And you thought you knew about Oscar!

[catalyst]

hi George
please give us some thoughts regarding bet selection for this progression. in the bet selection section. you have discussed lots of diferent types. i was thinking to use the bet selection you mentioned in your 'Tour de Force" thread. which one you are using for testing this progression? please tell us, so we can use the same one to compare our results with yours to find consistency in the progression.

at the end i want to say that you are the only one in the forum contributing heavily and maintaining the colourful attraction of Even Chances!
may god bless you and your family!

thanks
catalyst

[GLC]

I have been working on Horror session 3 that Bayes gave us.  It's not pretty at all.  It does well in the beginning, but then it goes into a stretch with 111 losses and only 49 wins.  That's a 30% to 70% win/loss ratio.  This system could handle that if there were some heavy concentration of wins, but unfortunately the wins are interspersed without any good clusters so we just keep digging a deeper and deeper hole.


If Fripper's "beating roulette with math" labby system can beat these horror sessions, I have to tip my hat to that system.  It's well worth learning.


If you play this Oscar system, you must set a realistic stop-loss.  I would say 300 units is reasonable.


These horror sessions are based on picking a single color (or any of the even chances) and betting on it exclusively.  Definitely not the way I would ever play the even chances.


I still play basically follow-the-last but switch to chops when I get 3 chops in a row.  This can have some streaks of losses also, but it also can find streaks on either color and also chops.  The most I've ever lost in a row with this bet selection method is 13 once and 12 once.  The rest have been under 10 losses in a row.


It's discouraging testing it against these really bad sessions.  It makes you lose confidence, but better to face reality than to get killed.


I wonder how the Whitaker progression Lucy and XXVV are working on would handle a 30/70 win to loss ratio?


I have tried to attach the file for your observation.  You can see that we are already down more than 1000 units and from checking out the upcoming spins, it looks like we're going to  drop another 1000+ before we get to a series that might have enough wins to pull us back to even.


I'm just putting out data as I can so we can all make the best decision we can.  I think these killer streaks are pretty rare, but when they hit, they can take the wind our of your sails.


Might be a good idea to go virtual after bad win to loss ratio starts and let it play itself out before we continue.  This isn't a foolproof method to handle it, but it can't hurt.


I think looking at some bet selection methods other than betting every spin with a mechanical selection method might keep our win/loss ratio much close to normal.  Pattern 4, Icey's dimension system.  Tomla021 told me that in his tests with Icaman's system he got a lot of clumping of wins.  That's a good thing with the method. 


Keep looking up.


GLC


I don't seem to be able to open my own attachment.  If you guys are having trouble, let me know and I'll just copy it in the next post.

[monaco]

hi George

on Bayes horror session 3, where you start getting in trouble (you actually noted down a 19 a loss when it should have been a win but i've corrected it in the example below), you could try the following -
after 2 losses, start a parallel Oscars on the opposite e/c. & bet the differential. In your example it would look like this:

    10   1   1   -1
    22   1   2   -2           
    4     1   3   -3            1 0 +1       (nobet -2: actual bet & running total)
1         1   3   -2            2 1 -1        (bet1b -3)
    32   2   4   -4            2 1  +1       (nobet -3)
    6     2   5   -6            3 1 +4       (bet1b -2)
3         2   5   -4            4 2 0        (bet2b -4)
    20   3   6   -7            4 2 +4       (bet1b -3)
3         3   6   -4            5 3 -1       (bet2b -5)
  0       4   7   -8            5 4 -6        (bet1b -6)
    10   4   8   -12          5 4 -1         (bet1b -5)
    8     4   9   -16          6 4 +5        (bet2b -3)
19       4   9   -12         7 5 -2          (bet3b -6)
    10   5   10  -17         7 5 +5         (bet2b -4)
    20   5   11  -22         8 5 +13       (bet3b -1)   
    24   5   12  -27         9 5 +22       (bet4b +3)
    End: +3u

the totals look a bit funny as the zero comes - betting the differential meant that when it came, you were only betting & losing 1 unit there instead of the 9 if you were betting the full amounts on both e/c's, but the Oscar numbers should carry on as normal. Betting the differential also keeps the bets even lower than a normal Oscars.

I've tried this on a few of Bayes' other sessions & its ok so far but we'll see..

lot of different variations you could try, such as start running the parallel Oscar after 4 losses, those 4 losses could be your 1-2-4-8 marty maybe, i'm trying running them both at the same time from the off as well.

[GLC]

hi George

on Bayes horror session 3, where you start getting in trouble (you actually noted down a 19 a loss when it should have been a win but i've corrected it in the example below), you could try the following -
after 2 losses, start a parallel Oscars on the opposite e/c. & bet the differential. In your example it would look like this:

    10   1   1   -1
    22   1   2   -2           
    4     1   3   -3            1 0 +1       (nobet -2: actual bet & running total)
1         1   3   -2            2 1 -1        (bet1b -3)
    32   2   4   -4            2 1  +1       (nobet -3)
    6     2   5   -6            3 1 +4       (bet1b -2)
3         2   5   -4            4 2 0        (bet2b -4)
    20   3   6   -7            4 2 +4       (bet1b -3)
3         3   6   -4            5 3 -1       (bet2b -5)
  0       4   7   -8            5 4 -6        (bet1b -6)
    10   4   8   -12          5 4 -1         (bet1b -5)
    8     4   9   -16          6 4 +5        (bet2b -3)
19       4   9   -12         7 5 -2          (bet3b -6)
    10   5   10  -17         7 5 +5         (bet2b -4)
    20   5   11  -22         8 5 +13       (bet3b -1)   
    24   5   12  -27         9 5 +22       (bet4b +3)
    End: +3u

the totals look a bit funny as the zero comes - betting the differential meant that when it came, you were only betting & losing 1 unit there instead of the 9 if you were betting the full amounts on both e/c's, but the Oscar numbers should carry on as normal. Betting the differential also keeps the bets even lower than a normal Oscars.

I've tried this on a few of Bayes' other sessions & its ok so far but we'll see..

lot of different variations you could try, such as start running the parallel Oscar after 4 losses, those 4 losses could be your 1-2-4-8 marty maybe, i'm trying running them both at the same time from the off as well.

excellent post monaco!  I think you have a safety brake there that improves this idea a bunch.  That's one of our objectives, to be able to survive through the worst series of spins we can expect to have thrown at us.

Good on ya mate!

I'm a little tied up right now, but will do a few tests when I can.  It's important for each of us to run tests on ideas so we know exactly what we're getting into.

GLC

[monaco]

another quick example just to illustrate on Smee's earlier sample, this time starting after 4 losses..

Win/Loss     Bet     Losses     Running Total

     L              1           1                  -1
     L              1           2                  -2
     L              1           3                  -3           
     L              1           4                  -4           
     L              1           5                  -5            1  0 +1         (nobet,-4: ActualBet,R.Total)
     L              1           6                  -6            2  0 +3         (1b,-3)
     L              1           7                  -7            3  0 +6         (2b,-1)
     W             1           7                  -6            3  1 +3         (2b,-3)
     W             2           7                  -4            3  2 0           (1b,-4)         
     L              3            8                 -7            3  2 +3          (nobet,-4)
     W             3           8                  -4            4  3 -1          (1b,-5)
     L              4            9                 -8            4  3 +3          (nobet,-5)
     W             4           9                  -4            5  4 -2          (1b,-6)
     L              5           10                -9            5  4 +3          (nobet,-6)
     W             5           10               -4             6  5 -3          (1b,-7)
     L              6           11                -10           6  5 +3          (nobet,-7)
     L              6           12                -16           7  5 +10        (1b,-6)         
     W             6           12               -10           8  6 +2           (2b,-8)
     L              8           13                -18           8  6 +10        (nobet,-8)     
     L              8           14                -26           9  6 +19        (1b,-7)
     L              8           15                -34           10 6 +29       (2b,-5)
     L              8           16                -42           11 6 +40       (3b,-2)
     L              8           17                -50           12 6 +52       (4b,+2)

End. +2u, highest bet 4u, highest drawdown -8

you could have made that last bet 3u if you just wanted to aim for the +1.

[GLC]

another quick example just to illustrate on Smee's earlier sample, this time starting after 4 losses..

Win/Loss     Bet     Losses     Running Total

     L              1           1                  -1
     L              1           2                  -2
     L              1           3                  -3           
     L              1           4                  -4           
     L              1           5                  -5            1  0 +1         (nobet,-4: ActualBet,R.Total)
     L              1           6                  -6            2  0 +3         (1b,-3)
     L              1           7                  -7            3  0 +6         (2b,-1)
     W             1           7                  -6            3  1 +3         (2b,-3)
     W             2           7                  -4            3  2 0           (1b,-4)         
     L              3            8                 -7            3  2 +3          (nobet,-4)
     W             3           8                  -4            4  3 -1          (1b,-5)
     L              4            9                 -8            4  3 +3          (nobet,-5)
     W             4           9                  -4            5  4 -2          (1b,-6)
     L              5           10                -9            5  4 +3          (nobet,-6)
     W             5           10               -4             6  5 -3          (1b,-7)
     L              6           11                -10           6  5 +3          (nobet,-7)
     L              6           12                -16           7  5 +10        (1b,-6)         
     W             6           12               -10           8  6 +2           (2b,-8)
     L              8           13                -18           8  6 +10        (nobet,-8)     
     L              8           14                -26           9  6 +19        (1b,-7)
     L              8           15                -34           10 6 +29       (2b,-5)
     L              8           16                -42           11 6 +40       (3b,-2)
     L              8           17                -50           12 6 +52       (4b,+2)

End. +2u, highest bet 4u, highest drawdown -8

you could have made that last bet 3u if you just wanted to aim for the +1.

Monaco,


This may be the tweak of the century. 


I know that there has to be a way for this to lose.  It looks like it's a lot of WLWLLWLWL, the same thing that hurts the original Oscars.  Is this the way you see it?


G

[GLC]

another quick example just to illustrate on Smee's earlier sample, this time starting after 4 losses..

Win/Loss     Bet     Losses     Running Total

     L              1           1                  -1
     L              1           2                  -2
     L              1           3                  -3           
     L              1           4                  -4           
     L              1           5                  -5            1  0 +1         (nobet,-4: ActualBet,R.Total)
     L              1           6                  -6            2  0 +3         (1b,-3)
     L              1           7                  -7            3  0 +6         (2b,-1)
     W             1           7                  -6            3  1 +3         (2b,-3)
     W             2           7                  -4            3  2 0           (1b,-4)         
     L              3            8                 -7            3  2 +3          (nobet,-4)
     W             3           8                  -4            4  3 -1          (1b,-5)
     L              4            9                 -8            4  3 +3          (nobet,-5)
     W             4           9                  -4            5  4 -2          (1b,-6)
     L              5           10                -9            5  4 +3          (nobet,-6)
     W             5           10               -4             6  5 -3          (1b,-7)
     L              6           11                -10           6  5 +3          (nobet,-7)
     L              6           12                -16           7  5 +10        (1b,-6)         
     W             6           12               -10           8  6 +2           (2b,-8)
     L              8           13                -18           8  6 +10        (nobet,-8)     
     L              8           14                -26           9  6 +19        (1b,-7)
     L              8           15                -34           10 6 +29       (2b,-5)
     L              8           16                -42           11 6 +40       (3b,-2)
     L              8           17                -50           12 6 +52       (4b,+2)

End. +2u, highest bet 4u, highest drawdown -8

you could have made that last bet 3u if you just wanted to aim for the +1.

I continued on with differential betting from where Monaco left off.  I didn't wait for any losses, just started with the very next spin to see what would happen through the WLWLWLWLW section.  I didn't reset unless I was at +1.  That means that if I was at zero I kept the progression going on both sides.


                                                           combination
  spin  bet loss   total  bet loss  total  total


     L     1    1       -1      1     0     +1     0
     L     1    2       -1      2     0     +2     +1    +1
     W    1    2       +1     1     1     -1      0
     L     2    3       -2      1     1     +1     -1
     L     2    4       -2      2     1     +2     -1
     W   2    4       +2      3     2     -3     -2 
     W   3    4       +3      3     3     -3     -2 
     W   4    4       +4      3     4     -3     -1
     W   5    4       +5      3     5     -3     +1    +1
     L     1    1       -1      1     0     +1      0 
     W   1    1       +1      2     1     -2     -1
     L     2    2       -2      2     1     +2     -1
     L     2    3       -2      3     1     +3      0
     W   2    3       +2     4     2     -4      -2
     L    3    4       -3      4     2     +4      -1 
     W   3    4       +3     5     3     -5      -3
     L    4    5       -4      5     3     +5      -2
     W   4    5       +4     6     4     -6      -4
     L    5    5       -5      6     4     +6      -3
     W   5    5       +5     7     5     -7      -5
     L    6    6       -6      7     5     +7     -4
     W   6    6      +6     8     6     -8      -6   
     L    7    7       -7      8     6     +8      -5     
     L    7    8       -7      9     6     +9      -3


So, playing this new differential way we went from Smee losing -116 units to Monaco and I actually braking even if he had only bet 3 so as to end up with +1 instead of +2.  I won 2 units and was down 3 units at the end.  That's a 116 unit swing in less than 50 spins.


The potential problem I see with this is that it could be quite a grind, but if your grinding leads to wins instead of losses, where's the downside?  Also, it will take some practice to get used to playing both sides.  It is easy to get confused.


Like Monaco said, we could tweak this in ways to make it more profitable without compromising the safety.


I'm going to test all of the horror 3 session and see how it looks after that.  We also have to remember that the bet selection is just one side of the even chance which is not my favorite method.  I'm using the horror sessions though because they represent what should be the worst hands we can expect to get in 200 spins.  If it can handle that, it should be able to handle anything better than that also.


G

[Smee]

Nice work guys....im very much watching with interest! Im really busy at the moment and dont have time to test sorry - ill try to do a little later tonight.

[Fripper]

Hey George


Nice work, keep it up.


Let me know if you want any more of these hell sessions. It doesn't have to be 200 spin sessions, it can be 100 or 50 or what you prefer. All can be done with bayes software which is posted in the forum, you may be aware of it, it's excellent.

Cheers

[monaco]

I know that there has to be a way for this to lose.  It looks like it's a lot of WLWLLWLWL, the same thing that hurts the original Oscars.  Is this the way you see it?


G

yes, i think that's right George,
I had a look at Bayes 65 reds in 200 spins session, firstly using the marty 1-2-4-8, then by starting both Oscars at the same time.

Quickly realised, starting both Oscars at the same time means this isn’t a hell session at all, but a complete heaven session, as it will take full advantage if there’s a large dominance of either e/c. see attach.2. Even the 1-2-4-8 could turn in a decent profit, albeit with longer games (attach,1).

It could well be that all the horror sessions produce decent results but i'm not sure that puts us in the clear.

The thing is these particular horror sessions might not be the real horror sessions for this method. Any set of spins where either e/c dominates to a large degree & i would expect any of these ways to produce a profit - it's when the results are choppy that it needs looking into, like the 2^nd half of Smee’s sample but on a much longer scale. How would we get on if we had 200 spins mainly containing chops/series of 2 or 3, things like that.


The good thing is that the differential betting keeps the bets so low, that you can stay in the game for a long time if needed. It's hard to say at this point how things would look without testing some choppy sessions.. will do that next.

possibly, you could try this on all 3 e/c's at the same time & you would only need 1 of them to contain a dominant e/c to overcome a choppy other 2.. if you had 2 containing a dominant half at the same time, it should be pretty good for profits, although it could be tricky to do in practice..

be interested to see your results, cheers 

[GLC]

"   The thing is these particular horror sessions might not be the real horror sessions for this method. Any set of spins where either e/c dominates to a large degree & I would expect any of these ways to produce a profit - it's when the results are choppy that it needs looking into, like the 2nd half of Smee’s sample but on a much longer scale. How would we get on if we had 200 spins mainly containing chops/series of 2 or 3, things like that.   "

This paragraph is the heart of the matter.  WLWLWLWLWL this pattern causes us to drop 1 unit for every 2 bets.  WLLWLLWLWLWLLWLL  this pattern lets us recover enough to pull ahead eventually.  WLWLWLWLLWLWLWLLWWLWLWLWLWLLWLWLWWLWL  this is the sequences that kill us.  We keep dropping 1 unit for each WL and we only recover 1 unit on each WLL OR WWL.  The problem I see is that we do keep our bets so low that it can take a long time to pull out of a hole.  As always a good WWWWWWLWWWWWW streak like this can pull us up pretty good.

The positive about the above is that I don't think we need to have that large of a buy-in to have a realistic chance to win 5-10 units.  I haven't been able to place a bet over 11 units so far, and looking at your sessions, neither have you.  I'm thinking a 50-100 unit buy-in should be adequate. 

If we have a good bet selection that hits in clumps, that would be ideal. 

Playing all 3 e.c.s at the same time could be pretty hectic, but if you had good tracker from Ophis that did all the minute calculations for you and you reset all three anytime the total reached +1 or more, this could bump up against the membrane separating long term winner from long term loser.

In spite of everything that we're seeing, I've been at this excited point numerous times in the past only to have the big hammer smash all our expectations to smithereens, so I'm staying cool for now.  Much work still to do.

One more thought, what if we do start with a 1-2-4-8 marty, that puts us pretty far down for a recovery to 1 unit. 

The safest way to play this thing may be to just play differentially from the get go and take a 5 unit win and call it a session.  A couple or three sessions should be adequate each day if you're betting a reasonably sized unit.

My next experiment is to play all 3 e.c.s at the same time and see what that looks like.  I can't believe one of them won't be streaking well enough to overcome a few single unit losses from the other two.  And if 2 e.c.s are streaky, it'll be short work to get to +5.

Thanks for all the help and interest, Monaco, Catalyst, Smee, Fripper.  You know as well as I do that if this keeps looking like a gold nugget more and more miners will start diggin' with us. 

If it gets too complicated, (like it already is with 3 e.c.s) people will be afraid of it because it's too much work for too little reward.  I know the feeling.  That's kind of how I feel about the labby Fripper and friends are working on.  That doesn't mean I don't keep my eye on their work, just in case!!  As soon as they send up the victory flag, no amount of complication will be too much!

Luck to all,

George

[amk]

Hello GLC,

I still have to research the method indepth......

It looks good........

I think you might have found the MAGIC to take you further.......

GLC:
"The safest way to play this thing may be to just play differentially from the get go and take a 5 unit win and call it a session.  A couple or three sessions should be adequate each day if you're betting a reasonably sized unit".

[Shadowman]

This is a great thread folks I have always had a soft spot for OG and it's various spinoffs.  I have used it differentially before but missed the obvious Monaco solution.

GLC , I notice throughout this thread that a recurring concern of yours is that it is a very slow grind. A possible solution may be to use the take out principle, whereby over a time period of your choosing, let's say every five spins in this example, you add a unit to your deficit so that when the series is finally closed there will be a whole bunch of extra units accumalated as well as the +1. So as an example say that it takes 20 spins to close the series and you add a unit every 5 spins then you will effectively make 5 extra units.  I hope this is of use.

Mike

[monaco]

One more thought, what if we do start with a 1-2-4-8 marty, that puts us pretty far down for a recovery to 1 unit.

yep, the 1-2-4-8 marty leaves us with too much of an uphill climb i think - if you hit a choppy period straight after losing it (as i did) then you could be in for a long night. I think the other benfits of playing it differently are outweighing it at the minute.

If we have a good bet selection that hits in clumps, that would be ideal. 

i feel i'm still learning the real impact of bet selection, but from what i've learned, we're not looking for one here that delivers a steady ratio of wins to losses, such as Authors System, maybe your own ftl & change to chops after 3 chops, or any kind of trend-follower - as these are designed to keep the number of wins & losses close together.

what we need as you say is one with a high variation of wins & losses delivering clusters of both, a high z score if im not mistaken, & is there one better than pick an e/c & stick with it?

(i did have a look at Gordonline's w/l's got through P4 & his 20 unit loss was turned in to a 20-odd unit gain this way - maybe something there with delivering an unequal ratio between wins & losses..)

i think also worth remembering here is your original tweak on the Oscar that started this thread, & trying to see if that can also help (I'm not sure if you're still using that in your results?).

best regards

[GLC]

 
i think also worth remembering here is your original tweak on the Oscar that started this thread, & trying to see if that can also help (I'm not sure if you're still using that in your results?).

best regards

I have been thinking about this and wondering if we need to keep track of the losses so we can escalate our bets to take advantage of the inevitable (usually) cluster of wins on one side or the other.  I haven't been using it in my tests because, to be honest, I've always reached +1 before I lost 11 times on either side.


My purpose in designing the escalating bet increases after a predetermined number of losses was for faster recovery with fewer wins in a row.  If we always increase by 1 unit after each win, when we finally start recovering, even though we will be winning on larger bets, they won't be as large as with my progression which recovers a lot more losses with each win.


The trade-off is that we can go into the hole faster and deeper also.  But, with the differential betting, I don't think we have to worry about this problem even if we stay at +1 unit after every win.  It would result in less tracking which will help if we try to play 2 or 3 e.c.s at once.