Hi, why is there code in the framework that requires a package 'human', which is not included? I would think it is great to have a human interface (which this...
... Oops! net.* is not supposed to be there (feel free to delete it, I will in the next version). I was building a human interface for people to play the game...
how is the tournament winner determined? is it by the total number of points won in the tournament? or by the number of games won (where a game is many...
I think that a natural metric is the total number of points in the tournament. This hopefully motivates you to maximize your own utility in each game. Of...
The results say that we'll receive and array back telling us what happened ("as [3,7,9]. ") and that our bot would be player 0 in this. What about the other...
Why not have two metrics: 1) Highest total points 2) Most 'wins', (say 2 for 1st, 1 for second, 0 for third). This could account for a strategy that says "Out...
The idea of the game is to learn about the other players in the course of a single session. I tried to eliminate a lot of information about the overall...
Should we have a formal poll? Please respond to this message yes or no (one per team) by Monday. If you have started a bot and hate the idea of changing the...
Practically, yes, but let me rephrase your question: Given a strategy object A which is in the game, is the value at index 1 of the array argument to...
OK, so it seems that there is no interest in rewriting the rules, but people are interested in clarifying details. Awesome! Please post any further questions...
OK, I added a bot which provides a GUI for a human to play against the computer. It is a bit rough, as I hadn't really planned to release it, but it should be...
I know it is getting a little close to the submission deadline, but I was curious about the game length. We have been talking some over here at Brown, and we...
I know it is getting a little close to the submission deadline, but I was curious about the game length. We have been talking some over here at Brown, and we...
First of all, glad to hear of your interest! Sadly, it is too late to change the rules for this year. My motivation for choosing such a short period of time is...
The deadline is almost here! I have heard from several people that they are working hard to develop bots. But there is still time! Below is an excerpt from the...
20
lemonadegame@yahoogro...
Dec 23, 2009 12:15 am
Enter your vote today! A new poll has been created for the lemonadegame group: Please, one vote per team. What is the chance that you will submit a bot? o...
There have been seven submissions. As stated before, there is a round-robin format. If all of the code works, there are 7 choose 3 triples of players=35 games....
OK, I have uploaded the agents that were given to me as a single zipped file agents2009.zip. NO WARRANTY IS EXPRESSED OR IMPLIED. This is code that was not...
With the source code distributed and some early results distributed, at this time I would like to open a fairly broad discussion here both about the...
Here are the experimental results when bots played against each other in every permutation. Soton:8.6328+-0.0117 RL3:8.5240+-0.0105 Pujara:8.5125+-0.0150 ...
Thanks to all who competed. This competition ran pretty smoothly. Here are the final rankings. I have totaled the standard error of the means of two adjacent...
Hi folks,
A couple of weeks ago, Marty posted three questions for discussion regarding the future of the lemonade game, and here I will respond to two of them...
Hey guys, a quick response to Marty's analysis of our strategy (Soton/EAsquared). He is spot on in saying that what we try and do is find a candidate "friend"...
I thought readers on this mailing list would be interested in checking out one of the outcomes of the competition - we're presenting our work on and with the...
Greetings, We are pleased to announce our paper discussing the lemonade game to be presented at the AAAI workshop Interactive Decision Theory and Game Theory...
