Relative probability

[YT2095]

I`m probably missing something simple here, but here goes:

 

you`re in your Lab and make Chemical X.

you don`t know what the properties of this chem X is.

you get a phone call that requires you to leave the lab and this chem.

 

now it`s either Stable or it isn`t, it`ll be just the same when you get back, OR it`ll be a disaster area!

you Don`t Know about Chem X though.

 

it`s one of 2 outcomes only.

 

are you taking a 50/50 chance by leaving it?

[insane_alien]

i have a question. is there a possibility that you can leave the container and chemical in a bath of liquid nitrogen?

 

if so then i would say its probably safe to leave the lab

if not i would say that you should probably do some tests first

 

although, in reality you can usually guesstimate from the structure whether its stable or not and infer the structure from the reactants and any other products that were produced/used in the making of X

[YT2095]

well, 2 points here, A) that works for Some chems, but I wouldn`t even THINK about that around a bottle of Nitro-Glycerine!

B) this is nothing at all to do with Chem, it`s a probability thing only.

it could well have been something else (that I can`t picture at the moment).

 

thnx anyway

[insane_alien]

nitro is stable at very cold temperature isn't it?

[YT2095]

erm... NO!!!!

 

13c or less and it`ll detonate without apparent reason at any time.

 

 

Dude, this is in the Maths Problems section for a reason too

[insane_alien]

oops, i never realised that this was the math forum. i just got here from "new posts".

 

yeah it probably is a 50/50 probability or close to it if its not.

 

P.S. i can't find anything about nitro randomly exploding below 13c. infact, i found loads of sites saying that it is cooled to 5-10c to desensitize it to vibrations. and 99/100 colder = more stable.

[YT2095]

""A serious problem in the use of nitroglycerin results from its high freezing point (13{degree} C [55{degree} F]) and the fact that the solid is even more shock-sensitive than the liquid. ""

 

http://www.britannica.com/nobel/micro/426_77.html

 

even Wiki: http://en.wikipedia.org/wiki/Nitroglycerin

 

now, back to Maths PLEEEEEZ!

[the tree]

The probability would depend on the amount of stable and unstable chemicals it could possibly be.

[YT2095]

well that`s just it, you DON`T know

 

it`ll do something Bad or it won`t.

[jordan]

I'd imagine you'd have to have some general idea of the number of stable vs. unstable chemicals out there. Personally, I know very little about chemistry so I can't really venture a guess. But I can't think of any mathematical way to show percentage of blow-upable-ness without some prior info on stable vs. unstable chemicals.

 

To me it's like you're asking the following:

There are 100 balls in a hat. Each ball is either red or green. I pick a ball from the hat. What are the odds it is red?

 

Well, if they're all green, then it's zero. All red, 100%. It all depends on the ratio of stable to unstable chemicals I guess.

[gcol]

I suggest there is insufficient data to answer the question, all chemicals having a different propensity to react violently at different temperatures. The knowledge and experience of the experimenter should be taken into account. What was the ambient temperature of the lab for example.

 

I will take a cynical guess and say he did it, the result was "unfortunate", and he is seeking a statistical excuse!

[the tree]

well that`s just it, you DON`T know

You should do. Randomly pick a hundred chemicals that could possibly be produced in your lab. The amount of them that are stable is the percentage probability of your chemical X also being stable.

 

The fact is, probability is based on knowlege.

[matt grime]

If you don't know what proportions of chemicals are stable/unstable, then you have no idea what the probability of a given chemical being stable is.

 

Mind you, if you want to play cards with me anytime please feel free, YT.

[swansont]

If you don't know what proportions of chemicals are stable/unstable' date=' then you have no idea what the probability of a given chemical being stable is.

 

Mind you, if you want to play cards with me anytime please feel free, YT.[/quote']

 

Deal me in, too. I mean, you can either win or lose, so the probability of winning is 50%, right?

[YT2095]

ok, if we stick to the bag and marbles idea, it`s a bit cleaner.

 

there is a bag of 100 marbles, they Could all be red, or they Could all be green, and infact any mixture of the 2 inbetween.

so from 100% probability right down to zero.

 

your saying its not a 50/50 chance of picking red one?

 

how about if there was no possibliity of a mixture, all the balls are either all reg or all green, you have to pick one, surely it`s just the same as a coin toss?

[ajb]

Doesn't classical probability theory require that you can in principle know all the possible outcomes? Without knowing the ratio of statble to unstable chemicals you don't have enough information to continue.

 

So I think your original question cannot be tackled using classical probability theory.

 

Maybe you would need to use topos theory? It is not something I know much about, but I know it can be applied to situations where classical probability theory fails. Maybe Matt Grimes, knows more about this....

[jordan]

ok' date=' if we stick to the bag and marbles idea, it`s a bit cleaner.

 

there is a bag of 100 marbles, they Could all be red, or they Could all be green, and infact any mixture of the 2 inbetween.

so from 100% probability right down to zero.

 

your saying its not a 50/50 chance of picking red one?

 

how about if there was no possibliity of a mixture, all the balls are either all reg or all green, you have to pick one, surely it`s just the same as a coin toss?[/quote']

It's pretty clear from the information that you've give us that we cannot give you a definite probably from the reactivity of that chemical. But I can tell you that the odds are almost zero that your chemical is guranteed to be non-reactive. I can also tell you that the odds are almost zero that your chemical is guranteed to be reactive.

 

What it seems to me you're trying to get from all this is that the odds are 50/50. That's not a valid conclusion. What I think would be fair to say is that the odds are probably best that that chemical has a 50/50 percent chance of being reactive. I could walk around your shop, look at your stuff and think 'yeah, that chemical's probably got a 50/50 chance of being reactive', but that's not a gurantee, it just seems like the odds tell me 50/50 are good odds. If I walked through a fireworks factory, I might say any randomly produced chemical might have a greater than 50/50 chance of being reactice (again, I don't know for sure, but the odds seem to be more in favor of that conclusion). If I walked through a plant that only had two chemicals, both known to be highly unreactive, I would say odds are that there's a less than 50% chance of getting a chemical reaction.

 

So I guess it all depends on the situation you're in to get information that can lead to the likelyhood of reactivity. We can't say for sure, but we can reason out what probability is more likely than others. That doesn't mean it's necessarily 50/50 though. (Knowing you, YT, I'd say the odds are pretty high that that chemical would be highly toxic or explosive )

 

[edit] wow, I re-read that and I'm not sure it'll make sense to anyone but me. Maybe someone can help clean that up a bit for me...

[BigMoosie]

The probability of an event occuring is based upon what information you have about it. For instance the probability of me dying this year is 1%, if I mentioned that I do not smoke or do drugs then the probability might drop to 0.9%. So it depends on what you know, if you know that 2 things can occur but have no idea in what ratio then the probability of either event is EXACTLY 50% until you attain more information.

[swansont]

The probability of an event occuring is based upon what information you have about it. For instance the probability of me dying this year is 1%, if I mentioned that I do not smoke or do drugs then the probability might drop to 0.9%. So it depends on what you know, if you know that 2 things can occur but have no idea in what ratio then the probability of either event is EXACTLY 50% until you attain more information.

 

 

No, the two scenarios you describe are not the same thing, and contradict each other. Since you can either die or not die, according to your second (incorrect) conjecture, your probability of dying is 50%. The correct answer, in that case, is "there's no way to tell," not some made-up number that gives the illusion that you have information about the topic.

[BigMoosie]

No, the two scenarios you describe are not the same thing, and contradict each other. Since you can either die or not die, according to your second (incorrect) conjecture, your probability of dying is 50%. The correct answer, in that case, is "there's no way to tell," not some made-up number that gives the illusion that you have information about the topic.

 

I was assuming in the first scenario that I had read some kind of statistical information.

[matt grime]

ok' date=' if we stick to the bag and marbles idea, it`s a bit cleaner.

 

there is a bag of 100 marbles, they Could all be red, or they Could all be green, and infact any mixture of the 2 inbetween.

so from 100% probability right down to zero.

 

your saying its not a 50/50 chance of picking red one?[/quote']

 

no, no, no. There is just no way to guess at all what the probability is. The point of probablity is that using knowledge you estimate what the chance of something is. If you have no knowledge you cannot estimate it at all.

 

you get knowledge by making reasonable assumptions, calculations and samples. you have none of these to deduce anything from.

 

In your marble scenario, if all possible mixes of marble are likely to occur in the bag you can work out the probability of picking one of a given colour. If you have no idea what the distribtution of the number of marbles is then you can't make any guesses at all with any confidence.

[matt grime]

So it depends on what you know, if you know that 2 things can occur but have no idea in what ratio then the probability of either event is EXACTLY 50% until you attain more information.

 

 

No it is not. And your response to someone else pointing this out also is incorrect.

[YT2095]

even if you know it WILL be A color of either red or green?

 

EDIT: perhaps if I break down my problem a little more it might help, in the bag could be Any amount of red or green to the value of 100 marbles.

 

from all red, then 99 red 1 green (so a 1 in 99 chance), then 98 red and 2 green (so a 2 in 98 chance) right the way down to 99 green and 1 red, then all green.

 

so there are 50 chances of it being in your favour to pick the right color, and 50 against.

a 1:1 surely?

[matt grime]

The probability is what the probability is. It's that *you* do not know what it is that matters.

Note, *you* are assuming that each of the distribution of balls, from 1 to 99 of one colour is equally likely: why are you doing this? (and awhy not allow for 0 balls of one colour?)