Jimmy Higgins
Contributor
- Joined
- Jan 31, 2001
- Messages
- 50,509
- Basic Beliefs
- Calvinistic Atheist
Cribbage, the manly man's game. I don't play human opponents much, so it is usually against the computer, which made me ponder about programming a cribbage opponent with different levels of skill. Cribbage is a bit like a sport, as there are different areas of mastering, unlike other card games. You've got:
Then I was pondering if you know how the player is going to play, do you modify how you play to defend it, making that style not as effective?
IE, when playing the cards:
But if you know all of this, doesn't it hurt, or should this just be expected because these are statistically sound ideas?
The other thing I ponder about are the best cut cards (ie increases your hand value) and the highest point card you can play. I was playing a game and pulled a cut card that increased my hand by 12 points, had two 3's and two 5's. And I was wondering, is there a better cut card for a hand, which on second thought would be a 7 if you have two 6's and two 8's (16 pt increase in hand value). Can that be out done?
Then there is play cards. There are couple of different angles, best single card play verses best point differential in a hand. Runs can lead to good points, 3,1,3,5,2,6,7,4 gets you 9 points without giving up subsequent runs. It'd be tough to get subsequent runs without giving any up like 1,3,2,5,4,7,6 would gain you 17 points for that hand of play. Of course, getting a sequence like this is near impossible and requires 99% luck and 1% awareness that the possible run exists. Of course, a 4 of a kind gets you 12, so maybe the most bestest single card play is 14. J, 5, 4, 4, 4, 4, but you give up points and the differential isn't as good.
Yes, I think way too much about this.
- Selecting which cards to hold on to
- Selecting which card to send to the crib
- Playing the card
Then I was pondering if you know how the player is going to play, do you modify how you play to defend it, making that style not as effective?
IE, when playing the cards:
- When playing the first card, I always play a card with a value less than 5, if I have one.
- I try not to put the play total to 21, and when possible, I try not to play a card that bumps the total to higher than 21
- When I can, my hopefully second to last card to 31 would be played in anticipation of a 10, so if we are at 16, and I have a 2 and 3, I'd play a 2, taking it to 18, they'd play a 10 to 28 and bam... 31!
- Avoid playing cards that can provide an opportunity for a sequence, unless you need the points to end the game in play... the other player could win first because you have the crib.
- I don't like 15's when I have to play a 5, unless I have a second 5 to triple it up. I commit ritual sacrifice if they play a fourth five.
But if you know all of this, doesn't it hurt, or should this just be expected because these are statistically sound ideas?
The other thing I ponder about are the best cut cards (ie increases your hand value) and the highest point card you can play. I was playing a game and pulled a cut card that increased my hand by 12 points, had two 3's and two 5's. And I was wondering, is there a better cut card for a hand, which on second thought would be a 7 if you have two 6's and two 8's (16 pt increase in hand value). Can that be out done?
Then there is play cards. There are couple of different angles, best single card play verses best point differential in a hand. Runs can lead to good points, 3,1,3,5,2,6,7,4 gets you 9 points without giving up subsequent runs. It'd be tough to get subsequent runs without giving any up like 1,3,2,5,4,7,6 would gain you 17 points for that hand of play. Of course, getting a sequence like this is near impossible and requires 99% luck and 1% awareness that the possible run exists. Of course, a 4 of a kind gets you 12, so maybe the most bestest single card play is 14. J, 5, 4, 4, 4, 4, but you give up points and the differential isn't as good.
Yes, I think way too much about this.