Odds And Probability

Summary

The way odds and probability are measured may appear to give us a certain reading orunder standing of how true something is, but if we look a little closer we find it’s not that straight forward.

Odds And Probability

Think about the word probability for a minute, or more specifically the word ‘probably’. I’ll probably go for a walk today. I’ll probably not eat any chocolate. The word probable is used loosely and doesn’t always convey facts, it actually usually conveys something we are saying we intend, or might do, but then again, we easily may not. But probability and odds are used all the time to determine lots of different things, and quite often they are used to determine some pretty serious things.

Should you take a lie detector test?

The quick answer to that question is, no. Many of us believe in the accuracy of lie detector tests and in fact they have been used to convict criminals, but lie detectors are not accurate, and here’s why. A lie detector test will produce a false reading 5% of the time. This may sound like it’s pretty accurate and that’s why it’s been used in such serious circumstances, but it is not accurate. If we interview 100 suspects and the lie detector test wrongly states that 5% of them are lying, when our criminal also lies we have a problem. Yes, he has been accurately tested as lying, but so have 5 other people. The wider we cast our suspect net the more difficult it gets to pin down who is lying and who isn’t. In situations like this we need certain results. The issue isn’t so much with the test but with the rarity of the situation we are trying to test. We can’t trust the results because they are almost accurate, but not completely, and in a rare situation this doesn’t yield certainty.

The surgeon’s odds

Let’s look at this issue in the context of an impending operation you need to have. You want the best surgeon, but how do you know who that would be? You’re going to want to know what your surgeon’s previous rate of success is so you can decide if you trust him, but just because he’s been successful 95% of the time unfortunately doesn’t mean he will be this time. Here, we need to separate the probability of something happening from the probability of finding it to be a fact. If you took a test to predict you getting an illness you may feel you can trust it, but let’s look at this in detail. If a patient has an illness and the test can tell us with 95% certainty that they do have it, does that then mean that a tested patient is 95% likely to have the illness? No. look at this way. If you can see that it’s raining outside the probability of you taking an umbrella out with you is high, but if I can’t see the rain and take an umbrella there isn’t the same probability that it will indeed rain

We can use odds and probability to try and predict things and to try and find answers, but we need to understand how the accuracy of these things work.