robert.loo
robert.loo
  • Threads: 1
  • Posts: 4
Joined: Apr 1, 2020
April 1st, 2020 at 6:01:29 PM permalink
I just learned that in Australia there is a tie bet available on their tiles table! As a seasoned player I’m very curious about the house advantage on this bet and would love to know the math behind it. The tie bet apparently pays even money with no commission taken and will pay you 25 to 1 if the low hand is at least a Wong and 20 to 1 if the low hand is a gong. On the outside the bet looks pretty amazing since ties are extremely common so I was just wondering if the bet is something that should be made all the time.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
April 1st, 2020 at 6:46:47 PM permalink
Roughly you tie about 40-42% (depending on how you play) and the chances of tying with such a good low hand is fairly small. So I worked out, based on a simple House Way and playing style it's about 86.59% (i.e. house edge over 13%).

Note that the Crown Melbourne doesn't even pay a bonus on Gongs (c.f. Table 8) which makes the House Edge 15.57% (ouch!)

i.e. Avoid.
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
April 1st, 2020 at 7:47:11 PM permalink
Quote: robert.loo

I just learned that in Australia there is a tie bet available on their tiles table! As a seasoned player I’m very curious about the house advantage on this bet and would love to know the math behind it. The tie bet apparently pays even money with no commission taken and will pay you 25 to 1 if the low hand is at least a Wong and 20 to 1 if the low hand is a gong. On the outside the bet looks pretty amazing since ties are extremely common so I was just wondering if the bet is something that should be made all the time.



Are you allowed to set your hand anyway you want? Or must you said at the official house way? If you are allowed to set it anyway you want, you can definitely increase the amount of times you will have a tie
robert.loo
robert.loo
  • Threads: 1
  • Posts: 4
Joined: Apr 1, 2020
April 2nd, 2020 at 12:52:53 AM permalink
Apparently you can set your hand however you like.
robert.loo
robert.loo
  • Threads: 1
  • Posts: 4
Joined: Apr 1, 2020
April 2nd, 2020 at 12:57:30 AM permalink
Wait is that really the house edge? regular tie pays 1 to 1 , tie with a wong or higher in the low pays 25 to 1 and 20 to 1 with a gong? yikes, I didn't think it would be that bad since in some cases you can set your hand where a tie is more probable.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
Thanked by
MrCasinoGames
April 2nd, 2020 at 3:19:40 AM permalink
Quote: robert.loo

Apparently you can set your hand however you like.

Assuming equal bets are allowed, that gives you roughly 4% back, thus the Tie Bet costs you about 10.24%. If you didn't have to make a base bet, then you can make plays purely for Tie purposes, the House Edge is about 3.35%. This assumes the 25/20/1 Bonus you mentioned.

In summary I get it that you can't overcome the House Edge with smart play and if the House only allows equal bets you need to learn a different "perfect" strategy to get to the figure above. It's not worth the bother.
Aussie
Aussie
  • Threads: 12
  • Posts: 415
Joined: Dec 29, 2009
Thanked by
robert.loo
April 2nd, 2020 at 1:47:49 PM permalink
Quote: robert.loo

Wait is that really the house edge? regular tie pays 1 to 1 , tie with a wong or higher in the low pays 25 to 1 and 20 to 1 with a gong? yikes, I didn't think it would be that bad since in some cases you can set your hand where a tie is more probable.





It’s 25-1 on a Gong or higher on the low. Minor correction but means a lower house edge.

And yes, can set however you like.

Only ever seen this bet in Crown Melbourne though.


EDIT: You CAN make this as a stand-alone bet. No need to play the normal bet.
Aussie
Aussie
  • Threads: 12
  • Posts: 415
Joined: Dec 29, 2009
April 2nd, 2020 at 1:51:44 PM permalink
Quote: charliepatrick



Note that the Crown Melbourne doesn't even pay a bonus on Gongs (c.f. Table 8) which makes the House Edge 15.57% (ouch!)

i.e. Avoid.




Unless the rules have changed in the last 8 weeks this is not correct. They definitely pay 25-1 on the tie if you have a gong or higher on the low hand.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
April 2nd, 2020 at 3:28:31 PM permalink
Quote: Aussie

Unless the rules have changed in the last 8 weeks this is not correct. They definitely pay 25-1 on the tie if you have a gong or higher on the low hand.

I can see the rules now say that, so I'm not sure what I was looking at a few days ago which suggested otherwise except I did see the same rule #8 (p26) so wasn't seeing things!

Using the rules that Gongs+ pay 25/1 the House Edge is about 0.65% less, about 2.70% if you just play the tie bet (but as I said you need to play differently - e.g. 6-6 6-5 4-4 4-4 you would not split the 8s).
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
April 2nd, 2020 at 3:28:32 PM permalink
Quote: Aussie

Unless the rules have changed in the last 8 weeks this is not correct. They definitely pay 25-1 on the tie if you have a gong or higher on the low hand.

I can see the rules now say that, so I'm not sure what I was looking at a few days ago which suggested otherwise except I did see the same rule #8 (p26) so wasn't seeing things!

Using the rules that Gongs+ pay 25/1 the House Edge is about 0.65% less, about 2.70% if you just play the tie bet (but as I said you need to play differently - e.g. 6-6 6-5 4-4 4-4 you would not split the 8s).
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
April 2nd, 2020 at 3:59:27 PM permalink
Quote: charliepatrick

I can see the rules now say that, so I'm not sure what I was looking at a few days ago which suggested otherwise except I did see the same rule #8 (p26) so wasn't seeing things!

Using the rules that Gongs+ pay 25/1 the House Edge is about 0.65% less, about 2.70% if you just play the tie bet (but as I said you need to play differently - e.g. 6-6 6-5 4-4 4-4 you would not split the 8s).



I think if you were solely playing for ties you could turn the bet +EV. (Just a guess). There are a bunch of hands that overdoing the high at the expense of the low would markedly increase ties. Play wong/3 instead of 9/5 as an example. There are many other examples.
Gee Joon/1 instead of 9/8. Etc....
robert.loo
robert.loo
  • Threads: 1
  • Posts: 4
Joined: Apr 1, 2020
April 2nd, 2020 at 5:21:58 PM permalink
Quote: charliepatrick

Assuming equal bets are allowed, that gives you roughly 4% back, thus the Tie Bet costs you about 10.24%. If you didn't have to make a base bet, then you can make plays purely for Tie purposes, the House Edge is about 3.35%. This assumes the 25/20/1 Bonus you mentioned.

In summary I get it that you can't overcome the House Edge with smart play and if the House only allows equal bets you need to learn a different "perfect" strategy to get to the figure above. It's not worth the bother.



3% ish sounds a lot better. I also came to the realization that both hands can’t win so I’m not really sure what the best strategy is in regards to the bet. Perhaps just ignoring it altogether might be better but since pushes occur so often It’s hard to resist the temptation to just use that as a primary bet.
Aussie
Aussie
  • Threads: 12
  • Posts: 415
Joined: Dec 29, 2009
April 2nd, 2020 at 6:16:12 PM permalink
This side bet isn’t played a whole lot and combined with the fact you don’t have to make the base bet to play it, that probably gives a good indication that it can’t be beaten.

In my experience Mostly people play the base bet with maybe one third also playing the bonus bet. Sometimes they also play the tie. I mainly see people playing the tie when they have run out of/are low on money as the minimum bet is typically much lower. EG $100 minimum on the base bet but only $10 or $25 on the tie.
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
April 2nd, 2020 at 6:34:34 PM permalink
Quote: Aussie

This side bet isn’t played a whole lot and combined with the fact you don’t have to make the base bet to play it, that probably gives a good indication that it can’t be beaten.

In my experience Mostly people play the base bet with maybe one third also playing the bonus bet. Sometimes they also play the tie. I mainly see people playing the tie when they have run out of/are low on money as the minimum bet is typically much lower. EG $100 minimum on the base bet but only $10 or $25 on the tie.



I'm going to play 100 hands using my 'tie' strategy. I'll report results. To remind me, if I have a gong in low hand and tie I win 20-1. If its a wong I win 25-1. Both assuming it results in a tie. I'll be back.

Just played 30 hands. Tie only 11. I gave up.
Aussie
Aussie
  • Threads: 12
  • Posts: 415
Joined: Dec 29, 2009
April 2nd, 2020 at 6:39:04 PM permalink
It’s 25-1 if a gong or higher on the low. Nothing extra if it’s a wong.
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
April 2nd, 2020 at 6:43:12 PM permalink
Quote: Aussie

It’s 25-1 if a gong or higher on the low. Nothing extra if it’s a wong.



Wong is higher than gong. If it says 'gong or higher' I'm sure wong is included.
Aussie
Aussie
  • Threads: 12
  • Posts: 415
Joined: Dec 29, 2009
April 2nd, 2020 at 6:46:26 PM permalink
Quote: SOOPOO

Wong is higher than gong. If it says 'gong or higher' I'm sure wong is included.




What I mean is that they will pay 25-1 if your low hand is a gong or higher. That covers times when your low hand is a Wong and also when the low hand may be a pair.
ksdjdj
ksdjdj
  • Threads: 94
  • Posts: 1707
Joined: Oct 20, 2013
April 2nd, 2020 at 11:57:57 PM permalink
Since there is not much to do at the moment, I decided to click the "random hand" button 300 times (see link below) just to see what it would show.

https://wizardofodds.com/games/pai-gow-tiles/calculator/
(I selected "Crown Casino" for the "dealer strategy")

Doing the above and using a "best chance to get a push" strategy, I got the following figures:

Push (average): 9895.7933

Total combos (according to the calculator): 20,475

Chance of a push: 48.33...%


Sorry, I don't understand pg tiles since I have never looked at it/played it before, so I don't know how many of those hands (if any) would have paid a bonus.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
Thanked by
ksdjdjMrCasinoGames
April 3rd, 2020 at 1:55:03 AM permalink
UPDATE
03Apr'20 19h42 BST: These figures are wrong as I had a bug in my code when splitting pairs: e.g. Hi8 Hi8 1-1 =9=. If you play there's 2200 chances of winning 25/1, if splitting there's 24208 chances of winning 1/1. However I was incorrectly using 24208 and giving the bonus.


Note I don't use Melbourne House Way but the one used in the UK so the figures might be slightly out. Here are the numbers just looking at maximizing the number of ties.

Note: I noticed I had to make a manual* adjustment as you occasionally need to split two pairs (e.g. with 6-6 6-6 4-4 4-4 you're better off splitting the pairs) as my original logic didn't look at alternatives for "TwoPairs", it did for other "OneWay" hands.

The chance of a tie this way is 46.85%.

Bonus Ties
1 216 624
25
30 415 600
Simple Ties
343 788 169
1
343 788 169
Not Tie
391 385 936
- 1
-391 385 936
736 390 729
0
-17 182 167
-2.333%

* - The above figures might be slightly out as I've just used the "Tie" figures for 6-6 1-1 {Hi8 Hi8, Lo8 Lo8, =9= =9=) instead of 6-6 6-6 ... ... or 1-1 1-1 ... ... rather than recoding the original program, so this is a very close estimate. However it still shows there's a House Edge.

Idea - PaiGow allow GeeJ to be 3 or 6, are you allowed to under-declare - for a hand like 4-1 2-1 could it be considered as a 1 rather than 8? I've assumed not!
Last edited by: charliepatrick on Apr 3, 2020
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
April 3rd, 2020 at 5:28:37 AM permalink
Quote: charliepatrick

Note I don't use Melbourne House Way but the one used in the UK so the figures might be slightly out. Here are the numbers just looking at maximizing the number of ties.

Note: I noticed I had to make a manual* adjustment as you occasionally need to split two pairs (e.g. with 6-6 6-6 4-4 4-4 you're better off splitting the pairs) as my original logic didn't look at alternatives for "TwoPairs", it did for other "OneWay" hands.

The chance of a tie this way is 46.85%.

Bonus Ties
1 216 624
25
30 415 600
Simple Ties
343 788 169
1
343 788 169
Not Tie
391 385 936
- 1
-391 385 936
736 390 729
0
-17 182 167
-2.333%

* - The above figures might be slightly out as I've just used the "Tie" figures for 6-6 1-1 {Hi8 Hi8, Lo8 Lo8, =9= =9=) instead of 6-6 6-6 ... ... or 1-1 1-1 ... ... rather than recoding the original program, so this is a very close estimate. However it still shows there's a House Edge.

Idea - PaiGow allow GeeJ to be 3 or 6, are you allowed to under-declare - for a hand like 4-1 2-1 could it be considered as a 1 rather than 8? I've assumed not!



Thanks for doing this. In my 30 Hand sample I did get one where I ‘wanted’ to count the Gee Joon as 6 to make the low hand worse. My favorite hand was teen, teen, day, 7. Played teen/day. 7/teen. It worked.
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
Thanked by
MrCasinoGames
April 3rd, 2020 at 10:24:48 AM permalink
Quote: SOOPOO

My favorite hand was teen, teen, day, 7. Played teen/day. 7/teen. It worked.

Interestingly it's also correct to split 6-6 6-6 1-1 4-4 as it's more than 28 times likely to get a Tie even though playing the hand normally would possibly get a 25/1 payout.

I might have another look at two pairs (since I assume you don't split them except for the ones I spotted earlier). However, for instance, with GJ GJ 6-6 6-6 you can't lose, so you might as well split them and hope the dealer has one hand>8 and one hand<8. I might have a relook at this sometime, although the chances of Two Pair are so low it's not likely to make much difference to the House Edge and also if you have low pairs you might win 25/1 if the dealer plays a higher pair.
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
April 3rd, 2020 at 10:41:34 AM permalink
Just played 30. Tied 17 Felt good. Lots of opportunities for silly plays to get that tie. Zero hands where I had a gong or wong as the low hand. When I play regularly I sometimes use information on tiles out to help.... (if i know there are no teens available for the dealer... well, you understand...) I'm guessing the value of the extra knowledge would be less in trying to get a tie...?
charliepatrick
charliepatrick
  • Threads: 39
  • Posts: 2946
Joined: Jun 17, 2011
Thanked by
robert.looMrCasinoGames
April 3rd, 2020 at 12:00:44 PM permalink
I have now rerun the figures looking at alternatives of spltting Two Pairs. It seems you don't split any Two Pair that is equal or lower than Hi8 Lo7. Higher hands are sometimes split, and sometimes not.

I have also changed the method of working out bonus hands so it considers the Low Hand for each way to play the hand.

(I had a bug where, e.g. Hi8 Hi8 1-1 =9=, it realised you were more likely to get a Tie by splitting, but it also then applied a bonus since it saw the Low Hand was Wong. As it happens you are best not splitting the pair and hoping to get a 25/1 payout.)

The chances of getting a Tie is now 46.81%.

The House Edge is 5.187%.

Bonus Ties
368 285
25
9 207 125
Simple Ties
344 256 376
1
344 256 376
Not Tie
391 656 339
- 1
-391 656 339
736 281 000
0
-38 192 838
-5.187%
  • Jump to: