Tuesday, 22 September 2020

The Scrabble challenge

In the intervals of lunch today, I fell to wondering about what the maximum score at Scrabble might be, in the sense of the sum of the scores of all the players. With our having trouble breaking out above 600, and not even making 500 reliably.

One could make an estimate of sorts by simply placing all the high value tiles on high value spots on the board, with the simple rule that if tile A had a higher value than tile B it deserved a better spot on the board. An estimate one could probably make in an hour or so with pencil and paper.

But to do better than that, I can't think of anything better than doing lots of more or less random trials - using the tools inside Microsoft Excel - and seeing what sort of distribution one got. Maybe one would get a fair idea of the maximum from that. 

Maybe some cunning trick will come to me one morning when I am up bright, early and bushy tailed.

A possible catch for me would be the need to come up with an Excel accessible list of permitted words - although I dare say a little surfing effort would deliver something for free, rather than putting some money in the way of reference 1.

PS: a little later, it occurred to me to simply ask Bing, rather than trying to compute the answer for myself. Lots and lots of hits, containing some very high scoring plays and games, albeit rather improbable. For example the 1,780 pointer at reference 2 - the trick being to get the good letters on a word spanning three triple word spots. So I have learned that theoretical maxima are not going to be a very good guide to what one is likely to achieve in practise.

Reference 1: https://www.wordfrequency.info/.

Reference 2: http://www.scrabulizer.com/blog/post/3. With thanks for the snap above. Not yet found out what the significance of the lower case 'n's is - if any. Nor how many of these words are not to be found in our old, if voluminous, OED. Some of them look a bit unlikely to me.

No comments:

Post a Comment