Drunk driving legalized-by-proxy in Florida

[BicycleTree]

How is that not what I said? I said all defendants, of any crime, have a right to any and all information about how the evidence against them was gathered and analyzed. Period.

Actually, I was referring to Moonbear and StatutoryApe, who were focusing on the rights of a single defendant and circumstances that might make a single defendant's trial more or less fair. The focus of your post did not appear to be so narrow.

To expand on that, it's not reasonable for a defendant to spend $10,000 in research to avoid a $500 fine or a couple days in jail (or whatever the penalty is). Something like that is only reasonable if the interests of many defendants are at stake. As hitssquad pointed out, some reasonable doubt is permissible in an isolated case for a DUI charge.

Why reverse engineer? The company should simply provide the information. If they refuse to, then the government should not be buying from them. PLain and simple.

Yes--it should. But reverse engineering is always going to be the integral part of it, because if you can't trust the company's product then you can't trust the company's plans.

 

[franznietzsche]

Yes--it should. But reverse engineering is always going to be the integral part of it, because if you can't trust the company's product then you can't trust the company's plans.

I wasn't claiming that they intentionally would make the product inaccurate, and then cover it up--that's a different manner entirely. Assuming that any flaws would be unintentional, them divulging their own specifications to which they built the product should be sufficient, assuming these are sufficiently detailed.

 

[BicycleTree]

Why would you assume that a scenario similar to the one I proposed in post #9 of this thread will never happen?


Also, if the device may be incompetently designed, it may be incompetently manufactured even from correct designs.

 

[Moonbear]

Actually, I was referring to Moonbear and StatutoryApe, who were focusing on the rights of a single defendant and circumstances that might make a single defendant's trial more or less fair. The focus of your post did not appear to be so narrow.

How was my post only applying to a single defendant? I gave an example of how an inaccurate machine might affect a certain defendant, but that was simply an example to illustrate why it is important to know how the machine works if it is going to be used to provide legal evidence. I don't think we're in disagreement on this issue, I just tried to give an example that doesn't appear quite as far-fetched as your example. Rounding errors would be far more likely than something that would detect different BAC in a breath test of people from different countries, but I realize you're just giving an exaggerated example for illustration there.

You would also need to show that any individual machine is accurate at the time it is used, not just that the design had been tested prior to production or that the specific machine was tested prior to shipment, but that it is regularly calibrated accurately and does not have an inherent bias in the programming that might not be detected by testing with a set of standards. Here's another example: gas pumps can be set to cheat the customer. For a long time, when whatever regulatory agency came around to check the calibration of the machine, they always measured fixed amounts of gas, like 5, 10 and 15 gallons to check that what was pumped is what the pump read and charged for. So what the unscrupulous did was set the pumps so they would "catch up" at 5, 10 and 15 gallons, and meet certification requirements. Now, a customer pulls up and pumps 7.5 gallons, and only gets 7 gallons, and that was not detected by the calibration/standardization method.

So, the machine might test accurately against a set of standards, but there may be inaccuracies outside of that range that could be determined if one examined the inner workings, or there could be a manufacturing problem in whatever the standards are that are used to calibrate the machine.

 

[BicycleTree]

All right--perhaps you did intend your comments to have broader scope than the rights of individuals in individual cases, and just didn't make that clear.


No, my example was not particularly exaggerated. Is there some technical reason why a racist couldn't design a breathalyzer that counts ethnic foods as inebriation?

But your example:

If someone gets to walk when they have a BAC of 0.099999999 and is convicted at 0.1, regardless if they were clearly impaired in their ability to drive, then how the machine handles rounding becomes an issue.

.1 - 0.099999999 = 0.000000001. So this is only an issue if the precision of the breathalyzer is on the order of 0.000000001. Given the information here: http://www2.potsdam.edu/hansondj/DrivingIssues/1093825780.html final rounding is definitely not an issue.

Research indicates that breath tests vary at least 15% from actual blood alcohol concentration

 

[Moonbear]

But your example:

.1 - 0.099999999 = 0.000000001. So this is only an issue if the precision of the breathalyzer is on the order of 0.000000001.

Yeah, that part I exaggerated. But say the rounding was at 0.01 for several calculations made to come up with the final result (not knowing how the things work, I don't know what calculations are involved), it could start being more significant.

Given the information here: http://www2.potsdam.edu/hansondj/DrivingIssues/1093825780.html final rounding is definitely not an issue.

Originally Posted by that site

Research indicates that breath tests vary at least 15% from actual blood alcohol concentration

Wouldn't that suggest it could be an issue? Not at the miniscule level I indicated, but at a larger level?

Either way, if what that site states is true, then there's even more reason to evaluate how the breathalyzer actually works to know if it's giving high readings; I would presume this would vary from manufacturer to manufacturer as well. We don't even know if the error is linear. Can you just subtract 15% to get the real value, or is someone with a lower BAC going to have a much greater discrepancy between their real BAC and the breathalyzer result than someone with a higher BAC, or vice versa? This certainly at least goes toward having the right to know how accurate the machine is, how it calculates BAC, and how it is calibrated to ensure that accuracy on a regular basis. Knowing how the machine operates also will indicate how much drift in the values can be expected over time. If the police are getting more convictions, they might not themselves question the machine, even if it's resulting in convictions of innocent people.

 

[Evo]

In a two year actual comparison of breath vs blood tests on drunk drivers, breathalyzer results were at or below the actual blood test results 94% of the time. The other 6% of the time it was only off by 0.01 g/210 L.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=3668478&dopt=Abstract

I guess if you're smart, you'll go with the breathalyzer.

If your breathalyzer test says your blood alcohol level is 1.2 and .08 is legally drunk, the small chnace it's slightly over won't matter. If your blood alcohol level is only slightly over the legal limit, they will usually do another breath test or do blood/urine to verify. You also have the right to get your own independant testing done.

 

[franznietzsche]

In a two year actual comparison of breath vs blood tests on drunk drivers, breathalyzer results were at or below the actual blood test results 94% of the time. The other 6% of the time it was only off by 0.01 g/210 L.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=3668478&dopt=Abstract

I guess if you're smart, you'll go with the breathalyzer.

If your breathalyzer test says your blood alcohol level is 1.2 and .08 is legally drunk, the small chnace it's slightly over won't matter. If your blood alcohol level is only slightly over the legal limit, they will usually do another breath test or do blood/urine to verify. You also have the right to get your own independant testing done.

Its irrelevant whether it actualy makes a difference. A defendant still has the right to all relevant information. Even if it doesn't change a single conviction, and the machines were 100% accurate, handed down on Mt Sinai, the defendant still has a right to that information, and unless that information can be provided the evidence should not be permissible.

 

[TheStatutoryApe]

I read your previous post Evo. I was just responding to hitsquad. Thank you, it was interesting.

And BicycleTree I think that the rights of any defendant to have access to information regarding the evidence brought against them is a much more broad argument than the narrow possability that there is a fault in the source code used in breathalizers.

As the information Evo has pointed out shows there is very little fault in the process of the machines workings. I doubt any of the judges involved in these cases would even question the reliability. The point here is the defendants right to question the evidence versus the breathalizer designers right to keep his designs secret.

 

[BicycleTree]

Wouldn't that suggest it could be an issue? Not at the miniscule level I indicated, but at a larger level?

There is a very big difference between rounding errors and total inaccuracy. The total inaccuracy was what was referred to in the article as 15%, which says nothing about rounding error.

 

[BicycleTree]

In a two year actual comparison of breath vs blood tests on drunk drivers, breathalyzer results were at or below the actual blood test results 94% of the time. The other 6% of the time it was only off by 0.01 g/210 L.

Actually, that's not what the article says. The other 6% of the time it was higher by more than 0.01 g/210 L.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=3668478&dopt=Abstract

 

[Evo]

Actually, that's not what the article says. The other 6% of the time it was higher by more than 0.01 g/210 L.

Obviously not much more than.

 

[Artman]

Its irrelevant whether it actualy makes a difference. A defendant still has the right to all relevant information.

Who says? These guys are just fishing for a loophole, is that really justice?

 

[BicycleTree]

Obviously not much more than.

True. But you should admit your mistake before you excuse it as minor.

 

[TheStatutoryApe]

Its irrelevant whether it actualy makes a difference. A defendant still has the right to all relevant information.

Who says? These guys are just fishing for a loophole, is that really justice?

The point is that all trials are supposed to be fair and there is a set manner in which it is determined whether or not a case is being made fairly. This should not be disrupted just because the person is not innocent. Even a person who is guilty deserves a fair trial.

 

[Moonbear]

The point is that all trials are supposed to be fair and there is a set manner in which it is determined whether or not a case is being made fairly. This should not be disrupted just because the person is not innocent. Even a person who is guilty deserves a fair trial.

In addition, the last time I checked, people are still presumed innocent until proven guilty. By deciding that these are just people looking for loopholes presumes they are guilty before they've had a fair trial. They could be looking to the details because they are certain they didn't have enough to drink to be over the legal limit and need to show there's a problem with the evidence against them.

This one is old too, but it shows that not all breath alcohol analyzers (I've realized that Breathalyzer is a brand name, so I'll be careful not to use that term generically now) are created equal:

J Forensic Sci. 1985 Oct;30(4):1058-63.

An in vitro study of the accuracy and precision of Breathalyzer models 900, 900A, and 1000.

Caplan YH, Yohman DT, Schaefer JA.

Ninety Breathalyzer instruments (Model 1000) and twenty instruments (Models 900, 900A) were studied using a protocol described by the Department of Transportation's "Standard for Devices to Measure Breath Alcohol." Although the mean of each of three concentrations tested (0.05, 0.10, and 0.15 g/210 L) compared favorably in both series, the standard deviation was consistently higher for the Model 1000 instruments. The Model 1000 instruments also produced a significant number of test results which exceeded the normally expected scientific deviation.

Unfortunately, I don't have access to the Journal of Forensic Sciences, so can't read the full length articles for this one or the one Evo cited. In the one Evo cited, I especially wanted to read the methods and see the results presented because the abstract mentions breath alcohol and blood alcohol tests were within an hour of each other but doesn't specify if the order was always consistent or how much variation there was.

Also, this following abstract raises some additional questions about interference with breath alcohol analyzers. The bold emphasis is mine. Again, I can't access the full article, so can only go by what the abstract suggests.

J Forensic Sci Soc. 1990 Nov-Dec;30(6):349-56.Evaluation of the Drager Alcotest model 7110 infrared breath alcohol analysing instrument.

Smith DJ, Laslett R.

Forensic Science Centre, Adelaide, Australia.

Until 1986, commercially-available infrared breath alcohol analysing instruments employed wavelengths in the region of 3.4 mu. The move to the 9.5 mu region in the Drager Alcotest 7110 promised greater discrimination against endogenous compounds such as acetone. The present study confirmed that acetone interference is insignificant and that in terms of in vitro accuracy and precision, the ten 7110 units tested were superior to the Breathalyzer 900, the instruments they will replace for evidential testing in South Australia. The new unit meets the South Australian Police demand for portability and its shielding prevents interference from any of the common radio frequency transmissions in Adelaide when operating as near to the source as possible. Comparisons of breath results (monthly averages) and their corresponding blood results accumulated during the first few months of operation showed no bias between the two techniques.

 

[Artman]

The point is that all trials are supposed to be fair and there is a set manner in which it is determined whether or not a case is being made fairly. This should not be disrupted just because the person is not innocent. Even a person who is guilty deserves a fair trial.

True, but a loop hole is not fair either. A fair trail is a trial that fairly judges guilt or innocence not one that allows a guilty person to walk on a technicality.

 

[BicycleTree]

Wouldn't that suggest it could be an issue? Not at the miniscule level I indicated, but at a larger level?

There is a very big difference between rounding errors and total inaccuracy. The total inaccuracy was what was referred to in the article as 15%, which says nothing about rounding error.

Weelll?10charac[/color]

 

[TheStatutoryApe]

True, but a loop hole is not fair either. A fair trail is a trial that fairly judges guilt or innocence not one that allows a guilty person to walk on a technicality.

I agree the loophole should be taken care of. You can make a law, or change the existing one(s), stating that the test is acceptable by the standards of the judiciary and will not be allowed to be questioned. You can form a compromise where the information regarding reliability, but not actual design, will be made available for any person standing trial for a DUI. Or you can tell the companies that produce these devices that the city will only purchase and use them so long as they are willing to provide design and reliability information.
When these people walk due to technicalities it shows us the loop holes and gives us the chance to fix them. Like I and Moonbear said previously these judges could have decided the person guilty based solely off the police officer's testimony if they wanted too. Perhaps part of the problem here was that the officers weren't present. Maybe that should be fixed too.

 

[hitssquad]

Perhaps part of the problem here was that the officers weren't present.

The case would be dismissed in that case, and therefore there would be no need to mount a defense.

 

[hitssquad]

Any person accused of a crime has the right to know by what means they have been found guilty of commiting one and how reliable it is.

Please clarify what you mean by what right a person has.

 

[hitssquad]

U.S. legal conviction thresholds

some reasonable doubt is permissible in an isolated case for a DUI charge

In the United States, reasonable doubt (not some reasonable doubt) is always (not sometimes) the threshold beyond which a person can be prosecuted for a crime.

 

[hitssquad]

Apparently at least some breathalyzers are not even fooled by oral feces.
http://www.newsoftheweird.com/archive/nw050529.html

--
A DUI suspect (unnamed in a March Toronto Sun report) put a handful of his own feces in his mouth in a police station in what officers said was an attempt to foil a Breathalyzer test. Said an official, "I don't think alcohol alone would make you do (that)." Nonetheless, said police, the man, who had been stopped on Highway 11 near Barrie, Ontario, still registered double the threshold for impairment. [Toronto Sun, 3-30-05]
--

 

[Moonbear]

Apparently at least some breathalyzers are not even fooled by oral feces.
http://www.newsoftheweird.com/archive/nw050529.html

--
A DUI suspect (unnamed in a March Toronto Sun report) put a handful of his own feces in his mouth in a police station in what officers said was an attempt to foil a Breathalyzer test. Said an official, "I don't think alcohol alone would make you do (that)." Nonetheless, said police, the man, who had been stopped on Highway 11 near Barrie, Ontario, still registered double the threshold for impairment. [Toronto Sun, 3-30-05]
--

I'd have to agree that guy had more problems than inebriation impairing his judgement.