Comparative Strategy Analysis

[heavy]

Over on the Tunica report-out thread we've wandered off into a discussion of the effectiveness of Steve Nelli's "Method" when compared to other betting strategies. I stated that I would like to see a side-by-side comparison of results of the Method compared to three or four different strategies. I specifically mentioned MP's $204 Across strategy, SIA's One Hit - Can't Miss strategy - and my old Heat Seeker strategy. The Heat Seeker can get a bit convoluted if you play it with all of the "qualifiers" I toss in. We might want to simplify that one a bit and make it a six and eight place strategy combined with continuous come betting or something like that. As for Nelli's Method - I'll confess that I have a copy of the old Craps Pro system that the Method is based on - and that a lot of the triggers Nelli uses in his Method are triggers I've used through the years (a lot - but not all). I'll also admit that I don't have 100% of the system so I'm not really qualified to run this exercise. Scribe and Buy The Four offered to do the comparison for us. MP suggested that we use a fair number of the rolls tracked at Tunica last week as the basis for our play. LagunaFlats tracked all of the rolls at the tables he played at while we were in Tunica, so I'm hoping we can coax him into providing those results. I do have some concern over the fact that every Method player I've met seems to play it slightly differently, so I would ask that whoever does this comparison stick as closely to the Method as taught as possible. I'll also state up front that the purpose of this comparison is NOT to reveal any proprietary information. Whoever takes up this project - I'll ask that you simply post the roll results and your plays but do not explain why you made the play. All we need to know is "Five rolled. Buying the $10 for $25" or whatever. Don't tell me "Fives bring out tens." Make sense?

Thoughts?

Any other strategies you'd like compared in this exercise?

[Mad Professor]

Maybe the parameters might be each participant has only $204 dollars to bet with.

I'm not trying to be a total asshole here, but if this little bet-method comparison is going to be rigged from the outset, such as only allowing a total-bankroll of $204; then count me out.

The same holds if you are only going to run it through a hand-selected 20-roll snapshot, and then try to point to one method as being far superior than the others.

That kind of approach would be totally preposterous. I thought you guys were serious about this.


MP

[wild child]

M P

You posted this approach to WAGER UP ENOUGH BANK ROLL $$$ TO BUILD A SIGNIFICANT $$$$$ AMOUNT SO THE TYPICAL D I BOARD MEMBER could build a LARGER BANKROLL thus having the underwriting to fund LARGER BUY IN thus LARGER $ WAGERS

Your $204 Across is a VERY STRONG APPROACH and I favor it with little tweaking.....
standing at the same craps table the % gain at day's end may be the way to compare.....

Your $204 Across allows for TWO NUMBERS HIT(or one # hit twice) and DOWN..as I comprehend.


If I am on the right page with my understanding,would there be a more apples to avocados comparison
then limiting the approach to TWO WAGERS per SHOOTER HAND

Better how about defining a SESSION AT THE CASINO as say TWENTY SHOOTER HANDS and each player can bet as many WAGERS for whatever amount and allow for trips to a VIRTUAL A T M

Then put the ideas to the test of BANK ROLL SURVIVABILITY say after THREE SESSIONS with allowing folks to go to THE VIRTUAL A T M to gather RE-BUY $$$$$$$$$$

Me thinks few other WAGERING IDEAS would cross the finish line netting as many apples or avocados

With THE M P $204 ACROSS after 20 hands WAGERING 2 and down could net a +40 AVOCADO gain
Playing with slices of avocado till the termination of hand could show differing results........

Would not the TYPICAL craps person wade in to perhaps THREE SESSIONS in A TYPICAL DAY

Would it be O K to do a virtual apple TO avocado comparison of net results after three sessions

How about a CASINO DAY YEAR of perhaps 100 CASINO PLAY DAYS

Me thinks THE M P $204 ACROSS... is in the winner's circle with the TOP THREE WAGER PLANS


Just me saying
W C

[Mad Professor]

Hi WildChild,

The MP $204-Across two-hit Regression w/Passive-Aggressive Pressing is not a two-and-down approach, but rather a two-hits-then-regress-then-aggressive-post-regression-pressing approach.

Our good friend, Al_Falcons, ran the MP-$204 and about a dozen other betting-methods through something like 20,000 dice-influenced rolls. To my mind, that entire Testing, Testing...Formulized D-I Betting-Strategies thread (including its ~1700 dice-influenced hands) really set the standard in terms of how to first CODIFY how each betting-method is to be played...and then TESTING their validity across a wide spectrum of variously skilled D-I's.

While I'm not suggesting that H undertake such a large project; I am suggesting that the roll-stat sample size be reasonably large, and each method's bankroll be reasonable large. After all, a typical group-shoot weekend encompasses upwards of 700+ dice-influenced rolls. Using a smaller sampling would seem somewhat disingenuous.


MP

[Maddog]

*** NOTE ***

This thread was started to possibly compare the Method strategy against the $204 Across strategy.

There were several posts that were off the original topic inserted here. These posts have been split to a separate topic and can be found here:

Ahigh Strategy Analysis

Discussions on that topic can continue there, while this topic can remain focused on the proposed experiment.

Cheers
/MD

[wild child]

.
M P

My bad understanding.......
.The $204 Two Hit then REGRESSION followed by PROGRESSION is STRONG and stands the test on it's own.......
I interjected my modification and that could be not as others would WAGER

When using it I STEEPLY REDUCE my $$$$ AT RISK and only play out the HAND with TABLE MINIMUM $
to pay the rental on the CRAPS TABLE SPACE I OCCUPY...Thus doing while locking up my winning wagers I am not fully taking advantage of the POTENTIAL of an EXTENDED HAND .......

That said , there seem to be folks using some criteria that gets them on or off by some planed instruction I comprehend not.... Seemingly with "MIXED RESULTS"

Could we monitor by the reduced need to replenish BUY IN from an A T M

A WEATHER REPORTER recently commented about Hurricane /Tropical Storm wind speed.
" It is less so the speed of the wind than what is blowing in THAT WIND" or words to that affect..

Just me saying

W C

[Mad Professor]

Hi WC,

You are absolutely correct on a number of points.

Let's start with the most obvious one.

~By CODIFYING each betting-method's step-by-step bet-trigger/bet-amount/bet-change processes before the whole roll-stats-revealed exercise starts; we not only reduce risk of cheating (not that anyone here would even consider anything like that), but we also make each betting-method much more 'testable' against FUTURE in-casino roll-stats.

Going back, I think it must be something like four years or so, our friend Scout (at least I think it was him...it may have been Shaver or AlleyOop, now that I think about it) ran a similar bet-method exercise similar to what H is proposing here.

The way he ran it was, "Here's this shooter's first four tosses from his first hand...X...X...X...X...how are you going to bet now?" He would then reveal the next couple of outcomes, and I would pre-state EXACTLY and PRECISELY how I'd handle my bets at each and every step along the way.

As I recall, I was the only one who participated in that little exercise, but I am told by many it was quite illuminating to see my bet-making thought-process explained and played out as each hand-segment unfolded.

If anyone wonders why I am so keen to have each participant codify their betting-process; it is simply so that it can be applied to FUTURE yet-to-be-rolled outcomes and not just ones that happened in the past (and where some participants might be privy to the outcomes ahead of time).

That is also why I hopefully suggested before the Tunica seminar, that someone in that gathering, roll-track the group's outcomes, and then, after all was said and done, apply the MP-$204 to those roll-stats on a wholesale basis, to see how well (or how badly) it held up when applied to yet another large group of variously-skilled shooters...and to compare those recent-results to all of the previous group-shoot roll-stats that the MP-$204 has already been applied to in the above-noted "Testing, Testing...Formulized D-I Betting-Strategies" thread.

I am hoping that we can do something similar in this exercise.

MP

[wild child]

QUOTE : "You are absolutely correct on a number of points "

The M P - Only A NUMBER........ you are mellowing


QUOTE :
" By CODIFYING each betting-method's step-by-step bet-trigger/bet-amount/bet-change processes before the whole roll-stats-revealed exercise starts; we not only reduce risk of cheating (not that anyone here would even consider anything like that), but we also make each betting-method much more 'testable' against FUTURE in-casino roll-stats."

THE M P $204.........is STRONG and will stand up belt buckle to any craps table hand on it's own strengths.....Further it can be played multiple times during a TYPICAL WEEKEND CRAPS JUNKET with minimal or even ZERO DRAW DOWN....By STRONG I am saying : it will not fracture a dedicated BANK ROLL and will increase the $$$ in reserve for future in casino exposure ....

The following may or may not figure into CODIFYING < > perhaps just afterthoughts........


On BUY IN:
How many shooters or exposure to time at table is a wise measure of a CRAPS WAGERING SESSION ?

ON SESSIONS:
How many sessions is it prudent to expose a BANK ROLL in a typical CASINO WAGER DAY

ON FREQUENCY:

This a BIGGIE...How many Casino Wager Days can one invest the time,effort and $$$$$ over a calendar year

"It may be less THE WIND than WHAT BLOWS IN THE WIND"

W C

[Blackcloud]

UNHH!!For BC that be bout 48

[Mad Professor]

UNHH!!For BC that be bout 48

UNHH!! Is that 48 rolls...48 hands...48 sessions...48 days...48 minutes...or 48 moons?

MP

[heavy]

Let me toss out some paramaters and see if everyone can agree.

Starting bankroll: $2500
Table odds: $10 Game - 20X odds - $5000 max bet
Number of decisions: TBD - based on a minimum of two laps around the table during the Tunica weekend with at least one 20+ number hand. I realize some of you would like 1000 plus rolls but I doubt that can happen. We'll consult with Laguna Flats on that.

Objections? Thoughts?

[Mad Professor]

I'm cool with that.

If someone wants to jump in and claim the simple PL w/max-Odds betting-option; you should do so soon...it ain't gonna hang around long.

If no one wants it, I'll run it in parallel with the MP $204-Across Regression.


Edited to add: How about picking an arbitrary number of consecutive shooters, with two laps being ~28 shooters.

MP

[Ahigh]

You know I am really confused. Why was my strategy split out? Am I not allowed to participate or did I misunderstand the challenge? I saw the post about the split, but now MP is saying "who wants line with max odds?"

I am just really confused.

But I can't help but wonder what the motivation was to not allow my betting strategy to go head-to-head with MP's?

I code up strategies too, and I'm no slouch as a programmer. I can do this stuff myself, but I figured it would be easy to code my strategy.

Even if I'm excluded, I can code up MP's strategy and do the whole experiment myself. But it would be absolutely fantastic if I wasn't continuously feeling like people are hatin' on me!!!

[heavy]

. . . it would be absolutely fantastic if I wasn't continuously feeling like people are hatin' on me!!!

LOL. Don't go getting a persecution complex on us. Maddog just wanted to see the comparison run with just the two strategies. I have no problem with us running multiple strategies as long as the results are manageable and can be posted in some sort of logical fashion that everyone can understand.

[Mad Professor]

I've got a suggestion on how to run this betting-strategy exercise, but it might be a bit too radical for you guys.

~Okay think in terms of being in something like the Craps Chatroom (BTW our next Craps Chat is set for Thursday, Sept. 6th at 9 pm Eastern).

~All participants in this little shindig (and anyone else who wanted to watch 'live-action' gaming) would sign-in on a given day at a given time.

~The moderator (I assume Heavy) would post one (1) outcome per minute.

~In between each new 'roll', each betting-participant would post what their next bet (if any) is.

~Each participant (and any neutral observers) would be responsible for calculating their running wins and losses.

I've got the Chat Room already set up. All we have to do is agree to a date and time.

Comments, thoughts, and suggestions welcome.


MP

[Mad Professor]

Hi Scribe,

What pretend-money betting-level would you be comfortable at?

The reason I ask is because we could (and most definitely should) calculate the results on a very equitable dollars-bet/dollars-won basis. That way, you could bet at wager-values you are totally comfortable with, all the while knowing that each method would only be judged on a return-on-investment basis.

And if this next fact adds at all to your degree of comfort, then all the better.

Historically, the MP-$204 generates an extremely conservative average ROI of ~10%; so the performance-bar is quite low.

Edited to add: How about if Heavy reveals each new outcome as outlined above; then I commence betting whenever I want, and after charting for as long or as little as you want; you start betting then?

MP

[DeadCat]

I though I posted these questions last night but don't see it now. Here it is again:

Are Put bets allowed?
Does the Field pay 3x on the 12?

Maybe I missed it but do we know anything about the shooter?


-DC

[heavy]

Deadcat - yes on the put bets. It never even occurred to me to include the Field bet but if you want it at 3X on the 12 that's fine.

I have been reminded by a board member that this particular sub-board (the betting systems and strategy forum) is on the "public" portion of the forum. so when we actually get around to doing this project I'll probably more it to one of the "members only" forums. At this point I'm looking for a book of rolls - preferably from Tunica's craps clinic. I spoke with Lagunaflats - he's back in Texas but his roll tracking book is still in Mississippi. Any of you other guys track rolls in Tunica? Seems like I saw FreddieB writing rolls down. AP? You out there? Help us out here, guys. If you could copy or scan your roll records and e-mail them to me I can take it from there. axispowercraps@gmail.com.

[Ahigh]

Deadcat - yes on the put bets. It never even occurred to me to include the Field bet but if you want it at 3X on the 12 that's fine.

I have been reminded by a board member that this particular sub-board (the betting systems and strategy forum) is on the "public" portion of the forum. so when we actually get around to doing this project I'll probably more it to one of the "members only" forums. At this point I'm looking for a book of rolls - preferably from Tunica's craps clinic. I spoke with Lagunaflats - he's back in Texas but his roll tracking book is still in Mississippi. Any of you other guys track rolls in Tunica? Seems like I saw FreddieB writing rolls down. AP? You out there? Help us out here, guys. If you could copy or scan your roll records and e-mail them to me I can take it from there. axispowercraps@gmail.com.

I'm not sure I understand how you can compare strategies with anything other than random roll data. At the point you use real roll data you are testing whether roll data is well-matched to a strategy and you can't make any conclusions about the system independent of the roll data.

The whole purpose of random roll data is to compartmentalize the strategy for the purpose of analysis without consideration for the bias of the roll.

I may be missing something, but using real roll data isn't going to get you be able to learn anything at all about the strategy itself.

Here's 32 across with the last 300 rolls I recorded from my roll. This has a 6.66% edge for the 4 and 10 (no pressure) and a 4.00% edge for the 5 and 9. Yet by avoiding the red, I still made money in 300 rolls. This is a horrible strategy, yet it looks great with my roll data.



So I think if you want to actually compare strategies, you need to compare them with unbiased data and on a really long timescale.

And I want don't pass with no odds .. no additional bets.

[heavy]

Well, the "theory" is that when dice influencers play they should never bet on random rollers. Admittedly, many "DI's" results are little more than random themselves.

Again, theoretically an advantage player should never wager on any negative expectation proposition. Of course, we do it all the time while trying to justify our actions by adding qualifiers and triggers that prompt entry into the game. Believe me, I'm one of the world's worst when it comes to this - or world's best depending on how you look at it.

At the end of the day - anyone can write a WinCraps script and run a strategy through a couple of million decisions. WinCraps will even let you "skew" your edge. Meawhile, this whole idea is getting way too complicated.

As for the Don't Pass - I don't disagree. But it's not the only bet I'd use.