better bots
 
AuthorMessage
tesla1889
17 Mar 2014 11:19 PM
there should be some more difficult bots to play against, especially with the lack of players nowadays
scooter1101980
18 Mar 2014 5:33 AM
Hello. I'm new to this site of course, and have never played, but got the old board game on e-bay. Old enough to know this game, but never did. Great game, but no one plays in my household. Playing my third game now, lost the first two big time, probably will the third. but . . . . gaining big time experience for months to come. I hope it's before months that I can play better, but who knows? Everyone has been very nice and encouraging on the site. I think what the site needs is a little more direction for the newbie, like myself. I think I know the rules, but do not know why people resign, what is that factor? Also, does the pie rule take effect in this game? If anyone could answer these questions, I would  greatly appreciate it, especially about the resigning. Also, about the better bots. maybe, but I know a few people who would come on to site if they could see rules. Maybe there are, I don't know. I do know, the people I have played have really taught me a lot by their moves, and I appreciate it. Thank you, and the goal of this site has to be to get more people involved. Thanks everyone.
yibyab
05 Aug 2014 2:53 AM
The pie (swap) rule is not implemented here.  The site owner (Mark) has done all the coding and hosts this site as a side-hobby, I think, so he can only make changes when he has time.  People have suggested enhancements in another topic, and Mark said he'd get to them when he could.  The pie rule was one of them, though I think the more important change ought to be a confirmation-of-move function.  It's way too easy to accidentally make a move, especially if you're not careful with a laptop touchpad.

But yeah, an option to swap or steal player #1's opening move would balance the game out.  As it is now, player #2 is at a steep disadvantage unless the opening move is graciously weak and intentionally neutral. 

I'm not sure about your question about resigning.  A player who realizes his position is hopeless can choose to resign rather than continue to stretch out a game, hoping for a mistake by his opponent.  The game will automatically end only when one player successfully connects a string from one border to the other.  But games are essentially decided well before that, so most of the time, the losing player will just hit 'resign' rather than waste time executing the moves all the way to automatic ending.

Someone has edited the Wikipedia page about TwixT in very detailed fashion.  It can be confusing to the new player because it delves into strategy when all the new player may want is basic play instruction.  Still, it's a good resource.  Here at TwixtLive, Mark has it coded much more like the physical board game, where a player can remove his own links in order to make a link that crosses an existing line.  (I didn't realize that when i first started playing, believing that once a link was laid down it could never be removed or crossed.)  Other sites use a convention where you can't remove a link but you can cross over your own.  Here, with link removal but no crossing, it's a little trickier if things start to get knotted up.

One of the things about playing TwixT or other turn-based games online that took me a moment to get used to was that you don't have to be online the same time as your opponent.  I originally thought this place was pretty vacant and sleepy when I first found it, since I always seemed to be here alone.  I thought I had to be here the same time as someone else in order to play.  But there are usually about dozen active players here, just not all logged on at the same time.  You learn that you can get engaged in a game with others yet never be logged on at the same time.  You make your move.  Then check back later or the next day to see if your opponent has made a move.  Games can take weeks to play, but it's sort of like post card chess that way. 

I don't know anyone who plays the timed games (2 mins; 20 mins to make a move).  Most play the no expiration games precisely because you never know when your opponent can respond to a move.  This results, though, in some games languishing forever unfinished because an opponent quits or forgets and never comes back.  I currently have four games open against players who haven't returned in more than a year, and there's no way to close those games out.  Another suggestion for update for Mark is to put in a maximum time before an opponent defaults or forfeits without making a move.

I've been away for awhile, but I'll probably start playing here more regularly.  I'm quite chatty and I do check the forum regularly, so if there are any questions I can help with, I'll try.  I'm just a visitor though.  It isn't my site.  Just so that's clear.  I can't make any changes, like making the Bot smarter in the self-play game.  Frankly, even at the most difficult setting, that robot play isn't very useful other than in learning the mechanics of the game.  I imagine making a Twixt Bot smart is a challenging coding process, so I wouldn't expect a fix there any time soon.  I'd lobby for other enhancements first.

/Bob (Yibyab) 
THX1136
16 Jun 2014 5:09 AM
I would agree with the "better bots" observation. All the choices are quite easy to beat - especially since you are always the first to move at the start of a game. It would be nice to see the choice to reverse the order of play so the computer goes first. I can understand when this is here as more of a hobby that Mark would not have the time to tweak out the site - especially in light of the amount of traffic it gets. Glad it's here though; no complaints!
 
 
 
Reply