Hounting Repeaters!

[daveylibra]

Hi Ben

Hope you don't mind me adding to an old thread. Your graphs are truly mind-boggling! Just trying to get my head around them. Does the 1st set (4 graphs) represent the same thing as the 2nd set (2 graphs)?
To make it simple for me, are you just saying, for example - take a look at the last 24 spins, you probably won't see less than 3 repeats or more than 15?
Also why do the graphs zig-zag, and not start at zero?
I reckon you might be on to something with this!

[Bigbroben]

I reckon, I just threw them there without explanation really...

Ok, let's take the one called Repeater15.  This chart represents: in the last 15 spins, how many numbers took 15 spins or less before to repeat?
If, say, the value at spin 41 is 2.  We know that the average for this ''time frame'' is 5.  One can expect a repeater to hit soon.

I don't know yet which ''timeframe'' is the best to use: is it 12, 15, 18, 24, 37 spins?  I produced them only for curiosity and some brainstorming.  Thought I could share them.  I wanted to see also if there would have been a regular ''wave length'' with repeaters.

It is just matter for reflexion.  If , say, with 24 spins one gets on average 18 nr to use and manages to see when to bet for repeaters and when repeaters are saturated, this is a good start.

[Bigbroben]

For Daveylibra.

Do whatever you want with it!  If you find a way to win with it, please share!

[Belgiangambler]

JEKHB76

i have a new angle with your beloved repeaters for you to test ( i don't have the time for it).

when a repeater drops put a unit on it
after 3 spins put an extra unit on it and so on and on
do this with every repeater you come across

at a new high start again, if not delete the dropped number and continue with the rest of them

[jekhb76]

JEKHB76

i have a new angle with your beloved repeaters for you to test ( i don't have the time for it).

when a repeater drops put a unit on it
after 3 spins put an extra unit on it and so on and on
do this with every repeater you come across

at a new high start again, if not delete the dropped number and continue with the rest of them

Hi,
I have tested this method in the past, with different results.
You can be lucky form a great deal of time, when the repeaters hit early in the game. But when they don't you are in big trouble. I've had more losses then winning with this, so for me it is a no go.
But thanks for your input.

[Bigbroben]

Ok, so this is a variation on the same theme as we've been buzzing with lately.
I've been thinking about safe ways to ride through tuff rides of unhits, when betting on hit nrs.  One notices that repeaters tend to come in bunches, especially when there are more numbers that have chips on.

With the approach of including the new unhit in our pool of nrs, one would think indeed repeaters will come in bunches eventually.  We do not know exactly when, but there will be a time.  So, this is point nr.1.

Point nr.2: we do not know when 2 or more hits come in batches.  Nor do we know when 1 hit will come.  What would make sense would be to bet the least when no hit and the most on a hit.  Sure, but we don't know.

So here's the deal.

Let's suppose the following situation with an agressive (+1 per loss) progression( amount of nrs betted on is not relevant here), with reset at new high:
L   L    L   L   L    L    L    L    L    L  W   L    L    L     W    L    W   W  New high
1   2    3   4   5   6    7    8   9  10 11  10  11  12   13  12  13  12
This represents a big dip before a new high.  It works, but dangerous and vulnerable.

Another way, with steady betting (steady, +1 after loss following a win if no new high):
L   L    L   L   L    L    L    L    L    L  W   L    L    L     W    L    W   W  New high
1   1    1   1  1    1    1    1    1   1   1   1    2    2      2    2    3     3
This is better.  Still exposed to long unhit runs.  Not as big a dip and new high anyways.

The third way, a safest way, I'd say.  ''Cruising'' bet is 1.  The bet will change after the first win to go to the ''supposed'' progression, and going back to 1 after a loss.

L   L    L   L   L    L    L    L    L    L  W   L    L    L     W    L    W   W  New high
1   1    1   1  1    1    1    1    1   1   1   1    1    1     1     2    1     3

Cons: -missing hits with higher stakes, losing with the high bet, if  L1 W1  L1  W1  L2  W1  L3  W1  L4.
           -could take longer to recover.
Pros:  No exponentially steepening downward curve.  Lesser br needed.

This method is better used with many nrs to start with.  It would be risky to start with one and wish for repeaters back to back early, although it can happen.
I've tried with the nrs from last spin until last repeater with some success ( see waitforhitprog).  A good example of recovery in the end.  Was crazy and frustrating but it came back.

Tried also with 18 nrs and it shows good, excellent preliminary results.  It's a win on first spin half the time, never got past 3u on nrs yet, since so many nrs: often a strike of repeaters.  I think I will pursue in this direction.

So:
1. Take last 18 nrs (not spins);
2. bet 1u on each;
3. Win: back to 1;
4. Lose: add new nr and spin, again until win;
5. When win, respin until new high or loss.
6. If loss, add new nr. Add and spin until win.
7. Win: add corresponding progression units on nrs and spin until new high or loss.
8. Loss: back to step 6. 


I'll keep updated.

[jekhb76]

Ok, so this is a variation on the same theme as we've been buzzing with lately.
I've been thinking about safe ways to ride through tuff rides of unhits, when betting on hit nrs.  One notices that repeaters tend to come in bunches, especially when there are more numbers that have chips on.

With the approach of including the new unhit in our pool of nrs, one would think indeed repeaters will come in bunches eventually.  We do not know exactly when, but there will be a time.  So, this is point nr.1.

Point nr.2: we do not know when 2 or more hits come in batches.  Nor do we know when 1 hit will come.  What would make sense would be to bet the least when no hit and the most on a hit.  Sure, but we don't know.

So here's the deal.

Let's suppose the following situation with an agressive (+1 per loss) progression( amount of nrs betted on is not relevant here), with reset at new high:
L   L    L   L   L    L    L    L    L    L  W   L    L    L     W    L    W   W  New high
1   2    3   4   5   6    7    8   9  10 11  10  11  12   13  12  13  12
This represents a big dip before a new high.  It works, but dangerous and vulnerable.

Another way, with steady betting (steady, +1 after loss following a win if no new high):
L   L    L   L   L    L    L    L    L    L  W   L    L    L     W    L    W   W  New high
1   1    1   1  1    1    1    1    1   1   1   1    2    2      2    2    3     3
This is better.  Still exposed to long unhit runs.  Not as big a dip and new high anyways.

The third way, a safest way, I'd say.  ''Cruising'' bet is 1.  The bet will change after the first win to go to the ''supposed'' progression, and going back to 1 after a loss.

L   L    L   L   L    L    L    L    L    L  W   L    L    L     W    L    W   W  New high
1   1    1   1  1    1    1    1    1   1   1   1    1    1     1     2    1     3

Cons: -missing hits with higher stakes, losing with the high bet, if  L1 W1  L1  W1  L2  W1  L3  W1  L4.
           -could take longer to recover.
Pros:  No exponentially steepening downward curve.  Lesser br needed.

This method is better used with many nrs to start with.  It would be risky to start with one and wish for repeaters back to back early, although it can happen.
I've tried with the nrs from last spin until last repeater with some success ( see waitforhitprog).  A good example of recovery in the end.  Was crazy and frustrating but it came back.

Tried also with 18 nrs and it shows good, excellent preliminary results.  It's a win on first spin half the time, never got past 3u on nrs yet, since so many nrs: often a strike of repeaters.  I think I will pursue in this direction.

So:
1. Take last 18 nrs (not spins);
2. bet 1u on each;
3. Win: back to 1;
4. Lose: add new nr and spin, again until win;
5. When win, respin until new high or loss.
6. If loss, add new nr. Add and spin until win.
7. Win: add corresponding progression units on nrs and spin until new high or loss.
8. Loss: back to step 6. 


I'll keep updated.

Nice approach, but More testing is needed.

[Bigbroben]

System rather ok, some stop-loss to be worked on.

I'll keep on testing.

[daveylibra]

Hi BigBroBen and thanks for the above.

This idea is very interesting, but I've started a new thread you might find of interest too
called "Turbo's Repeaters Simulation" which is more pessimistic.