Play it all again, Sam

RaleighCraps, I'm drastically altering the way I play craps in the casinos where I am known. I thought about playing all the numbers; place all 6 numbers across, hop the sevens and play the horn high yo, plus the ace-deuce. I know what you're thinking, and yes I know this is a loser. But it got me thinking, if someone really wanted to play this way, what would be the optimum bet spread? I calculated the HA for $15-$30-$36-$36-$30-$15 across, hop sevens for $36, $5 horn high yo and $1 ace/deuce, and came up with a 4.031% HA. Pretty ugly, but better than roulette.

So, what would be the optimum betting strategy to cover all the numbers? Assume 12 pays 3 to 1 in the field. This is certainly not a new idea, so it's probably already been determined.

The upside, if you can call it that, even though the strategy will always lose, is that there would be virtually no variance. I could also imagine, if you slowed the game enough, and if the box rated all the bets, or even only the place bets, you might earn significant comp benefits, especially on bonus point multiplier days.

Loose the hounds!

There would still be lotsa variance. The numbers rolled would be unlikely to distribute themselves perfectly under the bell curve. Of course, you wouldn't want that to happen anyway as an asymmetric distribution would be your only prayer for survival.

Your strategy (if, as you said, you can call it that) is very reminiscent of betting every number on the roulette layout. It seems equally pointless--you are essentially ENSURING a loss, though not with the absolute certainty of the every-number roulette bettor. What you are doing, in a nutshell, is hoping for as many "outside" rolls as possible, and as few "inside" rolls as possible. Given that, why not simply bet all the outside numbers, props, field, hardways, etc. and let the 6-7-8 be the province of the house? You don't actually gain anything by betting the inside numbers as well (you merely increase your mathematical loss/roll), so why not throw caution to the wind?

Quote: Headlock

So, what would be the optimum betting strategy to cover all the numbers? Assume 12 pays 3 to 1 in the field. This is certainly not a new idea, so it's probably already been determined.

I think the best betting strategy would be to bet the inverse of the odds on the payout (i.e. bet 5/6 of 1 unit on a bet with a 6:5 payout). It seems like the goal of a system like this is to minimize variance, and that's accomplished when each bet resolution has as close as possible to the same outcome.

If you're right and you lose 4% on average, the comp rate would have to be pretty high to offset it. I don't know enough about craps comps to say for sure, but I bet they would under-rate the theoretical loss pretty significantly.

Headlock, If you're expecting me to answer this one with math, you've incorrectly confused me with one of the site math wizards. I hang out here because I am trying to learn the math behind the HE, EV, mean, SQRT of all evil, and be a deviate.

Certainly the anti-hedge crowd will be screaming loudly (probably as I am typing this). Just looking at it off hand, it doesn't seem like there is any way possible to win with this. If any place number hits, you just win enough to cover the $42 you lose on the 1 roll center bets (hop,yo,A/D). You mention the field, but I don't see a field bet listed. If you are betting the field too, then the 1 roll loss on any place number is $47 on the 5,6,8, which is more than you would win on the 5/6/8. I could analyze all of the possible rolls and calculate the possible win or loss, but somehow I don't think that was the point of this thread, not to mention I am just too lazy to do it. ;-)

I'm not one to criticize anyone's bets. I support your right to play the Big 6/8, if that is what you want to do. As long as you don't expect me to play it with you, we will get along fine. I'll even cheer for you when you collect your $5 (while I pocket my $7 for my $6 bet). I have also learned not to try and educate someone who is playing the Big 6/8. They mostly look at you like you are a buzz-kill. I have found that if someone asks you a question, those are the ones you can spend time explaining the game to, and they are the ones who appreciate the advice.

Personally, I like to cover as many numbers as possible, although thanks to this forum, I am paying more attention to my bets now. My personal guideline is to only play the field if I get 3x on the 12. In spite of what the math says, I still prefer place bets to come bets. So I will play PL with odds and then place Inside or Across depending on how my BR is holding up. 2008 into 2009 this worked great for me 5 out of 5 trips. Last 8 trips -- gave it all back.

However, you mentioned comps, and my reckless play has meant my only expense for 4 to 6 days trips to Vegas has been my flight. My trips to MS have all been 100% free. However, even this value is being eroded, as they are practically giving away rooms these days. It is getting harder to justify playing craps now that I am back on my own money, even with the comps.

In my opinion none of it really matters for the durations of our sessions (my sessions will last for 8 hours if my br holds up). It still boils down to whether or not the numbers you are on are rolled when your bets are active. I played a short BR session in PA and got killed making the 'best' PL bet with 3x odds, while the idiots next to me made money only playing the hi,lo,yo every roll over a 3 hour period. (I'm not calling them idiots for making that bet, it was every other clueless thing they did that earned them the name.) I do think that point is lost on the crowd here. I don't think any of us will ever play enough hours of craps to ever be anything other than a short duration on the expected distribution curve. Therefore, any bet that makes you happy is a good bet, unless there is another bet on the table that does the same thing but pays better (like betting Any 7 instead of hopping 7s).

RaleighCraps, I know you're not one of the math guys. It's just nice to find a sympathetic ear sometimes.

I'm not saying I would ever employ this strategy, I just started wondering. I'm hoping one of the math guys can tell us what the optimum bet spread would be to collect on every roll (except the comeout roll) and minimize the loss. I would compare that loss to the potential comps.

In the bet spread I described, 2 or 12 loses $11, 3 or 11 loses $10, 4 and 10 lose $24, and 7 loses $14. A roll of 5,6,8 or 9 breaks even. There is no way to win on any roll.

I was looking at the iron cross as a way to get paid on every roll that was not a 7. You play the field and Place 5,6,8. It seemed like if you bet the correct amounts, this would be a decent system, but it just doesn't pan out. I ran simulations on WinCraps (thanks goatcabin) and it clearly shows that the 7 costs you more than you can win in between 7s.

As noted before, I doubt you would get a very good comp for your scheme, even though some of your bets have a high HE. I think they would rate you lower, since they can calculate almost exactly what you will lose. And as rdw said, I doubt the comps would be worth as much as you would lose on this play. Might be better off to pay for the room and go lay by the pool for 8 hours ;-)

Quote: RaleighCraps

I was looking at the iron cross as a way to get paid on every roll that was not a 7. You play the field and Place 5,6,8. It seemed like if you bet the correct amounts, this would be a decent system, but it just doesn't pan out. I ran simulations on WinCraps (thanks goatcabin) and it clearly shows that the 7 costs you more than you can win in between 7s.

As noted before, I doubt you would get a very good comp for your scheme, even though some of your bets have a high HE. I think they would rate you lower, since they can calculate almost exactly what you will lose. And as rdw said, I doubt the comps would be worth as much as you would lose on this play. Might be better off to pay for the room and go lay by the pool for 8 hours ;-)

I expect you're right. I'll just continue to hope one of the math guys will take the challenge. If not I can probably do it myself, but it will take a lot of trial and error. I figured out the Fire Bet hedge amounts myself (I'm patting myself on the back), but unfortunately haven't had the chance to put them to use.

Quote: Headlock

I expect you're right. I'll just continue to hope one of the math guys will take the challenge. If not I can probably do it myself, but it will take a lot of trial and error. I figured out the Fire Bet hedge amounts myself (I'm patting myself on the back), but unfortunately haven't had the chance to put them to use.

The thread on hedging against your 'in-the-money' Fire bet was great reading. It showed how to lock in an absolute profit which I hadn't really considered before. I hope to get to put it in action later this month in Vegas. :-D

Quote: RaleighCraps

The thread on hedging against your 'in-the-money' Fire bet was great reading. It showed how to lock in an absolute profit which I hadn't really considered before. I hope to get to put it in action later this month in Vegas. :-D

I hope you give us a nice long trip report. I always enjoy reading those. Be sure and let us know where you find the Fire Bet in Vegas, and good luck.

Quote: Headlock

I'm hoping one of the math guys can tell us what the optimum bet spread would be to collect on every roll (except the comeout roll) and minimize the loss. I would compare that loss to the potential comps.

I know you're looking for a "bet it all at once" spread, where you can sit back and collect after that, but any spread like that is going to suffer from the 7-wipeout problem. The way I've found to work around that is to slowly roll out your number coverage using multiple come bets. I came up with a strategy a long time ago where you'd bet pass and then keep making come bets, not take odds on anything until you had four different numbers established, and then you'd take odds all at once and wait for your numbers to hit. There was more afterwards to take advantage of a long hand, eventually working up to all 6 points with full odds.

That strategy won't directly serve your purposes because there is one situation where you're really exposed (when you put all four odds bets out) but you could modify the concept slightly to reduce your risk substantially. For example, you could make pass/come bets until you have decent coverage (3 or 4 numbers) and then use the ongoing winnings to fund additional wagers or odds bets. The problem is that any additional wagers are likely going to have a high vig, so greater coverage comes at a price. However, this approach minimizes loss-per-shooter compared to a big "bet it all at once" spread, because you've been taking money off the table as you work up to your coverage.

On the other hand, you did mention comparing the loss to comps, so that's something different to start with. If you know how a casino will comp you based on various bets, you can design a system to maximize your ROI, or comps-to-loss ratio. I've had decent success with $12 each place 6/8 + $50 no-4 and then jointly-pressing the place bets (and eventually taking down the no-4 at around ). Each time you make the bet you throw in 3 green and that's bound to get rating attention.

Quote: MathExtremist

I know you're looking for a "bet it all at once" spread, where you can sit back and collect after that, but any spread like that is going to suffer from the 7-wipeout problem. The way I've found to work around that is to slowly roll out your number coverage using multiple come bets. I came up with a strategy a long time ago where you'd bet pass and then keep making come bets, not take odds on anything until you had four different numbers established, and then you'd take odds all at once and wait for your numbers to hit. There was more afterwards to take advantage of a long hand, eventually working up to all 6 points with full odds.

That strategy won't directly serve your purposes because there is one situation where you're really exposed (when you put all four odds bets out) but you could modify the concept slightly to reduce your risk substantially. For example, you could make pass/come bets until you have decent coverage (3 or 4 numbers) and then use the ongoing winnings to fund additional wagers or odds bets. The problem is that any additional wagers are likely going to have a high vig, so greater coverage comes at a price. However, this approach minimizes loss-per-shooter compared to a big "bet it all at once" spread, because you've been taking money off the table as you work up to your coverage.

On the other hand, you did mention comparing the loss to comps, so that's something different to start with. If you know how a casino will comp you based on various bets, you can design a system to maximize your ROI, or comps-to-loss ratio. I've had decent success with $12 each place 6/8 + $50 no-4 and then jointly-pressing the place bets (and eventually taking down the no-4 at around ). Each time you make the bet you throw in 3 green and that's bound to get rating attention.

Good stuff, ME. I've never thought about the waiting for a bit to take odds on the come bets. That's a good strategy to lessen the blow.

Thanks for your input ME. I'm thinking about it.

One of my craps buddies places the 6 and 8 for $60 each if the point is one of the outside numbers. If the point is 6 or 8, he places the 5 and 9 for $50 each. On the first hit he takes it all down except one unit (ie down to $6 on the 6 and 8 if it's a $5 table). His 6 and 8 pay the same as my $5 come bet with $50 odds, but he gets comp credit for $60 while I get credit for only $5.

Teddy: If you grow weary of come bets as I did, another way would be to place bet as the number is rolled, similar to come bets as you are placing the number after it is rolled, and the 7 is really never your friend in come betting as you know. This way do not have the downside of come bets(losing on crap number clumps and losing on the 7 on the come out rolls).

I do agree that most sessions it is better to get your bets out there gradually as many PSOs and short rolls. One way I have had decent success is starting with pass-odds and placing the sister number so two bets only to start. Example if the initial point is 9 will have pass-odds on the 9 then also place the sister 5. Then after get a hit on the pass-odds or sister place bet, then start placing the number just rolled and its sister until all box numbers are covered or can go with less numbers covered. Need to get at least 4 numbers covered and preferably 6 numbers(based on a 10 shooter bankroll) on a good roll in a gradual way.

I've been looking into the Iron Cross, which apparently can be successful if that dreadful 7 doesn't insist on coming 1 out of every 6 rolls. Can you integrate hopping the sevens with the Iron Cross to reduce the house edge to near zero? And please, I understand this is a no-win strategy. I'm just looking for the loss/comps ratio.

I've been using the Wizard's craps simulator and don't see any hop bets available. Are they not there, or am I missing them?

The thing that I don't like with the slow outlay strategy is the first 3 or 4 non repeat numbers you don't get paid anything. So while you are protected against PSO, you are also looking at a loss if 4 different numbers are rolled and then 7. IMO, the slow out strategy may also have a negative affect on your comp rating too. The box hears my dealer repeat my $110 inside call. Is the box going to notice that your place bets have been steadily increasing this roll? Maybe, maybe not.

Here's a scheme that I ran through numerous simulations using WinCraps, that might interest you.
No PL/DP bet.
Wait for a point to be established then bet $430 across 50,75,90,90,75,50.
If you want to cover the 2,3,11,12 make a nominal field bet (it will cost you some winnings on the 5,6,& 8).
Either take it down to minimum after 1 roll, or wait until you get paid once, and then take down to minimum.

You will be up ~$100 after any 4,5,6,8,9,10 is rolled, and you will be putting from $22 to $44 back out as a minimum bet, thus guaranteeing min $56 to $78 win.
For this to work out, you need to get paid on the big bet about 5 out of every 6 times, depending on table minimums and how many more times you get paid after you drop back down.

WinCraps simulations actually show this to have quite a number of positive outcomes after 10,000 4 hour sessions (I can't remember, but I think it may have even had a positive skew), but it requires a large starting BR ($5 to $6K) for each session, AND total ruin happens much more often. On the plus side, each simulation always had at least one outcome that was well north of +$10k for the lottery types......
If only I had a BR large enough to give this a try......

Quote: Headlock

I've been looking into the Iron Cross, which apparently can be successful if that dreadful 7 doesn't insist on coming 1 out of every 6 rolls. Can you integrate hopping the sevens with the Iron Cross to reduce the house edge to near zero?

Heck no. You can reduce the variance to near zero but the house edge will be astronomical. The reason being you are making six of the worst bets on the table (hopping the sevens is really terrible), and each one of those bets is subject to the house edge. Any other of the strategies mentioned on this thread are better than that. And yes, I'm sure the casino would comp you generously if you played that strategy. You would be one of their best customers.

Quote: teddys

Heck no. You can reduce the variance to near zero but the house edge will be astronomical. The reason being you are making six of the worst bets on the table (hopping the sevens is really terrible), and each one of those bets is subject to the house edge. Any other of the strategies mentioned on this thread are better than that. And yes, I'm sure the casino would comp you generously if you played that strategy. You would be one of their best customers.

Whoops....

You're right, I meant how can I reduce the variance to near zero. But I see what you mean, adding 7's on the bounce, with an 11% house edge, can't be good.

Quote: Headlock

You're right, I meant how can I reduce the variance to near zero. But I see what you mean, adding 7's on the bounce, with an 11% house edge, can't be good.

You don't want to reduce the variance to near zero. Variance is the only way you can beat a negative game. Here's a system with near-zero variance: $1 12, $31 pass, $30 don't. Outcomes range between $0 and -$2, which on a $62 wager is very low variance. But you *never* win...

Remember, zero variance means every outcome is at the mean (house edge). If you put $1 on each spot in roulette, you'll bet $38 and lose $2 every time, which is exactly the house edge: -2/38 = -5.26%.

The quest is not to tame variance, but to shape it (and skew) to fit your playing preferences.