DP & Point bets

I stumbled across a dead simple betting strategy which I've found very little online about. It's a slight variation on the Doey-Don't (equal bets on the pass and don't pass lines) but, unlike that strategy, it can actually win money with very little risk.

Simply put, you bet the don't pass on the come out. If a point is established, you place an equal bet on the point. Adjusting slightly for the 6/8 or for buying the 4/10. You break even on the 7, and win a small profit if the point is hit.

I'm not sure I calculated the odds correctly, but I'm showing a house edge of just under 3.24% but could be much closer to the general EV for a DP line bet of 1.36%

Suppose we play the dark side for $100, we win on a 2 or 3, lose on 7, or 11, and push on 12. Easy peasy. But when a point is hit, that's where my math may be a bit off. I'm preparing for a cruise where the vig is collected up front on buy bets, so with this example, I'd buy the 4/10 for $105. I'd place the 6/8 for $102 to closely match the line bet. A matching bet of $100 makes sense for the 5/9. So in the following table I multiply the return if the point is hit by the number of combinations minus the loss from the line bet. Then do the reverse if the 7 hits (always 6 combinations) then divide the return by the total number of consequential outcomes for that point.

Point	Bet	Point - Line	Combos	PSO - Pt (✕ 6)	EV
4/10	105	100	3	-30	30
5/9	100	40	4	0	16
6/8	102	19	5	-12	7.54545

So if I plug in the expected values to include rolls which are resolved on the come out, this is what I get:

Roll	Return	Combos	EV
2	100	1	100
3	100	2	200
4	30	3	90
5	16	4	64
6	7.54545	5	37.72727
7	-100	6	-600
8	7.54545	5	37.72727
9	16	4	64
10	30	3	90
11	-100	2	-200
12	0	1	0
		TOTAL:	-116.54545

I believe these calculations themselves are accurate. I'm but not sure what unit to compare the return of ~$116.55 against the $3600 bet on the Don't Pass line for each of 36 possible rolls, or should I include the amounts wagered when a point is established? Which would make the house edge significantly lower.

In any case, this strategy is all about earning comps (e.g. free cruises) quickly over winning or losing much money. There's certainly a risk that the pit boss would notice your strategy and zero out your rating. It isn't my original strategy, but I found on an old youtube video by someone who claims to have been successful with it on a bubble craps machine.

Assuming that you are not rated on odds bets, then the simplest system with the best EV is just betting the don't on every roll. The house edge is 1.36%. Any system that uses any other bet will have worse EV and will be more complicated.

Are you really expecting more than 1.36% in comps?

Quote: ciabelle

I stumbled across a dead simple betting strategy which I've found very little online about. It's a slight variation on the Doey-Don't (equal bets on the pass and don't pass lines) but, unlike that strategy, it can actually win money with very little risk.

Simply put, you bet the don't pass on the come out. If a point is established, you place an equal bet on the point. Adjusting slightly for the 6/8 or for buying the 4/10. You break even on the 7, and win a small profit if the point is hit.

I'm not sure I calculated the odds correctly, but I'm showing a house edge of just under 3.24% but could be much closer to the general EV for a DP line bet of 1.36%

Suppose we play the dark side for $100, we win on a 2 or 3, lose on 7, or 11, and push on 12. Easy peasy. But when a point is hit, that's where my math may be a bit off. I'm preparing for a cruise where the vig is collected up front on buy bets, so with this example, I'd buy the 4/10 for $105. I'd place the 6/8 for $102 to closely match the line bet. A matching bet of $100 makes sense for the 5/9. So in the following table I multiply the return if the point is hit by the number of combinations minus the loss from the line bet. Then do the reverse if the 7 hits (always 6 combinations) then divide the return by the total number of consequential outcomes for that point.

I believe these calculations themselves are accurate. I'm but not sure what unit to compare the return of ~$116.55 against the $3600 bet on the Don't Pass line for each of 36 possible rolls, or should I include the amounts wagered when a point is established? Which would make the house edge significantly lower.

In any case, this strategy is all about earning comps (e.g. free cruises) quickly over winning or losing much money. There's certainly a risk that the pit boss would notice your strategy and zero out your rating. It isn't my original strategy, but I found on an old youtube video by someone who claims to have been successful with it on a bubble craps machine.
link to original post

(Tables Removed, Space Save)

What do you mean when you ask about, "House Edge?" Each individual bet has a House Edge, so I assume that what you want to know is the expected loss on your total action.

In order to determine that, the first (and easiest) thing we have to do is determine your average total bet. Using the bet amounts you have provided, here is our total bet for each outcome:

2, 3, 7, 11, 12: $100---> (100 * 12/36) = 33.3333333333

4, 10: $205---> (205 * 6/36) = 34.1666666667

5, 9: $200---> (200 * 8/36) = 44.4444444444

6, 8: $202---> (202 * 10/36) = 56.1111111111

Think of this like an expected bet amount, in total. Assuming we are going to behave this way every single Come Out, we can sum this up and determine our average expected total bet:

56.1111111111 + 44.4444444444 + 34.1666666667 + 33.3333333333 = 168.055555556

If we made it a binary by which we would bet $100 total 1/3rd of the time and $200 total 2/3rds of the time, then that would be an average of $166.67 (rounded) average bet, so this obviously checks out because of the little extra that is bet in a few instances.

EXPECTED VALUE OF INDIVIDUAL RESULTS

With that, we have to determine the EV of all of the individual possible results. After we have done that, we will be able to determine the total expected loss and compare it to our expected average bet amount.

The first thing that happens is that we lose $100 on rolls of 7/11, win $100 on rolls of 2/3 and Push on a CO of 12:

(1/36 * 0) + (100 * 3/36) - (100 * 8/36) = -13.8888888889

In other words, that is our Expected Loss as we isolate the Come Out roll.

Naturally, your betting system is very goal-oriented and its goal is to get a fair amount of action in whilst trying to reduce Variance. Otherwise, it would be an objectively terrible system because the entire thing forfeits the advantage on the DP bet you end up with on those occasions that the roll is a point number.

With that, let us look at the possible results for varying point numbers:

4 & 10

The first thing that we note is you establish a point of one or the other with a frequency of 6/36.

A result of seven results in a win for the DP, which is +$100, but then you lose the $100 from the 4/10 as well as the $5 commission because the commission is paid upon making the bet.

In the event that the 4/10 comes in, then you have lost $100 from the DP bet and have also lost the $5 for the commission, which I will treat as being part of the DP loss for simplicity. However, you have won $200 on the Buy bet, which essentially functions like an Odds Bet.

(3/9 * 95) - (6/9 * 5) = 28.3333333333

That is the expected result when we isolate that combination of bets, which we must now multiply by the probability of being able to make that combination of bets in the first place:

28.3333333333 * 6/36 = 4.72222222222

5 & 9

The first thing that we note is you establish a point of one or the other with a frequency of 8/36.

A result of seven is a win for the DP, which causes us to break even because we lose the $100 we have bet on the five or nine.

In the case of the 5/9 coming in, then you would get paid $140 for a profit of $40 because of the $100 DP loss.

(4/10 * 40) - (6/10 * 0) = 16

Once again, we have to have the opportunity to do this in the first place:

(8/36 * 16) = 3.55555555556

6 & 8

The first thing that we note is you establish a point of one or the other with a frequency of 10/36.

A result of seven is a win for the DP, which causes us to lose $2 because we are betting $102 on the Place 6 or 8.

A result of six or eight is a win of $119, but is a profit of $19 because we will have lost $100 on the DP.

(5/11 * 19) - (6/11 * 2) = 7.54545454545

7.54545454545 * 10/36 = 2.09595959596

THE RESULT

With that, we simply need to add all of our results with an expected profit to our expected loss on the CO roll:

-13.8888888889 + 4.72222222222 + 3.55555555556 + 2.09595959596 = -3.51515151516

That number reflects our expected loss on the overall proposition if we do it exactly as you suggest you will. We can compare this expected loss to our average expected bet and will arrive at an expected average House Edge for our total action:

3.51515151516/168.055555556 = 0.02091660405

In other words, you are expected to lose slightly more than 2% of all monies bet.

ANOTHER WAY

Another way we can confirm this is to simply multiply each bet amount, and the likelihood of our betting it, against the House Edge of each particular bet.

For example, we make the DP bet 100% of the time, so we would have:

100 * .0136 = 1.36

We are going to bet that 4/10, essentially for $105, 6/36 of the time. We also only care about bet resolved since we are playing these out to resolution:

105 * .0476 * 6/36 = 0.833

We are going to bet that 5/9 for $100 8/36 of the time.

100 * .04 * 8/36 = 0.88888888888

Finally, we are going to bet that 6/8 for $102 10/36 of the time.

102 * .0152 * 10/36 = 0.43066666666

Basically, this accounts for the expected loss of each of the individual bets and also factors in the probability that we make any of these bets in the first place, except for the Don't Pass bet, because we always make that. If we add these expected losses together, we get:

1.36 + .833 + .888888888888 + .430666666666 = 3.51255555555

This is slightly off from what we have above, but errors are due to rounding. It's within a fraction of a penny. If nothing else, what I did above is going to be substantially more accurate because, for this part, I was using House Edges that had already been rounded off.

The point is that the expected loss essentially matches also doing it this way and we would expect our average total amount bet to remain the same.

With that, I am confident to say that you expect to lose roughly 2.09166% of all monies bet.

TUTTIGYM DISCLAIMER: Just for Tuttigym, I would like to point out that it is highly unlikely that you will attempt this a sufficient number of times to end up with an actual loss of 2.09166% of all monies bet because that would require attempting what you are trying to do more times than you will actually live long enough to do. Of course, since you inquired about House Edge, then I can only assume that you also believe House Edge has both meaning and significance which would, in theory, mean that Tuttigym would not have any serious objections to what is taking place in this thread. Of course, math, in general, seems to sometimes offend Tuttigym (judging from his posts), so this disclaimer is for his benefit and to let him know that I am preemptively acknowledging that the above calculations likely cannot be physically proven by one individual person.

Quote: Mission146

With that, I am confident to say that you expect to lose roughly 2.09166% of all monies bet.
link to original post

Nice workup, Mission.

What is the point of hedging if it costs you 2.09-1.36 = 0.73% in house edge? A player is much more likely to blow out a decent sized bankroll hedging at a 2.09% cost compared to accepting slightly higher variance of just betting the dark side at a 1.36% cost. After all, the variance on even money bet is miniscule compared to most forms of slot play or video poker.

I doubt that the player can recoup 1.36% in comps, so why even do the play in the first place except to gamble it up?

Quote: Mental

Assuming that you are not rated on odds bets, then the simplest system with the best EV is just betting the don't on every roll. The house edge is 1.36%. Any system that uses any other bet will have worse EV and will be more complicated.

Are you really expecting more than 1.36% in comps?
link to original post

I don't know what he's expecting in comps, but I've certainly heard of more unusual things. For a brief time, I was aware of a Bubble Craps machine upon which you could play profitably right off the top on points multiplier days, and that was just on the value of the points, so any mail was extra.

In any event, I agree with you that this House Edge, which I believe to be about 2.09166% of total expected action (per CO roll) is significantly worse than the House Edge just sitting there and making DP bets.

I'm also not convinced that the variance is necessarily reduced that much. The only result that comes anywhere close to covering a Come Out Roll loss on one of our Don't Pass bets is the 4/10 being established and made, so it seems to me we are just trying to nickel and dime back our expected negative outcomes on the CO roll.

With that, I'm not sure how effectively this system even accomplishes its intended purpose. It seems like significantly more than expected (or maybe even just the expected amount) of Come Out roll losses are just as bad here as they'd be if we were only betting DP. Certainly, the HE relative to our total action is worse.

As the OP mentioned, there's also the possibility of the Craps Supervisor or Pit noticing what OP is doing and giving an average bet of zero, though perhaps as likely, just an average bet of $100 for the CO DP bets. I think what OP is doing and why should be fairly obvious; I just don't know why the house should have any reason to care.

I have not seen a positive EV craps play in a B&M casino for 10 years, but then I do not get out much. I am sure they are still around.

At this moment, I am playing a slot machine online right now that has a lower house edge than the OP's craps system. I am playing through $15K to earn a $1000 bonus. Clearly, it is possible to get large cashback percentages while playing good RTP games. But, I find that craps is explicitly excluded from almost every promotion from every online casino. The only craps I have played recently is at one casino where craps is not explicitly excluded from their regular 15% loss rebates. My craps sessions usually involve making and resolving a single point. Casinos slot executives seem really cautious about giving too much to craps players after having been burned too many times.

As for variance, I have been on a losing streak lately. I was stuck $13K in the first hour of play today and getting a bit grumpy. (No doubt exacerbated by jet lag coming back from Europe.) After another three hours of play, I am up $27K. Accepting variance is my superpower.

Quote: Mental

Quote: Mission146

With that, I am confident to say that you expect to lose roughly 2.09166% of all monies bet.
link to original post

Nice workup, Mission.

What is the point of hedging if it costs you 2.09-1.36 = 0.73% in house edge? A player is much more likely to blow out a decent sized bankroll hedging at a 2.09% cost compared to accepting slightly higher variance of just betting the dark side at a 1.36% cost. After all, the variance on even money bet is miniscule compared to most forms of slot play or video poker.

I doubt that the player can recoup 1.36% in comps, so why even do the play in the first place except to gamble it up?
link to original post

Thank you for the compliment!

Personally, I don't think there's any point in making this particular hedge. If you somehow thought that you would get more than an expected loss of 2.09% of all monies bet in comps, then sure, but I'd still prefer to only expect to lose 1.36%.

Additionally, if I wanted to come up with a system at all, then I'd come up with one that also incorporated some Odds on that DP bet. My logic goes:

Because Odds have a House Edge of 0%, and I will have to make less on the applicable Place/Buy bets than I am making on my Odds bets, I will have a somewhat lower expected loss relative to total action.

For example, the OP's 4 & 10 was $100 with paying a $5 commission, as we have already established:

Quote:

The first thing that we note is you establish a point of one or the other with a frequency of 6/36.

A result of seven results in a win for the DP, which is +$100, but then you lose the $100 from the 4/10 as well as the $5 commission because the commission is paid upon making the bet.

In the event that the 4/10 comes in, then you have lost $100 from the DP bet and have also lost the $5 for the commission, which I will treat as being part of the DP loss for simplicity. However, you have won $200 on the Buy bet, which essentially functions like an Odds Bet.

(3/9 * 95) - (6/9 * 5) = 28.3333333333

That is the expected result when we isolate that combination of bets, which we must now multiply by the probability of being able to make that combination of bets in the first place:

28.3333333333 * 6/36 = 4.72222222222

However, what if the table is 3/4/5x Odds? With that, I can make the $100 DP bet, $300 in Odds bets, and a $200 Buy 4/10 bet whilst paying $10 commission:

In other words, we have:

$100 DP
$300 Odds
$200 Buy
$10 Sunk Cost Buy Commission

If the number comes out a four, then I will lose $410, but I will make $400 on the Buy bet, therefore losing $10. If the number comes seven, then I lose the sunk costs of $210, but my DP makes $100 and I also make $150 on my DP Odds bet for a net profit of $40 on this.

(40 * 6/9) - (10 * 3/9) = 23.3333333333

23.3333333333 * 6/36 = 3.88888888888

We have a lower expected profit on this precise outcome, but we really have no reason to care about that if our goal is just to get action out there. For that reason, we can look at the total expected loss relative to our total amount bet each time we do this. We will ignore the initial DP bet for $100 because we make those anyway.

OP QUESTION:

105 * .0476 = 4.998

MISSION SUBMISSION:

300 * 0.000 = 0

210 * .0476 = 9.996

Isolating just what we do after a four or ten is established, the OP's question suggests a bet of an additional $105 total with expected loss of $4.998. My alternative to this is to actually bet an extra $510, with $210 being the same bet that the OP is making, however the expected loss on my total action is twice what the OP's is. However, most of my action is on odds, so when we look at the average House Edge of my total action:

9.996/510 = 0.0196

In other words, the average House Edge relative to my total action is 1.96%, which is still bad compared to just making a DP bet, but that's because the Buy 4/10 with paying commission on bet is such a terrible bet in the first place.

Also, to be honest, I could see where you would fade the variance of sixes and eights (the House Edge of a Place Bet on these isn't that much worse and can probably be made the same---or less---for average House Edge relative to total action if made in conjunction with Odds bets on the DP) and because those are 5/11 to come as opposed to 6/11 for the DP to win. I just don't understand why we would want to make 4%+ HE bets (especially without making additional Odds bets to trim the average HE on the additional action down) when our DP bet is at a HUGE advantage in the first place.

Not to be rude, but I'd have to say that if the variance of wanting a seven vs. a 4, 5, 9 or 10 is too much to sweat, then this probably isn't a worthy pursuit in any case.

Quote: Mental

I have not seen a positive EV craps play in a B&M casino for 10 years, but then I do not get out much. I am sure they are still around.

At this moment, I am playing a slot machine online right now that has a lower house edge than the OP's craps system. I am playing through $15K to earn a $1000 bonus. Clearly, it is possible to get large cashback percentages while playing good RTP games. But, I find that craps is explicitly excluded from almost every promotion from every online casino. The only craps I have played recently is at one casino where craps is not explicitly excluded from their regular 15% loss rebates. My craps sessions usually involve making and resolving a single point. Casinos slot executives seem really cautious about giving too much to craps players after having been burned too many times.

As for variance, I have been on a losing streak lately. I was stuck $13K in the first hour of play today and getting a bit grumpy. (No doubt exacerbated by jet lag coming back from Europe.) After another three hours of play, I am up $27K. Accepting variance is my superpower.
link to original post

You may not have.

Basically, the points converted to Free Play at a rate of 2% of total action becoming Free Play; it is clear to see why that is good with Bubble Craps.

Additionally, the machine counted any Odds bets towards this, so what you could get on Odds was just +2% (roughly) of whatever you were betting. I say roughly because you obviously have to run the FP through on something that has a House Edge itself, in most cases.

Also, everything about their system treated Bubble Craps the same way it would a slot machine, with only the exception that points converted to FP at 50% rather than the full amount, but that didn't matter, because they had 98%+ VP anyway.

It was pretty much the perfect storm. That's also as I remember it; this was fixed a long time ago and is probably five, or more, years ago...so my memory might be imperfect.

Yeah, online casinos are terrified of Craps. That's a holdover just from when their promotions were ridiculously good for the player. I guess some promotions are obviously still +EV, but the promotions of old, as I'm sure you actually had the opportunity to play, were ridiculously +EV for players.

Quote: Mission146

Quote: Mental

Assuming that you are not rated on odds bets, then the simplest system with the best EV is just betting the don't on every roll. The house edge is 1.36%. Any system that uses any other bet will have worse EV and will be more complicated.

Are you really expecting more than 1.36% in comps?
link to original post

I don't know what he's expecting in comps, but I've certainly heard of more unusual things. For a brief time, I was aware of a Bubble Craps machine upon which you could play profitably right off the top on points multiplier days, and that was just on the value of the points, so any mail was extra.

In any event, I agree with you that this House Edge, which I believe to be about 2.09166% of total expected action (per CO roll) is significantly worse than the House Edge just sitting there and making DP bets.

I'm also not convinced that the variance is necessarily reduced that much. The only result that comes anywhere close to covering a Come Out Roll loss on one of our Don't Pass bets is the 4/10 being established and made, so it seems to me we are just trying to nickel and dime back our expected negative outcomes on the CO roll.

With that, I'm not sure how effectively this system even accomplishes its intended purpose. It seems like significantly more than expected (or maybe even just the expected amount) of Come Out roll losses are just as bad here as they'd be if we were only betting DP. Certainly, the HE relative to our total action is worse.

As the OP mentioned, there's also the possibility of the Craps Supervisor or Pit noticing what OP is doing and giving an average bet of zero, though perhaps as likely, just an average bet of $100 for the CO DP bets. I think what OP is doing and why should be fairly obvious; I just don't know why the house should have any reason to care.
link to original post

Thank you for the detailed breakdown, Mission!

The idea was that if one was rated on ~$200 of play when a point is established, the comp value might exceed that of what a $100 DP bet alone would achieve, with much less variance than a $200 DP would entail. Odds bets are never rated, and while they lower the HE considerably, they also increase the variance considerably.

Another possible consideration is that once a point is established, you are effectively rooting for the point to hit which puts you more in line with the rest of the table. That means nothing in a vacuum, but prevents you from having to put on a poker face when the 7 hits.

The Eleventh Commandment of Gambling should be:

Thou shalt not play for comps. If your way of play generates comps then that’s great, but don’t ever alter your play just for comp purposes.

Quote: Ace2

The Eleventh Commandment of Gambling should be:

Thou shalt not play for comps. If your way of play generates comps then that’s great, but don’t ever alter your play just for comp purposes.
link to original post

Fair enough. Consider this strategy then as one to minimize variance while still maintaining a relatively low HE, allowing one to play for much longer -- barring a long stretch of 7/11's on the CO, which would be equally deleterious to the pure DP player. With a high variance, you'll most likely leave the table after a few minutes -- hopefully walking away with a lot more money than you started with, a slightly higher chance of walking away with a fair bit less than you started with ( these first two outcomes are assuming responsible bankroll management) or busted without a stop-loss and/or attempting a Martingale or similar betting system.

If you enjoy playing the game itself, and want to spend hours at the table, then minimizing your variance should be your goal. You still expect to lose money over the long run, but that's just the price for admission. With my example of earning free cruises in the OP -- I've earned three upcoming cruises so far (over $7000 in value) plus so FP and drinks just from the loyalty program from my local tribal casino in exchange for total losses of ~$1700. Since I also enjoy cruising and visiting new places, it's definitely a win for me.

Cruise comps are a different kind of animal in general as cruise lines make no revenue from an empty cabin. Similar to comped hotel stays with land based casinos only more so, as there are far more fixed costs (e.g. fuel, port fees) in operating a cruise ship versus a hotel. They could fill unsold cabins by lowering the fare, but then they run the risk of having folks who booked at a higher fare earlier demanding the lower rate as well. If they give away the cabin to someone who they expect will play in their casino, then they get a positive return as well.

If this low variance method causes me to play for 2-3 hours a night before hitting a stop-loss, or stop-win vs a few minutes if $100-200 was on the line with every roll then comps are a significant factor in how I will play. Yes, the 2.09% HE (Thanks again, Mission!) isn't the best bet on the table, but it's only surpassed by line/come bets and odds which are limited to 2x in cruise casinos. It pays better than optimal play (sans card counting) on blackjack, given the unfavorable house rules (BJ pays 6:5, etc.)

Factoring all that in, I believe this may be an instance where it definitely could be worth adjusting one's play to earn comps, assuming one's rating isn't zeroed out by the hedge,

You're very welcome.

If you're playing with money that you can ultimately afford to lose and believe the value of the comps that you perceive them to have exceeds your expected loss relative to total action, then I can understand why you would be interested in playing this way.

Personally, I would assign food, drinks and cruises an absolute value of $0.00, so this isn't something that I would do, but if you enjoy cruises and perceive them as having $7,000 in value, and can get them with less than $7,000 in total expected loss, then this seems fine.

And, at the end of the day, you're standing at the table and it's your money to do with as you please.

Quote: Ace2

The Eleventh Commandment of Gambling should be:

Thou shalt not play for comps. If your way of play generates comps then that’s great, but don’t ever alter your play just for comp purposes.
link to original post

What are the Ten Commandments of Gambling?

Quote: FatGeezus

Quote: Ace2

The Eleventh Commandment of Gambling should be:

Thou shalt not play for comps. If your way of play generates comps then that’s great, but don’t ever alter your play just for comp purposes.
link to original post

What are the Ten Commandments of Gambling?
link to original post

https://wizardofodds.com/gambling/ten-commandments/

Quote:

Thou shalt not cheat
No explanation necessary.

Thou shalt honor thy gambling debts
A true gentleman honors his debts, especially gambling debts. When making a bet with another person you are putting your honor on the line. If you lose, you pay. No excuses!

Thou shalt expect to lose
The Las Vegas Strip was not built by winners. Even with good rules and strategy the odds are still usually in the casino’s favor. So don’t get mad if you lose. Think of it as the price you pay for entertainment.

Thou shalt trust the odds, not hunches
If you want to maximize your odds then believe in mathematically proven strategies, not hunches. If hunches are so great why are there so many psychics working the Boardwalk in Atlantic City as opposed to playing?

Thou shalt not over-bet thy bankroll
Before you gamble determine what you can safely afford to gamble with - as entertainment money. Stick to your limits and don’t gamble with money you need for necessities.

Thou shalt not believe in betting systems
For every one legitimate gambling writer there are a hundred charlatans trying to sell worthless betting systems promising an easy way to beat the casinos. I know it sounds like a cliché, but if it sounds too good to be true it probably is.

Thou shalt not hedge thy bets
Hedge bets usually carry a high house edge. For example, never take insurance in blackjack and never bet the any craps or any seven in craps. Exceptions can be made for insuring life changing amounts of money.

Thou shalt covet good rules
Rules vary from casino to casino. To improve your odds know good rules from bad and then seek out the best rules possible.

Thou shalt not make side bets
Side bets are sucker bets. Period.

Thou shalt have good gambling etiquette
Gambling is a lot more fun when people are polite and respect each other. It is also good etiquette to tip for good service.