The debate around flat betting vs progressions comes up a lot.

I was never a big fan of progressions believing that flat betting would be good enough if you have a good bet selection. The problem with that however is that no 'holy grail' bet selection exists.

It doesn't matter how cute, clever or inventive you get, the run from hell is always lurking in the background waiting to pounce. My conclusion therefore is that progressions are ok as long as you know what you are doing.

This thread is going to show you some analysis between flat betting and progressions and will hopefully provide a few insights which may help with your game.

All the analysis will be based on short games of 100 spins which would roughly take 2-3 hours to play in a casino.

I am going to keep it real because nobody is going to go to a 24/7 casino and play a million games, and all that will do anyhow is to stop you from seeing the forest through the trees.

The next post will revolve around flat betting.

This post is going to examine some analysis of 600 rounds x 100 games. So that's 60k worth of games.

Remember this is all flat betting. The results will be for the game of Baccarat which has a lower house edge than either single zero or double zero roulette.

So what information would be useful to know.

*What are the worst and best things that could theoretically happen?

*What are the worst and best things that did happen?

*What were the average worst and best things that did happen?

*What typically happens if you don't get out well your ahead, but instead play the full round of 100 games?

*During 100 games, how often can you expect to be ahead 10, 20 or 30 units?

*In 100 games, how many times does your bankroll reverse from losing to winning?

*How often does it never reverse, but remains losing throughout 100 games?

Here are the answers....

The worst possible loss at the end of a 100 games.  -100 units.

Worst loss in 600 rounds of 100 games.  -34 units.

Average peak loss within a 100 games.  -8.13 units.

The average loss at the end of 100 games.  -1.38 units.

The average peak gain within a 100 games.  +6.80 units.

The best win in 600 rounds of 100 games.  +28 units.

The best possible win after 100 games.  +100 units.

So what useful information can be gained from this looking at the analysis above.

One thing is that you don't really need a 100 units to play 100 games. A third of that will suffice.

If you want to test that out, go to the 'Testmystrategy' website and see it with your own eyes.

The most useful information you can get out of these stats is knowing when to quit. In other words, having realistic expectations of what you can actually win.

Unfortunately, the average gain is +7 units and the average loss is about -8 units and this more or less equates to the house edge.  So if you get ahead by +7 units and are hopeful of making more gains in that particular 100 games, statistically speaking, it is maybe not such a wise move.

Back tomorrow with part 2.

Thanks for this analysis.  I hope I understand how you did your test.  If not the following will be nothing by gibberish.   

The interesting thing is the -34 units as the largest loss in a 100 game session.  That coincides almost exactly to the most losses times 2 to expect in 200 decisions.  Fripper's subject "Beating Roulette with Math" explored how to beat this 66 wins vs 134 losses in 200 decisions using a labby.

He discovered that as long as the losses were more front or middle loaded, it could be beaten.  If they are end loaded, it takes some pretty hefty bets with no guaranty that it won't skyrocket out of control.

An interesting challenge would be to come up with a progression that can beat a 100 spin session of 33 wins vs 67 losses no matter where the losses occur.  Of course if they all occur at the end of the 100 spins, nothing we come up with can over-come that.  Or is there something?

I have often considered that if we can come up with something that comes close to breaking even with that bad of a win to loss ratio, maybe we can increment after a bad 100 spin event and pull out ahead on the 2nd or 3rd series.

The ringer will be how many 33 vs 67 sessions can come in a row?

GLC

Quote from: GLC on Oct 11, 11:43 PM 2014
 
The interesting thing is the -34 units as the largest loss in a 100 game session.  That coincides almost exactly to the most losses times 2 to expect in 200 decisions.  Fripper's subject "Beating Roulette with Math" explored how to beat this 66 wins vs 134 losses in 200 decisions using a labby.

Well spotted GLC! I never thought of that, but you are right of course.

I noticed I forgot to answer the last few questions from my above post.

Here they are.

*During 100 games, how often can you expect to be ahead 10, 20 or 30 units?

*How often does it never reverse, but remains losing throughout 100 games?

*In 100 games, how many times does your bankroll reverse from losing to winning? (or vice versa)

and the answers....


Average % of games where the gain exceeds 10 units. = 25.

Average % of games where the gain exceeds 20 units. = 2.

Average % of games where the gain exceeds 30 units. = 0.

The highest % of games without a reversal. = 17.

(a reversal is where a losing position turns into a winning position or vice versa at some stage in the 100 games)

Average reversals per 100 games. = 3.48

So looking at the statistics above, forget about trying to win big flat betting. The most important thing is knowing when to get out with a profit.

That's it as far as flat betting goes. My next post and subsequent posts will show how you can fare much better with different types of progressive betting.

Now I am going to show you a progression and I will compare the stats against the flat betting approach and see how things stack up.

In this progression, I am going to raise the bet by 1 unit after every win up to a maximum of 4 units. After any loss, I revert back to 1 unit.

The bankroll is the same as the flat betting test. (100 units).

The previous flat betting results will be put in brackets next to these progression results.

Worst possible loss.  -100  (-100)

Worst loss in 600 rounds of 100 games.  -51  (-34)

Average peak loss within 100 games.  -15.09  (-8.13)

The average loss at the end of 100 games.  -1.57  (-1.38)

The average peak gain within 100 games.  +14.07  (+6.8}

The best win in 600 rounds of 100 games.  +63  (+28)

The best possible win after 100 games.  +250  (+100)

The average % of games where the gain exceeds 10 units.  55  (25)

The average % of games where the gain exceeds 20 units.  27  (2)

The average % of games where the gain exceeds 30 units.  9  (0)

The highest % of games without a reversal.  15%  (17%)

Average reversals per 100 games.  3.03  (3.48)

So you can see a lot of improvements here.

The most important one is that the average peak gain has doubled and we are still only risking a 100 unit bankroll.

Also, there are good improvements in the potential to be ahead by either 10, 20 or 30 units.

So the downside is not much worse and the upside is considerably better.

Part 3 tomorrow.

Quote from: GLC on Oct 11, 11:43 PM 2014
Thanks for this analysis.  I hope I understand how you did your test.  If not the following will be nothing by gibberish.   

The interesting thing is the -34 units as the largest loss in a 100 game session.  That coincides almost exactly to the most losses times 2 to expect in 200 decisions.  Fripper's subject "Beating Roulette with Math" explored how to beat this 66 wins vs 134 losses in 200 decisions using a labby.

He discovered that as long as the losses were more front or middle loaded, it could be beaten.  If they are end loaded, it takes some pretty hefty bets with no guaranty that it won't skyrocket out of control.

An interesting challenge would be to come up with a progression that can beat a 100 spin session of 33 wins vs 67 losses no matter where the losses occur.  Of course if they all occur at the end of the 100 spins, nothing we come up with can over-come that.  Or is there something?

I have often considered that if we can come up with something that comes close to breaking even with that bad of a win to loss ratio, maybe we can increment after a bad 100 spin event and pull out ahead on the 2nd or 3rd series.

The ringer will be how many 33 vs 67 sessions can come in a row?

GLC

If this is the case, why not end a session as soon as we are in +1 and start a fresh new 100 spins series?