MollysPoker.com

[Stig Holmquist]

Consider 50 draws of a Pick-4 game and assume that half the
digits have been drawn 16 times and the other half 24 times,
a highly unrealistic event.
Now test these results with the chi-square goodness of fit method
and you find the value 10x16/20=8. Now look up this value in a
chi-square table under 9 df and you find a probability of about 50%.
Does this make sense? What am I doing wrong?

Stig Holmquist

[Paracelsus]

"Stig Holmquist" <[email protected]> wrote in message
news:[email protected]...
| Consider 50 draws of a Pick-4 game and assume that half the
| digits have been drawn 16 times and the other half 24 times,
| a highly unrealistic event.
| Now test these results with the chi-square goodness of fit method
| and you find the value 10x16/20=8. Now look up this value in a
| chi-square table under 9 df and you find a probability of about 50%.
| Does this make sense? What am I doing wrong?
|


I can't see any error in your calculations.

Because the chi-square goodness-of-fit test is summative and
non-directional, it cannot 'see' the peculiarity of your hypothetical
scenario. Maybe there is a better test for this, but I've no idea what that
might me.

Or maybe the scenario you suggest is not really as unrealistic as it seems
intuitively??