Regression Math

Setting and influencing the dice roll is just part of the picture. To beat the dice you have to know how to bet the dice. Whether you call it a "system," a "strategy," or just a way to play - this is the place to discuss it.

Moderators: 220Inside, DarthNater

dognponyshow

Regression Math

Post by dognponyshow » Thu Oct 18, 2018 1:07 pm

I’ve recently resumed a serious attempt at DI after a two year absence. Long-time DI guy all the way back to Patterson’s early PARR days, and more recently have benefitted from numerous coaching sessions with Howard over past few years. Been mostly a lurker on this site for a long time, but perhaps it’s time to contribute.

After several recent casino re-acclimation sessions, I’ve finally come to accept that my SRR-for now- has settled in at the 6.8 to 7.3 range. I’ve set a target goal of 7.5, and if anybody can help get me there, it’s Howard. With that, I’ve finally given up on the idea of a steady diet of long hands as my profit-producer, and have come to the realization that Regression is my best path. Have been pouring through many of the Regression threads here, new and old, including re-reading MP’s Regression Avoids Depression for probably the third time in the past few years. Rather than hi-jack the good ISR threads elsewhere, thought it best to start this one that’s focused on the arithmetic.

I buy-in to MP’s concept, but since he’s no longer around for questioning, here’s where I’m hung up:

In Part 21 he lays out the bankrolls needed for various ISR wagers at various SRR’s. His SRR- 7 table shows, for example, that using 6 & 8 Place at a 2:1 Steepness Ratio that one can double a bankroll of $4,138 in 326 hands. He appears to be using $5 table examples throughout the series, ending up at $6 each. So since my play is predominantly AC for now, I’ve interpreted this to mean that for a $10 table he’s stating that one can double an $8,276 BR in same timeline. Heck, I can crank out 30 hands per day in a few hours with a bowl of Wheaties and a good night’s sleep, and decent shooting conditions. So I’m sipping MP’s Kool-Aid and thinking maybe I can double a BR in 10 or 11 playing days.

I’m translating his SRR-7, 2:1 6 &8 ISR, to an average profit per hand, as he likes to describe, of $25.38 for a $10 game. $4,138 BR, doubled to $8,276 for a $10 game, across 326 hands.

I’ve tried to break down what this strategy would’ve produced over my recent sessions, totaling 588 rolls. I do see the merits, because his approach indeed would’ve produced some nice, steady profits for these sessions, instead of the break-even, up and downs I experienced. But I’m just not seeing that same profit-per-hand average he’s using.

MP was known to over-state some things. As rock-solid as his arithmetic appeared to be over those 21 mind-numbing articles, that average profit per hand part just doesn’t seem to hold up. I don’t see from my numbers how these ISR’s could double a BR in 10 or 11 days. Any thoughts as to the validity of his bankroll-double projections? Or perhaps I’m misinterpreting somewhere?

Moe Bettor
Posts: 1596
Joined: Mon Dec 21, 2015 2:31 pm

Re: Regression Math

Post by Moe Bettor » Thu Oct 18, 2018 2:19 pm

It's an ideal. Heavy's perfect casino. But it doesn't work in real life because even if you have an SRR of 7..which is huge..you just have those
PSO's and four roll hands mixed in with the 20 rollers even if you are premiere DI. Or sometimes you don't get the 20 roll hands at all for a weekend. If you don't regress, you are so far in the hole that it takes huge rolls to get you just even. In the long run..over millions of rolls maybe, MP's theory probably will get close to working..but we don't roll like that. Regression, in my opinion, is the closest thing to a sure way of making money playing craps. That and "Bring me down" when you get ahead. The big rolls are great, but you never know when you are getting one or when you are in the middle of one. Personally I hate leaving money on the table. Although sometimes I forget to use it, the three rolls and down also aids in preservation of capital. That is..if you as a shooter or another shooter does not hit one of your numbers in 3 rolls..forgetaboutit! So you say..but you missed a sixty roll hand because after three rolls the shooter went on to hit your numbers 20 times! Ok. Fair enough. But..my chip rack, though fairly small..is in the win column anyway. Yes..I don't have the $1000 others scored on that hand. I have $75 bucks. Tell me about your $1000 hand tomorrow and we'll see who is doing what.

User avatar
heavy
Site Admin
Posts: 10559
Joined: Mon Nov 22, 2010 11:46 am
Location: Texas
Contact:

Re: Regression Math

Post by heavy » Fri Oct 19, 2018 1:30 am

Yeah, Nick pretty much nailed it, and from the tone of your post I think you have a whiff of it as well. Irish would way you're understating MP's penchant for overstatement, but we'll move on beyond that.

Even thought I used a 2-1 regression for years - $24 six and eight - one hit - regress to $12 each six and eight - I struggled to make more than an average of $40 - $50 per session. Hard to grow a bankroll at that rate when it costs $70 to fill my tank. These days I tell the students in the seminar to consider doing a 5 - 1 regression. So on a $10 game you'd be running from $60 six and eight down to $12 six and eight. Of course, first you have to have sufficient bankroll to afford to put $120 on the table.

At the end of the day I'd suggest you look - not at MP's regression plays but John Patrick's regression plays. Of course, John would tell you to "screw regressions - play one hit and down on every shooter."
"Get in, get up, and get gone."
- Heavy

DanF
Posts: 546
Joined: Sat Oct 01, 2016 4:33 pm

Re: Regression Math

Post by DanF » Fri Oct 19, 2018 4:32 am

The truth is you need both to win in this game.

You have to regress and you have to press bets on big hands. Which is totally an opposed way of playing.

You have to be conscient when the table is giving less then 10 roll high hand that regression is the only way you can make money. You also have to know if you see a good % of hand you would win on to cover your losses over time.

But you also need to be aware that the big hands are the only way you can be profitable on a long term play. They need to happen or you will bleed money, so you need to have a bet on table you can build from after your regression. That will profit long term play. You need a betting scheme that takes care of it so you don’t have to think and just react. Or else you will miss out on those big wins to cover your red days and bring you profits.

Me, I know I can lose 500$ one day and be ok with it. Cuz I know I will average 300-400 profit with 1k+ days more often the I lose. It’s in my betting process/game control.

I lost 451$ my last session and had a 460$ loss this month also. But I am up 2k on the month. Cuz I won some big hands and controlled bleeding on bad ones.

I recently realised the one move that brings the most profit on 10$+ table early in the game is: up one unit aka the kelly system. But man I love to parlay bets. It brings soo much power to your game when it streaks.

It all comes back to one thing. You need to have your bets pay themselves early, depending on what table gives you.

User avatar
mssthis1
Posts: 1254
Joined: Thu Oct 23, 2014 6:59 pm

Re: Regression Math

Post by mssthis1 » Fri Oct 19, 2018 9:32 am

Lots of good info here. One thing I'd like to add is don't get hung up on SRR. I've had 720 roll books where my SRR was 9 or above where I make little profit and books around 6 or even lower than 6 where I made a boatload of money.

The difference? Come out sevens. After many thousands of rolls they blend into whatever your SRR is. In one session though they are a huge factor and often the difference between profit and loss.

dognponyshow

Re: Regression Math

Post by dognponyshow » Fri Oct 19, 2018 2:35 pm

These are the type of thoughtful responses I was hoping for, so thanks all. One thing that's dawned on me is the need to customize a betting pattern tailored to my abilities and tendencies. And even though I'm not a "true" SRR-7 just yet, I think it's a valid base point. So one thing I know, from both recent play, and previous experience that spans 20 years, is that an SRR-7, or thereabouts, shooter can expect to see a steady diet of hands in the low to mid-teens, several uneventful mid single digits, and of course the omnipresent PSO or 2 and 3 roll hands.

For me, having a rough idea of my primary and secondary hit foundation frequencies, I'm seeing the merits of something like this (based on $10 table, reality for my geography): $44 inside, take three hits to "pay for these,", then follow a one-third pressing series (Sharpshooter suggests this in his book), which translates to "up a unit," through roll 10. So let's say I catch the third inside hit at roll 5 or 6, I'm now in what I call my "power zone." In this situation, I'm clearly relaxed because I know I've covered my outlay, and I've developed a rhythm that includes good breathing.

After roll 10, I'd regress to $12 6 and 8. By this point, if I've made a PL win or two, I am typically very good on the C-O with an all 7's set of catching a nice cluster of "junk". I'll typically take a $2 any crap and rattle off a 7, a yo and one or two craps on a good C-O . So I'd invest these profits in either pressing the 6/8 after point established, or perhaps spreading to 5 and 9. From there, I'll just bang away with what typically results in a mid to upper teens roll. If I could string together a handful of these, and weather the inevitable short rolls, I envision some modest, yet more consistent sessions than I've had.

Intuitively, and replaying all these years of shooting, I think this has merits. It incorporates Regression, but at a level that suits my current comfort zone. I've found previously that delving into that ISR 5:1 zone just causes me to tighten too much at the beginning of the hand. We'll see how it unfolds.

User avatar
Raider
Posts: 418
Joined: Mon Oct 01, 2012 2:20 am

Re: Regression Math

Post by Raider » Sat Oct 20, 2018 7:59 pm

DnP, have you considered tracking you by hand SSR? I have found that I need at least 10 rolls to get my toss dailed in for that session. On my first hand of the session, its one unit DP/PL, and no other bet until the 6th roll, and the bet is a field/come combo. When the 7 comes ( and it does 80% of the time on the first point) my warm up is either free, or cost me a unit. I only shoot three hands, If I can''t do it in three, I am not going to, on that day.

Just a thought
Physics Trumps Probability

dognponyshow

Re: Regression Math

Post by dognponyshow » Sun Oct 21, 2018 8:26 pm

Yes Raider, I also typically need 8 to 10 rolls to warm-up. I do NOT count these rolls in my "official" SRR calculations, nor do I bet anything more than minimum PL with no odds. Quite often, my toss will look good from the start- proper rotation, target landing zone, no splatter. That's usually a sign of a decent session to follow. Conversely, there are some days where it takes me quite a while to get dialed in. Maybe 5 hands or more. That's when the "rent" needed on the PL for warm-up can start getting costly. It's also a sign that this just might not be my day.

That inconsistency is all the more reason I am still striving, after all these years, to find the proper betting strategy.

wild child
Posts: 1522
Joined: Thu Jul 28, 2011 8:29 pm

Re: Regression Math

Post by wild child » Mon Oct 22, 2018 1:51 pm

DnPS
You done good because
your post is of strong interest to practitioners of D I at the Craps Tables...
The post stimulated imo well thought out responses...

H R n R maintains highly accurate roll tracking ,
both in practice and belly to the table live sessions...

Howard could point you toward a Game Plan
well tailored to your future casino adventures...

Please keep us updated on you future progress with all things Dice Influence.

just me saying
w c

User avatar
Raider
Posts: 418
Joined: Mon Oct 01, 2012 2:20 am

Re: Regression Math

Post by Raider » Mon Oct 22, 2018 3:49 pm

DnPs,

I put a white to the side the PL along with the DP, rent is one buck. since I live in a one shop town, it a $10 min. too much rent for a hand that goes 6-8 rolls, LOL I have taken a page from heavys routine this go round. I train before I go to work, then run a head to head game when I get home from work. The current plan is to play when I get off work. I have decided that this training plan will included pratice games games after work and after training. train the way you fight. Like you, I am getting back to it, after a very long lay off. The lessons of the past have taught me that tracking all of my rolls is too distracting and its easier to look at the by hand ssr and the money won from the games . still need to find a covert way of tracking my rolls at the table, thats the only tracking I am intrested in at this time.
Physics Trumps Probability

User avatar
heavy
Site Admin
Posts: 10559
Joined: Mon Nov 22, 2010 11:46 am
Location: Texas
Contact:

Re: Regression Math

Post by heavy » Mon Oct 22, 2018 9:41 pm

MP suggested a 5:1 or even a 6:1 regression many moons ago.
True MP wrote about it. But like many of MP's writings - I think you'll find they had origins in other works. Look, regression theory has been around as long as there's been gambling. Many of you first encountered it via MP - much like I first got serious about it after reading John Patrick. Just saying. As much as I appreciate his writings and contributions to DI in general - it looks to me like he's closed the store front and is no longer out there. I can't blame him for that. After all, we're having the same old discussions - in many cases with the same old people - that we had damn near 20 years ago. Eventually somebody has to learn something. But hey, I'm waxing long at the end of a long day.

Bottom line - it's never wrong to take a profit off the table. EVER. Anyone who tells you otherwise just hasn't hit the wall on losses yet. His time will come.
"Get in, get up, and get gone."
- Heavy

User avatar
mssthis1
Posts: 1254
Joined: Thu Oct 23, 2014 6:59 pm

Re: Regression Math

Post by mssthis1 » Tue Oct 23, 2018 12:04 pm

heavy wrote: Mon Oct 22, 2018 9:41 pm
Bottom line - it's never wrong to take a profit off the table. EVER. Anyone who tells you otherwise just hasn't hit the wall on losses yet. His time will come.

Excellent words of advice. When I have more time I'll post some roll between sevens charts from random rolls to show the huge differences in sessions as to where the seven may or may not show up. A regression will not overcome random dice. It will allow you to have the excitement of betting larger than you probably should at times and still having action on long rolls.

If you're playing a session and have made regression several times in a row you should consider taking your small win and bail. If you don't the Sheriff will get you eventually and it takes many small wins or a roll over 20 or so to overcome a couple PSO's.

wild child
Posts: 1522
Joined: Thu Jul 28, 2011 8:29 pm

Re: Regression Math

Post by wild child » Tue Oct 23, 2018 6:07 pm

DnPS

One ploy I picked up from my days of associating with personas I met at
a Leonard Benson event from the way back ancient times ( early 1990's )...
.
Over the years it has shown profit as a go-to RECOVERY / buy in survival move...
.The beauty is the thing may adapt to tinkering
..........Tailored to Individual Appetite.....
.
AS AN EXAMPLE , for the purpose of this discussion:
.
A) Place $96 across
B) After POINT is established make serial D/C wagers( at Table $ Minimum )
C) TAKE DOWN PLACE BOX NUMBER(s)
and
$ winnings as a D/C covers that BOX NUMBER

D) Should TWO BOX NUMBERS COVERED BY D/C
.... BE LOST TO A REPEAT or one Box # lost twice......

..... PULL NOT HIT NUMBERS DOWN and make ZERO further D/C WAGERS..
Take a breath before Place BETTING any Box Numbers or MOVING ON.

In a nut sack :
This is my version of The Strategic Withdrawal
"Blended Recovery of Buy In Distress Play"

You are likely to become unpopular with
the PURE BRIGHT SIDE (majority )Right Side

as your Buy In inches into recovery and on to the PLUS SIDE OF RECOVERY...
.
The Down Side is anything may and often does occur during a craps hand
and this could lose causing you to call a halt to play at-least for a few moments.

Also with any & all game plans it may not be palatable to a number of critics

(whom may harshly critique your Buy In /Bank Roll Survival effort...)

A CAUTION : it could deflate and show at times disappointing results ....
................In REALITY WORLD ,this is occurring in a casino at a craps table
.
Should you elect to goto this life-vest during rough seas at a Dice Table
give it a number of "Test Drives " against recorded sessions
and live AT HOME PRACTICE SESSIONS....
.....numerous times prior to walking it around
......in-front of God and everyone...
......in a real live casino craps pit.....
AND
it may protect your U S D $ until
an identifiable HOT HAND SHOWS
or
a WELL IDENTIFIED PROLONGED
POLAR BEAR ICE COLD TREND VISITS THE CRAPS PIT...
.........( also a HARVEST for the PREPARED)

just me saying
w c

memo
Posts: 872
Joined: Sat Jul 23, 2011 4:47 pm

Re: Regression Math

Post by memo » Wed Oct 24, 2018 3:40 pm

Any time you have a discussion about the early writing by MP, one must put it into historical perspective. We as a DI community, have evolved. Even though it is true that it is the same old stuff, rehashed over and over, our focus has changed...As we matured, our expectations also became a bit more grounded.

SRR was the 'Holy Grail' of success. Somewhere, along the way MP even made claim to a SRR of 23, (Which, later, he tried to walk back). The consensus being that one needed SRR of 8 to really be a valid DI and you really should be moving to 9. Hence all the tables that were put forward by MP of expected results if having SRR of 7, 8, 9..Or even higher. In essence, 'The Emperor had no clothes'.

We were so enthralled by the possibilities it was easy to overlook the small problem that nobody could really do it. Attitude was..Since everyone is doing it, if I really work at it.. (practice till you puke), I ought to be able to achieve SRR 7.5 or even 8..
No one was challenging the achievability of SRR's of 8, 9, 10 and on to infinity. Buzz Lightyear was also popular at about that time.
There was one voice out there pounding away, telling us to hold on...Rethink. If this sounds too good to be true, it might be...
His initials are Irish Setter. Actually, he was a bit more blunt about it.

Don't get me wrong. I am a fan of MP..He published a huge volume of information and motivated us to continue in more ways than you can count.
That SRR rabbit hole was a misconception. We should have been talking 6.5, 6.75, 7.2. I believe it had to do with a lot of disillusionment..It could have even sparked the biased dice mess.
Additionally, we should have de emphasized it and put more focus on other parameters.

So, in reference to ISR..
Do not pay attention to the 8, 9 values.
Think 6.5-7 SRR and adjust your regression point accordingly.
Also remember that MP was always painting everything in a positive (overly) light. Use your real world practice results and war game to see if you can come out ahead. If you are too uptight with a 5:1...Go to something that fits your gut and not mathematical formulas.

Memo

User avatar
heavy
Site Admin
Posts: 10559
Joined: Mon Nov 22, 2010 11:46 am
Location: Texas
Contact:

Re: Regression Math

Post by heavy » Wed Oct 24, 2018 4:10 pm

Without getting too deep into MP bashing . . .
Somewhere, along the way MP even made claim to a SRR of 23 . . .
I believe the actual claimed number was 28. And yes, I defended that many times arguing that he was talking about a short-run SRR and that anyone can walk up to the table and toss a 28 number hand and thereby have an SRR of 28 - for that 28 rolls. Then it's do-over time.

MP also had some issues early on with the "originality" of some of his writing. Back in the original dicesetter.com days Irish received several complaints from authors about articles of theirs being . . . and I will put this delicately . . . borrowed from in great detail without proper permission or attribution . . . on Irish's website. I recall reading something he wrote once and thinking, "Man, that was great. I really liked it the first time I read it . . . right after I wrote it." But eventually he got beyond all that and became an extremely productive writer. And in that great volume of material there are certainly some real gems. But the guy you have to tip your hat to is Stanford Wong. If Wong had not been researching dice control and stumbled onto MP's article base on Irish's site then the Mad Professor's Crapshooter's Bible would never have been written. He would have faded into Internet obscurity as a guy only a few of us remember and that would have been the end of that.

It's kind of funny. About the time that Wong decided that he was going to back away from DI as a viable advantage play worth pursuing he offered to sell me a great many of MP's books for pennies on the dollar. Just before MP did his disappearing act I got the same offer from him. I suppose I should have offered to buy them by the pound. I could probably have made a few bucks selling them on Amazon as I'd probably be the single source of them at that point - unless you wanted a used one. LOL.

And now that we've thoroughly hijacked this thread . . . back to regressions.

$640 across. One hit. Go to $96 across. One hit. Go to $32 across. One hit. Take me down. Just don't get nailed by that PSO.
"Get in, get up, and get gone."
- Heavy

wild child
Posts: 1522
Joined: Thu Jul 28, 2011 8:29 pm

Re: Regression Math

Post by wild child » Thu Oct 25, 2018 10:32 am

heavy wrote: Wed Oct 24, 2018 4:10 pm
And now that we've thoroughly hijacked this thread . . . back to regressions.

$640 across. One hit. Go to $96 across. One hit. Go to $32 across. One hit. Take me down. Just don't get nailed by that PSO.
Heavy

DnPS encouraged this thread of substantial interest.
I again thank him
and thank you for directing the DnPs's post back on subject..

Returning to the original track the train of thought
that rolled in and stimulated this conversation:

Entering a hand with the craps wager game plan

The $640 exposure may stand
out of reach of a great number of the craps player community....

We know not with total surety when THE SEVEN (7) will take down a dream.

Bold concept of blanketing the six box numbers with a $ VALUE large enough
to allow ISR to insure a profit after a successful first roll
and
reducing the USD $ exposure to loss on following rolls/tosses ..
.
Were a player to stand down for the C/O toss,
would the game plan offer equal risk ?
Would it be wiser to PULL DOWN after one more roll of the dice
or hang in for a third or greater number exposure at a subsequently lesser risk ?

Just me mulling
w c

User avatar
Raider
Posts: 418
Joined: Mon Oct 01, 2012 2:20 am

Re: Regression Math

Post by Raider » Thu Oct 25, 2018 4:23 pm

If it sounds too good to be true, it is. The first time I tried the across the board ISR, the next roll was a seven. I didn't like the odds of trying it six more times, one and down, just to break even. thats aside from the fact that I did not have the bankroll to try it six more times. I have tried $85 on 5,6,8, regress on the first hit, press on the second, and try to see the field numbers coming. Like every thing else, some times it works, some times it doesn't
Physics Trumps Probability

memo
Posts: 872
Joined: Sat Jul 23, 2011 4:47 pm

Re: Regression Math

Post by memo » Fri Oct 26, 2018 9:26 am

wild child wrote: Thu Oct 25, 2018 10:32 am
heavy wrote: Wed Oct 24, 2018 4:10 pm
And now that we've thoroughly hijacked this thread . . . back to regressions.
Heavy

DnPS encouraged this thread of substantial interest.
I again thank him
and thank you for directing the DnPs's post back on subject..
w c
I do not really think we got off subject with this thread...
There were strong references to MP and John Patrick. They both discuss ISR, however, they are not the same. I believe JP will stand the test of time, however, I do not feel the same about MP....Partly because of reasons cited above..A discussion that Would take us off topic.

DnPS expressed some disillusionment in the contrast with his results vs. his expectations based on the 'regression vs. depression' articles by MP. Luckily, he is seasoned enough at this, to look for clarification. Someone that is new to DI and to this board, however, may not.

Most of us O'l Dudes here will re read MP's articles, distill the information down, take away the brilliance that is there. At the same time disregard the, Build it and they will come.. (practice and you will profit), Mr. Rogers, rosy mentality which implies that high SRR's are easily achieved, maintained, and stable. Furthermore, one only need to plug himself into the formula to see what amount of profit you will steadily receive from the casinos.

Anyone can read a JP book and put his ideas into practice immediately...I do not think the same is true of MP...At least, not without putting a significant bank roll at risk. Caution advised. JP wrote for the random game, MP wrote for DI hopefuls.

One way or the other, I have incorporated MP's teachings into most all of my craps philosophy. But any discussion that points to and gives credence to his writing while not giving insight to his over optimistic viewpoint can be a bit irresponsible.

Memo

wild child
Posts: 1522
Joined: Thu Jul 28, 2011 8:29 pm

Re: Regression Math

Post by wild child » Fri Oct 26, 2018 2:35 pm

Memo
Thank you ..Points well taken....

In the mind of me J P slanted more so on
the making wager part of the craps game
( with hopes of a $ U S D win than emphasis (if any )
to actually touching/tossing the cubes.....
.
An aspiring player could view the videos
&
read his books (especially the Advanced )
and
in a relatively short time span function somewhat............
in a live session at an actual craps pit possibly loosing small
or walk away with a POSITIVE USD $ experience.....

Consistent repeatable positive experience when actually
tossing an educated shoot generally requires learning, coaching
and
the dedication to practice to get it somewhat right more than tossing failures..
which takes a bit greater time and effort ..

To lift a line from Star-trek : Live Long & PROSPER

Just me saying
w c

DanF
Posts: 546
Joined: Sat Oct 01, 2016 4:33 pm

Re: Regression Math

Post by DanF » Sat Oct 27, 2018 4:52 am

640 across one hit regress will work, but to play it you need 30-60k BR & a next to perfect toss, which means it ain’t no BR growing plan.

154 inside or 134 even is a good start with a 5-10k roll.
What did good to me tho was downtown/uptown 144.

Collect 98 & down to 44 inside. Fullpress first hit & collect twice. Then to what moon your throw takes you ;)

Post Reply