Statistics terminology

roundthebend

If we're rolling a die, there's a 3 in 6 chance of rolling an even number.
What terms describes the "3" and "6" here?

Likewise, there's a 1 in 2 chance of throwing heads on a 2 sided coin.

Or, a 13 in 52 chance of picking a diamond card at random from a standard pack of cards.

paul_c2

Not sure of the proper terms, but I think I'd use "desired outcome" and "outcome".

Garthy

Odds or chances?

A5D5E5

You are describing a probability space.  This has three parts - the sample space (range of possible outcomes), some events (rolling an even number) and the probability measure (chance of the event happening = 3/6= 0.5).

https://en.m.wikipedia.org/wiki/Probability_space

stickyfiddle

if you think of it as a fraction, 3 in 6 is 1/2, so it's effectively the numerator and denominator. But I don't think there's a specific word for each 

roundthebend

A5D5E5 said:

You are describing a probability space.  This has three parts - the sample space (range of possible outcomes), some events (rolling an even number) and the probability measure (chance of the event happening = 3/6= 0.5).

https://en.m.wikipedia.org/wiki/Probability_space

So, when rolling a die, the sample space is numbers 1-6 inclusive. The probability measure is 1 (you can only roll 1 number).

That's pretty good. But what about a horse race? The sample space I care about is the number of horses in the race, not the names of the horses, or their race number etc.

The chance of Shergar winning a 7 horse race is 1 in 7. The sample space is 7?

A5D5E5

roundthebend said:

A5D5E5 said:

You are describing a probability space.  This has three parts - the sample space (range of possible outcomes), some events (rolling an even number) and the probability measure (chance of the event happening = 3/6= 0.5).

https://en.m.wikipedia.org/wiki/Probability_space

So, when rolling a die, the sample space is numbers 1-6 inclusive. The probability measure is 1 (you can only roll 1 number).

That's pretty good. But what about a horse race? The sample space I care about is the number of horses in the race, not the names of the horses, or their race number etc.

The chance of Shergar winning a 7 horse race is 1 in 7. The sample space is 7?

Not quite.  The event is the outcome you are interested in.  The probability measure gives you the chance of the event happening (1/6th say for rolling a given number).  This can vary with different events - think of it as the odds of each horse winning.

Try it out "Good day.  Tell me what is your current probability measure for the event "Shergar wins" from the sample space "The 3:20 at Doncaster"."

I'm sure any bookie worth his salt will give you a knowing nod and take your money with trepidation.

LoFi

A5D5E5 said:

Not quite.  The event is the outcome you are interested in.  The probability measure gives you the chance of the event happening (1/6th say for rolling a given number).  This can vary with different events - think of it as the odds of each horse winning.

Try it out "Good day.  Tell me what is your current probability measure for the event "Shergar wins" from the sample space "The 3:20 at Doncaster"."

I'm sure any bookie worth his salt will give you a knowing nod and take your money with trepidation.

scrumhalf

roundthebend said:

The chance of Shergar winning a 7 horse race is 1 in 7. The sample space is 7?

Is there some secret you've been hiding for 30-odd years?

I had to study statistics as part of my accuntancy qualification. In 35+ years nobody has come to me with a bag of blue and red balls and asked what the probability was that the next ball he/she would pick would be blue. If you want a blue ball, open the bloody bag and pick one.

roundthebend

I'm trying to prepare a little training exercise for a group of novice Software Developers. One thing that always trips people up is naming stuff. So, it would be good to know the proper names for things and that can be one option put forward for people to consider.