Comparative Strategy Analysis

[Ahigh]

And as much as I hate to admit it, I record ALL of my rolls and don't omit them. I had a great month in August from my recorded rolls, but September gave it all back and then some.

[Ahigh]

The interesting thing though, is that around 450 rolls, the buy the 4 and 10 started stomping the other strategies on my rolls with no signs of coming out of the stratosphere.

So it comes back to the 4/10 still wins in the long haul, and I am sure the edge has a lot to do with that.

But the SHAPE of the ah280 is absolutely beautiful. It's just so triangular with only two slopes. It's either up $50 a roll or down $280 a roll. Two possible outcomes on every roll gives you this really awesome chart appearance.

So even if it's a totally unusable strategy due to the high edge and dependence on a high SRR, I think it's worth examining in terms of betting specifically against the red.

I still like searching for a strategy that will work for even my shitty rolls I had after the 300 I recorded in August. I know I will record some better rolls in the future, but I'm not going to mislead myself or others to think that I always shoot good, so I have to throw my terrible rolls in there too!

[Ahigh]

So I verified this with the charts, but the AH280 is tracks near identically to a 12 unit iron cross (22 * 12). So congratulate me for reformulating the iron cross that requires 12 times as much money. It shouldn't be a surprise as everyone should recognize the iron cross as the definitive high-edge "no seven" bet strategy.

[heavy]

This fine thread has over 14000 views. Much math is involved. Results are damned interesting. Comments welcome. Dive in.

[London Shooter]

Some great stuff in here. Love the analysis of the MP204 v the AH280, even if not comparing apples with apples.

Also the view that SRR of 6.25 is the breakeven point for either of the above.

Finally, the fact we should all know, yet some still seek to not believe - if you are playing the random game over enough rolls you will lose money. No strategy/system/method can be mathematically proved to beat the random game by overcoming the house edge.

[London Shooter]

22i - for you if you haven't read this.

[Bankerdude80]

Fascinating read.

[220Inside]

Good stuff. Thanks, LS.

[mssthis1]

Since I'm a keep it as simple as possible type of person, I'll stick to the MP 204 across variation that I have tweaked for my own personal preferences.

The AH280 makes my head hurt and I'd be afraid of a dealer coming across the table to choke me if I played it regularly.

Note on the last charts that the winning sample is only approx 300 rolls and the sample where the 4/10 strategy is an outlier is around 1200 rolls. Both of those samples are too small to bear much weight in my opinion. Size matters if you are defiantly trying to prove or disprove a strategy.

This is good, useful information. It's good to see different strategies charted against each other even in small samples.

[rhythm roller]

Since I'm a keep it as simple as possible type of person, I'll stick to the MP 204 across variation that I have tweaked for my own personal preferences.

MSS,
Curious what your tweaks are to MP 204 if you don't mind sharing. I like that betting as well---only on myself.

[mssthis1]

I just bet standard 3 unit bets across and regress to 1 unit after 2 hits instead of making staggered bet amounts.

As an example. $15 on each number $18 6/8 and regress to $5 on each number $6 6/8 after 2 hits. From there press as you see fit or same bet, whatever fits your style of play.

[rhythm roller]

Mss,
Thanks for the answer to my question. One more question. Then 3X odds on the point instead of placing it?

[mssthis1]

Mss,
Thanks for the answer to my question. One more question. Then 3X odds on the point instead of placing it?

3x odds and regress to 1 unit after 2 hits. Playing at the $15/5 level, a pass line with double odds and regress to just the pass line would be ok too. That would depend on bankroll and how long you intend to play.

You are going to have ups and downs.


if I play for any length of time having $700 invested is not unusual for me. I'm a slow starter some days. Last time I played a $100 buy in would have been sufficient.

[AlamoTx]

I don't often post here, but I felt like, especially for new players, I needed to get my two cents worth on the board.

First, I’m sure this post won’t be popular on this board, but if it helps one new person save some money at the game, then I’m glad I’m posting it.

My point: Regression betting is ultimately a losing proposition for your bankroll. Here is my argument. By way of disclaimer, I am not a mathematician, so I won’t try to prove this based on the math; but, I will make some hypothetical game assumptions I think most of us who have played a lot of craps will agree are reasonable based on empirical experience.
Using a recent example, let’s test the famous $204 Across, Two hit and Regress method.
To illustrate outcomes, I will posit a come better at 3,4 and 5x odds vs the regression player in 2, single game scenarios, ending in outcomes that I suggest are not that unusual in real play. I am not going to assume even a 15-roll hand. Those are rare, statistically, regardless of how good you think you are. A shooter with an SRR of 8 (damn good for all but the self-deluded) will have an over 15-roll hand 17.6% of the time, meaning that out of 100 attempted hands, only 18 of those will be 15 or better. (Read Wong on Dice) Regrettably, for place bet regression players who slap a large sum down across the box, many of those games will be point 7s and rolls under 5, some of which never hit a box number or only hit one. That is the Achilles heel of place bet regression systems. Put in a positive light, regression betting allows you to win a little while risking a lot. It takes money to make money. If you're going to risk it, let it do its thing. If that $204 stays at work and gets 4 hits, you'll be damned near even and ready to make some real money if the dice keep making numbers...but I digress.

[BTW, pardon my calculations below, if I don’t get the math exactly right. The point made will be the same at the end]

Game One: First, the regression player establishes a point of 6 at a $10 table ($10 tables are realism in Vegas). The shooter takes $35 odds behind the six, and immediately places $30 on the 4,5,9 and 10 and $42 on the 8 - $207 Across.
Roll 2: 8 – Shooter racks $49 and now has $158 at risk;
Roll 3: 5 – Shooter racks $42 and now has $116 at risk;
Shooter regresses to $15 on the 4,5,9 and10 and to $18 on the placed 8 and reduces odds on the point 6 to $10.
Shooter has pulled down $99 from the board (could have done more by going to $10 units) and now has only $22 at risk and all the box numbers. This is as good as it gets for the regression shooter.
Roll 4: 11 – No blood
Roll 5: 4 – Shooter racks $27 and is now up $5;
Roll 6: 3 – No blood
Roll 7: 5 – Shooter racks $21 and is now up $26
Roll 8: 9 – Shooter racks $21 and is now up $47
Pretty soon this shooter needs to decide whether to begin a progression or keep racking hits. Even a small progression at this point will cut into his or her accumulation rate (Statistically a shooter with an SRR of 8 has an average hand length of 11 rolls before 7 out happens). This shooter decides she is a fabulous dice controller and will press until at least 4 numbers are green chips.
Roll 9: Another 4 and shooter buys the 4 for $25 and racks $17 of the $27 win – up $64
Roll 10: 8 – Shooter bumps the $18 8 to $30 and racks the net $8 – up $72
Roll 11: 8 – Shooter racks $35 and smiles. She is on a roll and will collect 4 hits on her pressed numbers before increasing any more – now she is up $107
Roll 12: SEVEN – Damn! She had it going too. This is such a good system, she is going to try it again when the dice come back to her.

Game One: Progression player. To illustrate a point, all the hits will be the same for the progression come bettor as though she were playing side by side with the Regression player.

Roll 1: Point of 6. Shooter places $50 odds behind the passline and makes a $10 come bet – mas o menos $50 at risk given that the come bet somewhat protects against the point – 7 out;
Roll 2: 8 – Shooter moves her come bet to the 8 and puts $50 odds behind it and puts out another come bet – mas o menos $110 at risk with the second come bet protecting against the 7;
Roll 3: 5 – Shooter goes to the 5 and puts $40 odds on the 5 – At risk $170; the shooter’s strategy will be to hold what she’s got until she makes one of her numbers, then she will go to 4 bets with full odds and try to build on that.
Roll 4: 11 – No blood
Roll 5: 4 – No blood
Roll 6: 3 – No blood
Roll 7: 5 – Shooter racks $70 retrieves $40 in odds and puts a come bet out there – At risk $70;
Roll 8: 9 – No blood, but shooter goes to the 9 with full odds and puts a $10 come bet out – at risk $110
Roll 9: 4 – No blood;
Roll 10: 8 – Shooter racks $70 and is off and on the 8 – at risk $40;
Roll 11: 8 – Shooter racks $70 and is again off and on – Shooter is up $30 and still has a come bet out there.
Roll 12: 7 – SEVEN. Same expletive-laced reaction. Shooter wins on the come bet and everything else comes off the board – Shooter is up $50.
The come better did half as good as the place bettor… this time.
Now, we will look at something that happens all too often unless, of course, you are bullshitting yourself or have overblown expectations about your game.

Game Two: Regression player
Roll 1: 7 – Winner for the passline – shooter is now up $117.
Roll 2: 4: This establishes the point. Bettor puts $20 behind the number and places the rest of the box as follows: 5 - $30; 6 and 8 - $42 each; 9 and 10 - $30 each - $204 across. Shooter rubs her hands together enthusiastically, she is already up $112, and this ‘show ‘em your ass and duck quick’ system is going to kill the house tonight.
Roll 3: 5 – Shooter racks $42. Game on. One more hit, and she starts hitting the house again.
Roll 4: 11 – No blood
Roll 5: SEVEN – What the F! I’m better than that. This can’t be! Regrettably, the shooter passes the dice. Shooter is now DOWN $45.

Game Two: Progression player
Roll 1: 7 – Winner on the passline – shooter is now up $60.
Roll 2: 4 – Shooter puts $30 behind the 4 and makes a come bet - $30 at risk
Roll 3: 5 – Shooter moves to the 5 and puts $40 odds behind it and has a third come bet out there – at risk $70.
Roll 4: 11 – Shooter racks $10 on the come – at risk $60;
Roll 5: SEVEN – What the F! I’m better than that. This can’t be! Regrettably, the shooter collects on her last come bet and passes the dice. Shooter is down $40

Conclusion:
Both games represent a common scenario we all see between those long rolls we get some of the time. The progression bettor does not risk as much to start with. Point 7 will not hurt the progression player nearly as much as it will hurt the regression player. The progression player can survive a few lousy hands like those shown above. Keep in mind that after a winning first hand, the regression player ended down $45 after the 2d hand. Not terrible, but now, to feed her regression system, the next time she bets $204 across, she must win $249 just to remain even. Heaven help her if she has a game 3 where she just can’t seem to hit any numbers within the first 5 rolls. With an SRR of 8, the shooter will average 11 rolls, and ‘average’ means there are some shorter and longer than that. If the regression player repeats a bad hand for game 3 with $249 on the line, when she goes out there for game 4 with her bold $204 across, she then potentially goes in the hole for the $249 she has already dropped plus the new $204 she has to go with to continue the system - $453...to get two hits and then start trying to recover with $15/$18 units?

Do your own calculations on this stuff, but I caution everyone, especially novices, about over estimating your skills. Delusion: “At home, I hardly ever 7-out before making two box numbers”. Craps is still a game of chance, even if you influence the dice positively, and how you bet will determine your outcome. My years of experience playing with all kinds of systems has taught me that showing the gambling gods your big bets across the box is a prayer to those gods to rip that easy money out of your hands and make you earn it. As one of my mentors once taught me: “Experience is what you get when you were expecting something else.” If you are expecting to bet big across, regress quickly and then kill the house, I suspect you are going to get some ‘experience’ along the way.

Entertainment is another thing altogether. If you are looking for adrenaline, place betting big across and then hoping to beat that train across the intersection will definitely give you a thrill. If it is adrenaline you want, just realize you are going to pay for it. If it is winning you want, expect to be a little bored sometimes, but in the long run a lot richer.

As far as the game examples I posited...I'm telling you...that kind of shit happens...and it happens a lot.

Alamo

[Moe Bettor]

"Point 7 will not hurt the progression player nearly as much as it will hurt the regression player. The progression player can survive a few lousy hands like those shown above. "

Let's see here..the progression play has a PL bet with odds and $204 across and it's POS time. PB (ProgressionBetter)is now down a ton. Maybe $250? I don't know..Maybe not much for a person who buys in for a few thou. This means that the PB has to win..starting out at least $250 to even get flat! The RB(Regression)..$10 place bet. Rolling dice gets a 9. Shooter (me) throws for 6's and 8's..so I go $18 6 and 8. Exposure is $46. One hit of $21 and I can either regress to $12 6 and 8 which leaves my exposure now at $34-$21..$13 bucks. and ready to rip and roar on pressing. Next hit $14..I can same bet and actually be ahead a buck or I can take both bets to $18 each. Look..not every craps player is smart about betting and DI'S go PSO often..or just have rotten days..or the variance catches them by the short curlies and won't let go. Your examples are potent, but illustrate fairly stupid play..which, yes, people do. I would guess that most on this forum who've thrown out $204 across either don't do it anymore or have achieved such enormous amounts of money that it's nothing to them. Or maybe one hit and regress to the minimum. But I could be wrong.

[House of Orange]

Well the $204 across was meant for a guaranteed 7 SRR or whatever it was. Not for the schmuck with a home table SRR of 7. Huge difference

[Moe Bettor]

I think there is a misconception here about SRR and maybe Irish can say something about it. SRR is not a guarantee of something happening in your favor at a particular time. It is an over-all measurement..which is to say thousands of rolls produce this SRR number. At any particular time, a good DI can go PSO five times in a row. Yeah..I can see $204 across as perhaps one of the better strategies. What were the others? Was $204 across at the top of the best or was something else?

[London Shooter]

Ha, ha Irish mentions flat betting and uses it the same way I would - as in same bet - but don't tell Heavy

Boring but as a baseline it's going to out-perform almost all other strategies no matter how fancy they are.

What about doing 88 inside and just same betting it? Hopefully you are in MS where you can buy the 5 and 9, or alternatively buy the 4 and 10 instead. Compare that to MPs 204, AH280 or whatever. It will definitely hold its own. But are you well enough funded to even be betting at this level? I know I'm not.....

[memo]

Alamo,

Thanks for posting and for the thoughtful illustration. I find it refreshing to look at these betting strategies to compare pros and cons...
The statistics you are stating are spot on and it all makes sense. It is, however a short term look, which in a discussion like this is probably all that realistically can be done...Unless you use pure probability and test as Ahigh did previously..He did use large samples and predicted what kind of SRR would be necessary for success.

What would you say is the average wager for the progression better, and buy in, as opposed to the regression better?
I am thinking that they should be closer to being the same...
and there is the rub...One is starting smaller and building up while the other is starting larger with a sudden decrease, then building back up....Hard to compare. I guess one would have to make an assumption of the average hand length and assume how many non box numbers would come up. Otherwise the average bet would vary quite a bit, which would make it a difficult comparison if one pointed to the monetary result of a hand or two to make a determination of which style is more efficient.

That being said, I like your argument and the thinking used to base your opinions. I am reviewing my betting strategies and this will factor in greatly.

Memo

[skasower]

All,
This was a great discussion. I am wading in as a relative beginner in DI but not a beginner to Craps. In my opinion, there are so many exogenous variables impacting the individual outcomes of each toss that it is observable that sometimes randies who slam the dice into the far wall have pretty good rolls where they hit their numbers even twice. So, while I love mathematics and statistics, my betting is timid. I am trying to improve two areas of my Craps play: 1. on axis tosses more consistently, 2 better observations of those small portions of time when a "hot" roll is upon me at the table.

As for 1., it requires practice and real play over time to improve. As for 2, I believe that the timid progression betting is superior to large initial "across the board" bets. Could this be demonstrated statistically. My gut feeling is that a statistical evaluation showing trends within a longer term set of tosses (like 40,000) is not something I have seen. But, really folks that is the reality of our Craps "gamble" each time we approach the table to play. Note: I am not advocating a "Scoblete" 5 count, but some count is warranted to effect the observed or "gut" feeling that betting is potentially beneficial. I watched Howard R&Rr "chart a table full of tossers, and that approach was also interesting for informing one''s "gut" feelings.

Sorry to veer into the mystical and away from the scientific, but really I believe we need some of that Hogwarts magic to succeed in Craps.

Again, thank you all for this discussion. It provides great "food for thought."

skasower...aka...