My Math Forum (http://mymathforum.com/math-forums.php)
-   Probability and Statistics (http://mymathforum.com/probability-statistics/)
-   -   Help with Probability (http://mymathforum.com/probability-statistics/344450-help-probability.html)

 Devans99 June 21st, 2018 11:56 AM

Help with Probability

Sorry in advance I did not do it at school. Could someone please take a quick look at my calculations below? I’m not sure whether my method is ok...

Proposition 1 - There was a creator god.

Big Bang is evidence for creator at 60% probability, so combining probabilities:

50% + 50% x 60% = 80%

Fine tuning is evidence for the creator 75% probability, so:

80% + 20% x 75% = 95%

Prime mover is evidence for the creator 25% probability, so:

95% + 5% x 25% = 96.25% chance of a creator god is

To double check, I’ve done the inverse proposition below:

Proposition 2 - there was not a creator god

Big Bang is evidence against no creator 40% probability so combining probabilities:

50% x 40% = 20%

Fine tuning is evidence against no creator 25% probability, so:

20% x 25% = 5%

Prime mover is evidence against no creator 75% probability, so:

5% x 75% = 3.75% chance of no creator god.

 MarkFL June 21st, 2018 12:45 PM

The number of grains of sand in the Sahara Desert is either prime or composite...should I therefore assume a 50% probability for either event because there are two possibilities?

If I roll a standard 6 sided die, I will either roll a 1 or I will roll a number other than 1...are the two events equally likely?

 Devans99 June 21st, 2018 01:02 PM

I don’t think you read the whole of my post. For any question with two possible answers you start with 50% probability and then adjust that number with evidence for or against.

For evidence for you use addition and for evidence against you use multiplication. Both the examples above are evidence against so use a different calculation to mine.

 Devans99 June 21st, 2018 02:44 PM

Just adding a similar example as I think it might clear up my approach:

I find it helps to think of the probability space as a box. Let’s start with the proposition ‘the dog is nice’. Let’s assume you know nothing about this or any dog then the chance of the dog being nice is 50%. So imagine the probability space cut 50% / 50% dog is nice / dog is nasty.

Now we can add a peice of evidence FOR the proposition. The owner says the dog is nice and we trust him 75%. So we already know that 50% of dogs are nice what about the 50% of dogs unknown? Well we can multiply that 50% by 75% and add it to the 50% we already had for dog is nice: 50% + 50% x 75% = 87.5%. Think of the original 50/50 probability space growing to 87.5/12.5 dog is nice / dog is nasty.

So above is how you compute ‘evidence FOR’. ‘Evidence AGAINST’ is a different calculation:

Starting with dog is nice 50%

Now add a piece of evidence AGAINST: ‘the dog bit me’. 90% chance dog is nasty so that’s a 10% chance the dog is nice. So we take 50% x 10% = 5% chance dog is nice. NOTICE THE CACULATION IS DIFFERENT DEPENDING ON WHETHER THE EVIDENCE IS FOR OR AGAINST THE PROPOSITION.

 MarkFL June 21st, 2018 03:11 PM

Yes, I read the entire post...I just addressed the first issue I found. You need to begin with a proposition that is falsifiable. There is no compelling evidence for or against gods. It looks like you're just pulling numbers out of the air. For example, how did you determine that the Big Bang is evidence for or against a magical creator being, specifically 60% for?

 Devans99 June 21st, 2018 03:21 PM

‘There is no compelling evidence for or against gods‘

So what probability do we assign to the proposition ‘do god(s) exist’ bearing in mind it is a yes/no question and we have admitted no evidence either way? I fail to see how you can argue against 50% ?

I have indeed pulled numbers out of the air; rather hoping to concentrate on whether the method was sound rather than the actual peices of evidence and their % probability. My fault for using too emotive an example...

 MarkFL June 21st, 2018 03:51 PM

Without evidence, we cannot assign any kind of probability. All we can say is there either are or are not gods. If we make any kind of claim for or against, then we have a burden of proof that cannot be met. So, the only reasonable action is to reject any claims regarding gods.

This is why you see textbooks on stats/probability using testable events, like drawing cards rolling dice, etc. You won't find them looking at unfalsifiable hypotheses.

 Devans99 June 21st, 2018 03:59 PM

NO! Without any evidence, when confronted with a two-valued proposition, we must assign a 50% / 50% outcome. Think of the example of tossing an unweighted coin.

I think you are letting my emotive choice of example (is there a creator) cloud your thinking. Take a look at the dog example I give above - for any unknown proposition start at 50% / 50% outcome and then adjust the percentages up down for each piece of evidence.

 MarkFL June 21st, 2018 04:05 PM

Quote:
 Originally Posted by Devans99 (Post 595777) NO! Without any evidence, when confronted with a two-valued proposition, we must assign a 50% / 50% outcome.
Incorrect. You're confusing probability and possibility. Just because there are two possible outcomes does not mean we assign equal likelihood to both outcomes. Without evidence we can say nothing regarding probability.

I think you're trolling at this point, and so I bid you farewell.

 Devans99 June 21st, 2018 04:33 PM

I’m not trolling, you are not getting my examples. I’ll try one more:

- the murderer had blood on his clothes. Makes him 70% likely to be the killer
- the murderers prints are on the knife. Makes him 90% likely to be the killer
- taking BOTH pieces of evidence into account, how likely is it that he is guilty?
- 70% + 30% x 90% = 97% likely he is guilty
- You see how two pieces of evidence combined make the killer more likely to be guilty than each individual piece of evidence alone does.

All times are GMT -8. The time now is 03:47 PM.