Probability
Post Reply
 
Thread Rating:
  • 3 Votes - 3 Average
  • 1
  • 2
  • 3
  • 4
  • 5
15-03-2013, 09:12 PM
RE: Probability
I don't want to read all this. Someone sum up 19 pages of what was discussed for me.

"Ain't got no last words to say, yellow streak right up my spine. The gun in my mouth was real and the taste blew my mind."

"We see you cry. We turn your head. Then we slap your face. We see you try. We see you fail. Some things never change."
Find all posts by this user
Like Post Quote this message in a reply
15-03-2013, 09:19 PM
RE: Probability
(15-03-2013 09:35 AM)TheBeardedDude Wrote:  If you want a mathematical demonstration for why you are wrong, here goes.

Let's say I have a bucket of marbles of different colors (I know the true number of colors, but telling would defeat the purpose of the whole bloody experiment). I will however, tell you that the bucket contains a finite number of marbles at 11,111 (Just the way the number comes out after I add up the colors of marbles and the numbers I assigned them)

I can calculate the true probabilities because I know the actual numbers of each, you don't.

So, I allow you to start drawing a random sampling of marble from the bucket of 11,111. The only number of draws that allows you to actually say what the true probability is, would be 11,111 (or to put it another way, to observe them all).

That is not how reality works unfortunately. Typically we have to do what we can with the information and observations we can access. So, in a case like this, we would do something called rarefaction in order to come up with an estimate for the number of colors at any given sample size (a rarefaction curve) as a way of guessing what the best estimate for the number of colors in the bucket is, but also realizing that this estimate is not going to be true, since I can't actually discern the true number without observing them all directly.

Here is the rarefaction curve I have generated (it is still running but takes quite a while since it does 10,000 runs)



So, based on the rarefaction curve, how many colors would you say there are? (I am at run/sample size 440 and the number is 3.44)

We might assume the number is 4. Is that true? Is 4 the magic number? My hint is, no. But the rarefaction curve and the probabilities we generate don't tell us the true diversity in the bucket, only the diversity of our samples.

After I have that information, I might say something like:
The most commonly drawn marble was white, with the next most commonly drawn marble being black, a small number of red, and the rare blue marble. The probabilities for each draw having been white is 90%, black = 9%, Red = .9%%, and Blue = .09%

Those probabilities only describe what I drew, not what is in the bucket. I know this to be true since I know that these are in fact not the only colors in the bucket, and we know that in reality, we are also not likely taking a sampling of everything. Which is why we don't use statistics and probabilities to make predictions, only observations of our data.

All you demonstrated is that the probability of there being more than 4 colors will always be less than 1 until you draw all the marbles. You haven't shown that probability of there being more than 4 colors is not increasing with draws always resulting in one of the 4 colors without ever drawing a new color.
Find all posts by this user
Like Post Quote this message in a reply
15-03-2013, 09:24 PM
RE: Probability
(15-03-2013 09:12 PM)Buddy Christ Wrote:  I don't want to read all this. Someone sum up 19 pages of what was discussed for me.

I am trying to convince people that the more marbles you observe that are white while never observing black ones, the more likely it becomes that all marbles are white.

The fail is happening by people claiming I could only make that statement if I observe all the marbles.....which is silly.
Find all posts by this user
Like Post Quote this message in a reply
15-03-2013, 09:30 PM
RE: Probability
(15-03-2013 09:24 PM)Heywood Jahblome Wrote:  
(15-03-2013 09:12 PM)Buddy Christ Wrote:  I don't want to read all this. Someone sum up 19 pages of what was discussed for me.

I am trying to convince people that the more marbles you observe that are white while never observing black ones, the more likely it becomes that all marbles are white.

The fail is happening by people claiming I could only make that statement if I observe all the marbles.....which is silly.

I read that part in the first post. Are you telling me that you haven't gotten anyone to concede to your original post after 19 pages? Was there a follow-up statement if they had agreed?

"Ain't got no last words to say, yellow streak right up my spine. The gun in my mouth was real and the taste blew my mind."

"We see you cry. We turn your head. Then we slap your face. We see you try. We see you fail. Some things never change."
Find all posts by this user
Like Post Quote this message in a reply
15-03-2013, 09:32 PM (This post was last modified: 15-03-2013 09:38 PM by Chas.)
RE: Probability
(15-03-2013 08:05 PM)Heywood Jahblome Wrote:  
(15-03-2013 06:57 PM)Chas Wrote:  No, you are confusing two things - the common definition and the mathematical definition.

Just like 'it's just a theory' is a confusion of definitions.

I don't have a problem using the definition bearded dude supplied. No where in that definition does it say a probability must be calculated. It says a probability is a quantitative description. Quantitative, meaning related to or expressed in measured numerical values.

In a nod to girlyman I change my criticism to you are conflating probability to outcome.

What do you think quantitative means? It means you calculate it. You are making qualitative assessments of chance.


Sorry, no. I am not the one conflating anything.

Skepticism is not a position; it is an approach to claims.
Science is not a subject, but a method.
[Image: flagstiny%206.gif]
Visit this user's website Find all posts by this user
Like Post Quote this message in a reply
15-03-2013, 10:02 PM
RE: Probability
(15-03-2013 09:32 PM)Chas Wrote:  
(15-03-2013 08:05 PM)Heywood Jahblome Wrote:  I don't have a problem using the definition bearded dude supplied. No where in that definition does it say a probability must be calculated. It says a probability is a quantitative description. Quantitative, meaning related to or expressed in measured numerical values.

In a nod to girlyman I change my criticism to you are conflating probability to outcome.

What do you think quantitative means? It means you calculate it. You are making qualitative assessments of chance.


Sorry, no. I am not the one conflating anything.

Quantitative does not mean it has been calculated. It means it is expressed as quantity.
Find all posts by this user
Like Post Quote this message in a reply
15-03-2013, 10:05 PM
RE: Probability
(15-03-2013 10:02 PM)Heywood Jahblome Wrote:  
(15-03-2013 09:32 PM)Chas Wrote:  What do you think quantitative means? It means you calculate it. You are making qualitative assessments of chance.


Sorry, no. I am not the one conflating anything.

Quantitative does not mean it has been calculated. It means it is expressed as quantity.

Which you cannot do in your example.

Skepticism is not a position; it is an approach to claims.
Science is not a subject, but a method.
[Image: flagstiny%206.gif]
Visit this user's website Find all posts by this user
Like Post Quote this message in a reply
15-03-2013, 10:23 PM (This post was last modified: 15-03-2013 10:27 PM by Heywood Jahblome.)
RE: Probability
(15-03-2013 10:05 PM)Chas Wrote:  
(15-03-2013 10:02 PM)Heywood Jahblome Wrote:  Quantitative does not mean it has been calculated. It means it is expressed as quantity.

Which you cannot do in your example.

X is a quantity. An unknown quantity but a quantity never less.

In my example i compute the quantities of X so your claim above is 100% false. I then show these quantities to be increasing while drawing only white marbles. My example would work with any finite number of marbles. Therefore I can credibly claim that the probability of all the marbles being white increases with each draw of a white marble while not drawing a black one regardless if am lacking actual details of the composition of the bin which would allow me to compute X. Show the example is wrong Chas.

Your wasting everyone's time by simply asserting that it is wrong.
Find all posts by this user
Like Post Quote this message in a reply
16-03-2013, 05:35 AM
Re: Probability
The actual ratio of each color does not change, but the observations do indeed show changing probabilities based on the observations. It just so happens that my example moves quite slow based in the number of colors and the quantities of each.

You did read my updates showing how the number of predicted colors continued to approach 4, right?

Had I let it keep running, it would have found a fifth color and the probabilities would have necessarily changed based on a new observation.

How many colors were there? You can't answer it because the statistics don't say a damn thing about it.
Find all posts by this user
Like Post Quote this message in a reply
16-03-2013, 07:02 AM
RE: Probability
(15-03-2013 10:23 PM)Heywood Jahblome Wrote:  
(15-03-2013 10:05 PM)Chas Wrote:  Which you cannot do in your example.

X is a quantity. An unknown quantity but a quantity never less.

In my example i compute the quantities of X so your claim above is 100% false. I then show these quantities to be increasing while drawing only white marbles. My example would work with any finite number of marbles. Therefore I can credibly claim that the probability of all the marbles being white increases with each draw of a white marble while not drawing a black one regardless if am lacking actual details of the composition of the bin which would allow me to compute X. Show the example is wrong Chas.

Your wasting everyone's time by simply asserting that it is wrong.

You are wasting my time with your continued confusion between mathematics and perception.

Skepticism is not a position; it is an approach to claims.
Science is not a subject, but a method.
[Image: flagstiny%206.gif]
Visit this user's website Find all posts by this user
Like Post Quote this message in a reply
Post Reply
Forum Jump: