Wednesday, September 22, 2010

Class notes: 9/22/10

JB will be returning the score and/or word of each player at the end of the round

We did a run:
Dan (g6): We happened to bid 5 for a Q, which we think to be a relatively low bid, then 95 for an E. We are assuming that nobody else will bid high
Daniel: Highest scoring word without 7 letters is 32. 95 is way too much
Blake: Maybe getting something for 0 points and having a huge rack is a good thing this time

Flavio: We have a lot of letters right now, but we've been bidding pretty low. We plan to get as many letters as possible, hoping that we'll eventually get a 7 letter word.

KAR: So what have we learned so far?
Daniel: G1 from last year is really good.
Neha (g5): We have been bidding pretty close to the true value of the word, without thinking about the value to others (we're not getting things).
Zach: Hidden letters form a burden for last year's players, more so than for us
Neerja: We are trying to do more upfront
John: Everyone is driving up the cost of common letters more so than they should be
Eva: We should adjust bids based upon the frequency of letters remaining
KR: But a few people overbidding can still drive everything up

Monica: But some letters may not come up at all in the game
Daniel: An agressive/passive strategy will play differently in a game of different size. We need to take that into account.

KR: In a two player game, maybe you should never bid 0.

David: With 6 hidden letters, people should want different things, so maybe bids will be lower

Monica: Actually, you can pay a LOT now, because you only need 2 letters!

Deliverable for Monday

A player that is robust under these various kinds of configurations.

Final deliverable due at noon on Wednesday.

No comments:

Post a Comment