Scrabble Data

[whatnoloan]

This post was sparked by a comment getofftheoven made. It seems rather innocuous, so I'll go ahead and share it:


"Noticed that John played sixes on six different turns in that game. That's got to be pretty rare."


I wondered how rare this actually was. This stat, and many other interesting data, could be gathered by playing Quackle against itself ad infinitum. Indeed, this method has already been used (e.g. Superleaves). While getofftheoven's data is just a "fun" stat, there are plenty of stats that could really push Scrabble strategy forward.

For a time, I participated in a project called BOINC. This meant to gather scientific data using processing power of people worldwide. (Scientific progress goes BOINC, hehe.) It seemed a neat idea.

Now imagine a BOINC-like tool applied to Scrabble. Scrabble players the world over could contribute part of their processing power to gathering this data. I think this would be really cool, though I don't know how to execute it. What do you, the readers, think of this?

[synodalhaj]

I think that it's overkill and suffers extremely from diminishing returns--the total amount of strategy that there is remaining to be gleaned could be done from ~100M-1G runs, which you could get done fairly quickly without a massively distributed setup.

If you want to do BOINC related stuff, check out World Community Grid:
http://www.worldcommunitygrid.org/

Like SETI@home, but actually useful.