Subcultural engagement for the 21st Century...
Barbelith is a new kind of community (find out more)...
You can login or register.

Underground > Conversation >

Expressing probability precisely

Page: (1)2

12:08 / 26.08.06

As part of a training programme I'm working on, I could use some help with an exercise. I've used something like this for years in live training but am currently trying to put together an e-learning version. Means I can't use the exercise as I usually do.

What would work though, would be for me to gather several other people's answers to this quiz and calculate the average responses. That would give people a standard for comparison, although these would not be the right answers, just the average of some kind guinea pigs.

Here are ten statements. What I'd like is for you to translate the words into a precise percentage that indicates probability, e.g. There's an even chance might convert to 50%.

1. It’s likely
2. It may happen
3. It’s highly probable
4. There’s a chance
5. It’s very likely
6. There’s a real chance
7. There’s a significant likelihood
8. It’s foreseeable
9. It’s as likely as not
10. It’s virtually certain

Cheers. If you can arsed with this, it will be another bit of the programme dealt with and I might make my deadline.

This is just one small part of an e-learning package on assessing and managing Clinical Risk. If any of you have any snappy quotes or anecdotes on the subject of Risk, feel free to voice them and I may put them to good use and incidental illustration in the text.

Couple of interesting things I've learnt this morning: Did you know that Monday is the most frequently chosen day for killing yourself? Or that 5310 injuries were perpetrated by trousers on their wearers in 2002, according to RoSPA?

Be careful out there.

Kiltartan Cross

12:37 / 26.08.06

1. It’s likely - 75%
2. It may happen - 50%
3. It’s highly probable - 80%
4. There’s a chance - 40%
5. It’s very likely - 90%
6. There’s a real chance - 60%
7. There’s a significant likelihood - 75%
8. It’s foreseeable - ??
9. It’s as likely as not - 50%
10. It’s virtually certain - 90%

I think I'd need to know more of the context of the "it's foreseeable" one before saying.

12:44 / 26.08.06

1. It’s likely - 60%
2. It may happen - 40%
3. It’s highly probable - 70%
4. There’s a chance - 10%
5. It’s very likely - 90%
6. There’s a real chance - 75%
7. There’s a significant likelihood - 65%
8. It’s foreseeable - 10%
9. It’s as likely as not - 50%
10. It’s virtually certain - 99%

13:19 / 26.08.06

1. It’s likely - 60%
2. It may happen - 35%
3. It’s highly probable - 75%
4. There’s a chance - 20%
5. It’s very likely - 80%
6. There’s a real chance - 40% (I think this is my cynicism; the phrase reads like doomed-ish optimism to me.)
7. There’s a significant likelihood - 70%
8. It’s foreseeable - 40%
9. It’s as likely as not - 50%
10. It’s virtually certain - 95%

14:12 / 26.08.06

1. It’s likely - %70
2. It may happen - %40 - %50
3. It’s highly probable - %80+
4. There’s a chance - %30
5. It’s very likely - %75 - %80
6. There’s a real chance - %55
7. There’s a significant likelihood - %50
8. It’s foreseeable - %75
9. It’s as likely as not - %50
10. It’s virtually certain - %95

I'm not sure I properly understand probability - thus my vagueness on some of the answers. I suppose I tend to think in probability zones or something. - %60 - %70 rather than %65, though I suppose this is rather dodgy, mathematically speaking.

I found it strange how I responded to the various comments. I 'there's a real chance' a relatively low %55 because it's something people say when something's not really too likely, but they want to convince you/themselves that it could. There's a 'real chance' for anything. Doesn't mean it's probable.

'It's forseeable' seems somewhat more sure of itself. Forseeable is a big word, so must be more factual, eh?

'It's a significant likelihood' technically means 'A majority chance of it happening', but it sounds like something politicians say. Politicians are liars.

Kiltartan Cross

14:35 / 26.08.06

I'm not sure I properly understand probability - thus my vagueness on some of the answers. I suppose I tend to think in probability zones or something. - %60 - %70 rather than %65, though I suppose this is rather dodgy, mathematically speaking.

From a physicist's point of view, probability zones are usually much more useful than exact values.

When we flip a coin, for instance, the physical result is one of a vast number of possible states - the coin (not that, strictly speaking, there's any such thing as a coin - it's the name we give to a huge cloud of particles which changes slightly from instant to instant anyway) could land in any one of a huge number of slightly different positions; it's our interpretation of this which simplifies things and brings it back to mathematical theory... physically speaking we've got two broad zones (heads and tails) which encompass lots of possibilities.

15:04 / 26.08.06

1. It’s likely-70%
2. It may happen-40%
3. It’s highly probable-80%
4. There’s a chance-45%
5. It’s very likely-75%
6. There’s a real chance-85%
7. There’s a significant likelihood-90%
8. It’s foreseeable-95%
9. It’s as likely as not-50%
10. It’s virtually certain-99%

16:26 / 26.08.06

1. It’s likely 70%
2. It may happen 10-20%
3. It’s highly probable 80%
4. There’s a chance 5-10%
5. It’s very likely 80%
6. There’s a real chance 5-10%
7. There’s a significant likelihood 40%
8. It’s foreseeable 80%
9. It’s as likely as not 50%
10. It’s virtually certain 90%

16:47 / 26.08.06

::reads everyone else's answers:: God, if I ever hear a doctor saying 'There's a real chance' to me, ze's going to be surprised when I burst into tears!

16:59 / 26.08.06

1. It’s likely 70%
2. It may happen 30%
3. It’s highly probable 85%
4. There’s a chance 15%
5. It’s very likely 90%
6. There’s a real chance 20%
7. There’s a significant likelihood 60%
8. It’s foreseeable 50%
9. It’s as likely as not 50%
10. It’s virtually certain 95%

17:07 / 26.08.06

1. It’s likely - 65%
2. It may happen - 30%
3. It’s highly probable - 85%
4. There’s a chance - 25%
5. It’s very likely - 80%
6. There’s a real chance - 65%
7. There’s a significant likelihood - 75%
8. It’s foreseeable - 50%
9. It’s as likely as not - 50%
10. It’s virtually certain - 90%

17:18 / 26.08.06

Is there any consistency at all here? I think we may not be helping, Xoc.

17:18 / 26.08.06

"It is virtually certain" translates as 90% for a bunch of you guys?

Weird.

18:09 / 26.08.06

1. It’s likely - 65%
2. It may happen - 40%
3. It’s highly probable - 80%
4. There’s a chance - 20%
5. It’s very likely - 90%
6. There’s a real chance - 60%
7. There’s a significant likelihood - 60%
8. It’s foreseeable - 10%
9. It’s as likely as not - 50%
10. It’s virtually certain - 99%

18:13 / 26.08.06

1. It’s likely - 50
2. It may happen - 10
3. It’s highly probable - 80
4. There’s a chance - 10
5. It’s very likely - 70
6. There’s a real chance - 20
7. There’s a significant likelihood - 30
8. It’s foreseeable - 10
9. It’s as likely as not - 50
10. It’s virtually certain - 95

18:17 / 26.08.06

1. It’s likely - 75%
2. It may happen - 20%
3. It’s highly probable - 90%
4. There’s a chance - 10%
5. It’s very likely - 90%
6. There’s a real chance - 0%, lying politician-speak
7. There’s a significant likelihood - 40%
8. It’s foreseeable - Meaningless, too conditioanl
9. It’s as likely as not - 50%
10. It’s virtually certain - 99%-0%, depends who says it.

18:23 / 26.08.06

I decided not to read the others before posting. God, we're a cynical lot! Glad I'm not alone in thinking "very real chance" is not helpful.

18:26 / 26.08.06

Happy happy happy! Thank you, lovely people. Will start calculating mean, mode etc when the weekend's over and I put away the vodka bottle.

How much do I love you? Quite a lot. Especially you, just there. Yes, I mean you. Look at you: cute as a very cute button.

18:28 / 26.08.06

Oh, and maybe it's just me, but there's something about "virtually" to me which expresses the states "as near as dammit certain" and "has no certainty" (my inner pedant says "it's certain or it's not, so to qualify it means 'not certain'") at the same time.

Kiltartan Cross

19:37 / 26.08.06

However do we manage to communicate with one another?

22:12 / 26.08.06

1. 59%
2. 25%
3. 75%
4. 11%
5. 79%
6. 61%
7. 71%
8. 49%
9. 50%
10.80%

22:30 / 26.08.06

Xoc, do yours.

23:11 / 26.08.06

1. It’s likely - 80%
2. It may happen - 50%
3. It’s highly probable - 90%
4. There’s a chance - 50%
5. It’s very likely - 80%
6. There’s a real chance - 80%
7. There’s a significant likelihood - 75%
8. It’s foreseeable - 30%
9. It’s as likely as not - 60%
10. It’s virtually certain - 95%

Xoc, what are the 5000 trouser injuries? Surely not all dudes catching themselves in the fly?

00:28 / 27.08.06

1. It’s likely - 80%
2. It may happen - 50%
3. It’s highly probable - 88%
4. There’s a chance - 48%
5. It’s very likely - 75%
6. There’s a real chance - 60%
7. There’s a significant likelihood - 88%
8. It’s foreseeable - 35%
9. It’s as likely as not - 50%
10. It’s virtually certain - 92%

ORA ORA ORA ORAAAA!!

12:15 / 27.08.06

it may be too late, now, but here are mine:
1. It’s likely 60%
2. It may happen 40 %
3. It’s highly probable 80%
4. There’s a chance 5%
5. It’s very likely 80%
6. There’s a real chance 70%
7. There’s a significant likelihood 60%
8. It’s foreseeable 40%
9. It’s as likely as not 50%
10. It’s virtually certain 95%

all of them should be ranges, really. But that's less helpful.
I am interested to see that most people break their chance-ratings into 20 5% chunks... Gary Gygax would be proud.

12:18 / 27.08.06

1. It’s likely - 75%
2. It may happen - 30%
3. It’s highly probable - 90%
4. There’s a chance - 10%
5. It’s very likely - 90%
6. There’s a real chance - 60%
7. There’s a significant likelihood - 70%
8. It’s foreseeable - 50%
9. It’s as likely as not - 50%
10. It’s virtually certain - 95%

Kiltartan Cross

12:42 / 27.08.06

I am interested to see that most people break their chance-ratings into 20 5% chunks... Gary Gygax would be proud.

I think it's because in the absence of probability ranges, we don't want to overspecify, and imply more precision than we feel; twenty broad grades implies a higher degree of vagueness than one hundred narrow ones. Like if I said, oh, I dunno, that I had five pounds in my pocket, as opposed to saying that I had five pounds and three pence. With the first I'm implying I have five pounds or thereabouts, with the second I'm implying that I know exactly how much cash there is. Something along those lines, anyway... what are other people's thoughts?

20:32 / 29.08.06

Thank you one and all. That's that all tidied up at the start of the section on communication of risk information. You had a great range in your answers, which is the most useful thing for people to see in the exercise. They never use figures to express risk but will daily try to communicate how likely certain foreseeable and unforeseeable events are.

This exercise gives rise to lots of reflection on the language they choose and what they should say to be most effective. Kay's points are particularly apt because it's all about "about" values. It's all about imperfect predictive tools.

The statistics tell you that the man in front of you is three times more likely to kill himself than if he were a woman. A simple, actuarial prediction on paper. But he could be the one that bucks the trend and how to discern that? Should you deprive him of his civil rights to keep him safe, if he's not going to kill himself? If he's the woman who bucks the trend, do the figures make her more or less safe?

I think 42 is the answer.

21:08 / 29.08.06

I'd actually quite like a digest of your findings per phrase, Xoc, and if there's any interrelation you've discovered; looking at it again, certain of the phrases do seem paired almost. I like p-math.

08:33 / 30.08.06

This is irrelevant to Xoc's questions, but I'm curious about something I read in a novel yesterday: I never really understood probability, which is why I need to check, but I do think I'm right about this one.

Okay, so there's a safe with a four-digit combination. The number of possible combinations is 10,000, isn't it? You can try all the combinations by putting in all the integers from 0000 to 9999? So someone who said there were 'billions' of possibilities would be talking nonsense?

Thanking you.

08:37 / 30.08.06

Yes, it's 10,000 possible combinations, Deva.

08:38 / 30.08.06

Yes, absolutely right Deva.

08:41 / 30.08.06

Is there a formula for figuring that out?

09:01 / 30.08.06

Well, yes. But the formula is just a way of counting all the numbers from 0000 to 9999....kinda. Its 10 independent possibilities for each number, and since there are 4 numbers in the combination that makes 10 times 10 times 10 times 10. Which is 10,000.

Sjaak at the Shoe Shop

09:05 / 30.08.06

So in mathematics:
Formula = X to the power N
Where X is the number of digits
N is the number of wheels

So if you have a safe with letters instead of digits and 5 wheels you end up with 26 to the power 5 combinations = 11.8 million possibilities

Page: (1)2