I had a nice long post typed up on what I think the bracer would do and probability, but it got complicated and instead I'll illustrate with an analogy.
Consider the following question:
"What's the probability that I'll flip "heads" on this coin that I'm holding?"
You might think the answer is obviously 50%. But that's not at all obvious.
First, there's the question of whether I'll flip the coin at all. Maybe I won't. To answer that, you first have to predict the probability that I'll flip this coin at all.
OK, so lets refine the question. "If I flipped this coin right now, what's the probability that it'll land heads?"
You might answer 50-50 to that - or whatever the probabilities are, coins aren't *actually* symmetric, they're pretty close but not exact.
What about "If I flipped the coin I'm holding right now, a 2005 "Kansas" US Quarter, what's the probability that it'll land heads?"
That'll be slightly different than the previous answer. Now you have exact information on what coin it is.
What about "If I flipped the coin I'm holding right now, a 2005 "Kansas" US Quarter (and the "Heads" side is currently facing up, ready to flip), what's the probability that it'll land heads?" That's yet different. Because of the dynamics of coin flips, it's more likely by about a percent for a coin to land on the same side it started out on.
What about "If I flipped this coin I'm holding right now, a 2005 "Kansas" US Quarter (and the "Heads" side is currently facing up, ready to flip by a Persi Diaconis tails-calibrated coin-flipper ( http://news-service.stanford.edu/news/2 ... is-69.html
) ), what's the probability that it'll land heads?" That is VERY different.
What about "given the quantum-mechanical wavefunction for all of the atoms in the room I am in, what's the probability that this coin will land heads?" That'll give pretty much an exact answer.
And yet, if you don't think of all of those individual factors, you'll get an answer from the bracer to the question "Hey' what's the probability that guy flips heads?" and it'll be a completely useless answer because you won't know which of those questions it is answering.
(For the record, I did flip the coin, it started out heads, and despite not owning a Persi-Diaconis machine I intentionally flipped it in such a way that it would land tails. If you could accurately predict my thought process, the actual flipping of the coin would have had essentially no randomness in it.)
The relevance to the bracer is as follows. The information you get from it is EXTREMELY dependent on
1) Exactly what it is asked.
2) What you already know about the subject at hand.
3) What the bracer knows about the subject at hand.
With enough information available to it, you can get a near-certain answer to any question. With little enough information, you can get a useless answer to any question.
Asking vague questions about subjects about which you have no knowledge is equivalent to asking "what's the probability that it'll land heads" with no information. Sure, you'll get 50-50 - but there's a million different ways that reality could be biased towards either heads or tails, and they'll all average out and you won't get any information about them. You could ask "what's the probability that a dwagon will pop in this hex" and you'll probably get a generic probability that doesn't give you any information besides "dragons pop at a rate of 1 dwagon per 200 hexes".
The bracer is most useful when you can narrow down exactly what you're asking about. "What's the probability that a stack of 7 dwagons can defeat that stack of 4 warlords and 30 bats?" - now THAT you can use to plan a battle.
The relevant question is "what information does the bracer have access to?" In real-life terms, if I ask it about some guy flipping a coin, does it have access to the quantum mechanics of all the atoms, allowing it to calculate what'll happen to within the heisenberg uncertainty principle? Or will it average over all coin flips everywhere and give 50-50?
I think it is evident that the bracer has access to all the "built-in" erfworld mechanics. Stats, etcetera. If Parson asks it "what's the probability that a dwagon is croaked by 5 bats", he doesn't need to look up the strength and levels of dwagons and bats personally.
But beyond that, what information does it have the ability to use? I suspect it has to have access to all of the information Parson has, and his estimates about the world. It knows as much as he does, otherwise he would have to explicitly describe all sorts of the details of the current situation to it, and he'd have to worry about accidentally leaving out information that he knows but hasn't told the bracer explicitly.
Beyond that - MAYBE it has access to the information that other commanders or units on Parson's side do? MAYBE it has access to information from others that it's making predictions for. (Like, it was allowed to access Charlie's plans for what to do with the information in deciding whether the deal was worth it for him.) I would suspect not.
But it doesn't have to be powerful as a predictamancy artifact to be powerful as a mathamancy artifact. Being able to put exact numbers on situations you know is HUGELY useful to a good Warlord.
So you're saying that you think Parson DID know all the necessary variables to predict the future, but we just didn't see him figure them out because the audience wouldn't get it? I think that is implausible. If Parson were capable of such a feat he would be a seer in his own right,
So to that I would answer yes. Parson's a great warlord - the whole POINT of being a good warlord is to "predict" what the enemy is going to do and what your options are to counter it, and to be good at guessing the answers to those sorts of questions! I find it quite reasonable that if Parson had to come up with a guess as to whether the deal is worth it for Charlie, he'd run through a few possible uses of the information and a few possible ways it could leak out, and come up with "well, probably not." He wouldn't be sure, because there's a lot of possible scenarios in either direction, and even if he can visualize them each individually, putting numbers on something you know is hard anyway, and combining them is harder still. But the bracer lets him put exact numbers on the situations he knows about - that's immensely powerful, but not in the way you're imagining.