On Thu, 24 Jul 1997, Rita Norr wrote:

 Recently one of the LA players asked me what I'd play with the opening rack:

 ABGHLSZ.  I said I'd probably play ZAG (he had played BLAHS).  I offered to
 run a simulation in Maven.  I am, however, perplexed by Maven's 2nd choice
 ... after 2274 two-ply simulations, it showed:

   ZAG    38.35 pts.
   ZAGS   28.74 pts.
   BLAH   26.86 pts.
   BLAHS  22.87 pts.
 
 I can't imagine why putting the S on ZAG would simulate so well.  The only
 possible explanation I could think of was that you are leaving 4 consonants
 with ZAG and only 3 with ZAGS.  Of course it does score two extra points,
 but certainly the S is more valuable than that!
 Any of you J---s (fill in the blanks John, Jim, James, Joel, etc.) able to
 shed any light on this??  Rita

JG weighing in,

Um, did you misread the table? ZAGS doesn't really do "so well". Burning the S simulates almost 10 points behind ZAG. In a linear analysis, it should only fare 5 points worse (7 for the S - 2 extra pts). The difference is probably attributable to the fact that in the ZAG scenarios, if your opponent doesn't have an S he ain't scorin shit. In this situation the S is probably better than 7 points next turn, tho I'm not sure I'd daresay 12. Probably dangling an additional thru letter and allowing 7 letter singulars to now go down are responsible for a coupla the points.

Your thinking of it doing "so well" was probably visavis comparison with BLAH etc. These moves just aren't as good as the tight ZAG. The Z doesn't really play well yet unless you can drop it on D8, so getting the 20 is better than getting the 6 for a B etc.. In fact I'll go out on a limb and say the H is the second best letter on the rack, and is in fact worth keeping.

Anyway this analysis is all off the cuff; I'm waiting for Fred to check in with some award-winning insights.

Last Modified 9/26/00

Back to Scrabble Correspondence Page
Back to Scrabble Main Page
Back to Jim's Main Page