I ran another simulation which cut the initial bet schedule in half for my WiWo simulation. Just to recap, the player comes in with 2 hands, and places waiting bets of 1 hand if the count drops.
The initial SCORE was 43.23. After the bets were reduced by half the SCORE surprisingly increased to 44.72, while IBA and TBA both decreased. Does this make sense? I would expect SCORE to remain unaffected because each bet was reduced by an equal proportion.
See image below.
MJ
It's certainly possible. Reducing all bets by half shouldn't affect SCORE by much, unless other factors were involved. A change of 43.23 to 44.72 is nothing.
> It's certainly possible. Reducing all bets by half
> shouldn't affect SCORE by much, unless other factors
> were involved. A change of 43.23 to 44.72 is nothing.
SCORE increased by 3%, which is the same gain from using RA indices. Do you think standard error caused the increase?
I just looked at the SE for each of the sims. The sim with 3 billion rounds had an SE of 0.03. The sim with 2 billion rounds had an SE of 0.021.
As our sample size increases, shouldn't sampling error decrease?
I'd have to see the sims. But you can't call it 3 billion rounds when the player in question played 565 million hands.
> SCORE increased by 3%, which is the same gain from
> using RA indices. Do you think standard error caused
> the increase?
> I just looked at the SE for each of the sims. The sim
> with 3 billion rounds had an SE of 0.03. The sim with
> 2 billion rounds had an SE of 0.021.
> As our sample size increases, shouldn't sampling error
> decrease?
Posting Permissions
-*-
Bookmarks