Drug sniffing dogs are barely better than a coin-toss

[Read the post]

Drug sniffing dogs

Well of course their accuracy might suffer a little if they’re sniffing drugs.


anecdotally, I’ve read of cases where just about ANY behavior by the dog is interpreted as an “alert” so the degree of training of the dog isn’t the biggest issue.


If any of these dogs are laid off as a result of this investigation, I’d be happy to adopt!


When they’re being directed by the trainer of course they’re going to be just as randomly correct.


So, they are less “dogs” and more “probable-cause-on-demand generators”. It’s a feature, not a bug.


We as citizens don’t know what behavior a given dog exhibits as an “Alert”
We don’t know what a given dog has been trained to “alert” to.

We DO know that dogs are incredibly easy to train to do almost anything.
We also know that police have shown time and time again that they’ll bend or break the rules, or have them rewritten in their favor in order to persecute who they thing the “bad guys” are no matter what an individual’s guaranteed rights actually are.

Is it such a stretch to believe that the police might be training their dogs to alert on cue rather than in the presence of contraband? Or both? How can we know? How can we effectively defend against it?


Now if coins could fetch, it might be a toss-up


There was a mythbusters where they tried to hide drugs from a drug-sniffing dog and were unable to fool it. So there must be at least some that do their job correctly.

1 Like

They’re way better than a coin toss, unless the coin is rigged.

There was a time when thanks to patriotic judges ‘lie detectors’ were thrown out as evidence, will we also remove presumption of good faith in trained mobile probable cause generators and see SCOTUS rule out use of dogs as probable cause?
It is as bad as the 60s through the 80s in Oregon(Multnomah county sheriff’s and court’s policy) when being inside any VW vehicle was considered probable cause.

1 Like

Lots of sniffer dogs are actually trained with substitute odours, rather than the actual substances they’ll be looking for; there are commercial substitute scents available for cocaine, heroin, LSD*, and marijuana**. Also several varieties of corpse scent; fresh, post-putrefaction, and drowned.

*Can a dog really detect the scent from the fraction of a milligram of LSD in a single dose? Presumably this is mainly for detecting labs, rather than users.
**Not specified whether this applies to hashish as well.


If there are drugs, dogs are great at finding them.

If there are no drugs, handlers are great for using dogs to say there is probable cause and search whatever they want without a warrant.

Because dogs are great at finding drugs when they are there and because half the time they come up empty, shouldn’t one conclude that half of the time it’s a waste of everyone’s time and the dog is being taken away from actually searching for real drugs?


I don’t know how to link to a specific comment on the WaPo article, so quoting one Mark Z. wholesale because I think he makes such a good point:

Drug dogs are a weird edge case. The most consistent legal category to put them in would be that they’re a measuring instrument for detecting evidence beyond the reach of human senses, like a microscope or a mass spectrometer. The “dog alert” is then an observation that the handler uses to draw a conclusion, rather than being treated as evidence. Of course then the handler would have to be qualified as an expert, and could be held accountable for her track record. Instead of “Lex has a 40% false positive rate” we’d be saying “Lex’s handler has a 40% false positive rate”. And then the police would get excoriated for putting such an unreliable “expert” on the stand, and the search of the vehicle would be inadmissible.

The way the Seventh Circuit approaches the case is to treat the dog as the expert witness. They talk about the dog’s training history, and they conclude that there’s not enough evidence to say the dog was improperly trained. Problem: You can’t cross-examine a dog. So they’ve created a category of magical expert witnesses whose reliability can never be challenged because they can’t speak.


This topic was automatically closed after 5 days. New replies are no longer allowed.