Suppose that five people (A, B, C, D, and E) are being tried for breaking some law. What B did is slightly worse than what A did, what C did is slightly worse than what B did, and so on...

Suppose that A and E are tried first, and that A is found not guilty and E is found guilty.

If the remaining three are tried in the order B, C, D, then B will likely be found not guilty since what they did is similar to what A did, and C will likely be found not guilty too since what they did is similar to what B did. We don't know about D since their action is similar to both C (who was found not guilty) and E (who was found guilty).

However, if the remaining three are tried in the order D, C, B, then D will likely be found guilty since what they did is similar to what E did, and C will likely be found guilty too since what they did is similar to what D did.

This means that C is more likely to be found guilty if the cases are tried in the order D, C, B than the order B, C, D.

Obviously, a system where the order in which people are tried determines the outcome of the trial isn't fair, so why does it make sense to give any value to legal precedent when determining whether someone is guilty or not?

• I think this assumes that they are tried by the same judge/jury, if they are all treated as separate trials, you may not be able to draw those conclusions. Commented Dec 5, 2018 at 14:55
• This is a great question, though I'd argue that there's an even stronger logical argument you could use in your favor than the one you make. In your argument, you say that the five relevant parties in order of increasing guilt are A, B, C, D, and E, and that A and E are tried first, leading to a condition where B and C are implied to be not guilty, and D guilty, because of some sense of the "line in the sand" between guilt and non-guilt lying somewhere in the middle. Instead, I would have used a different example: Commented Dec 5, 2018 at 15:53
• What if E and A were tried first, and E was found not guilty, and A was found guilty? If you are more guilty than a guilty man, then you too should be found guilty. Likewise, if you are less guilty than a not guilty man, then you too should be found not guilty. Then, you have conflicting precedent, whereby the logic established in the case against E implies that A through D should all be not guilty, but the logic established in the case against A implies that B through D should all be guilty. This makes a stronger case than saying "out of B, C, and D, the first few shouldn't be guilty". Commented Dec 5, 2018 at 15:59
• From what I can tell, your question boils down to asking "why does 'legal precedent' make sense?". Your premise about trying people in a certain order is based on a false premise, and it doesn't contribute anything. As for whether stare decisis "makes sense", it's only a feature of common law systems. So are you asking what brought about that legal principle? Or are you just covertly arguing against SD, phrasing it as a rhetorical question. It seems to me that your badness-comparisons are a distraction. Commented Dec 6, 2018 at 2:08

Suppose that five people (A, B, C, D, and E) are being tried for breaking some law.

why does it make sense to give any value to legal precedent when determining whether someone is guilty or not?

I gather that you mean that A-E broke the same law. Even if we naively assume the judiciary acts with integrity and that each A-E's position is litigated competently, the paradox you describe departs too much from the essence of stare decisis to actually challenge it.

First, fact-finding is not a [mathematical] continuous function. Both fact-finding and laws themselves are replete with "discontinuities". Whether A-E broke the same law or different ones, laws and legal doctrines contain subtleties, exceptions, and conditions. Thus, the slight difference between A's conduct and B's conduct could be one of those decisive subtleties contemplated in the law, thereby leading to A's acquital and B's sentencing.

Second, fact-finding is not an n-dimensional function. It is not even a metric space. Thus, not all aspects of A's conduct and B's conduct are susceptible to assessment under one same, consistent metric that could encompass every facet at issue.

Third, legal precedent is aimed at controlling many more cases than those involving n consecutive verdicts during one same trial. Thus, even if it were accurate to portray rulings or verdicts as being sequence-dependent, it would fall short of dismantling the notion of stare decisis over a much more general class of matters: those involving unrelated parties and unrelated offenses.

Fourth, an inconsistency (here in the form of sequence-dependence) of verdicts at one same trial is not a deficiency of legal precedent, but of the fact-finder's subjective factors such as bias for or against certain defendants, greater sensitivity to a condition, tiredness, or even self-awareness that at some point it needs to assign more (less) severity in the course of going from A (from E) to E (to A).

Fifth, a fact-finder will not necessarily judge defendants A-E exactly in the sequence A-E. If the offenses are almost indistinguishable, defendants B-D are likely to succeed in their request not to be judged in a sequence A-E or E-A.

Legal precedent is a good thing. Inconsistencies and denial of justice have much more to do with judiciary's debauchery and arbitrariness than with this type of thought-provoking but unlikely hypotheticals. Indeed, justice would be served more often (and benefit us who have been unlawfully harmed) if case law (and underlying statutes) were consistently applied.

This assumes that there is an objective and numerical measure of "degree of guilt". The law simply does not work like that. There is no way to say that:

What B did is slightly worse than what A did, what C did is slightly worse than what B did, and so on.

with any assurance that all possible fact-finders would agree, or indeed that any would. Whether these people A-E are part of the same set of acts, or merely committed crimes that are in some way similar to each other, the premise is not to be relied on.

If these people were tried separately, especially before different fact-finders, no attempt will be made to calibrate the verdicts or sentences of A thru E.

Even if they were tried together, other things may come into the decision. There may be mitigating or aggravating factors. Youth or age, intelligence, and motive may be considered both in making a guilty/not guilty decision, and in the sentence of those found guilty. There is also the possibility of conviction of a lesser included offense such as second degree murder or manslaughter when the charge was first degree murder.

There is also the possibility of bias on the part of the fact-finder, for or against some of the defendants. Then too, some judges or juries will honestly consider a given violation particularly serious, others will have different views. Some juries may even refuse on principle to convict for particular offenses. Others will not.

In short, the level and kind of consistency you are looking for is not achieved, and indeed is not attempted, in criminal court proceedings.

When penalties or results are wildly inconsistent, some courts will find this to be unjust. In the US this has occasionally resulted in convictions being overturned on due process grounds, particularly when racial bias is alleged. But the situation must be rather blatant before this will be done.

Your question shows a lack of understanding of precedent in common law systems and the difference between "facts" and "law" in a court case and indeed how crimes are prosecuted.

Here's the potted summary -

Within any given legal system there is a law or laws that makes certain acts or omissions a crime. For example, most (but not all) criminal offences in New South Wales, Australia are codified in the Crimes Act 1900.

Have a look at DIVISION 10 - OFFENCES IN THE NATURE OF RAPE, OFFENCES RELATING TO OTHER ACTS OF SEXUAL ASSAULT ETC. The various sections in that division define the offenses of Sexual assault (s61I), Aggravated sexual assault (s61J), Aggravated sexual assault in company (s61JA), Assault with intent to have sexual intercourse (s61K), Indecent assault (s61L), Aggravated indecent assault (s61M), Act of indecency (s61N), Aggravated act of indecency (s61O) and Attempt to commit offence under sections 61I-61O (S61P), it also abolishes the common law crimes of rape and attempted rape (s63). Each of these is a different crime with different elements carrying different punishments.

A given fact pattern may suggest that one (or more) of these offences may have been committed. Police will investigate and prepare a brief of evidence for the Director of Public Prosecutions (DPP). The DPP will consider the elements and consider which elements of which crimes they believe that they can convince a jury beyond reasonable doubt that the accused committed. For example, the offence of Sexual assault (s61I) requires the DPP to prove i) there was sexual intercourse ii) without consent and iii) the accused knew there was no consent - if they do so the defendant is liable for imprisonment of up to 15 years. To prove Aggravated sexual assault (s61J) they need to prove i) to iii) above plus iv) that there were "circumstances of aggravation" (as defined) - if so, the maximum penalty is 20 years.

Now, in a common law system, the bare statements in the Crimes Act 1900 are not the totality of the law. Courts in these systems also look to precedent - which comes in binding and persuasive types. Courts are bound to follow a precedent from higher courts in their hierarchy. Court hierarchies are complicated but if you work in the system you know how this fits together. Persuasive precedents come from courts at the same level or below in your hierarchy and from courts in other hierarchies and jurisdictions with similar laws. For example, the NSW District Court may, in the absence of a binding precedent, be asked to consider a decision from the Supreme Court of Queensland or the High Court of England etc. or, of course, to an earlier decision that the particular court itself made.

The role of common law precedent is to provide consistency on the basis that similar facts should have similar outcomes. Therefore, much of the legal argument in a case is often about distinguishing the facts in this case from the facts in the case that set the precedent - that is, the facts are not similar so different outcomes are OK.

Note that parliament can always write a new law that overturns existing precedent. Indeed, the newer the law and the less precedent there is the more litigation tends to happen under that law - that's because, in the absence of precedent, no one is entirely sure what the law actually is.

When the case comes to trial there is a trier of law and a trier of fact. The trier of law decides what the law is based on statute and precedent. The trier of fact decides what happened. Usually in a criminal trial the trier of law is the judge and the trier of fact is the jury. In a civil trial (except in the USA which still uses juries in a lot of civil cases), the judge is usually both.

The job of the trier of law is to tell the trier of fact what the law is and that if they decide the facts are one way, they must convict and if they decide they are the other way they must acquit. They then leave the jury to get on with deciding the facts. For example, one of the "circumstances of aggravation" for Aggravated sexual assault is "the alleged offender breaks and enters into any dwelling-house or other building with the intention of committing the offence or any other serious indictable offence". Assume that the prosecution alleges that the accused did break and enter but the defense maintains that they were invited into the building. The judge will instruct the jury that they are to consider the evidence and decide beyond reasonable doubt if the defendant broke and entered or not and that if they decide they did then (assuming the other elements are also proved) they must decide that the defendant is guilty of Aggravated sexual assault and if they don't that they are not guilty.

Precedent can only inform the law, not the facts. Each jury decides for itself what the facts are are and what "beyond reasonable doubt" means. Therefore, the same evidence can result in different juries making different decisions. That said, in mock trials in academic studies there is remarkable consistency in verdicts - the jury system seems to work.

Crimes are all or nothing - the defendant is guilty or not guilty - there is no "worse" in the conviction. If the jury decides beyond reasonable doubt that the facts of the case are that the defendant committed all the elements of the offence then they are guilty - if not, they're not guilty.

Where "worse" does come in is when there has been a conviction, the judge decides what the sentence should be. There is a completely different law that governs this - the Crimes (Sentencing Procedure) Act 1991 and yes, this is also subject to precedent. I won't go into details here, but a judge has wide discretion to decide what the appropriate penalty should be including aggravating and mitigating circumstances and appeals are only upheld when the sentence is manifestly unjust (either too harsh or too lenient). Its interesting to note that academic studies of sentences find that, in general, judges give harsher penalties than lay people when presented with the same facts - it is only consistently the other way when the victim is a child.