The Portuguese Police thought the dogs were 100% accurate.

[John]

The Portuguese Police thought the dogs were 100% accurate.

It is evident that the Portuguese changed tactics following the introduction of the cadaver dogs into the equation.  In the following article Kate explains how they were effectively tricked into allowing their possessions and the car hire to be submitted to forensic examination.


Ricardo Paiva (Portuguese Police liaison officer) played a more prominent role in the interrogation this time, giving me his spiel about the dogs. “These dogs have a 100 per cent success rate,” he said.

“Two hundred cases and they’ve never failed.” I just stared at him, unable to hide my contempt. These dogs had never been used in Portugal before, and he knew little more about them than I did.

Could the police have been so naive?



Anguish ... Kate, with Cuddle Cat, and Gerry in tearful appeal for help days after disappearance.


KATE McCann tells today how she wrecked a bed as she kicked out in rage after the first day of the shambolic police hunt for her abducted daughter Madeleine.

http://www.thesun.co.uk/sol/homepage/news/3569557/Kate-McCann-I-smashed-bed-in-rage-at-cops.html

[debunker]

Just as naive as many [ censored word] who insist on believing that to this day!

They all made the same mistake- Grime said that never in 200 trials had Eddie alerted to pork and othe foodstuufs- very different to being 100% right about cadaver odor.

[DCI]

So Kate wouldn't answer the questions?

Carlos had advised me not to answer any of the questions put to me. He explained that this was my right as an arguida and the safest option. Any responses I gave might unintentionally implicate me in some way.

[Carana]

Ricardo Paiva (Portuguese Police liaison officer) played a more prominent role in the interrogation this time, giving me his spiel about the dogs. “These dogs have a 100 per cent success rate,” he said.

“Two hundred cases and they’ve never failed.”

Did Paiva really not understand?

[AnneGuedes]

Please, John, could you insert (...) between the different elements of the above quotation ?

[icabodcrane]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

[debunker]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

Explain please

[icabodcrane]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

Explain please

The dogs are more often accurate, than inaccurate  ...  considerably more often

If you can present evidence, statistics,  or reviews that suggest otherwise, please do

[debunker]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

Explain please

The dogs are more often accurate, than inaccurate  ...  considerably more often

If you can present evidence, statistics,  or reviews that suggest otherwise, please do

80% accuracy is not very high, and that is at the high end of estimations. One indication in five in error. Would you bet your house on it?

[icabodcrane]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

Explain please

The dogs are more often accurate, than inaccurate  ...  considerably more often

If you can present evidence, statistics,  or reviews that suggest otherwise, please do

80% accuracy is not very high, and that is at the high end of estimations. One indication in five in error. Would you bet your house on it?

80%is not very high !

If you had to bet the farm on it, which odds would you go for ...   10 to 8  ... or 2 to 10 on  ?

The fact remains that, statistically,  the dogs were far more likely to have been right in the McCann case than to have been wrong

[AnneGuedes]

“Two hundred cases and they’ve never failed.” I just stared at him, unable to hide my contempt. These dogs had never been used in Portugal before, and he knew little more about them than I did

But others knew more about them, the ones who suggested they be brought in and the ones that had workes with them, so that is a bit of an odd statement.

And if KM knew nothing about these dogs, where did her contempt for them come from? Or was it contempt for Mr Paiva, ah, must have been, after all wasnt he the Fu......To..... she referred to in her book?

This is an odd statement because the quotation is inaccurate, as I said above.
Here is the correct one :
If I’m honest, I’d been quite nervous about seeing the videos of the dogs. I had no idea what to expect, although I was quite sure something couldn’t be quite right about the results they had apparently produced. We knew from Bob Small that the responses of specialist dogs were, or ought to be, classed as intelligence, not evidence, but in my head I’d built up these film clips into the most damning ‘evidence’ imaginable; the ‘I rest my case, Your Honour’ finale. Now Ricardo was giving me his spiel about the dogs. ‘These dogs have a 100 per cent success rate,’ he said, waving an A4 document in front of me. ‘Two hundred cases and they’ve never failed. We have gone to the best laboratory in the world using low-copy DNA techniques.’ His emphasis suggested this was the gold standard. I just stared at him, unable to hide my contempt. What did he know about low-copy DNA? I was so tempted to ask him to elaborate. These dogs had never been used in Portugal before, and he knew little more about them, either, than I did.

[debunker]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

Explain please

The dogs are more often accurate, than inaccurate  ...  considerably more often

If you can present evidence, statistics,  or reviews that suggest otherwise, please do

80% accuracy is not very high, and that is at the high end of estimations. One indication in five in error. Would you bet your house on it?

80%is not very high !

If you had to bet the farm on it, which odds would you go for ...   10 to 8  ... or 2 to 10 on  ?

Neither- chance is far too great.

Your statement about the matter is mathematically flawed.

You said:

"There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case"

(I am quoting it here so that it cannot be watered down in future answers.)

The likelihood that the results in PdL were 100% accurate is going to be of the order of one in fifty. My math is rusty but if the likelihood is 80%, there is a normal distribution about that point which would predict the actual likelihood that the alerts were 85%, 90%, 95%, 99%, 100% accurate; the 100% mark would be at about 3 standard deviations giving a likelihood of 2 chances in 100. I can't remember the exact formula but that is what the math predicts.

[debunker]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

Explain please

The dogs are more often accurate, than inaccurate  ...  considerably more often

If you can present evidence, statistics,  or reviews that suggest otherwise, please do

80% accuracy is not very high, and that is at the high end of estimations. One indication in five in error. Would you bet your house on it?

80%is not very high !

If you had to bet the farm on it, which odds would you go for ...   10 to 8  ... or 2 to 10 on  ?

The fact remains that, statistically,  the dogs were far more likely to have been right in the McCann case than to have been wrong

You see, I was right to predict your back shuffling. You have already changed your claim to:

"The fact remains that, statistically,  the dogs were far more likely to have been right in the McCann case than to have been wrong"

from the original:

"There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case"

[icabodcrane]

There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case

Explain please

The dogs are more often accurate, than inaccurate  ...  considerably more often

If you can present evidence, statistics,  or reviews that suggest otherwise, please do

80% accuracy is not very high, and that is at the high end of estimations. One indication in five in error. Would you bet your house on it?

80%is not very high !

If you had to bet the farm on it, which odds would you go for ...   10 to 8  ... or 2 to 10 on  ?

Neither- chance is far too great.

Your statement about the matter is mathematically flawed.

You said:

"There is, of course,  a statistically high likelihood that the dogs were  100%  accurate in the McCann case"

(I am quoting it here so that it cannot be watered down in future answers.)

The likelihood that the results in PdL were 100% accurate is going to be of the order of one in fifty. My math is rusty but if the likelihood is 80%, there is a normal distribution about that point which would predict the actual likelihood that the alerts were 85%, 90%, 95%, 99%, 100% accurate; the 100% mark would be at about 3 standard deviations giving a likelihood of 2 chances in 100. I can't remember the exact formula but that is what the math predicts.

Statisically there is a much higher chance of the dogs having been accurate than there is of them having been inaccurate

That is a fact,  however to choose to dress it up

[debunker]

Statisically there is a much higher chance of the dogs having been accurate than there is of them having been inaccurate

That is a fact,  however to choose to dress it up

However much you choose to back track you mean.

80% reliability is pretty poor in the real world.

Would you fly in a plane that had a 80% chance of not crashing?