As far as I know, before now only three people have created automated video poker strategy makers:
1. TomSki -- creator of Video Poker Strategy Master software.
2. ? -- Whoever helped with Frugal Video Poker.
3. ? -- Whoever helped with Video Poker for Winners.
Usually video poker strategies have to be created by hand with a lot of trial and error. It is not easy getting a computer to concisely create a strategy based on the 2.6 million possible starting hands. It is like the Mount Everest of video poker programming. So, my congratulations to JB for this accomplishment.
So, video poker players, please give it a test drive and let us know what you think.
I've used the strategy guides at VPgenius, but JB's calculator looks more thorough.
I made a (much simpler and less accurate) strategy calculator for one project, but mine was useful to me because I was making hundreds of paytable adjustments to find a decent strategy for a very complex game.
Quote: AyecarumbaVery cool. I assume JB considered defaulting the paytable to single coin instead of 5 coins. I can understand why 5 coins was set as the default, but it is unusual for me to see and adjust the table X5, especially when the machines are typically referred to with their single coin payouts (e.g., "9/6 JoB"). At the end of the day, it is better for it to default to 5 coin.
The ones I tried were defaulted to 5 coins.
However, the strategies created by vpgenius.com are not entirely accurate: they are created by identifying the types of plays made, and then sorting the plays by their expected value. This is a fast method of creating a strategy, but not entirely accurate (especially with lower-valued plays).
The strategies generated by the new calculator on WizardOfOdds.com use the correct (albeit slower) method, which is to compare each line against all other lines to figure out where it belongs in the list. It is correct for some lines to appear higher on the list than others even though they have a lower EV. The classic example is King vs. Suited King-Ten in 9/6 Jacks or Better: a King by itself has a higher EV than Suited King-Ten, yet Suited King-Ten is typically the better play. The reason why the EV for a King is higher is because its EV is derived only from hands where the King by itself was the better play.
The correct method to compare two plays (such as King vs. Suited King-Ten) is to only look at hands where the best play is one of the two plays being compared; only then can you see that the EV for a King is lower than the EV for Suited King-Ten among hands where the best play is one of those two plays. However, even after you've done that, there is still a choice to make when assembling the strategy list: do you arrange the plays in the way that yields the highest overall return (when exceptions are ignored), or do you arrange them in the way that produces the fewest exceptions? For this calculator, I chose to arrange them in the way that produces the highest return (when exceptions are ignored).
I did not know this. Major props. One of my favorite gambling sites on the web, besides the Wizard's. Very happy to hear you were behind it.Quote: JBI'm not sure if many people know this or not, but I am the person who created vpgenius.com. (In fact, that's how the Wizard found me.) I sold it shortly after I finished it though, and am no longer associated with it.
Oh! Awesome. Thanks for your work, sir; I've found it very useful, and I appreciate it greatly.Quote: JBI'm not sure if many people know this or not, but I am the person who created vpgenius.com. (In fact, that's how the Wizard found me.) I sold it shortly after I finished it though, and am no longer associated with it.
Paytable
Natural Royal Flush 250 4000 (1 coin and 5 coin respectively)
Four Deuces 500 2500
Wild Royal Flush 25 125
Five of a Kind 15 75
Straight Flush 8 40
Four of a Kind 4 20
Full House 3 15
Flush 2 10
Straight 2 10
Three of a Kind 1 5
Using the new Video Poker Strategy Calculator on WizardofOdds.com, I put in the 5 coin amount and the Total Return amount was 5.007, which I don't think is correct. So I put in the 1 coin in value and the Total Return amount was .9895.
Therefore, my question is, is this game really a 100% payback game?
Quote: SONBP2Using the new Video Poker Strategy Calculator on WizardofOdds.com, I put in the 5 coin amount and the Total Return amount was 5.007, which I don't think is correct. So I put in the 1 coin in value and the Total Return amount was .9895.
Therefore, my question is, is this game really a 100% payback game?
Yes. You didn't use the correct prize for the royal flush (4000/5 = 800, not 250, even though the single-coin prize is 250).
With the correct paytable, perfect play returns 100.1496% and the basic strategy returns 100.1438%. For what it's worth, this is the default paytable that comes up for Loose Deuces in the calculator.
This would be easy to fix - when an analysis is complete, redirect to a new URL specific to that paytable. Something like https://wizardofodds.com/games/video-poker/strategy/calculator?game=ddb&rf=1240&fl=5 (encoding all non-default payouts in the URL), or even https://wizardofodds.com/games/video-poker/strategy/calculator?game=384592 (referring to an internal database ID).
This would also allow users to bookmark games they've analyzed for future reference.
Thanks again for the excellent tool.
3 to a Royal Flush TJA; TJQ; TJK; TQA; TQK; TKA; JQA; JQK; JKA; QKA
2 Pair 2233; 2244; 2255; 2266; 2277; 2288; 2299; 22TT; 22JJ; 22QQ; 22KK; 3344...
1 Pair JJ; QQ; KK
But every single conflict between RF3 (AHT) and two pair (no AA) is listed as an exception to the basic strategy:
Right Play Wrong Play Pattern(s)
2 Pair: TTJJ 3 to a Royal Flush: TJA (more)10♣ 10♦ J♣ J♦ A♣
2 Pair: TTQQ 3 to a Royal Flush: TQA (more)10♣ 10♦ Q♣ Q♦ A♣
2 Pair: TTKK 3 to a Royal Flush: TKA (more)10♣ 10♦ K♣ K♦ A♣
You can increase the accuracy of the basic strategy and reduce the number of exceptions by separating out the weaker RF3's:
3 to a Royal Flush TJQ; TJK; TQK; JQA; JQK; JKA; QKA
2 Pair 2233; 2244; 2255; 2266; 2277; 2288; 2299; 22TT; 22JJ; 22QQ; 22KK; 3344...
3 to a Royal Flush TJA; TQA; TKA
1 Pair JJ; QQ; KK
This is what my desktop software does, and I was surprised that JB's calculator didn't do the same. Intentional, or bug?
Pat Hand Four 5's through K's; Four 2's, 3's, 4's with any A,2,3,4; Four Aces with any 2,3,4; Straight Flush; Royal Flush
4 of a Kind 2222; 3333; 4444; 5555; 6666; 7777; 8888; 9999; TTTT; JJJJ; QQQQ; KKKK; AAAA
Pat Hand Four 2's, 3's, 4's
Obviously that last line is superfluous. (The output for regular DDB doesn't do this.) A little further down:
3 of a Kind + 1 Kicker
5552; 5553; 5554; 555A; 6662; 6663; 6664; 6665; 666A; 7772; 7773; 7774;
7775; 7776; 777A; 8882; 8883; 8885; 8886; 8887; 888A; 9992; 9993; 9994;
9995; 9996; 9997; 9998; 999A; TTT2; TTT3; TTT4; TTT6; TTT7; TTT8; TTT9;
TTTA
With the following exceptions listed:
Right Play Wrong Play Pattern(s)
3 of a Kind + 1 Kicker: 6662 3 of a Kind + 1 Kicker: 6663 (less)6♣ 6♦ 6♥ 2♣ 3♣
6♣ 6♦ 6♥ 2♣ 3♦
6♦ 6♥ 6♣ 2♣ 3♠
6♥ 6♣ 6♦ 2♠ 3♠
3 of a Kind + 1 Kicker: 6664 3 of a Kind + 1 Kicker: 6663 6♦ 6♥ 6♣ 3♠ 4♣
3 of a Kind + 1 Kicker: 6665 3 of a Kind + 1 Kicker: 6663 6♦ 6♥ 6♣ 3♠ 5♣
3 of a Kind + 1 Kicker: 666A 3 of a Kind + 1 Kicker: 6663 6♦ 6♥ 6♣ 3♠ A♣
Much simpler to put "3 of a Kind + 1 Kicker: 6663" on a separate line of the basic strategy. Also, while it makes no difference, odd that the calculator was able to collapse suit symmetries when displaying example exceptions for 6664, 6665, 666A but not 6662.
(JB, should I be posting this minor stuff in the forum thread, or should I email or PM you?)
$1 Progressive Double Double Bonus:
Royal Flush Jackpot was $4500 so I entered 900 into the calculator
Straight Flush 50
Four Aces with any 2,3,4 Jackpot was $2025 so I entered 405 into the calculator
Four 2's, 3's, 4's with any A,2,3,4 160
Four Aces 160
Four 2's, 3's, 4's 80
Four 5's through K's 50
Full House 8
Flush 5
Straight 4
Three of a Kind 3
Two Pair 1
Jacks or Better 1
Maybe the calculator is not designed to do this.
Quote: SONBP2Can the Video Poker Strategy Calculator be used to calculate progressives? I was hoping to use the calculator to determine, based on the pay table, the house edge for vp progressives. I would like to know at what point the game's return is over 100%. I entered the following paytable but it keeps returning an error message.
$1 Progressive Double Double Bonus:
Royal Flush Jackpot was $4500 so I entered 900 into the calculator
Straight Flush 50
Four Aces with any 2,3,4 Jackpot was $2025 so I entered 405 into the calculator
Four 2's, 3's, 4's with any A,2,3,4 160
Four Aces 160
Four 2's, 3's, 4's 80
Four 5's through K's 50
Full House 8
Flush 5
Straight 4
Three of a Kind 3
Two Pair 1
Jacks or Better 1
Maybe the calculator is not designed to do this.
I got it to work. The payback according to the pay table given is 97.07%. The progressives would need to be very large to reach 100%.
Quote: fivespotMinor issues with the output analyzing QQ DDB. Top of the chart:
Pat Hand Four 5's through K's; Four 2's, 3's, 4's with any A,2,3,4; Four Aces with any 2,3,4; Straight Flush; Royal Flush
4 of a Kind 2222; 3333; 4444; 5555; 6666; 7777; 8888; 9999; TTTT; JJJJ; QQQQ; KKKK; AAAA
Pat Hand Four 2's, 3's, 4's
Obviously that last line is superfluous. (The output for regular DDB doesn't do this.)
The reason for this is that, the "4 of a Kind 2222; 3333; etc." line refers to natural 4 of a kind, and the "Pat Hand Four 2s, 3s, 4s" refers to quick quad hands such as 222AA, 3332A, 44422, 4443A. One type uses 4 cards, the other uses 5 cards, and this fact basically "requires" that they be listed separately, because the program will not group together plays of different types -- and internally, these two plays are completely different even though they pay the same in the front end.
Quote: fivespotA little further down:
3 of a Kind + 1 Kicker
5552; 5553; 5554; 555A; 6662; 6663; 6664; 6665; 666A; 7772; 7773; 7774;
7775; 7776; 777A; 8882; 8883; 8885; 8886; 8887; 888A; 9992; 9993; 9994;
9995; 9996; 9997; 9998; 999A; TTT2; TTT3; TTT4; TTT6; TTT7; TTT8; TTT9;
TTTA
With the following exceptions listed:
Right Play Wrong Play Pattern(s)
3 of a Kind + 1 Kicker: 6662 3 of a Kind + 1 Kicker: 6663 (less)6♣ 6♦ 6♥ 2♣ 3♣
6♣ 6♦ 6♥ 2♣ 3♦
6♦ 6♥ 6♣ 2♣ 3♠
6♥ 6♣ 6♦ 2♠ 3♠
3 of a Kind + 1 Kicker: 6664 3 of a Kind + 1 Kicker: 6663 6♦ 6♥ 6♣ 3♠ 4♣
3 of a Kind + 1 Kicker: 6665 3 of a Kind + 1 Kicker: 6663 6♦ 6♥ 6♣ 3♠ 5♣
3 of a Kind + 1 Kicker: 666A 3 of a Kind + 1 Kicker: 6663 6♦ 6♥ 6♣ 3♠ A♣
Much simpler to put "3 of a Kind + 1 Kicker: 6663" on a separate line of the basic strategy.
I'm not sure why it didn't.
Quote: fivespotAlso, while it makes no difference, odd that the calculator was able to collapse suit symmetries when displaying example exceptions for 6664, 6665, 666A but not 6662.
That's not an error. There is only one unique pattern where the (666A|6664|6665) vs. 6663 exception applies, but there are 4 unique patterns where the 6662 vs. 6663 exception applies.
Quote: fivespot(JB, should I be posting this minor stuff in the forum thread, or should I email or PM you?)
Here is fine. Thank you for the PM alerting me to it though, as I hadn't checked this thread in a while.
I play Deuces Wild exclusively and almost exclusively play it at the Tampa Hard Rock (not the best place for good returns but there is no real competition in the area- which is a topic for another post). Basically, I chose Video poker because of what I read about it regarding returns from the Wizard at wizardofodds.com I have enjoyed it quite a bit and have also profited from it somewhat. I started off by using the optimal strategy for Deuces Wild (DW) on the wizardofodds.com website. Soon I realized that the paytables at the Tampa Hard Rock are nowhere near Full Pay and I needed another source for optimal DW strategy. This is how I found VPgenius.com, which is one of the most amazing websites on the planet for VP players. Over the last year or so I have used vpgenius.com a lot in trying to determine optimal play for the three or four different DW paytables available at the Tampa Hard Rock. I have used the training feature also to help me memorize the optimal strategy and to learn about penalty cards. This brings me to my questions, which may be on many other people's mind:
1) The vpgenius site seems to be down in that it shows a runtime error "Server Error in "/" Application" every time you click on a game to get it's strategy. This has happened before, but the website has always recovered before. However, this time, the site has been down for an extended period of time and shows no sign of coming back up. I can't contact anyone since all of their contact links don't work correctly either. Any idea on what's going on? I know that you are no longer associated with this website but you may have some connections and would be able to find out when/if this website will be running correctly in the future? Also, you created the website so maybe you can shed some light on this error that is showing up.
2) I understand that the vpgenius.com strategies are not entirely accurate and the one you made for wizardofodds.com are better (which kind of worried me a little since vpgenius.com is what I have been following for a while). I had noticed in the past a discrepancy between the optimal strategy laid out by vpgenius.com compared to the optimal strategy laid out by wizardofodds for full play DW (in the lower value plays) and I was fixing to "ask the wizard" regarding this, but now I know the answer now after reading your post. Also, I understood that vpgenius.com optimal strategy did not take into account penalty cards. Having said that, I feel that the interface that you created for vpgenius.com just looks better and, in some ways, is easier to follow. Plus, the training feauture is awesome and allows you to learn about penalty cards and hone your skills at a specific pay table. The wizard of odds strategy guide that you created is nice too, but it is much harder to learn about the lower value hands and penalty cards because you have to go through the extensive exception list to learn (at least for DW it's extensive). So my second question is if you are planning to create a similar interface to vpgenius.com for wizardofodds.com
3) Speaking of exceptions and penalty cards, I know from reading what the Wizard wrote that it's a painstaking process to go through the exceptions and penalty cards to derive a perfect optimal stragegy for DW (or any VP game for that matter). In fact, the wizard had provided us with a great optimal strategy for Full Play DW that takes into account exceptions and penalty cards. However, full play DW is mostly a thing of the past and most VP palyers do not have access to this game. Here in Tampa it is nonexistent. The wizards full play DW strategy is a great place to start learning about DW but it is not practical because most players out there are not playing this paytable. Even "not so ugly ducks" DW is often not available. Here at Tampa the best you can get is "Ugly Ducks" DW and lesser paytables. As you know, these lesser pay tables do have differences in optimal stategy compared to full pay that are important to recognize (ex. flush and two pair are higher value in may pay tables). So my question is if it is possible to have optimal strategy, including exceptions and penalty cards, available to us on the wizardofodds website for pay tables other than Full Pay and "not so ugly ducks" (maybe the wizard can answer this)
4) Could you or the wizard recommend for us a website or program for free or for money that is closest to the interface of vpgenius, with the training tools, but has the more accurate strategy calculator available on wizardofodds? Also, could you recommed the best VP iphone app?
Hopefully, you made it this far and hope to hear from you soon. Once again thank you JB and the Wizard for your incredible work.
Quote: SONBP2Can the Video Poker Strategy Calculator be used to calculate progressives? I was hoping to use the calculator to determine, based on the pay table, the house edge for vp progressives. I would like to know at what point the game's return is over 100%.
The problem with trying to calculate when a progressive would be > 100% is, you need to know the "per hand" rate of increase of the progressive jackpots - and even then, the calculations would probably have to be under the assumption that the strategy wouldn't change. (For example, if the progressive is more than 2425x the total bet on, say, X/6 JoB, or 2418x on X/5, then with a king-high SF, you discard the nine to try for the royal.)
100.5400% is the perfect return for a $5814 royal.
100.5393% is the basic strategy return for a $5814 royal.
https://wizardofodds.com/games/video-poker/strategy/a-1-b-74-c-1-d-0-d-1-d-2-d-3-d-4-d-6-d-9-d-25-d-50-d-1162.8/
Now the strategy linked above is different than the strategy for non-progressive 9/6 Jacks or Better. All 3 to a Royal draws are held over a high pair for example.