 |
|
Your understanding of it is flawed. What's the expected
Posted on: January 29, 2024 at 19:40:46 CT
zounami MU
Posts:
72138
Member For:
11.19 yrs
Level:
User
M.O.B. Votes:
0
outcome if you flip a coin 100 times?
50 heads, 50 tails.
A model would *predict* that outcome based on a binomial distribution, because it's the most likely.
And that would be the *right prediction*, regardless of what the *actual outcome* turns out to be.
If you run a trial and get an extreme outlier – like 95 heads – it doesn't mean the *prediction* of 50 heads was "wrong" per se, it means the outcome was statistically improbable – "lucky" or "unlucky" depending on whether heads or tails is favorable.
In basketball, if a team goes 10-0 in games decided by a single possession, humans consider that to be "lucky" because it defies the odds in a favorable way. It's not an "error", it's just a statistically improbable outcome.
50 heads, 50 tails.
A model would *predict* that outcome based on a binomial distribution, because it's the most likely.
And that would be the *right prediction*, regardless of what the *actual outcome* turns out to be.
If you run a trial and get an extreme outlier – like 95 heads – it doesn't mean the *prediction* of 50 heads was "wrong" per se, it means the outcome was statistically improbable – "lucky" or "unlucky" depending on whether heads or tails is favorable.
In basketball, if a team goes 10-0 in games decided by a single possession, humans consider that to be "lucky" because it defies the odds in a favorable way. It's not an "error", it's just a statistically improbable outcome.
Please explain why this message is being reported.
MESSAGE THREAD
- 
Your understanding of it is flawed. What's the expected - zounami MU - 1/29 19:40:46