Aaron, Russell, and Drew:

Perhaps John and/or Jason could enlighten us, but it sounds like 
you've all said things that are wrong. 

You guys are all trying to explain why DICrOTIC sims 6 points better 
than DIChOTIC, but the fact is that it doesn't.  It only does that 
much better in the championship player report that I ran.  When you 
sim them head to head, they're within a point of each other.  

Aaron, you're on the right track, but, as Drew said, DICrOTIC simmed 
better.  Drew, yours is an interesting theory that the R ensures that 
we each have a hot spot; one takes the triple the other may bingo 
to/thru the R.  But again, they sim the same! 

Finally, Russell, this is where we could get John and Jason to chime 
in, but I don't think that's how the sim works.  It picks a random 
rack, then tests all the potential plays against that rack.  It does 
that for each iteration.  And according to Brian Sheppard, author of 
Maven, 1,000 is enough iterations for most simulations.  You 
certainly don't need 50,000 iterations unless you're simming to the 
end of the game on each turn. 

QUESTIONS FOR JOHN AND JASON: How does the simmer work on the 
championship player? Specifically: 
1) How many plies does it do? 
2) How many iterations does it do? 
3) Does it use the same rack to test each play in a given iteration? 

It seems to me that the 6 point difference suggests that it either 
goes to the end of the game, which can induce huge variability, 
and/or, it fails to run enough iterations.  

Please enlighten us.  I'm not sure I can stand too much more 
speculation... ;)

Mike Eldeiry








--- In [email protected], Russell Honeybun <[EMAIL PROTECTED]> wrote:
>
> 
> >it sims the competing plays with the same racks
> 
> Incorrect.  Quackle draws random tiles for each
> potential play it identifies (which are not
> comprehensive)
> 
> The 6 point difference isn't due to triple-triples,
> it's just due to random bonuses played.  At most 1000
> sims would have been processed, which is <1% of the
> required sims to produce results with any certainty.
> 
> Conclusion: don't waste time with the reports.
> Compare plays with 2-ply if you are going to sim at
> all (endgame is a different case)
> 
> You are smarter than a computer program.
> 
> TcheQ
> 
> 
> 
>        
> 
______________________________________________________________________
______________
> Pinpoint customers who are looking for what you sell. 
> http://searchmarketing.yahoo.com/
>


Reply via email to