Proposed test on Numbers/cycles - KTF/Random Thoughts inspired

[falkor2k15]

Inspired by Priyanka's Quad/Dozen cycles framework and her tracker for KTF, I propose to carry out an interesting experiment on numbers:

*Play sleepers (0s) and potential repeaters (1s) for several repeats/extended cycles (at least 12 ala Law of the Third) keeping a separate BR for each bet type. If the sleepers come out with the most profit then it's +1 for them otherwise they get a -1 and the repeats get a +1. So for 12 cycles we might get results that look like:
+s +s +s +r +r +s +r +r +r +r +r +r

After those results have been generated we need to see what the most common row looks like as the average winning trot - and it has to look like that across multiple data sets. If that passes then I can then proceed to the next part of the test, but I am not sure what that will be yet.

We need to end up knowing whether to bet on 0s or repeaters every cycle based on the most likely path judging from the stats. And I need to find out how many cycles to play for in order to get maximum profit from flat-betting. If I even get that far - providing there are patterns across multiple data sets - then I might get to test if the profit can be extended by knowing when to switch to 2s and when to possibly add in a progression (potentially riding on any imbalances).

What do you think? Any feedback? How far will I get? What kind of results should I expect?

[MoneyT101]

Sounds interesting....can you do a short session to get a little more understanding so we can attempt to test it out?

[falkor2k15]

See attached... Here I am betting just sleepers and ignoring any negative results where the repeaters won:

Judging by the first test results for cycles with 12 repeats the sleepers do well during the middle of the trot around repeat 6/7, but around repeats 1-2 and 11-12 they seem to lose more often.

[falkor2k15]

I did 100K spins x 2 for both sleepers and repeaters. I never realised they could both lose, but a clear pattern emerges from the first repeat:
Sleepers 1st repeat:
-12253
-14883

Repeaters 1st repeat:
-296
-3029

So, clearly, if you had a choice of what to stick to - out of sleepers of repeaters - you would begin with the repeaters!

I've now figured out how the tests should proceed next, based on the first split of the tree:
Repeat 1 (1s) > Win > Repeat 2 (0s or 1s?)
Repeat 1 (1s) > Lose > Repeat 2 (0s or 1s?)

[falkor2k15]

There's a clear pattern emerging here across all data sets:   
Repeat 1 (1s) > Win > Repeat 2 (1s)
-2734
-3768

Repeat 1 (1s) > Lose > Repeat 2 (1s)
33
-398

Repeat 1 (1s) > Win > Repeat 2 (0s)
?
?

Repeat 1 (1s) > Lose > Repeat 2 (0s)
?
?

[TurboGenius]

So, clearly, if you had a choice of what to stick to - out of sleepers of repeaters - you would begin with the repeaters!

Some Genius mentions that all the time lol - eh what does he know.

[falkor2k15]

Some Genius mentions that all the time lol - eh what does he know.

Well, without some suitable test I wouldn't have known either way...

Repeat 1 (1s) > Win > Repeat 2 (1s+)
-2734
-3768

Repeat 1 (1s) > Lose > Repeat 2 (1s+)
33
-398

Repeat 1 (1s) > Win > Repeat 2 (0s)
-4795
-3516

Repeat 1 (1s) > Lose > Repeat 2 (0s)
-2482
-2169

If we are in profit by the time of the first repeat then we should still follow the repeat trail - albeit virtually perhaps - and I guess in a few repeats time we'll be ready to switch to sleepers.

[TurboGenius]

There's never a reason to play sleepers.

[falkor2k15]

There's never a reason to play sleepers.

We'll see... it seems the gap is closing depending on the accumulated success rate, but I could be wrong... will be carrying out further tests tomorrow to confirm.

[falkor2k15]

Repeat 1 (1s) > Win > Repeat 2 (1s+) > Win > Repeat 3 (1s+)
-270
-293

Repeat 1 (1s) > Win > Repeat 2 (1s+) > Win > Repeat 3 (0s)
-3697
-3448


Repeat 1 (1s) > Win > Repeat 2 (1s+) > Lose > Repeat 3 (1s+)
-651
-1012

Repeat 1 (1s) > Win > Repeat 2 (1s+) > Lose > Repeat 3 (0s)
-938
-643


Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Win > Repeat 3 (1s+)
-1296
1271

Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Win > Repeat 3 (0s)
-286
-2751


Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Lose > Repeat 3 (1s+)
85
343

Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Lose > Repeat 3 (0s)
-778
-1089

[Priyanka]

There's never a reason to play sleepers.

Very wise words. And notto, please dont jump on it. Sleepers are different from non-hits as per your definition.

[falkor2k15]

The non-hits (0s) already won above in 1/8 situations, so the next series of tests is as follows:
Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Win > Repeat 4 (1s+)

Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Win > Repeat 4 (0s)


Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Lose > Repeat 4 (1s+)

Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Lose > Repeat 4 (0s)


Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Lose > Repeat 3 (1s+) > Win > Repeat 4 (1s+)

Repeat 1 (1s) > Lose > Repeat 2 (1s+) > Lose > Repeat 3 (1s+) > Win > Repeat 4 (0s)


Repeat 1 (1s) > Win > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Win > Repeat 4 (1s+)

Repeat 1 (1s) > Win > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Win > Repeat 4 (0s)


Repeat 1 (1s) > Win > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Lose > Repeat 4 (1s+)

Repeat 1 (1s) > Win > Repeat 2 (1s+) > Win > Repeat 3 (1s+) > Lose > Repeat 4 (0s)


Repeat 1 (1s) > Win > Repeat 2 (1s+) > Lose > Repeat 3 (0s) > Win > Repeat 4 (1s+)

Repeat 1 (1s) > Win > Repeat 2 (1s+) > Lose > Repeat 3 (0s) > Win > Repeat 4 (0s)


Repeat 1 (1s) > Win > Repeat 2 (1s+) > Lose > Repeat 3 (0s) > Lose > Repeat 4 (1s+)

Repeat 1 (1s) > Win > Repeat 2 (1s+) > Lose > Repeat 3 (0s) > Lose > Repeat 4 (0s)

Soon I bring the 2s into the fray, but I don't think it's the right time yet.

[nottophammer]

Very wise words. And notto, please dont jump on it. Sleepers are different from non-hits as per your definition.

But is it wrong

[nottophammer]

Priy
i'll tell you whats not wrong that KTF tester at Westspiel casino, using those 71'000 spins, 5'160 done in 60 spin games +3249

[Priyanka]

But is it wrong

I didn't say it's wrong.