How could the charges be structured so that the jury feels the most serious charge is overreach (lowest probability of conviction), but the less serious charge (highest probability of conviction) then seems reasonable? Doesn't that give the jury a low hanging fruit sense of accomplishment in administering justice because they eliminated the charges with lowest p-values? Doesn't that effectively guarantee a conviction of *something*, given enough charges? For instance: charge someone with DUI, DWI, reckless driving, speeding, ignoring commands to pull over, driving on the shoulder, rolling through a stop sign (albeit very slowly), 10 counts of making a turn without signaling, and driving too slowly (at a time different than when speeding). Isn't the jury now tempted to make a decision by finding some sort of median in this Gaussian of charges? Suppose your jury does not need a unanimous decision --- how does that shift the effective median? What has to happen so that the median says "acquit of everything"? Doesn't that lead the jury not to do that because then the defendant could "get away with it"? Don't acquittals behave, roughly speaking, according to a geometric distribution on the number of charges? On 5/10/19 20:46 , Brent Meeker via math-fun wrote:
Interesting. In California DWI and driving with a blood alcohol level above 0.08% (formerly 0.1%), called DUI, are separate offenses. I was on a jury in which the defendant was charged with both. The cops said he had ignored their red light behind him for a couple of blocks. He didn't pass their walk-along-this-line test, but he passed the others. His BAC tested at 0.1% by the urine method which was the limit at that time. The defense had a former police lab employee testify. I asked him what was the uncertainty in the a measurment (in CA a juror can ask a question by writing it out and giving it the judge who passes it to whichever lawyer he thinks appropriate)? He said it was 0.01% So on the grounds that 0.099 to 0.101 was not above 0.1 beyond a reasonable doubt, I voted to acquit on the DUI count. But we convicted on the DWI count. I was convinced that if he didn't notice a flashing red light in his rearview mirror for two blocks he wasn't fit to drive.
I think you were right to want to convict. Surely he knew he had diabetes and was more susceptible to alcohol.
Brent
On 5/10/2019 7:54 PM, James Propp wrote:
I've written another essay (a little shorter than usual), which I'll either call A Mathematician in the Jury Box, or, “But how should we define ‘intoxicated’?” or “But how should we define ‘intoxicated’?”, or, A Mathematician in the Jury Box
(I can't decide which is better). See http://mathenchant.org/048-draft2.pdf
Comments are welcome!
Jim _______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun