That having been said, one major problem with the Exclusionary Rule, as implemented, is that it is generally only effective in cases where evidence was obtained in a fashion which no reasonable person could judge to be reasonable. I would suggest that it needs to be bolstered by affirming the right of defendants to have jurors examine the factual matters surrounding the legitimacy of evidence, and have the jurors instructed to disregard any evidence they find was not gathered in legitimate fashion. To be sure, this check by itself would not eliminate the need to have judges enforce the Exclusionary Rule (since jurors may in some cases be loath to ignore evidence of a heinous crime, no matter how patently illegitimate the means by which it was obtained), but it would also provide some protections the exclusionary rule does not (e.g. if based on the cross examination of a cop who sought and received a search warrant, a jury concludes that the cop knew or should have known the his basis for “probable cause” was bogus, the jury could disregard the evidence gained thereby; a judge examining the issue, by comparison, would probably only strike the evidence if there was no possible way the cop could have believed his basis for “probable cause” was legitimate).

https://docs.google.com/file/d/0B61j4W5ufFbcNmJjN2IxODItMjNiOC00OTgwLWFmZjItNjMyYjkwYzhiZTFi/edit?authkey=CLrg3LkK

As any fool knows when flipping coins with the state the flip always hinges on the states: heads I win, tails you lose doctrine.

Which is more “just”: penalizing two guilty people for their crimes, or allowing two guilty people to escape without penalty because the second crime was obtaining evidence in an illegal manner?

Good idea, but then how would we convince the kids we are the
most wonderful and most free country in the world, stuff we drill
into their heads starting at age 3? With the drug dog “alert” system
we can have both the phony 4th amendment for show/indoctrination
AND be able to rifle through anyone’s stuff, anywhere…

I am going to try to be an apologist for the judge here, at least in a limited sense. Here is his mitigation:

He doesn’t want to let a guilty person go free. That is understandable. I looked it up. This was a suppression case. If he suppresses, then the guilty person goes free. If Green didn’t have cocaine, and he sued police for tossing his SUV then police might well have gone the other way with respect to Bono’s ability to give probable cause.

Now, I understand that the judge isn’t supposed to treat innocent litigants different from the guilty ones. But it is hard not to (it might be easy if one doesn’t think cocaine should be a criminal offense — but imagine that it was evidence of some crime that personally makes you angry that was sitting hidden in Green’s SUV). I think the judge gave into the understandable temptation not to let the guilty man walk. I also think that this temptation is too much for judges. Not all of them, but many of them — their intelligence and integrity notwithstanding.

This is why I often try to convince Mr. Balko (and now Ken White) that we should get rid of the exclusionary rule. It is the exclusionary rule that is making a j/k out of 4a jurisprudence and making the legal system a mockery. Now there obviously need to be other remedies to lower the incidence of unreasonable searches by police. It is fun to imagine what those remedies could be, but I will save that discussion for other thds.

For now: GET RID OF THE EXCLUSIONARY RULE! This here case under discussion shows how it leds to bad results and alwys will.

Anyone who would believe anything a cop says would buy a bridge in Brooklyn.

It would at least be more honest.

I think Bramblyspam is off a bit as well.

Missing the big picture you are, explain I will.

This has nothing to do with the dog. You could replace the dog with a divining rod, a hat with two magic stones, or a stuffed animal dressed like a drug sniffing dog.

It’s all about the handler. And the handler would get the same results with a lighter. “The lighter lit! That’s a sure sign that drugs are in the car!” Or a stick. “The stick has detected drugs!” Or hell, a Corn Dog. “The Holy Corn Dog has vibrated in the presence of illegal drugs!”

This has absolutely nothing to do with statistics or mathematics… it’s all about the perceived “accuracy” of drug sniffing dogs, along with the preconceived notion that the handlers of these dogs are honest.

In addition, there is a prescribed notion among the “Law and order” types that egregious abuses can’t happen to them, thus… it’s ok that dirtbags and lowlifes are subject to searches. After all, they’re dirtbags and lowlifes. They SHOULD be searched. These people will be happy to wink approval to otherwise illegal searches.

Stop using statistics or math in defense of the dog. It’s reasonable to assume that a perfectly trained and handled dog will be accurate 100% of the time. If the actual statistic is LESS, than that’s on the handler. Always focus on the handler.

The public needs to be trained to expect 100% accuracy from drug dogs. Not this sorry ass “Well! It was shake!” or “Uh… Drugs WERE there.” or “Derp! 50% is fine!” crap.

A dog makes a decision, either Positive or Negative, about every vehicle it is presented. Generally, we care more about how accurate those positives are *relative to the entire universe of tests*. In this case, the article tells us the number of Positives, and the number of those later determined to be True Positives (22/85 = 26%), and hence false positives (63/85 = 74%). But to understand how accurate the dog is, we need to know how many Negative indications the dog made – which the article does not tell us.

A large number of negatives would make a huge difference on the dog’s specificity (“to the ability of the test to identify negative results”), or false positive rate (the percentage of negative vehicles identified as positive), since those all include the number of true negatives in the denominator of the calculation.

Assume every Negative result is true – which seems reasonable given how well dogs can detect drugs. If the dog did not alert on 1000 vehicles over the same period, the specificity would be 1000 / (1000 + 63) = 95%, and the false positive rate would be 63 / (63 + 1000) = 6% — and that would be doggone remarkable, especially compared to the medical diagnostic tests on which we make life-and-death decisions, and in my mind definitely sufficient for probable cause.

But if there were only, say, 50 Negative results, the specificity would be 44%, the false positive rate would be 55%, and that would be atrocious.

So, I say, does anyone know where to find the number of negatives? Without it, any discussion is statistically deficient.

That statement assumes that the judge is ignorant on this issue. I believe the judge knows that drug sniffing dogs are misused by the cops and supports that misuse.

If drama is what it takes to educate the public, I’m all for it. It’s a lot better than nothing.

Given that that’s the case, the large group of ‘negatives’ that should lead to a large number of false positives becomes far less significant. So the real ‘problem’, as has been argued, is that they’re using the dog as a way to fake probable cause for a search in a group that they think they *should* be getting a much higher percentage of positives from. That they still have such a crappy rate of true to false positives means there is something else confounding the system.