Author Topic: We can go far from any expectation  (Read 622 times)

Jesper

  • Hero Member
  • ******
  • Posts: 1141
  • Thanked: 549 times
  • Gender: Male
We can go far from any expectation
« on: January 29, 2017, 11:56:49 AM »
We can not be sure to win regardless of the expectation. In a finite number of spins we can still be far from the theoretical.

I have many times seen terrible  skewed outcomes. They are terrible if it goes against us, and often when it is with us, we bet low, and opposite  when against.

I played  434 spins with target to make 100. It went rather good except from a part when my dozen did not show up more than about once in 20 spins, total in a part of play 7 in 100. (it would not help to change this can allways happen regardless of bet selection).

I did use a very conservative progression, which did not put too hard at the bank. This time the played dozen come well back and it was possible to ride it out.  I accepted I was going far behind (in units it was just cent play), and reset to one unit. About 150 spins later it was on track. Still the win compare to turnover was not  bad around 5%. My highest bet was 20, most of the time between 1 and 3 (average about 5).   I went flat  one for the first 50 spins.

If we play a negative progression, the casino do opposite. If we play one straight up the casino bet against as we should cover 35 numbers. The difference is the casino has an edge, and in a large set of play (and many players

spread out the carpet) wins. The player can win and that is due to the outcome are in favour over the expection, which happen every so often in a narrow space.
« Last Edit: February 12, 2017, 09:40:59 AM by kav »


 
The following users thanked this post: kav, Reyth

Bayes

  • Veteran Member
  • ****
  • Posts: 653
  • Thanked: 536 times
  • roulettician.com
Re: We can go far from any expection
« Reply #1 on: January 29, 2017, 01:04:12 PM »
I have many times seen terrible  skewed outcomes. They are terrible if it goes against us, and often when it is with us, we bet low, and opposite  when against.


True, but if we bet for those terribly skewed outcomes to level off (which they always do - it cannot be otherwise), it reduces the strain on our bankroll. I've been doing it for years with great success (and it's not gambler's fallacy either).

Just finished a session in which I made 63 units and no stake over 6 units, which is fairly typical.

I bet the even chances because the swings are manageable that way. Betting higher odds results in longer drawdowns with attendant frustrations.



For the best ways to attack these kinds of sequences see Sputnik's CPM threads.
 
The following users thanked this post: Reyth

kav

  • www.Roulette30.com
  • Administrator
  • Hero Member
  • *****
  • Posts: 1657
  • Thanked: 698 times
  • Gender: Male
Re: We can go far from any expectation
« Reply #2 on: February 12, 2017, 09:42:25 AM »
Bayes,
Would you be so kind as to explain how you do this?
 
The following users thanked this post: Reyth

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3318
  • Thanked: 979 times
Re: We can go far from any expectation
« Reply #3 on: February 13, 2017, 03:37:53 PM »
Btw, that dozen can even go double what it did.  I have seen 40 and simulated 41 consecutive misses.

I know you already know this Jesper because you are the king of skewed results!


Jesper

@Bayes:  I agree wholeheartedly, even with Sputnik's CPM!
« Last Edit: February 13, 2017, 05:32:35 PM by Reyth »
 

Bayes

  • Veteran Member
  • ****
  • Posts: 653
  • Thanked: 536 times
  • roulettician.com
Re: We can go far from any expectation
« Reply #4 on: February 14, 2017, 01:11:14 PM »
Bayes,
Would you be so kind as to explain how you do this?

Kav, in a nutshell, I bet for strong deviations to level off and constantly switch from one EC to another, one pattern to another, and use a mild progression. There really isn't anything I'm not telling you, no secrets I'm holding back. That's all there is to it.

The software tracks 1024 bet selections for each EC and orders each one every spin by its performance. It selects the worst performing and automatically switches when there have been a number of wins, which ensures you will never be chasing losses for too long (unless you check the "Fix Ptn", in which fixes the current pattern until it's unchecked).

The radio buttons L2, L3, L4 let you switch between different "types" of deviation. For example, L3 searches for the worst performing BS in terms of the number of spins since a double win, so for H/L the number 32 in 10/32/10 means that the selection hasn't achieved a double win for 32 spins. L2 represents a BS which has gone for 10 spins without at least 2 hits in the last 8 spins. Can't remember what L4 counts, but you get the idea. Higher counts mean stronger deviations. There isn't just one way to define "losing" (or winning); one BS might be doing "well" in terms of one criteria but "badly" in another (although sometimes they do coincide so e.g. L2 and L4 may give the same pattern).

The key principles are diversification + betting at the "limits". And by the way, I've tried betting with the "trend" (downwards) and the results are not nearly as consistent.

 
The following users thanked this post: kav

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3318
  • Thanked: 979 times
Re: We can go far from any expectation
« Reply #5 on: February 14, 2017, 03:44:36 PM »
The radio buttons L2, L3, L4 let you switch between different "types" of deviation. For example, L3 searches for the worst performing BS in terms of the number of spins since a double win, so for H/L the number 32 in 10/32/10 means that the selection hasn't achieved a double win for 32 spins.

Nice!  I have simulated 55 consecutive spins until a double win, did you get that number too?

Quote
L2 represents a BS which has gone for 10 spins without at least 2 hits in the last 8 spins.

10 spins without at least 2 hits in the last 8 spins, would be something like LWLLLLLLLL, or WLLLLLLLLL or WWLLLLLLLL so isn't it not necessary to count 10 spins, but only 8?

 

Bayes

  • Veteran Member
  • ****
  • Posts: 653
  • Thanked: 536 times
  • roulettician.com
Re: We can go far from any expectation
« Reply #6 on: February 14, 2017, 05:05:16 PM »
Nice!  I have simulated 55 consecutive spins until a double win, did you get that number too?

I haven't done any simulations but the worst I've ever seen (on BV) was 82.  :o

That's 5.23 Standard deviations below the mean, which is pretty consistent with results I've seen and simulated for any bet on the table (approx 5 std).

Quote
10 spins without at least 2 hits in the last 8 spins, would be something like LWLLLLLLLL, or WLLLLLLLLL or WWLLLLLLLL so isn't it not necessary to count 10 spins, but only 8?

I probably didn't explain this very well. The count works on a  rolling basis. So e.g.

L
L
L
W
L
L
L
L start count < 2 W in last 8
L +1
L + 2
L + 3
W +4  - still only 1 win in last 8
L +5
L +6
L +7
L +8
L +9
W reset count (at least 2 in last eight)

Actually I think it might be at least 2 in the last 6 spins, not 8. Can't remember and I've lost the source code for this program.

It's not the only software I use but the principles are the same for all.
 
The following users thanked this post: kav, Reyth

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3318
  • Thanked: 979 times
Re: We can go far from any expection
« Reply #7 on: February 21, 2017, 09:19:59 AM »
I have many times seen terrible  skewed outcomes. They are terrible if it goes against us, and often when it is with us, we bet low, and opposite  when against.


True, but if we bet for those terribly skewed outcomes to level off (which they always do - it cannot be otherwise), it reduces the strain on our bankroll. I've been doing it for years with great success (and it's not gambler's fallacy either).

Just finished a session in which I made 63 units and no stake over 6 units, which is fairly typical.

I bet the even chances because the swings are manageable that way. Betting higher odds results in longer drawdowns with attendant frustrations.



For the best ways to attack these kinds of sequences see Sputnik's CPM threads.

@Bayes:

1) Isn't this like trigger betting?
2) Aren't you concerned that waiting until a certain extreme event occurs, just makes the chances of the event continuing for a longer period more likely?
3) Are the odds of hitting greater after an extreme event has occurred?
 
The following users thanked this post: kav

Bayes

  • Veteran Member
  • ****
  • Posts: 653
  • Thanked: 536 times
  • roulettician.com
Re: We can go far from any expection
« Reply #8 on: February 22, 2017, 11:34:56 AM »
1) Isn't this like trigger betting?

I suppose so, although it's a "fuzzy" kind of trigger.  :)

Quote
2) Aren't you concerned that waiting until a certain extreme event occurs, just makes the chances of the event continuing for a longer period more likely?

It doesn't make it more likely, but it's true that by seeking out strong deviations you will definitely find them. You see more 4-5 std events than when following the "standard" even chances, but the approximate "limit" is still around 5 std. I don't immediately start a negative progression after the trigger, but wait for a hit or two (depending on the particular pattern being tracked) and only then progress cautiously.

Quote
3) Are the odds of hitting greater after an extreme event has occurred?

Not mathematically - the odds always remain fixed, but the rationale for starting an attack after 3+ std is that the potential losing run cannot go on for as long as when you're starting to bet at some random entry point.
 
The following users thanked this post: kav, Reyth

MrPerfect.

  • Veteran Member
  • ****
  • Posts: 642
  • Thanked: 398 times
Re: We can go far from any expection
« Reply #9 on: February 22, 2017, 11:53:22 AM »
 Somehow people tend to do " unmotivated" triggers.
   Trigger itself should be an event we can measure or rely upon.  For event to be qualified as a trigger , it has to posses sertain qualities:
  1. Being logically motivated.
 Bird farts in other continent are definitely affect roulette outcome on where you play, buy no reliable way to measure them or correlate to final result.
 2. Being statistically significant.
      Whatever your trigger is, it has to correlate to ball behavior or final result by the degree wich is beyond the doubt. 
 3. Being consistent in its affect of outcome.
      Need to form hypotesis priory or based on data collected previously,  to verify trigger in future data taken. Easiest way to test your triggers is to separate your sample in many samples and crossest them...  bigger part could be used for hypotesis formation and smaller for testing. There are lots of possibilities.
  4. Being not random. Need to verify if other situations in game can produce same or similar result without relying on your trigger.
    If all of these conditions satisfied, congratulations!!! You got yourself a trigger ;).
 
The following users thanked this post: Reyth

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3318
  • Thanked: 979 times
Re: We can go far from any expection
« Reply #10 on: February 22, 2017, 11:55:33 AM »
The rationale for starting an attack after 3+ std is that the potential losing run cannot go on for as long as when you're starting to bet at some random entry point.

YES!  See THIS is what I am talking about.  THIS is the secret to beating roulette!!!

Identifying and finding a way to consistently dwell with these extreme events, limits our negative exposure!

Everyone has thought of "wait for 10 reds in a row" and bet black but this fails because:

1) 10 reds in a row is insufficient to approach the worst the wheel will bring (not extreme enough)
2) Waiting for 10 reds in a row can turn you into a skeleton with cobwebs

.
Female roulette player waiting for 10 red
« Last Edit: February 23, 2017, 08:28:49 PM by Reyth »
 

kav

  • www.Roulette30.com
  • Administrator
  • Hero Member
  • *****
  • Posts: 1657
  • Thanked: 698 times
  • Gender: Male
Re: We can go far from any expectation
« Reply #11 on: February 23, 2017, 05:16:10 PM »
Totally agree with Bayes.
My problem is a practical one.
Would it be possible to identify this kind of deviations in a brick and mortar casino environment?
The good thing in "real" casinos is that you have many tables at your disposal at the same time so more chances to develop extreme events.
The problem is that tracking is harder in comparison to sitting in  in your home in front of your pc with your tracking programs.
« Last Edit: February 23, 2017, 08:29:51 PM by kav »
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3318
  • Thanked: 979 times
Re: We can go far from any expectation
« Reply #12 on: February 23, 2017, 08:33:04 PM »
Totally agree with Bayes.
My problem is a practical one.
Would it be possible to identify this kind of deviations in a brick and mortar casino environment?
The good thing in "real" casinos is that you have many tables at your disposal at the same time so more chances to develop extreme event.
The problem is that tracking is harder in comparison to sitting in  in your home in front of your pc with your tracking programs.

This is so true.  I know in my heart that I will never be able to go inside a real casino and play and there is no way for me to ever make it so I will and still have the same statistical luxury that the software affords. 

It feels like I am vulnerable because I am limiting my choices but I would rather have that and be certain of the statistics I am facing.

There are "real" casinos online too, so I am not totally out of the game. :D
 
The following users thanked this post: kav, december