That was funny, I got offended by Zax's ANALytical reference (real Freudian! But according to Transactional Analysis, I'm OK.) and totally missed the whole thread If you read the thread from bottom to top, you really wonder about the squirrels at first, too. Kevin
On Sun, May 9, 2010 at 2:50 PM, Kevin Leeds <[email protected]> wrote: > > There is a logging feature that you can turn on somehow to see what is > being considered. (Search in the newsgroup for more information) > > There seems to be a bug in some available version of Quackle. I have > managed to download some open source tools and successfully built Quackle > from the source code in so now I also have a version that runs in Windows XP > without the bug. It seems to be adding the extra points from the final rack > two times instead of once (so it is the total points on the rack times > four). O'laugh has asked about development help. > > (I remember I emailed JKB and Olaugh about helping at the beginning of the > project, but my abilities aren't quite up to their level. I had a course in > AI once! we learned the A* algorithm for heuristic searching! It can solve > the 15 puzzle! Also I used formal logic to find out who shot JR, and later I > wrote a theorem prover in IBM Basic!) > > I don't want to say for certain what version has this bug, or whether it > exists in Championship Player or simulations or both or neither, but it > shouldn't be very hard to figure that out. > > Thanks to Jason and John for Quackle, by the way! > > Kevin Leeds. > > > > > On Sat, May 8, 2010 at 4:51 AM, David B.Lewis <[email protected]>wrote: > >> >> >> Toward the end of a game, I thought I had made a play that shut down the >> last bingo lane and assured a win. When I load the position into Quackle and >> ask Championship Player about the situation, it suggests a few different >> moves with a 100% win but not the move that I chose. When I specifically >> enter that move and then simulate, the moves that Quackle had suggested show >> up as winning 100% of the time, and the move that I chose shows up as >> winning 99.96% of the time. >> >> For the life of me, I can't figure out what could happen in that 0.04% of >> the situations. There was only 1 tile in the bag at the time, so I've tried >> all of my opponent's possible racks, and nothing turns up that is a win. >> >> Is there any way to get at the information that Quackle uses during a >> simulation, so that I could either determine that a mistake was made on >> Quackle's part (and that my move was a sure thing) or find out just what >> terrific move my opponent may have had? >> >> Thx in advance. >> >> PS: ideally, I'd like Quackle to say "Great move, but you missed the >> remote possibility of your opponent's disconnected thirteen here". On a >> related note, in a previous post, I thought it would also be useful to show >> where moves played, so that I could see, for example, that 75% of my >> opponent's responses to my move use a hook that I had opened up or a >> hot-spot that I had created or a spot I didn't block. >> >> >> > >
