#1
|
|||
|
|||
King Louis
The game is King Louis, a form of five card draw in which kings are wild as well as the lowest card in each players hand is wild.
So my question is if you hold a duece is it best to keep it? For example you hold 2579K, should you draw 3 to the K2 because you have a better chance of hitting another 2? or draw 4 to your K hoping to net more wild cards that way? If anyone knows the math that would be great too. |
#2
|
|||
|
|||
Re: King Louis
AKA 'Kings and Little men' when played as a pot-matching game. No reason to keep a deuce since you're always going to have a 'smallest' card.
|
#3
|
|||
|
|||
Re: King Louis
Right, what I am asking is if your attempt is natuarally to increase the number of wilds in your hand is it more likely that you will pair a 2 by drawing three to a K2, or is it more likely that you will draw a low pair that is lower than your other to cards if you draw four to a K?
|
#4
|
|||
|
|||
Re: King Louis
Guesses: I think you should draw 3 to K2 instead of 4 to a King, but you should probably draw 4 to a King if your lowest card is above 7 (complete guess), but I've never heard of this game or done the math. Also, I suspect drawing to a straight or flush is virtually always wrong ,unless SF with 2 wild? You probably need Jacks full to bet for value after the draw vs someone trying to play well.
|
#5
|
|||
|
|||
Re: King Louis
The other question is are you playing Five of a Kind as a valid hand.
|
#6
|
|||
|
|||
Re: King Louis
My guess is that you shouldn't keep the 2, but let's work it out. Rather than trying to figure out "how often will I get my lowest card to be a pair", let's calculate "how often will I make quads or better". (Drawing 3AAA is no worse than getting 337A.)
Starting with 2579K and drawing three to 2K: This is the easy case. Let x = one of 3-Q or A (41 cards). Then the possible 3-card draws that give you quads or better are: 222: 1 combination 22K: 9 combinations 2KK: 9 combinations KKK: 1 combination PPP (P=3467TJQA): 32 combinations PPP (P=579): 3 combinations PPy (P=3467TJQA, y != 2PK): 1776 combinations (=8 * 4C2 * 37 remaining cards for 'y') PPy (579): 684 combinations 22x: 123 combinations ( 3C2 * 41 ) 2xK: 369 combinations xKK: 123 combinations 2xx: 2460 combinations ( 3 * 41C2, note we didn't count 2PP earlier.) xxK: 2460 combinations ( 3 * 41C2 ) Total = 8050 combinations Possible = 16215 combinations (47C3) Quads or better = 0.496 Starting with 2579K and drawing four to the K: This is more complicated. For each possible lowest card (I'll use 3 here) we want to calculate the number of possibilities for: 3333: 1 333K: 12 33KK: 18 33xx (x=4-Q,A): 3996 3xxK: 7992 3xKK: 444 3KKK: 4 3PPy (y=x except P): 6768, note we've already counted 33PP and 3PPK 3PPP: 124 This gives 16683 combinations with 2 as the lowest card (again) 19359 combinations with 3 as the lowest card, 15299 combinations for 4, 8877 combinations for 5, 9477 combinations for 6, 5211 combinations for 7, 5041 combinations for 8, 2511 combinations for 9, 1911 combinations for T, 859 combinations for J, 207 cominbations for Q, and 35 combinations with A as the "lowest card" (AAAA, AAAK, AAKK, AKKK) This gives a total of 85550 combinations in 178365 (47C4) possibilities, or 0.480 of getting quads or better. So, it appears that holding onto the deuce really is slightly better. But any errors I made in the calculation could easily swing things the other way. '2' is the most likely lowest number whether we draw 3 or 2 cards. Notice that in many cases where we would get quads by drawing four, we would also get quads by hanging onto the two. (Drawing one fewer only matters if it's the fourth card that pairs.) I didn't work out how often we get five of a kind in the two scenarios, though; there might be a difference. |
#7
|
|||
|
|||
Re: King Louis
[ QUOTE ]
PPy (579): 684 combinations [/ QUOTE ] You made a slip here: 3x3x38 = 342 combinations (you can also tell this is too many from the above entry: 1776/8 = 222 per specific unseen rank, so 3x222 = 666, but for unseen ranks, the number must be less than 666). Also, you forgot to mention that a straight flush can be made (it's better than quads). Suppose we start with precisely Ks 9s 7h 5d 2c. There are then 31 spade sf's, 28 heart sf's, 32 diamond sf's and 46 club sf's for a total of 137 straight flushes. Thus, there are 7845 combinations that make quads or better in the case of Ks 9s 7h 5d 2c and the probability of making quads or better is about 0.483811286. DRAWING FOUR -------------- I think the best way to think about this part is to consider regular draw and what the chances are of drawing four to a high card and making one of: quads, full house, trips, two-pairs or a pair. You can see that for each case in regular draw, you will make at least quads in "King Louis". It just so happens I have a table for drawing four for regular draw: C(47,4) = 178365 combinations quads: 52 boat: 288 trips: 4102 two pairs: 8874 pair: 76392 The total here is 89708, so to make quads or better AND to make at least a pair in regular draw, the odds would be about 0.502946206. Here, I haven't even computed the combinations that will make a straight flush or better. It's clear that you should draw FOUR to a king rather than keep the deuce in "King Louis". |
#8
|
|||
|
|||
Re: King Louis
If you redo the calculations to have only 2 kings left in the deck will that change it back to keeping the 2?
|
#9
|
|||
|
|||
Re: King Louis
No.
From my calculations, if you start with Ks 9s 7h 5d 2c and the deck is missing the Kh, you should still draw four to the king. I get the probability of making quads or better drawing three to K2 as 6810/C(46,3) = 0.448616601. Even forgetting about making a straight flush when drawing four cards to the king, you will make quads or five of a kind with a probability of 74528/C(46,4) = 0.456708644. |
|
|