Mike, Thank you for the opinion, it means a lot.
I would love a progressive built around it, but see it as a table spot.
The progressive does risk a brutal beating at the Hands of Murphy's Law. The first night that EZ Pai Gow went live at the Fiesta, it gave up Five Aces for $2,500. That looked bad, but was totally survive-able.
How much would the progressive need to be? Maybe it could be added to an existing progressive game? I'd be more likely to play the $1 on a progressive game paying 75/100/500/5%/10%/100% for quads, SF, RF, cracked quads, 5A, 7card SF. I don't know how unlikely cracked-quads really is, but would replacing the full-house payment with a 5% payment for cracked-quads be about right? (if it also paid if they were split)
Some Pai Gow hands would never loose: Five aces (because of the aces on top), the 7card SF, and the Royal, because they are not only so rare, they'd need a rarer hand to beat them. I don't think a 7c SF has ever lost to a higher 7c SF.
Now, split Four of a kind is actually quite weak; it forms two pairs, essentially, and if they are low, then any higher two split pair, split full house, or three-pair hand beats it.
Three pair is not a part of the bet, but might be, and would fit into a progressive:
Bad Beat Progressive: (illustration, not accurate)
Three pairs....................: 50:1
unsplit full house............: 20%
unsplit Quads or better...: 100%
Actually, three pairs might be tougher to lose, because it always has a pair up, while a straight seldom does. To be researched.
Note the unsplit full house. Players always seem to split up full houses, but if you have a full house with pairs 5's or less and an AK/AQ for the top, keep the FH together and play the AK for the top. Because this is rare, unsplit full houses are as rare as quads.
Unsplit quads loosing is rare, as it occurs every 500 hands, so to be beat by a stronger hand, it would be 500^2 or once in every 25,000 hands.
This is about as often as the AKQ of spades winning the Three-card progressive, so I feel it's spot-on.
One concern I have is with players whose 5-card side is straight or better and it loses, but their 2-card side prevents the loss, and their looking at the prize "that got away" because of the stronger two-card side. Should the progressive be based on the five-card side only?, as
Bad beat progressive: high hand (illustration, not accurate)
unsplit full house...: 20%
quads or better.....: 100%
Pretty simple and I cannot find any negative feedback about this side bet.
Did you check this side bet still available?!?!
This will work!
If I do this. I will start with trips or better beaten by better hand. just for higher pay out.
3 pairs beat or something higher could be little bit complicated....i guess.
And try to avoid non-split concept and highest possible hand vs dealers highest possible hand bad beat thing.......grrr.. this is more confusing...
Anywayz... Good Idea Dan. See u at G2E.
We got some interest today from a game distributor on it. They see it purely as a progressive, and not as another Pai Gow table game or bet; Mike was right.
This does make sense, because as a table bonus type bet, it'll have to start at a very low level (A pair of Jack's or Aces losing as a bad beat), in order to have a decent hit frequency. From there, they'll be too many bets on a cluttered layout. And besides, there are enough Pai Gow tables out there (SMI's Fortune, Galaxy's Emperor's, DEQ's EZ Pai Gow [yeaa!], Pai Gow Thrill, Pai Gow Mania, Pai Gow Bi-Polar, Pai Gow Geisha, Pai Gow Bar Girl, etc.)
Another new Pai Gow table cannot be crammed in, but a good one can displace others out, but this side bet can't do that as a new "table," - and it won't be added as a fourth or fifth side bet now........
But a good progressive can go atop ANY Pai Gow table that's out there!
It IS a bit complicated, true...REALLY true.
The bet is based on a strong hand losing, but many strong hands can be broken up into a slightly weaker but WAY better balanced hands, such as the full house 5552276, played as 22/55576. This will lose to a higher full house split, or to a straight or flush with a pair of 3's or better, not that uncommon. Quad 7's with no ace will be played as split pair 7's, and lose to 10' & 8's, just an ordinary hand. Yet if you have 55522AK, you play AK/55522, keeping the full house together (as the AK is about as strong as a pair of 2's, but the full house is considerably stronger than trips on the 5-card side!) Also, 7777A92 is best played as A9/77772, even KQ/77772! (DON'T break up a monster four of a kind if you got a decent top, unless four K's or something!) What happens quite often in PGP is that people needlessly break up monsterously strong hands to play a luke-warm two pair hand, 77/77A92.
So....the requirements for a bad beat are:
1. A Strong hand on the five-card side (a straight or better)
2. That was not expected to lose as a combined hand.
That's it. Simple, straightforward rules that anyone can grasp, with monster progressive payouts on a hit.
We CAN make it simpler, where if ONLY the 5-card side is beaten, and the hand pushes instead of loses, the bets wins, regardless of the two card side. (I covered all this in the provisional). This will simplyfy it, but greatly reduce the pay table, - really dilute it, as pushes are extremely common with these strong five-card hands, as opponents with weak hands will put up strong two-card sides, having a larger pool to draw the two-card side from. (The hand 6544322 is bet played as a straight for the push: 42/65432, though some disagree.)
With the full-hand loss rules (which is really the definition of a bad beat), we have to calc the frequencies of the two-card sides combined with the five-card sides.
Here we come up with probabilities of the two-card side win based on the two cards left over from a five-card side of a straight or better - all using up five cards on that side, so only a static two-card pool for the two-card side - easy enough.
But when I started at a pair of Aces or better losing, to develope a "high hit-frequency" bet that started at a low win level, it was MESSY! I had:
1. Average Strength of two card side for one pair (Aces), a five-card pool for the best two-card hand;
2. Average strength of the two-card side for two pairs unsplit, - a three card pool to compose the best two-card side;
3. Two pairs split, where I had to use the average strength of the lower pair.
4. Trips, where I had a four card pool to compose the two-card side.
5. unsplit full houses, and straights or better, - use the average strength of two remaining cards (including considering full houses with an EXTRA pair!);
6. Split full houses, using the average strength of a pair in the range of 2's to Aces (8's), at 89.87%, according to Mike's probabilities in his PGP appendix.
This was nuts. Mike has the speadsheets in his PGP appendix - a great help, but an awkward spreadsheet. (Charles Mousseau - HELP!)
Since both the dealer and the player would need to field five-card montser hands on their respective high sides, they would then be equal in getting either a stronger or weaker two-card side between them, but where the dealer wins copies at 0.3927% of the time. (0.5236%*.75; there are 191 2-card combinations equally between the two, minus one of the four matching sets out in the other's hand. Essentially, this is the same rate as getting two pairs in exactly four cards. I might be wrong...please check, I'm a little foggy...) So the five-card side loss rate is multiplied by the two-card side loss rate of 50.39%. (Otherwise in PGP, hands copy much more frequently, because the pool isn't two-cards, it averages close to five cards, for a 1 in 40, or 2.5% copy rate, weighted towards AK,AQ,AJ,KQ, and KJ. THOSE two-cards sides copies at about 5.3% of the time, using Wongs "Optimal Strategy for Pai Gow Poker" appendices. (I have a two-pair splitting strategy that uses adjusted tables for those two-card sides...)
Anyway, the progressive table will be based on having any monster five-card side hands where the combined hand loses.
Players will NOT "adjust their hand settings" to win the progressive, because if they did, they'd push instead of win so many winning hands, they would lose more, and the house would win more, if they tried. However, they may FINALLY play correctly the hands like A9/77773, and AK/55533, etc!
I didn't take the time to finalize prelim math until now because I hadn't gotten a bite from a distributor, and I was working on other things. But I only asked Dan D. and Russell C. at SMI about it, and they passed at this point. SMI passes on a number of winners here and there, I feel, but they got their hands full with other items and their own stuff. Earlier documents were written to describe approximate ranges; we'll zero in exactly now.
To get a perfect house edge and possibility of how player play their hand depend the progressive jackpot amount and so on...
This is only a side bet. However, it could be more than 20k calculation for GLI... :)
The most important part of this side bet is SIMPLIFY.
I am sure you know what to do with it and this is my opinion.
John and I were discuss about this just because we have to kill time while we having coffee. :D
"Adjust their hand setting" possibility is the main issue of our conversation. Even simple math will be enough to explain the player WILL set their hands correctly, but it's possible that they will get shock when they actually can get a real good payouts if they play differently depends on their main bet amount and progressive amount. It could be only $5 difference, however players will get frustrated and those frustration could be same as BlackJack's Lucky Lady happen when they don't bet on it.
Frustration could be usable as psychological way of gaming. So it's better to get involve in some way but. the psychological point of view of Pai-Gow Poker game, Player already have those frustration when they set the hands which makes the game more interesting. This keep telling me SIMPLIFY the side bet. We think the side bet should not effect the main game for any reason. To avoid the tiny possibility of changing play of actual Pai-Gow, simplify is the only answer of the matter. Which means try to set the side bet with only high hands or all hi and low hands together which is I recommend. This will reduce the actual concept of the side bet you were thinking about, but it will simplify the bet.
Such as flush or better beats by dealer's higher hand gets 8 to 1, and quad beaten by higher gets 100 to 1, straight flush gets 10% straight flush without joker gets 100% so on.. the side bet will be located beside the bouns bet.
Well... this will reduce too much juice from the actual idea which is not my point Dan.
I will share if I pop up any idea that could help you.
'Wish' you the best!
First - Good idea for the progressive. Definitely just go with the 5 card hand losing to dealer's five card hand. I think I would start being paid out when 3 of a kind gets beat, which might not happen enough for the desired frequency, BUT is fairly simple:
I'm sure this math doesn't work, but an example:
5 Card hand loses pay table
3 of a kind beat: 10:1
Straight beat: 15:1
Flush beat 25:1
Full House beat 100:1
Quads beat: 10%
Straight flush beat: 100%
Also, I played EZ Pai Gow at Ameristar St. Charles about a month ago. Saw that the black/red bets were gone, made me sad. Was down $250 and not happy. Hit a 9 high pai gow for $200, made me happier.
True, the jackpot will have to be Quads or better losing; a straight flush losing is considerably rarer. And Quads would make it "look more possible."
The big questions is "do we pay on only the five-card side losing" or do we pay if the full hand loses.
I'm begining to think now if the five-card side only loses. - what if you had a guy with a straight flush lose to a higher straight flush - and his two-card side just edges him out - for a zero dollar gain instead of the jackpot. He'd be screaming!