracle, ever on stand-by to retrieve the ball, should Alf Wood or whomsoever decide to shank one into the Severn.
Comment
-
There is a Camarthen* Coracle and Netsman's Association with a page on Facebook that you might find interesting, HP. They also sell hand made miniature coracles from time to time, if your coracle lust can be sated by a mantelpiece sized version.
* - market town in west Wales.
Comment
-
Originally posted by Jah Womble View Postad hoc - OTF's 'coracle oracle'.
The (British) coracle was traditionally used widely in Wales and Shropshire: there was one guy, the late Fred Davies, who earned fame/infamy during the seventies for watching Shrewsbury Town matches from his coracle, ever on stand-by to retrieve the ball, should Alf Wood or whomsoever decide to shank one into the Severn.
It's alleged that one foggy day he pursued a 'ball' some distance down the Severn before it revealed itself to be a swan and fucked off. That's a story that's entertaining the first 64 times some old-timer tells you.
As Sits points out, the great thing about coracles was that in their day (the Bronze/Iron Age and before Sky launched the Premier League) they could be taken on long journeys, constructed, as they are, from hide stretched over a wicker frame.
There you go HP.Last edited by ChrisJ; 09-03-2018, 13:03.
Comment
-
Originally posted by Hot Pepsi View PostYES! Thanks. I want one. Or maybe learn how to make one.
https://www.coraclesociety.org.uk/buildingcoracles
Comment
-
Originally posted by Benjm View PostThere is a Camarthen* Coracle and Netsman's Association with a page on Facebook that you might find interesting, HP. They also sell hand made miniature coracles from time to time, if your coracle lust can be sated by a mantelpiece sized version.
* - market town in west Wales.
Comment
-
Originally posted by Various Artist View PostCarmarthen, in case you can't find it under the name Benjm's given you there, HP
Comment
Comment