CSI and cadaver dogs - some facts and statistics.

[OxfordBloo]

Irrelevant. They sent their best dogs. Why are they the best? I suggest you research the actual dogs involved in this case.

Aside from Grime's advertising their is no reason to believe they were the best dogs or anything out of the ordinary.

That is just a desire on your part, you have no evidence.

Until you produce clear and compelling scientifically justified EVIDENCE that they were superior to other dogs, we must continue to assume that they bad a similar competence to any other scent dog.

Over to you for a cite to support your belief.

[slartibartfast]

No. Consider a fair coin. One toss has the possibility of fifty per cent for heads. Two serial tosses have the probability of 25% for two heads.

Consider a pair of dice. On the first roll the chances of throwing a six is about 16%. Two rolls of the dice, the chances are 0.025.

Probabilities are transitive.

So for a serial test involving 80% probability of correctness each, the probability is 8/10 times 8/10 or 64%.

You are using an arbitrary definition of "correct".

[OxfordBloo]

You are using an arbitrary definition of "correct".

No I am not.

[slartibartfast]

No I am not.

Yes you are.

[Alice Purjorick]

No. Consider a fair coin. One toss has the possibility of fifty per cent for heads. Two serial tosses have the probability of 25% for two heads.

Consider a pair of dice. On the first roll the chances of throwing a six is about 16%. Two rolls of the dice, the chances are 0.025.

Probabilities are transitive.

So for a serial test involving 80% probability of correctness each, the probability is 8/10 times 8/10 or 64%.

Is that transitive in the mathematical sense or in the grammatical sense?

[pathfinder73]

Aside from Grime's advertising their is no reason to believe they were the best dogs or anything out of the ordinary.

That is just a desire on your part, you have no evidence.

Until you produce clear and compelling scientifically justified EVIDENCE that they were superior to other dogs, we must continue to assume that they bad a similar competence to any other scent dog.

Over to you for a cite to support your belief.

The police who have worked on cases with these dogs. Go and ask them and SY why they are using the unreliable April Jones springer spaniels in PDL.

A reality check.

[sadie]

The police who have worked on cases with these dogs. Go and ask them and SY why they are using the unreliable April Jones springer spaniels in PDL.

A reality check.

Impressive, agreed


But, I think, same as bodies, meat under water will bloat as it decomposes and bubbles will come up?


Am not sure about that but it seems to make sense to me.


Never-the-less impressive.




But what are you trying to prove, Pfinder? 
We now know that none of the alerts by Eddie and Keela PROVED anything.  No forensic back up and also obvious errors.

[OxfordBloo]

Is that transitive in the mathematical sense or in the grammatical sense?

Mathematics.

[Alice Purjorick]

Mathematics.

And you believe the transitive property always holds?

[OxfordBloo]

And you believe the transitive property always holds?

No.

But in statistical analysis of unrelated but serial events the probability is carried transitively from test to test, resulting in a reduction of probability over a series.

If dogs were 100% accurate (which we know they are not) then serial tests would remain 100%
but for lesser certainties the fractions multiply and reduce the probability

90% means 80% with two tests

80 means 64

70 means 49. (no better than a coin flip)

[slartibartfast]

No.

But in statistical analysis of unrelated but serial events the probability is carried transitively from test to test, resulting in a reduction of probability over a series.

If dogs were 100% accurate (which we know they are not) then serial tests would remain 100%
but for lesser certainties the fractions multiply and reduce the probability

90% means 80% with two tests

80 means 64

70 means 49. (no better than a coin flip)

..and if they are 20% inaccurate the fractions multiply and twice means 4% probability of inaccuracy etc.

[OxfordBloo]

..and if they are 20% inaccurate the fractions multiply and twice means 4% probability of inaccuracy etc.

No.

If they were looking for the same thing then a second alert would increase the likelihood as you say; the accuracy would increase. But in the case of searching for cadaver it requires two separate alerts to two separate cases- Keela is failing to react to anything and Eddie is alerting potentially to cadaver. So in this case the probabilities multiply and reduce the accuracy.

Have you studied statistics or are you working on common sense. The one thing that you learn about statistics is that common sense often leads you astray in dealing with such calculations.

[slartibartfast]

No.

If they were looking for the same thing then a second alert would increase the likelihood as you say; the accuracy would increase. But in the case of searching for cadaver it requires two separate alerts to two separate cases- Keela is failing to react to anything and Eddie is alerting potentially to cadaver. So in this case the probabilities multiply and reduce the accuracy.

Have you studied statistics or are you working on common sense. The one thing that you learn about statistics is that common sense often leads you astray in dealing with such calculations.

Yes, I have a degree in it which is why I am questioning you arbitrary selection of the probability you want to combine.

[OxfordBloo]

Yes, I have a degree in it which is why I am questioning you arbitrary selection of the probability you want to combine.

If you have a degree in statistics you should understand the argument above; anyone can claim anything on here and I do not accept your expertise. Try working it through from first principles.

Two blood dogs each with an average of 80% reliability.

Parallel Case

Dog one alerts giving one reason to believe that there is the possibility of 4 in 5 cases being correct.

Dog two alerts giving one reason to believe that there is a possibility of 4 in 5 cases being correct.

As you are repeating the same test, the second test increases the accuracy and two alerts increase the chance of the dogs being correct.

Serial Case

Dog one alerts giving one reason to believe that there is the possibility of 4 in 5 cases being correct.

Dog two alerts giving one reason to believe that there is a possibility of 4 in 5 cases being correct.

As the two rests are not connected the second test introduces inaccuracy and the accuracy is decreased.

[G-Unit]

As has been observed, statistics are fine for an overview, but don't have any value in an individual case. I've smoked since i was 15 and have no health problems whatsoever.