Jotto: uses for sick processor speed

[mattlistener]

So, I've written Jotto: Guess My Own Word, set it up to play solitaire Jotto on every word in my 3200+ dictionary 100 times, and set it off. My desktop unix box at work (a Sun Blade 100) is taking about 13 seconds to guess a word 100 times. Yowza! It should be done in 12 hours.

The output will be the average number of times it takes to guess that word, making random guesses under queue's Modified Jotto rules. Evil Jotto will have several possible uses for this information. Heh heh.

A quick glance at the output shows that words with common letters tend to be harder to get than words with rare letters. Interesting: max number of guesses I've seen is 20! For reference, Queue and I haven't required more than 9 guesses in 7 games each.

Another consequence here is that I will shortly have highly-calculated data on the logically-hardest-to-guess 5-letter words, which gives me an unfair word-choosing advantage, so I need to talk to Queue about evening that up for our ongoing game. Queue, what would you think about switching to keeping score via guesses over/under par? (That would bring "easy" words back into our guessing-space, too.) I'd also be happy to share my par-enhanced dictionary with you.

Comments

[queue.livejournal.com]

Queue, what would you think about switching to keeping score via guesses over/under par? (That would bring "easy" words back into our guessing-space, too.) I'd also be happy to share my par-enhanced dictionary with you.

I'm happy keeping things how they are.

I can already guess what some of the most difficult words are: words with easy letters and 4+ anagrams.