Short Term SAP

[XDN]

On 11/10/2016 at 5:37 AM, McVince said:

SAP needs to be re-vamped.   No one will see the true value of this tool unless old theories are thrown out and new, more innovative theories are brought in........Find the guys that you can trust and do it as a team. 

McVince since you wrote this a few days ago i've been giving some thought to this subject and IMHO I think you are right.

I've come up with an SAP idea (not proven yet) that more closely follows what we are discussing on the board these days. We all know that the shoe biases don't last, worse, many times they change extremely fast, within 3-8 events. Thats that is the main reason for the "Follow The Shoe" / Short Term NOR / Last 3 Event Methods development. But we haven't done this to SAP yet. My "Short Term SAP" thoughts are as follows....

1. Reset SAP Counts on a loss trigger (say 3 LossIAR on the bias you are playing) or on an obvious shoe change (SS goes ZZ).

2. Change the count method from addition to multiplication for short term tracking. Since this would be a short term method we need to generate a MC/LC spread quickly, within just a 2-4 events. This will make the numbers bigger quicker, so it might be unwieldy if a bias lasts the whole shoe but really it shouldn't be....the numbers will just be bigger. We could multiply by the existing count instead of simply adding the counts. I think this works logically because if we are tracking/looking for a concentrated set of events in a short term "set" of events. In other words if we see  2122  2's are WAY over represented in those 4 events, we need a multiplier to build a spread quickly to represent that overage of 2's in that small set of events. Example using 2122

First       2 Event = SAP Count = 2   (Starting Point)

****Ignore the 1 event for this example***************

Second  2 Event = SAP Count = 4   (Previous SAP Count(2) X SAP Multiplier(2))

Third      2 event = SAP Count = 8   (Previous SAP Count(4) X SAP Multiplier(2))

etc.....

Note... We may have to change the multiple for 1s as multiplying 1x anything is always 1, we would never get a multiplier effect 

This would be a companion SAP method to FTS. I'll try to get some testing in on this in the little free time I have. If anyone wants to test some shoes to see if this works, please do so.

Edited November 17, 2016 by XDotNet

[brad01]

will check it out and let you know

[XDN]

34 minutes ago, brad01 said:

will check it out and let you know

I'm going to try to get a few shoes in this weekend as well to test this.

[wolfat]

I am obviously interested in this but before start some testing IMO it should be better fix some rules.

I suggest to check the sap ct IE, on the last 4 events instead resetting the count, if an event is still hot throughout the whole shoe we can play it, if it cools down, we can switch easily, if the 4 events are tied, you can play reverse sap, etc

I'd also discuss about weighting or not events. Currently I don't do this, simply put an event in relationship with bigger events, you can do this with a blink of an eye, no need tracking and keeps things easier.

I usually divide my sap card in sections of 4 events, a MC event, to be played, need to be within the last 4 ev, if it's not RECENT better wait, IMO

Thoughts?

A

[ECD]

There are ongoing discussions to disclose my version of SAP.  I will say that in my opinion SAP does not need to be changed "Much".  What will need a huge change is on HOW IT IS PLAYED.  A problem with SAP which was pointed out by some of our talented players is that SAP just shows you what has happened in the past.  I guess SAP worked well when there was regular cards.  Unfortunately with pre-shuffled cards, the shoes that are totally biased are few and far between.  I use SAP for future events that have to happen.  I also use SAP in addition to NOR ( Another obsolete system according to another forum).  I have said this before, SAP should be used IN CONJUNCTION with whatever method you are playing.  It really is a great tool.  

I use a 24 shoe cycle.  In the past five cycles which includes live play, online play (The most of all) and old shoes (Yes I practice every day), I have an average of a  83.5% win rate over the five cycles.  

I am not disclosing without the consent of the administrators.  

Edited November 20, 2016 by ECD
Adding information

[wolfat]

54 minutes ago, McVince said:

There are ongoing discussions to disclose my version of SAP.  I will say that in my opinion SAP does not need to be changed "Much".  What will need a huge change is on HOW IT IS PLAYED.  A problem with SAP which was pointed out by some of our talented players is that SAP just shows you what has happened in the past.  I guess SAP worked well when there was regular cards.  Unfortunately with pre-shuffled cards, the shoes that are totally biased are few and far between.  I use SAP for future events that have to happen.  I also use SAP in addition to NOR ( Another obsolete system according to another forum).  I have said this before, SAP should be used IN CONJUNCTION with whatever method you are playing.  It really is a great tool.  

I use a 24 shoe cycle.  In the past five cycles which includes live play, online play (The most of all) and old shoes (Yes I practice every day), I have an average of a  83.5% win rate over the five cycles.  

I am not disclosing without the consent of the administrators.  

...one sample?

thanks

A

[XDN]

18 hours ago, wolfat said:

I am obviously interested in this but before start some testing IMO it should be better fix some rules.

I suggest to check the sap ct IE, on the last 4 events instead resetting the count, if an event is still hot throughout the whole shoe we can play it, if it cools down, we can switch easily, if the 4 events are tied, you can play reverse sap, etc

I'd also discuss about weighting or not events. Currently I don't do this, simply put an event in relationship with bigger events, you can do this with a blink of an eye, no need tracking and keeps things easier.

I usually divide my sap card in sections of 4 events, a MC event, to be played, need to be within the last 4 ev, if it's not RECENT better wait, IMO

Thoughts?

A

Pretty please? 

[Pando]

I would love to see it to, please

[wolfat]

ok, here is a sample of my sap chart.

I track a line every 4 events, on avg you should have 2110 or 2101 or something... It helps you to stay focused on RECENT EVENTS.

#1 is classic one but with events NOT weighted. you must put the event you are interested in relationship with bigger events to realize how is running.

#2 is the one I currently use, it fits my needs. 1s are only circled, I want only know if they come in multiples or singles.

2s and 3s work together and show me how OTBL is going.

Groups of 4 or more are separate, I want to know how long they are and in what side they fall.

This is it! 

I find it helpful in tracking the shoe, you must use the data to improve your play.

A

 

[avion]

Thank you wolfat for sharing this "perspective" of how you created and use this Sap chart.  In my humble viewpoint, it is easier to see the flow of the shoe and helps to clearly see what the shoe is doing and isn't doing.  

[wolfat]

17 hours ago, avion said:

Thank you wolfat for sharing this "perspective" of how you created and use this Sap chart.  In my humble viewpoint, it is easier to see the flow of the shoe and helps to clearly see what the shoe is doing and isn't doing.  

...this is the main use for it!

A

[LOU30]

A picture speaks a thousand words! Thanks Wolfat that explains your last few post perfectly.

[ECD]

Very nice.   Again, everyone should be good enough on this forum to take whatever method of play that fits their eyes and make it there own.  

 

Kudos

[Pando]

I want to add another dimension to the SAP discussion.

Here is a shoe I played yesterday;

P6341131115821211431132151111

Its a great shoe for me (which is why I played it in the first place).

My trigger was the 6IAR at the start, I knew I was in for some good profit. $$$

But that's not why I am writing this

Lets split out the Player and Banker sides

P641118114131111

B31315221312511

The weighted SAP for the whole shoe is 16/6/16/24

You can see it has low 2's, nothing unusual there

Now lets do a weighted SAP for each side

P 10/0/4/16

B 6/6/12/8

By splitting the shoes into separate sides with separate SAP's I can see the trends/bias much more clearly than if they are in one running SAP count

For example, the Player side has NO 2's but had more 4IAR+ than banker

Conversely the Banker side had all the 2's and more 3's, but was lower in 1's and 4IAR+

Its true sometimes we can see this information on the tote board, but splitting the running SAP count, it helps me anyway.

I now do it routinely on my scorecard and combining it with splitting the shoe into P & B, I easily see the patterns in the shoe

Some of those patterns in this shoe (with MDB+ in mind) are for example, the single 1's in Banker, and the 2 x 3's going to 4 in Banker

You get the idea. I have not done any testing on the veracity of playing MDB+ on each side but it looks promising. I suspect that the data is stronger on each side rather than the shoe as a whole. But that is yet to be proven

regards Pando

[avion]

Great post Pando.....you are really progressing in your observation skills of reading the shoe.  

[wolfat]

How I said before, often the two sides tell different stories, chose the tell you like more...

PANDO: this is a shoe I played two days ago, nothing more than a 3 iat, thinking about you! LOL

A

[pauld]

Hi, I thought I would contribute to the Short-Term SAP discussion – another way to look at it.  I too thought SAP was of value, but fail to see how events 40+ hands ago impact what is happening now.  So I took the usual weighted SAP count and using the 4D-Analyser-LivePlay (B2) spreadsheet, tried various ideas.  I was disappointed with the results, and then someone on this forum mentioned not using the weights, just use the number of events, that has produced better results. 

If you look at the attached file, NMCexample.pdf (NMC stands for ‘Near MC’) I added some columns – you will see that columns AR, AS, AT, and AU are the number of events (1s, 2s, 3s, 4+s).  I added columns AV, AW, AX and AY – these are the number of events MINUS a number we ‘look-back’ to – in this instance it is looking back 7 hands and subtracting what the number of events were 7 hands ago.  For example, hand 22 the events are 9-1-3-1, but if you subtract 7 hands back, you get 3-0-2-1, in this instance the MC and NMC are the same – but not always – you can see from the last hand, that betting flat 1 unit bets, that MC ended +4, while NMC ended +7. 

I also looked at LC but cannot figure out a way for it to be of value – except maybe to bet opposite!

I analyzed different ‘look-back options’ looking back from 3-5-7-9-12 hands back to subtract from the MC and it seems that looking back 7 hands and subtracting is the best way to go. 

However, one thought I have had is that perhaps there is a different best ‘look-back number’ for each of the 1-2-3-4+ in-a-row groupings, instead of looking back 7 hands for all of them – not sure how to research that.

How is it doing?  I've only ran 55 shoes, of varying length, with no rules the NMC is averaging 1 unit per shoe on the 1 unit bets so far.  Using a 2-3-4 negative progression with limit of loss max to -8 per shoe, averaging 3.2 units per shoe. Probably would do better if I didn’t let the losing shoes go to -8.  Some MC/NMC rules that are valid: if MC by itself, don’t bet, if MC and NMC say same, bet it, NMC by itself, bet it. 

I have no idea how this idea could be implemented in the casino environment – too much numbers to play with.  So, just another idea to think about regarding near-term SAP and the value of that.

Maybe Keith can have this data/concept incorporated into his master file so it could be examined for a lot of shoes?

My perspective on shoes: I always say that every shoe has a couple to a few ‘patterns’ in it (like SS groupings), or a NOR pattern, and maybe it is a repeating pattern or there are different patterns in the shoe, with ‘chaos’ in between the pattern sets. There won’t be a lot of patterns, 2 to 5 or so.  Goal is to figure out what pattern you might be in or going into, how long it will last, how long the chaos will last.

Just another contribution.

Paul

 

 

RESULTSv30.pdf

NMCexample.pdf

[wolfat]

Paul, great job!

I must admit that didn't understand everything but it seems a great analysis of SAP.

I like the idea to look for MC and NMC matching, maybe it's a good way to go.

 

"it seems that looking back 7 hands and subtracting is the best way to go"

I didn't make your deep analysis but, in a previous post on this argument, I speak about looking for an "updated" SAP count on the last 4 events, that's equal to 8 hands on avg.

I also think that a MC event to be playable must be RECENT. I like the idea of matching it with pure MC event.

A

[pauld]

Yes, your 'looking back' 4 events is very similar to the looking back on the SAP numbers (I guess they are really only event numbers). Basically, if 7 hands ago, the 1-2-3-4+ event numbers were 4-2-2-2 and this hand they are 9-3-2-2, you subtract the counts from 7 hands ago and the Near MC count is 5-1-0-0.

[Pando]

On ‎28‎/‎11‎/‎2016 at 3:44 PM, wolfat said:

How I said before, often the two sides tell different stories, chose the tell you like more...

PANDO: this is a shoe I played two days ago, nothing more than a 3 iat, thinking about you! LOL

A

Yes you did say that Wolfat and I thank you for it

It has added a new dimension to my game, its well worth playing that way