It seems to me that the current legal system works like this:

Prosecution presents objective facts. Defense refutes facts and conclusions based on them. Jury decides guilt/innocence based on dependant's shoes.

Okay, perhaps I'm being a little spurious here... in reality the lawyers are more likely to discuss the defendant's shoes than objective facts. (substitute the words subjective viewpoint for defendant's clues to get any idea what I am talking about.)

This system has basically remained static since... well a long time ago... although the mechanics of it may have changed a great deal, but the idea seems to be that the jury has to get a general idea of the guilt/innocence of the defendant and see if this is more or less than some undefined arbitrary reasonable doubt. This system evolved in a time before probability or statistics hardly existed...

So I propose the following: Evidence is presented in specific seperate events by the prosecution, which the defense attempts to refute. The jury reaches a mutually acceptable probability for each of the events occuring, and thus calculates the overall probability of the defendant's innocence based on these probabilities. Then this probability is compared with a standard level used for reasonable doubt- which is essentially the number of innocent people that it is acceptable to send to jail to preserve justice- I would suggest about 5% (i.e. 5 innocent people wrongly convicted for every 95 guilty). This is also the level usually used by biologists for statistical tests.

In addition I would consider the following: That these probability lists be published, so that appeals could be directed perhaps at individual events rather than the whole case.

That rather than a clean cutoff point of 5%, have a gradiated sentencing system based on the calculated doubt in the assertion of guilt- i.e 2% chance of innocence, 10 years; 4% chance of innocence, 5 years; etc...
-- radicalllama, Aug 30 2004

How can your subtitle be "Justice For All" when you suggest that 5 innocents imprisoned for every 95 guilty is acceptable?
Doubt is not a hard entity. How can someone determine the difference between being 5% doubtful or 6%? If you are applying statistics based on an intangible, the results are not repeatable. So where you say that current juries use "some undefined arbitrary reasonable doubt" there is also no definition of the arbitrary level of "5% doubt".
-- reap, Aug 30 2004


I think that your opposition is based on my presentation of the idea rather than the substance of it...

reap- I am not suggesting that people try to quantify their doubt at all! The whole point is that rather then using these vague feelings of doubt etc, the jury estimates probabilities of events occuring given the presumed innocence of the defendant...

And UnaBubba- I would ask you what you would consider to be an acceptable figure then. But be aware that the figure must be used from both sides, i.e. if you say that 1% is acceptable, that means we have to be 99% confident that the defendant is guilty to convict them. And 99% confidence is a high level to reach.

Selecting a figure does not present a flaw in my idea, rather, my idea just brings to light the reality that this is necessary in any justice system, that if you wish to punish the guilty you have to accept that some innocent will be punished. In the present system we are protected from facing this reality by not using definite figures but vagaries. I also proposed to lessen the burden on the innocent by linking confidence in the conviction to the sentencing (although the price of this is lessening the punishment on some guilty parties). That is why, although perhaps I cannot claim absolute justice for all, I can claim greater justice for all then at present.

In our present system people use their overall feeling to arrive at a decision, which will be influenced by their prejuduces (sp?) -e.g. If the defendant in a murder case is shy and quiet the jury member might think (if only subconciously) of the quiet person who is really the murderer in a number of films... or, being shy and quiet themselves they might bond with the defendant, both of which will bring bias.

So, can I have a crossaint now? :)

P.S. I would be interested to know the true figure of % of innocent people convicted. It would be disturbing whatever it was but I think it might be over 5%.
-- radicalllama, Aug 31 2004


//It's my understanding that you have to be 100% certain. Your moral stance would certainly be different if you found yourself part of the 5% (or 1%) unjustly convicted.//

But your position is not tenable in reality, as you will be forced to let off everybody except those who confess- and even some of these may be innocent, and covering for someone else, so if you opt for 100% certainty you must by definition never find anyone guilty.

Do you dispute my point that to have a justice system of the form that we do you have to wrongly convict some innocent people? Although it is a hard reality to stomach I must repeat that if you take it as an unacceptable evil to wrongly punish the innocent then by definition if you were on a jury you could never find a defendant guilty.

Instead, society holds that it is the lesser evil to wrongly punish a small proportion of innocent people in order to punish a large proportion of guilty people. All I am doing is seeking to protect the innocent by calling for these proportions to be standardised.

Hmmm, just thought... when I said 5 out of every 100, what I guess I really meant was 5 out of every 100 who we are 95% confident are guilty...

Anyway, am I bringing you around?
-- radicalllama, Aug 31 2004


There is much wrong with this.

//Evidence is presented in specific seperate events by the prosecution...//
I am not sure what this means. Does it mean that evidence is led discretely on each contributory event in a trial ("The Purchase of the Weapon", "The Argument Preceding Attack", "The Accused's Behaviour After The Alleged Incident" etc.) and the finder of fact (as it is not always a jury) decides on the substance of that issue separately, before moving onto the next discrete issue for proof, finally doing some sort of statistical analysis based on the probabilities generated by this system?

If this is what you mean, this is a terrible idea:
1. Events leading up to, during and post incident will have a wide range of levels of bearing on any traditionally derived verdict.
Your system, as I read it, makes no account for this, and therefore the prosecution's job becomes to lead a wide range of easily provable (or admitted) 'events' to skew the probability calculation in their favour. If account is to be taken for this, then the probability calculation becomes massively complicated and, if we're frank, obfuscatory.
2. //The jury reaches a mutually acceptable probability for each of the events occuring//
So the jury's job is further complicated. Not only do they have to calculate a probability for every discrete 'happening' and then perform some calculation to determine overall guilt, they also have to arrive at some statistical probability for the total sequence of events, and compare that, like an actuary, against the value derived from the evidence led. First thing taught in Scottish Law Schools - never underestimate the stupidity of juries.

The adversarial system is absolutely necessary for criminal trials (which is what we are taking about here) as it affords a significant (but by no means totally effective) measure of protection on the individual. It's assymetric warfare - the might of the state vs the limited resources of the individual. "Beyond reasonable doubt" is the keystone of these protections. Your idea, as posted, rather than removing the subjectivity of the current system, increases it by making the finder of fact responsible for a significantly larger number of subjective judgements. Furthemore, the complexity of the proposed system affords the more greatly experienced prosecutors another advantage over their opponents, however recidivist they may be.

Penultimately, //so that appeals could be directed perhaps at individual events rather than the whole case//
In my jurisdiction (and, if I remember correctly, in most other western legal systems), appeals can only ever be on a point of law; the finding of fact is final.

And, for those of you still awake, finally // you will be forced to let off everybody except those who confess ... if you opt for 100% certainty you must by definition never find anyone guilty.//
Is horsefeathers. The fact of the matter is that only a tiny fraction of charges ever come to court and of those that do make it to trial, the vast majority of accuseds will change their pleas to 'guilty.' Of those that do go to full trial, almost every single one of the men & women found guilty will have been advised by their representation to plead guilty, but they have refused out of some criminal hubris. The remainder will have been banking on their lawyers managing to get some vital scrap of evidence declared inadmissible.
-- calum, Aug 31 2004


Perhaps the results of the Jury's deliberations could be broken down into the proportion of the objective facts find in favour / not-in-favour of the defendant and the proportion of subjective issues in favour / not-in-favour of the defendant.
-- Brett-Blob, Sep 27 2007


//It is better for a hundred guilty men to go free than for an innocent man to be found guilty.//

I'd much rather have one innocent man hang with 100 convicted mass murders than to let the whole lot back out on the streets.

Heck, if killing me gets 100 mass murderers the death sentence, then off me right now!
-- ye_river_xiv, Sep 30 2007


//The jury reaches a mutually acceptable probability for each of the events occuring// In what relevant way does this differ from the current system of allowing the jury to weigh all the evidence and find a "mutually acceptable" probability of guilt?
-- MaxwellBuchanan, Sep 30 2007



random, halfbakery