As a former DUI prosecutor, I worked on 1000's of cases; the main piece of evidence was typically a chemical test, which was either a blood draw or a breath sample. In Michigan, the breath sample is collected with the Datamaster machine. Under Michigan law, the trier of fact is allowed to assume that the blood alcohol content result collected via chemical test was the same as when the driver was operating the vehicle.
As a former prosecutor my favorite lines were "over the legal limit" "more than double the legal limit" "he/she failed the test". I would push and push that chemical test being over the legal limit - what else was there to consider, the person was guilty! Anything that silly defense lawyer tells you is just smoke and mirrors - there was a test, and they failed it, end of story, find them guilty.
Many prosecutors try to over-complicate the issue and act as if they need to defend the chemical test number - the more complicated you make it, the less a jury focuses on the "failed test" concept. As a prosecutor, I did the very opposite - make it as simple as possible - juries don't want to think too much, and if you spend so much time defending something, they lose confidence in the issue and question the reliability of the evidence.
As a former prosecutor I had a different view of the blood vs breath results. In my experience, a blood test is considered to be more reliable by the "criminal law community", because well it's tested at the Michigan State Police lab, and there's supposedly less variables that can impact the result, because it is coming directly from your body. The person testifying about the result (lab tech) is also supposedly more knowledgeable and a better witness than a police officer who does the Datamaster machine.
Both prosecutors and defense lawyers tend to give the blood result more weight and if it's over the legal limit for an OWI (0.08) or Super Drunk (0.17) then there's less incentive to challenge it. I completely disagree with this perception, and bring this mindset to my criminal defense practice. Here's why:
In Michigan, let's say the police stop my client at 2 am. A Datamaster result would usually happen within the hour of this traffic stop, and makes for a better argument when you're telling a jury to believe the BAC level is the same as when the person was driving the car. A blood test in contrast happens hours later at the hospital, or you have to wait for someone to come to the jail to draw the blood. Typically double or triple the time goes by - a good defense lawyer will jump on that argument.
Once the blood is drawn, it's supposed to be sealed in a tube for testing, but that sealed tube sits around and is transported to the lab over the course of a few days. Just picture that sealed blood sample bouncing around the mail truck as it heads to Lansing in an uncontrolled environment. Throw in freezing temperatures of winter or 95 degrees in August, and that sample is compromised. The best is when the lab tech brags about the sample being stored in a controlled environment prior to testing (but after bouncing around the mail truck in 95 degree weather for 3 days) - all I have to say is REASONABLE DOUBT.
It's a pain to get the lab tech to come to court to testify and limits your flexibility to set the case for trial. It's so much easier to have the police officer available to come in and testify.
Sure a Datamaster has it's faults, but cops are trained not to know enough to be challenged. They have a procedure and if they follow it, that's it. They are not required to know how the machine works or why it does X, Y and Z but rather to follow steps. If the jury likes the cop and they believe he followed the steps then they will probably buy into the result and not find reasonable doubt. There's also usually two tests (not required), but the second result can confirm the first result vs a single blood test result.
As a criminal defense lawyer, I find that prosecutors are more confident in the blood draw, but at trial those are easier to challenge. And the goal is not to show the test is wrong, faulty or an error, but rather just raise enough doubt to get a not-guilty verdict, and the door is wide open on the blood test.
With the Datamaster, I am looking for numbers close to the limits of 0.08 and 0.17 as there are built in margins of error listed in the training manual for the operator. I am also looking for a police officer who is going to be chatty and give me open windows. Officers are trained not to discuss how the machine works or WHAT IF's but human nature makes most officers answer questions and be open to possibilities. If the officer answers YES IT"S POSSIBLE to enough questions then that's my reasonable doubt. After all it's a machine and the person pressing the buttons doesn't know how it works, why should be believe the result?