Unsolved Madeleine McCann * Current Trial of Main Suspect Christian Brueckner

[faible]

verification of THE dog.

You've missed handler error.

 

[Angry Red Bull]

Glad someone mentioned it.

Probablity decreases with each subsequent step in the chain of events. It doesn't increase.

Yes but we are talking two types of probabilities here. The probability that it did have a cadaver at each location on two locations and probability as inverse of those events.

The probability of two events happening in this case errors is multiplied each event.

The math is correct and in a sense you are right. The probability of a negative the 3% is still multiplied in a chain.

The probability of of error on one is 3% . The probability of errors twice in two separate events is the multiplier

 

[Angry Red Bull]

You are really digging a hole for yourself. I suggest you just stop.
I MULTIPLIED 0.97 by 0.97 (0.97 being the probability of success). The probability of success in two successive events is 0.97 TIMES 0.97 = 0.941. Therefore the probability of failure - 0.059 = 5.9%.

A win is the dog making a mistake 33-1

Two wins in a row two mistakes in a row is probability of each multiplied

33 X 33

 

[31550]

A win is the dog making a mistake 33-1

Two wins in a row two mistakes in a row is probability of each multiplied

33 X 33

But they only need to make one mistake for the evidence to be meaningless. You are relying on the dog being correct in BOTH locations.

 

[Angry Red Bull]

But they only need to make one mistake for the evidence to be meaningless. You are relying on the dog being correct in BOTH locations.

Yes but we have two results and BOTH gave indications so therefore it is multiplied

 

[31550]

Yes but we have two results and BOTH gave indications

With a probability of 5.9% that one or both are false.

 

[Angry Red Bull]

With a probability of 5.9% that one or both are false.

I've explained consistent with multiplication theory of probability. It's right.

A WIN is two mistakes in a line each linking to the next. If probability of one mistake is 33-1 then for you to WIN you have to multiply because the probability is harder by each events probability

Back to dogs it means if they gave an indication at BOTH locations (and they did) the chance they are wrong by doing THAT (not your hypothetical alterative you choose) is the multiple of the individual probabilities of them being wrong.

 

[Kurve]

 

[Angry Red Bull]

It's important and that's why I actually did the math and laid it out. At days end at 1 in 6806 it makes it highly improbable to be unrelated mistakes of the dogs. I simply can't accept that.

Can you say a single dog had error rates as part of the 97% mean? Sure you can BUT Eddie's history was a 100% record success in over 200 cases. That's why he was chosen. So it's highly unlikely he would depart statistically from the scientific study

 

[jason_recliner]

At a more fundamental level, a paper finding dogs were successful 97% of the time doesn't mean it's 97% certain a body was in the unit or car or wherever. The study you keep referencing doesn't factor in the biases and errors specific to Maddie's kidnapping and murder. You can't take one study and apply it to a real event and expect the same results. It might be applicable (due to chance) but it almost certainly isn't (particularly considering all the evidence indicating Maddie's parents aren't involved).

 

[31550]

It's important and that's why I actually did the math and laid it out. At days end at 1 in 6806 it makes it highly improbable to be unrelated mistakes of the dogs. I simply can't accept that.

Can you say a single dog had error rates as part of the 97% mean? Sure you can BUT Eddie's history was a 100% record success in over 200 cases. That's why he was chosen. So it's highly unlikely he would depart statistically from the scientific study

Your continual reference to "1 in 6806" is straight out misleading.

The studies seem to suggest that cadaver dogs have a mean success rate of 97%. This equals a failure rate of 3%. In 3 out of 100 cases they will either falsely detect or fail to detect.

Even though Eddie may have been considered 'successful' in over 200 cases, this doesn't mean he is 100% accurate every time. How would you know for sure he was correct in those 200 cases? Maybe they only brought him in where they already strongly suspected he would find something. If Eddie didn't detect anything, how would they possibly know if he had missed something?

 

[faible]

I've posted this prior, obviously reddit post so take it or leave it:

I am a cadaver dog handler and I think there's a very high chance of false positives in this case. The first thing to consider is that there were two dogs, but there was only one handler, and most false alerts are handler error.

As others have noted, there were huge breaches of search protocol in the vehicle search. Any dog will eventually give a false alert if you keep telling it to work the same area over and over, which is what happened. It's been awhile since I read it, but I recall that the pattern of alerts in the apartment also made me suspicious that they were reworking the dogs over the same areas over and over again in there. So they were basically asking for a false alert.

Even if the alerts were correct, though, we're talking about hotel rooms/vehicles where who knows what could have occurred. Someone crashes their bike and bleeds all over their clothes, then drops them in the trunk of the car and the fluid soaks into the fibers of the upholstery (or behind the sofa, or anywhere else the dogs alerted)? That could be enough to get a cadaver dog alert even though it has no relevance to the McCann case.

Dogs are pretty amazing and I have a tremendous amount of faith in mine, and all of their alerts have been backed up by forensic evidence. I'd still never convict someone just based on the evidence they provide, especially since on a few occasions that forensic evidence showed that, while the alert was correct (there was actually blood there), it had nothing to do with the actual mission we were on.

I guess my answer to that would be that dogs are always right until they're not. This was a pretty unusual case with a lot of media scrutiny and pressure on the handler to find something, and I do think that played a role. If you watch the video another commenter posted, it's not even slight changes. I will admit I hadn't even actually watched it until yesterday (just relied on written reports), and it was actually worse than what I was picturing. The handler moved on quickly from all of the control vehicles when his dog didn't show interest, but called him back like 4-5 times to the McCanns' vehicle, tapped it, gestured towards it, etc. even after the dog had repeatedly moved on. I think even my younger dog, who has a notorious, "**** you, I know my job" attitude so it's hard to get in her way, would give a false alert if I did that.

Dogs' noses can definitely be more sensitive than current forensic collection techniques, at least from my understanding, but I couldn't tell you exactly why. We're really still just starting to understand exactly what dogs are detecting. But forensics teams are usually looking for specific compounds, like blood, trace DNA, etc. Dogs' noses are extremely sensitive and can detect more than that, or samples that are too small to get any valuable information from.

A 97% success rate in a single controlled study is not necessarily representative of real-world performance nor does it allow for handler error. As this poster notes, he has serious issues with the way the handler conducted the search, and given the high-profile nature of the case it's quite possible the handler was acting in an unusual fashion which prompted the dogs to react. Human error.

 

[Angry Red Bull]

Your continual reference to "1 in 6806" is straight out misleading.

The studies seem to suggest that cadaver dogs have a mean success rate of 97%. This equals a failure rate of 3%. In 3 out of 100 cases they will either falsely detect or fail to detect.

Even though Eddie may have been considered 'successful' in over 200 cases, this doesn't mean he is 100% accurate every time. How would you know for sure he was correct in those 200 cases? Maybe they only brought him in where they already strongly suspected he would find something. If Eddie didn't detect anything, how would they possibly know if he had missed something?

That is the probability based on statistical accuracy scientifically tested. Ive not misled anyone. I understand completely that a dog a single dog won't necessarily have mean data.....so his history is relevant. We would need to delve into parameters of the statistical analysis and YES how the handlers measured success of his dog has bearing on that dogs relationship to the study outcomes.

I suspect the error rate is of incorrect indications given not where there were none. So situations where he indicated but there was no DNA material there. That is how they train the dogs to search and find KNOWN positive indications. It enables them to then determine whether the dog has attained a standard of accuracy to be put in investigative work.

I've watched training videos of Eddie & Keela

If anything if there is a bias in my analysis it is away from finding involvement, by all those qualitative considerations I didn't factor in in quantitative way. So the probability is significantly more likely that those indications are Maddie than what I state. An actuary could quantify them. I can't

 

[faible]

Ive not misled anyone.

there is a 97% chance a dead body was behind the sofa.

99.99% that there was a cadaver in both places

 

[Angry Red Bull]

At a more fundamental level, a paper finding dogs were successful 97% of the time doesn't mean it's 97% certain a body was in the unit or car or wherever. The study you keep referencing doesn't factor in the biases and errors specific to Maddie's kidnapping and murder. You can't take one study and apply it to a real event and expect the same results. It might be applicable (due to chance) but it almost certainly isn't (particularly considering all the evidence indicating Maddie's parents aren't involved).

I would love to hear you outline the biases and errors. I would also like you to list the evidence indicating non involvement. Please do. Point form is fine please.

 

[faible]

I would love to hear you outline the biases and errors. I would also like you to list the evidence indicating non involvement. Please do. Point form is fine please.

1. Multiple police investigations over 17 years and zero convictions or charges.

 

[jason_recliner]

I would love to hear you outline the biases and errors. I would also like you to list the evidence indicating non involvement. Please do. Point form is fine please.

It would be a pointless waste of my time. More or less every position you have taken on this forum has proven to be wrong yet you're so deluded by self grandeur that you cannot see you are wrong about this one as well.

Faible has summed it up perfectly succinctly above.

 

[Angry Red Bull]

It would be a pointless waste of my time. More or less every position you have taken on this forum has proven to be wrong yet you're so deluded by self grandeur that you cannot see you are wrong about this one as well.

Faible has summed it up perfectly succinctly above.

You're as bad as CW on CB. Believe me I know he is guilty.

In your case ......believe me I know the McCanns weren't involved. Then explain it to me please.

 

[jason_recliner]

You're as bad as CW on CB. Believe me I know he is guilty.

In your case ......believe me I know the McCanns weren't involved. Then explain it to me please.

 

[jason_recliner]

You're as bad as CW on CB. Believe me I know he is guilty.

In your case ......believe me I know the McCanns weren't involved. Then explain it to me please.

I (and others) have literally and clearly explained, both at a high level and in detail, how you are incorrectly and inappropriately trying to correlate the findings of a study with an event and you willfully, arrogantly, delusionally ignore it all. Why should I waste any more time?

 

[Angry Red Bull]

"According to several feline behaviorists and child psychologists, an adult cat's IQ is comparable to that of a two to three-year-old child since both species learn through imitating, observing, and experimenting"

Well that probably explains why I'm unable to hold a sensible conversion with you. Impressive you can type though

 

[31550]

Mods. Clean up on Aisle 3.

 

[FR0GGY]

Mods. Clean up on Aisle 3.

Behind the couch you say?!!

 

[Sweeetleftfoot]

1. Multiple police investigations over 17 years and zero convictions or charges.

Because the uk police have never investigated the parents. They have investigated the abduction route from day 1.

Surely any investigation needs all
Angles covered at a minimum.

UK police guilty of flawed tunnel vision in hunt for Maddie McCann answers, former top cop says

Scotland Yard's six-year investigation into Madeleine McCann's disappearance was a poisoned chalice laced w...

www.9news.com.au

 

[bomberclifford]

Yes but we are talking two types of probabilities here. The probability that it did have a cadaver at each location on two locations and probability as inverse of those events.

The probability of two events happening in this case errors is multiplied each event.

The math is correct and in a sense you are right. The probability of a negative the 3% is still multiplied in a chain.

The probability of of error on one is 3% . The probability of errors twice in two separate events is the multiplier

To be blunt, this reads like you're doing a bit. It's so convoluted and wrong that the only way to explain it is you're doing an intentional parody.