Wild52 at the Flamingo in Las Vegas
February 20th, 2011 at 10:07:51 AM
permalink
I was in Las Vegas for Super Bowl weekend and staying at the Flamingo. Luckily, I only came home with a mild flu and no bedbug bites or norovirus. While I was there, I saw a new table game named Wild52' rel='nofollow' target='_blank'>http://lasvegasblog.harrahs.com/las-vegas-casinos/flamingo/its-time-to-get-wild-52-at-flamingo-las-vegas/"]Wild52 (link here). The game is a seven card stud game of you + five of your closest friends vs. the dealer. The deck contains one fully wild card that substitutes for any card in the deck that's not already in your hand (sorry, no flushes with two aces allowed) except when it makes a five of a kind. Each person including the dealer is dealt five cards and two community cards are dealt face down. To play, you must ante (table minimum was $10 when I was there), once you receive your cards, you can choose to bet by placing double your ante in the bet circle or fold and lose your ante bet. One of the community cards is then turned over and you now get to either raise by placing double your ante in the raise circle or stand (no bet). The dealer the turns the second community card over and evaluates his own hand. If he doesn't have a pair of fives or better (a little less than a quarter of the time) then he returns your "Bet" and "Raise" but the "Ante" still plays. The dealer must have a pair of fives or better to qualify your bet and raise.
There are two bonus bets, one is for the joker appearing in either your hand or the two community cards and has about a 11.6% casino advantage and the other bonus bet is a pay table based on your final hand that is about 10.5% to the casino's advantage.
I'm just beginning to put together an engine to calculate a betting strategy for this game. My initial goal was to try to figure out which hands to fold and which hands to automatically bet and raise. So far, a pair of Queens or better is an automatic bet and raise. There is no easy formula for figuring out when to fold .. yet.
Here's a spreadsheet of my' rel='nofollow' target='_blank'>http://www.dottygale.com/p1x/wild52_bet_anaysis.xlsx"]my initial Monte Carlo simulation. It shows the starting five card hand, the ranks of that cards within it (in code), How many wins, losses, push, How many times the players hand won but the dealer didn't qualify, same for losses, the Ante (as a negative number), wins/losses based on a single bet, wins/losses based on a bet and a raise and then some calculated columns to decide whether to fold, bet and/or raise.
So far, the casino advantage is 12.5%. When I add logic for betting four card flushes it may improve a little bit but in my initial opinion, this game stinks!
There are two bonus bets, one is for the joker appearing in either your hand or the two community cards and has about a 11.6% casino advantage and the other bonus bet is a pay table based on your final hand that is about 10.5% to the casino's advantage.
I'm just beginning to put together an engine to calculate a betting strategy for this game. My initial goal was to try to figure out which hands to fold and which hands to automatically bet and raise. So far, a pair of Queens or better is an automatic bet and raise. There is no easy formula for figuring out when to fold .. yet.
Here's a spreadsheet of my' rel='nofollow' target='_blank'>http://www.dottygale.com/p1x/wild52_bet_anaysis.xlsx"]my initial Monte Carlo simulation. It shows the starting five card hand, the ranks of that cards within it (in code), How many wins, losses, push, How many times the players hand won but the dealer didn't qualify, same for losses, the Ante (as a negative number), wins/losses based on a single bet, wins/losses based on a bet and a raise and then some calculated columns to decide whether to fold, bet and/or raise.
So far, the casino advantage is 12.5%. When I add logic for betting four card flushes it may improve a little bit but in my initial opinion, this game stinks!
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 20th, 2011 at 11:05:27 AM
permalink
Does that mean they took out Pokerjack?
Edit: I just found this link:
http://www.totalrewards.com/casino-gambling/table-finder/table-finder-home.shtml
Pretty informative, if it's up to date. (And no, Pokerjack isn't on there).
Edit: I just found this link:
http://www.totalrewards.com/casino-gambling/table-finder/table-finder-home.shtml
Pretty informative, if it's up to date. (And no, Pokerjack isn't on there).
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice."
-- Girolamo Cardano, 1563
February 20th, 2011 at 1:17:15 PM
permalink
I actually kicked off an exhaustive analysis of the Ante bet (including the Play and Option bets) last night, but it probably won't be done for a couple of weeks. There are 150,891 * 48 * 47 * 1,370,754 = 466,618,500,732,384 unique scenarios to analyze, so obviously it is going to take a while to finish.
What didn't take long was an analysis of the two sucker side bets. The house edge on the Joker bet is 11.3208% and the house edge on the Bonus bet is 10.4271%. I won't know the house edge (or element of risk) of the Ante bet until the exhaustive analysis is finished.
I also made an assumption that I hope is correct. Consider the following scenario:
Community cards: Ac Ad
Dealer has: Joker, Kc, Kd, 2s, 3h
Player has: Ah, Ks, Kh, 2h, 3s
Both the dealer and the player have Aces full of Kings as their best 5-card hand. I made the assumption that, for the ante bet, only the rank matters, and that it does not matter whether it is natural or wild. Hopefully that is correct. The Bonus paytable makes a distinction between Natural and Wild hands (preferring Naturals), but I assumed that for showdown purposes, only the poker rank matters, and that in the above scenario the player and dealer tie, and therefore the player's Ante, Play, and Option bets would push.
What didn't take long was an analysis of the two sucker side bets. The house edge on the Joker bet is 11.3208% and the house edge on the Bonus bet is 10.4271%. I won't know the house edge (or element of risk) of the Ante bet until the exhaustive analysis is finished.
I also made an assumption that I hope is correct. Consider the following scenario:
Community cards: Ac Ad
Dealer has: Joker, Kc, Kd, 2s, 3h
Player has: Ah, Ks, Kh, 2h, 3s
Both the dealer and the player have Aces full of Kings as their best 5-card hand. I made the assumption that, for the ante bet, only the rank matters, and that it does not matter whether it is natural or wild. Hopefully that is correct. The Bonus paytable makes a distinction between Natural and Wild hands (preferring Naturals), but I assumed that for showdown purposes, only the poker rank matters, and that in the above scenario the player and dealer tie, and therefore the player's Ante, Play, and Option bets would push.
February 20th, 2011 at 4:41:55 PM
permalink
It took my humble PC three days to go through all 154,143,080 hands in a 53 card deck just to see if my joker evaluation engine was working correctly. Thanks to the Flying Spaghetti Monster, it worked on the first try. I like the Monte Carlo method because I can get a good feeling for trends without having to evaluate trillions of scenarios. As you can see, I came pretty close to your numbers on the sucker bonus bets.
I've added extra granularity to my spreadsheet for each 5 card dealt hand, whether it has a wild, four card flush or an outside straight. I'm going to expand my definition of an outside straight to include a situation like 3,5,6,7,9 where it takes one of two card ranks to complete the straight. Previously, I was only capturing the four cards in a row. Once I get a few million hands generated, I'll add a new link in this thread.
I've added extra granularity to my spreadsheet for each 5 card dealt hand, whether it has a wild, four card flush or an outside straight. I'm going to expand my definition of an outside straight to include a situation like 3,5,6,7,9 where it takes one of two card ranks to complete the straight. Previously, I was only capturing the four cards in a row. Once I get a few million hands generated, I'll add a new link in this thread.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 20th, 2011 at 5:09:55 PM
permalink
Quote: s2dbakerIt took my humble PC three days to go through all 154,143,080 hands in a 53 card deck just to see if my joker evaluation engine was working correctly. Thanks to the Flying Spaghetti Monster, it worked on the first try. I like the Monte Carlo method because I can get a good feeling for trends without having to evaluate trillions of scenarios. As you can see, I came pretty close to your numbers on the sucker bonus bets.
I've added extra granularity to my spreadsheet for each 5 card dealt hand, whether it has a wild, four card flush or an outside straight. I'm going to expand my definition of an outside straight to include a situation like 3,5,6,7,9 where it takes one of two card ranks to complete the straight. Previously, I was only capturing the four cards in a row. Once I get a few million hands generated, I'll add a new link in this thread.
I'm curious about your simulator. What strategy are you using the for Play and Option bets?
For example, some hands should be folded before seeing the first community card. Some hands you should make the Play bet but not the Option bet. Other hands you should make both the Play and Option bets. This is one of those games where you need to analyze what you would do later on in order to determine what you should do earlier on.
Are you determining optimal strategy on the fly as the cards are dealt?
February 20th, 2011 at 5:49:24 PM
permalink
With me, these thing evolve. First, I got MS-SQL to randomize a deck of cards ( later cut to the first 40 cards to speed it up ), previously I had programmed a series of functions that could evaluate a 5,6 or 7 card hand for "Texas Hold'em Bonus" and so I converted those over to Wild52 to consider the possibility of a wild card. Then I wrote a script to deal out two 5-card hands and 2 community cards to see if I was doing that right. I expanded the script to store all of the pertinent information into a table and expanded it to six player hands and one dealer hand per deck. I set it to run in a loop for awhile. I occasionally pluck the info out of the table and store it in a spreadsheet to see if I can make a determination about strategy.Quote: JB...Are you determining optimal strategy on the fly as the cards are dealt?
That's where I am now.
Here's the SQL statement that generates the data I put in' rel='nofollow' target='_blank'>http://www.dottygale.com/p1x/wild52_bet_analysis2.xlsx"]in this spreadsheet (link fixed):
select c_index,
wild_5 is_wild,
Four_Card_Flush,
Outside_Straight,
SUM(win) Win,
SUM(loss) Loss,
SUM(push+dnq_push) Push,
SUM(dnq_win) DNQ_Win,
SUM(dnq_loss) DNQ_Loss,
sum(win+loss+push+dnq_win+dnq_loss+dnq_push)*-1 Ante
From wild52.dbo.wild52_test
Group By c_index,
wild_5,
Four_Card_Flush,
Outside_Straight
Order By 1,2,3,4
After a little analysis, I'll create another table that stores wins and losses for bets made per the instructions I give it. For example, I'll tell it to bet on any hand that has a pair of Queens or better, Has a joker in it, four of a suit or an outside straight. Then I'll include the first community card into the evaluation and see where I am. I'll probably tell the engine to raise if the 6 card hand evaluates to a pair of Queens or better. If I see the percentages for the player improve, I know I'm guessing better on strategy. My goal is to get a strategy that's easy to remember that minimizes the casino's vig.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 20th, 2011 at 7:44:24 PM
permalink
Quote: JBI also made an assumption that I hope is correct. Consider the following scenario:
Community cards: Ac Ad
Dealer has: Joker, Kc, Kd, 2s, 3h
Player has: Ah, Ks, Kh, 2h, 3s
Both the dealer and the player have Aces full of Kings as their best 5-card hand. I made the assumption that, for the ante bet, only the rank matters, and that it does not matter whether it is natural or wild. Hopefully that is correct. The Bonus paytable makes a distinction between Natural and Wild hands (preferring Naturals), but I assumed that for showdown purposes, only the poker rank matters, and that in the above scenario the player and dealer tie, and therefore the player's Ante, Play, and Option bets would push.
Funny you should voice that concern. I had an email exchange with "Ken" from the Wild52 web site and I gave him a similar scenario only in Flush vs. Flush with joker. He said that the two hands would push. I'm sure he wouldn't mind me posting the redacted email trail:
Quote: KevinHi Ken,
Last weekend, when I was at the Flamingo, I didn’t get the opportunity to ask the dealer about certain situations that may arise with the use of the wild card. I’m hoping that you can clear something up for me. My understanding is that the wild card will substitute for any card in the deck. So according to the rules of Wild52, which hand would be superior:
Joker, Ace Diamonds, 9 Diamonds, 7 Diamonds, 3 Diamonds for a Flush Or
Ace Clubs, King Clubs, 9 Clubs, 7 Clubs 3 Clubs for a Flush?
Since the Joker substitutes for any card in the deck, if you have four Aces and a Joker, for what card is the Joker substituting to make 5 of a kind?
I’m a computer programmer and want to get my simulation of your new game as correct as possible.
Thanks You,
Kevin
His reply:
Quote: KenHi Kevin,
Thanks for the e-mail. To answer your questions, both flushes are equal. The joker would be a king of diamonds. Between the dealer and player, the hand would be a "push". If the player held the joker and placed a joker bet, he or she would be paid 4 to1 or 10 to 1, if joker was one of the community cards.
Regarding the other question, the joker would be recognized as a fifth ace and make the highest hand in the game of Wild 52!
Thanks again,
Ken N.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 20th, 2011 at 8:04:07 PM
permalink
I will definitely have a look my next visit to the Strip.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
February 21st, 2011 at 5:45:39 PM
permalink
I'm going back to Las Vegas from April 29 through May 2. Maybe I'll have all the bugs worked out of my Wild52 simulator by then. Meanwhile, here's 14+ million more random hands.
edit:
It looks like I found a strategy (Hand 2) that lets me be up after 2600 hands. That's probably more than I'll ever play in a lifetime. It's pretty easy to follow too. I'll crunch the numbers some more. I'm sure that Hand 2 will come back to reality sooner or later.
edit:
It looks like I found a strategy (Hand 2) that lets me be up after 2600 hands. That's probably more than I'll ever play in a lifetime. It's pretty easy to follow too. I'll crunch the numbers some more. I'm sure that Hand 2 will come back to reality sooner or later.
Hand win win_played loss loss_played push_count fold_count win_count loss_count vig
1 3077 991 3181 1665 2 1096 1372 1283 -0.01661872802
2 3361 1049 3178 1606 2 1007 1384 1269 0.02798593057
3 3252 1080 3518 1574 3 805 1335 1318 -0.03929098966
4 3459 1187 3717 1469 1 526 1344 1310 -0.03595317725
5 3816 1290 3933 1363 4 280 1362 1291 -0.01509872241
6 3707 1305 4324 1344 8 0 1305 1344 -0.07682729423
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 21st, 2011 at 9:08:16 PM
permalink
I seem to have stumbled on to a strategy that yields a mere 2.9% to the house. I'll have a better read on it in the morning.
Bet with:
a Wild Card in your Hand Or
A,J,10,x,x or better Or
Four cards to a Flush Or
and Outside Straight
Otherwise fold.
Raise with:
a Wild Card in your Hand Or
J,J,Q,x,x or better if the exposed community card is not a Wild Card Or
J,J,J,x,x or better Or
( Four cards to a Flush AND An Outside Straight ) <- Include the community card
Bet with:
a Wild Card in your Hand Or
A,J,10,x,x or better Or
Four cards to a Flush Or
and Outside Straight
Otherwise fold.
Raise with:
a Wild Card in your Hand Or
J,J,Q,x,x or better if the exposed community card is not a Wild Card Or
J,J,J,x,x or better Or
( Four cards to a Flush AND An Outside Straight ) <- Include the community card
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 22nd, 2011 at 11:15:18 AM
permalink
I think I'm begining to dislike games that require the dealer to qualify.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 22nd, 2011 at 7:48:12 PM
permalink
I take it all back.Quote: s2dbakerI think I'm begining to dislike games that require the dealer to qualify.
I can squeeze the vig down to 2.1% with a relatively easy to remember strategy.
First off, You will fold 3 out of ever 10 hands. If you hate folding a hand and giving the casino your ante then this game isn't for you.
You will bet when you have:
The Wild Card in your 5 card hand OR
You're dealt any pair or better OR
You have a four card flush OR
You have an outside OR
You have an inside straight OR
Your hand has at least four cards of rank 8 or higher.
You will raise a little less than 4 out of ever 10 hands that you bet.
You will raise when you have:
The Wild Card is in your original 5 card hand OR
( Your six card hand has a four card flush AND an outside straight chance ) Or
Your six card hand is three of a kind or better regardless of where the Wild Card is Or
Your six card hand is equal to or better than a pair of Queens with a King kicker and the exposed community card is not the Wild Card.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 22nd, 2011 at 8:10:53 PM
permalink
Here's some geekery to help instill confidence. I simulated a single table with six players against a dealer. Each player got a chance to play about 200,000 hands or so (it's still running). I set each player to bet the same strategy. Here are the hands, 1 through 6 and an aggregate (99):Win is the units won
Win_Play is the amount of hands with a winning outcome for the player
Loss is the units lost
Loss_Play is a losing outcome for the player including Folding
Raise_pct is the percentage of hands raised after the bet, not all hands
Win_Count and Loss_Count are results regardless of how thehand was played out ( you can see I had a lucky dealer )
Vig is the percentage loss (win-loss)/(win+loss)
Hand Win Win_Play Loss Loss_Play Push_Count Fold_Count Fold_pct
1 255297 82493 265428 117562 190 58503 0.29237173784
2 254383 82153 266049 117897 195 58595 0.29281980550
3 254929 82369 265683 117681 195 58503 0.29236151020
4 253907 82051 266472 117992 203 58623 0.29301537477
5 254907 82387 264434 117640 218 58935 0.29454392795
6 255444 82298 265195 117755 192 58803 0.29386366021
99 1528867 493751 1593261 706527 1193 351962 0.29316265870
Hand Raise_Count Raise_pct Win_Count Loss_Count Vig
1 53107 0.37506267876 100050 99858 -0.01945556675
2 53010 0.37459985442 99861 100050 -0.02241599286
3 52994 0.37424612646 100030 99880 -0.02065645816
4 53020 0.37484534624 99627 100238 -0.02414586291
5 52860 0.37448460546 99851 100020 -0.01834440184
6 53224 0.37667374380 100080 99831 -0.01872890813
99 318215 0.37498512267 599499 599877 -0.02062503523
Win_Play is the amount of hands with a winning outcome for the player
Loss is the units lost
Loss_Play is a losing outcome for the player including Folding
Raise_pct is the percentage of hands raised after the bet, not all hands
Win_Count and Loss_Count are results regardless of how thehand was played out ( you can see I had a lucky dealer )
Vig is the percentage loss (win-loss)/(win+loss)
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 23rd, 2011 at 2:58:08 AM
permalink
I make one small change to the strategy and knocked the vig down to just over 1.8%
Instead of raising on Q,Q,K,x,x, I changed it to Q,Q,A,x,x
Changing it to A,A,x,x,x made it worse. I'll try the K,K,x,x,x next.
Instead of raising on Q,Q,K,x,x, I changed it to Q,Q,A,x,x
Changing it to A,A,x,x,x made it worse. I'll try the K,K,x,x,x next.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 23rd, 2011 at 6:51:04 AM
permalink
Quote: s2dbakerI make one small change to the strategy and knocked the vig down to just over 1.8%
You should keep at it until you find the player advantage ;)
Seriously, thanks for the analysis and strategy. Socccer bets aside (sucker bets? <shrug>) it doesn't seem like such a bad game to play.
Donald Trump is a fucking criminal
February 23rd, 2011 at 8:50:43 AM
permalink
Thanks Nareed. I'll keep trying. I'm going to apply some Contract Bridge type thinking to the raises next. It helped with the bets so why not?
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 23rd, 2011 at 9:16:50 AM
permalink
You're welcome. I see the house edge keeps dropping the more you look at it. Even if it's a lousy game, the analysis is very much appreciated.
Donald Trump is a fucking criminal
February 23rd, 2011 at 10:02:37 AM
permalink
Quote: s2dbakerThanks Nareed. I'll keep trying. I'm going to apply some Contract Bridge type thinking to the raises next. It helped with the bets so why not?
If you don't want to go through the exhaustive analysis (it looks like JB already has that cooking), why not try a hybrid? Instead of guessing at the right strategy, input a starting hand, then evaluate each of the possible strategies (fold/bet, stand/raise) through some number of randomly-picked hands. You'll quickly converge on an optimal strategy for each starting hand, then save that and run the simulation under that strategy.
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice."
-- Girolamo Cardano, 1563
February 23rd, 2011 at 10:17:28 AM
permalink
I think there's over 7000 unique starting hand evaluations. I don't think I can squeeze anymore out of Fold/Bet without making it complicated. Then there's 7000 more unique 6 card hand evaluations where all you have to do is stand or raise. The major difference between the two bets is that you have 2 chances to improve your hand on the initial bet and only one more chance on the raise. With that in mind, it's a simpler game to figure out than Texas Hold'em Bonus which I also like. I think I'm just one or two more guesses away from wrapping this one up.Quote: MathExtremistIf you don't want to go through the exhaustive analysis (it looks like JB already has that cooking), why not try a hybrid? Instead of guessing at the right strategy, input a starting hand, then evaluate each of the possible strategies (fold/bet, stand/raise) through some number of randomly-picked hands. You'll quickly converge on an optimal strategy for each starting hand, then save that and run the simulation under that strategy.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 24th, 2011 at 8:40:31 PM
permalink
I will have some minor adjustments to this, one of them being that if the Wild Card pops up as the first Community Card then you raise with three of a kind 8s or better otherwise check. Also I'm trying some Bridge type valuations for betting where Aces get 4 points Kings 3, Queens 2 and Jacks 1. So far, that's not dramatically improving anything and just making it more difficult to remember.Quote: s2dbakerYou will bet when you have:
The Wild Card in your 5 card hand OR
You're dealt any pair or better OR
You have a four card flush OR
You have an outside OR
You have an inside straight OR
Your hand has at least four cards of rank 8 or higher.
You will raise a little less than 4 out of ever 10 hands that you bet.
You will raise when you have:
The Wild Card is in your original 5 card hand OR
( Your six card hand has a four card flush AND an outside straight chance ) Or
Your six card hand is three of a kind or better regardless of where the Wild Card is Or
Your six card hand is equal to or better than a pair of Queens with a King kicker and the exposed community card is not the Wild Card.
Here is the latest' rel='nofollow' target='_blank'>http://www.dottygale.com/p1x/wild52_bet_analysis4.xlsx"]latest version of data in an excel spreadsheet. The inappropriately named "Sheet 3" has the analysis for the first community card being the wild card.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
February 26th, 2011 at 8:00:20 PM
permalink
I came up with 11.3193%. My calculation is as follows:Quote: JB...The house edge on the Joker bet is 11.3208%...
Select 1.0-((((5.0*5.0)+(11.0*2.0))/7.0)*(1.0-((52.0/53.0)*(51.0/52.0)*(50.0/51.0)*(49.0/50.0)*(48.0/49.0)*(47.0/48.0)*(46.0/47.0))))
I'll break that up:
1 - ((52/53) * (51/52) * (50/51) * (49/50) * (48/49) * (47/48) * (46/47)) Is the chance of a wild car appearing in any one of seven places on the table where it pays for it to do so or 13.207757323%
Your return is 5 units if it's in one of 5 spots and 11 units if it's in the other two remaining spots. Average return of 6.714285 units.
Take the return times the probability and you get 88.680656 units back for every 100 units you bet leaving the house advantage at 11.3193.
I don't want to split hairs since that difference probably wouldn't cover the tip for one drink in a month of play but I'd like to know if I'm missing something from my calculation so that I can get the best calculations possible.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
March 2nd, 2011 at 8:19:39 PM
permalink
Here's 10 Megs Of Spreadsheet. It's the latest analysis of Wild52 (probably the last ). I don't think I can squeeze any more out of this game. Fiddling with the hand strengths may make a little difference here or there but it won't amount to much. Here's my method for getting the house edge down to 1.78%
In SQL language:
In English:
Evaluate your 5 card hand and bet if you have:
1) The Wild Card Or
2) Four cards of the same suit, an outside straight, inside straight or any combination thereof Or
3) Equal to or better than A, J, T, 8, 6 and your fourth highest card rank is 8 or better Or
4) Equal to or better than A, K, Q, 8, 3
Otherwise fold. You will fold over 30% of these hands.
If you made it past the betting hurdle then raise your bet if you have:
1) The Wild Card in your original five card hand Or
2) Four cards of the same suit and an outside straight with the six cards Or
3) Three of a kind 7s with at least a Jack and a 3 ( Q and 2 will work, 10 and 4 will not )
The next four items apply when there's no wild card in the first community card:
4) Pair of Jacks with a Queen kicker and Four suited with the six cards
5) Equal to or better than Q, Q, K, J, 10 and an Outside Straight with the six cards
6) Equal to or better than K, K, A, Q, 10 and an Inside Straight with the six cards
7) Two Pair 6s and 3s with the six cards
The next two items apply only if the wild card flips in the first community card spot
4w) Three of a kind deuces with at least a 10 kicker And four cards suited or an outside straight ( this is a really rare situation that's hardly worth mentioning but the wild card can be key to making your four card flush or outside straight )
Otherwise check. You will Raise less than a third of the time after you bet.
12 rules to memorize. I tried to keep it simple. If you want to make it really simple then ignore rules 4, 5 and 6 and the house edge rises to around 1.91%
In SQL language:
If @p_c_index > 11413120802 Or
( @p_c_index > 11411100805 And @l_4th_highest_card > 7 ) Or
53 In (@p_card1,@p_card2,@p_card3,@p_card4,@p_card5) Or
@l_four_card_flush + @l_outside_straight + @l_inside_straight > 0
Set @p_bet = 2
Else
Set @p_fold = 1
Select @l_four_card_flush = dbo.f_eval_small_flush(@p_card1,@p_card2,@p_card3,@p_card4,@p_card5,@p_community_card_1,0)
Select @l_outside_straight = dbo.f_eval_outside_straight(@p_card1,@p_card2,@p_card3,@p_card4,@p_card5,@p_community_card_1,0)
Select @l_inside_straight = dbo.f_eval_inside_straight(@p_card1,@p_card2,@p_card3,@p_card4,@p_card5,@p_community_card_1,0)
If ( @p_c_index_6 > 30606030300 And @p_community_card_1 < 53 ) Or
( @p_c_index_6 > 21313141209 And @l_inside_straight > 0 And @p_community_card_1 < 53 ) Or
( @p_c_index_6 > 21212131109 And @l_outside_straight > 0 And @p_community_card_1 < 53 ) Or
( @p_c_index_6 > 21111120000 And @l_four_card_flush > 0 And @p_community_card_1 < 53 ) Or
( @p_c_index_6 > 40202021000 And @l_four_card_flush + @l_outside_straight > 0 And @p_community_card_1 = 53 ) Or
@p_c_index_6 > 40707071102 Or
53 In (@p_card1,@p_card2,@p_card3,@p_card4,@p_card5) Or
@l_four_card_flush + @l_outside_straight > 1
Set @p_raise = 2
End
In English:
Evaluate your 5 card hand and bet if you have:
1) The Wild Card Or
2) Four cards of the same suit, an outside straight, inside straight or any combination thereof Or
3) Equal to or better than A, J, T, 8, 6 and your fourth highest card rank is 8 or better Or
4) Equal to or better than A, K, Q, 8, 3
Otherwise fold. You will fold over 30% of these hands.
If you made it past the betting hurdle then raise your bet if you have:
1) The Wild Card in your original five card hand Or
2) Four cards of the same suit and an outside straight with the six cards Or
3) Three of a kind 7s with at least a Jack and a 3 ( Q and 2 will work, 10 and 4 will not )
The next four items apply when there's no wild card in the first community card:
4) Pair of Jacks with a Queen kicker and Four suited with the six cards
5) Equal to or better than Q, Q, K, J, 10 and an Outside Straight with the six cards
6) Equal to or better than K, K, A, Q, 10 and an Inside Straight with the six cards
7) Two Pair 6s and 3s with the six cards
The next two items apply only if the wild card flips in the first community card spot
4w) Three of a kind deuces with at least a 10 kicker And four cards suited or an outside straight ( this is a really rare situation that's hardly worth mentioning but the wild card can be key to making your four card flush or outside straight )
Otherwise check. You will Raise less than a third of the time after you bet.
12 rules to memorize. I tried to keep it simple. If you want to make it really simple then ignore rules 4, 5 and 6 and the house edge rises to around 1.91%
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
April 21st, 2011 at 8:16:20 PM
permalink
I'd like to see the Wizard's analysis of Wild52. I'm going to Las Vegas next weekend and I'd like to know if I was in the ballpark with my own analysis.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
April 21st, 2011 at 10:42:19 PM
permalink
Quote: s2dbakerI'd like to see the Wizard's analysis of Wild52. I'm going to Las Vegas next weekend and I'd like to know if I was in the ballpark with my own analysis.
Sorry, I haven't analyzed it. This analysis will be somewhat involved, so I'd like to see a few more placements before I bother with it.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
May 16th, 2011 at 2:37:49 PM
permalink
People are scared of this game. I was in Las Vegas a couple of weeks ago and noticed that the Wild52 game had 2 or three people playing even when everything else was packed at the Flamingo. That gave me a chance to play it for a little while. The Goddess Fortuna was not with me that evening as the dealer kept coming up with straights, flushes and full houses whenever I had a kick-ass hand. My full house should never get beaten by the dealer's full house!!!! Grrr!!!! I don't have the exact amounts but I think I lost $100 betting the minimum $5 over the course of an hour.
Here are some observations: The dealers encourage you to make the sucker bonus bet and if you aren't betting it, and are more than happy to point out how much money you would have won had you been betting it. The dealers make a lot of errors. I got paid on a straight when the dealer had a flush. I had to bite my tongue hard on that one because the other players saw it too. We all got paid on the Ante because she picked up on the pair of 5s. In the course of an hour the dealers made several mistakes, sometimes in our favor and sometimes against, the latter we were more than happy to correct.
I found favor with Fortuna by switching to Texas Hold'em Bonus. I won enough to cover dealer and cocktail waitress tips. If it's still there the next time I go to Las Vegas, I'll play Wild52 again. It feels like you're more engaged than with Texas Hold'em Bonus.
Here are some observations: The dealers encourage you to make the sucker bonus bet and if you aren't betting it, and are more than happy to point out how much money you would have won had you been betting it. The dealers make a lot of errors. I got paid on a straight when the dealer had a flush. I had to bite my tongue hard on that one because the other players saw it too. We all got paid on the Ante because she picked up on the pair of 5s. In the course of an hour the dealers made several mistakes, sometimes in our favor and sometimes against, the latter we were more than happy to correct.
I found favor with Fortuna by switching to Texas Hold'em Bonus. I won enough to cover dealer and cocktail waitress tips. If it's still there the next time I go to Las Vegas, I'll play Wild52 again. It feels like you're more engaged than with Texas Hold'em Bonus.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
May 20th, 2011 at 10:17:54 PM
permalink
I don't know why this game fascinates me. I re-engineered the entire database so that I could try something different. A quick review of the betting rules is in order:
There are three bets; Ante, Bet and Raise. The Bet and Raise are twice the amount of the Ante. The Bet and Raise are returned to the player if the dealer doesn't have a pair of 5s or better. The player can fold after the Ante but does not have to raise.
So the decision points are, 1) Do I bet or fold after looking at my initial 5 cards? 2) Do I raise or stand pat after seeing the first community card?
Let's start with 2. You want to Raise your bet when your hand's potential wins more than enough to cover the times that you don't get paid because the dealer doesn't qualify. It has to be a pretty strong hand. What is that hand? I don't care, because I played out 100 million hands and stored the information in a table. I'll worry about simplifying it later but right now, I can use that table as a base of knowledge for the first decision.
When to fold and when to play. You want to play when you are going to win or at least lose less than if you fold. But how do you know when you are going to lose less than if you just fold? Using the data from the first 100 million hands, I can get a good idea with what I should raise and with what I should stand pat. So I started up the simulator again and using the tables I've already built (and continue to build upon) I'm figuring out where it makes sense to fold.
I call this the psychic time traveller method since it requires knowing 11,358 (more or less) possible changing datapoints in advance and if it changes, changing the past. The data is constantly being saved and re-evaluated so it's non-deterministic. A hand that you would have raise with 20 minutes ago may have had an unfortunate run of luck and now it's not a hand with which to raise. So with that in mind, you may be betting 5 or just 3 but not always consistently, it all depends on the state of the data at that moment . Then you look at the losses and if they exceed the total number of hands (including ties because you would have folded that too), then the advice is to fold otherwise you play on. Again, 20 minutes from know, a hand you would have folded, you now bet. I'm using the best strategy based on the data gathered so far.
The more data I get, the less flopping around of what gets raised and what gets folded. I think I'll let this simulation go another 100 million hands.
There are three bets; Ante, Bet and Raise. The Bet and Raise are twice the amount of the Ante. The Bet and Raise are returned to the player if the dealer doesn't have a pair of 5s or better. The player can fold after the Ante but does not have to raise.
So the decision points are, 1) Do I bet or fold after looking at my initial 5 cards? 2) Do I raise or stand pat after seeing the first community card?
Let's start with 2. You want to Raise your bet when your hand's potential wins more than enough to cover the times that you don't get paid because the dealer doesn't qualify. It has to be a pretty strong hand. What is that hand? I don't care, because I played out 100 million hands and stored the information in a table. I'll worry about simplifying it later but right now, I can use that table as a base of knowledge for the first decision.
When to fold and when to play. You want to play when you are going to win or at least lose less than if you fold. But how do you know when you are going to lose less than if you just fold? Using the data from the first 100 million hands, I can get a good idea with what I should raise and with what I should stand pat. So I started up the simulator again and using the tables I've already built (and continue to build upon) I'm figuring out where it makes sense to fold.
I call this the psychic time traveller method since it requires knowing 11,358 (more or less) possible changing datapoints in advance and if it changes, changing the past. The data is constantly being saved and re-evaluated so it's non-deterministic. A hand that you would have raise with 20 minutes ago may have had an unfortunate run of luck and now it's not a hand with which to raise. So with that in mind, you may be betting 5 or just 3 but not always consistently, it all depends on the state of the data at that moment . Then you look at the losses and if they exceed the total number of hands (including ties because you would have folded that too), then the advice is to fold otherwise you play on. Again, 20 minutes from know, a hand you would have folded, you now bet. I'm using the best strategy based on the data gathered so far.
The more data I get, the less flopping around of what gets raised and what gets folded. I think I'll let this simulation go another 100 million hands.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
June 6th, 2011 at 8:35:38 PM
permalink
Just got back from staying at the Flamingo. My friend tried this game out. Dealer said that its likely on its way out of the casino as not many people were playing it.
June 8th, 2011 at 3:29:40 AM
permalink
I have that feeling too.Quote: Biggle10Just got back from staying at the Flamingo. My friend tried this game out. Dealer said that its likely on its way out of the casino as not many people were playing it.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
