Alright, first off, I just recently came to grips with this fact. A system that shows a profit even after 5,000 spins, but lost more bets than it won (for simplicitys sake lets say that system only bet on even chance bets), isn't actually a successful system.
I thought and believed that if I could make a system that showed a profit after 100,000 spins, even if it lost more bets than it placed, was a successful system. Technically, it is a successful system if you walk away from the table a winner....however, if that system does lose more of the bets it places, eventually you'll take a loss. The thing is, you'll never know when you sit down at the roulette table, if it's poised just at that perfect point where you're going to lose.
So, I've been spending quite a ridiculous amount of time trying to find/make a system that wins more of the bets it places than it loses. I figured the best place to start from would be the even chance bets, as they are known for being the absolute most difficult in terms of making something successful.
Well, I've attached a screen shot of a system I just recently have been testing. It only bets on Red and Black, and it does use a progression. However, the progression is actually an up as you win progression, not an up as you lose progression.
The progression it uses is "The Midas Touch" progression. The thing with the midas touch progression is it just bets on one even chance bet regardless of what the tables doing. So, it can have drawdowns of over 10k sometimes. So, I modified it to actually vary how it bets. However, the point of this test wasn't so much to see what kind of profit could be made, it was more so to see if the systems bet selection could win more bets than it lost.
Well, it has won more bets than it lost (only betting on red and black), but I'm not sure what to make of the results. What I mean is, how successful is what I've made? Is this common? I'm not being coy, I'm really curious. I just don't want to invest much more of my time in something that is actually a common thing.
I'd like some input from anybody who doesn't mind giving me some. If this is/isn't really that good, what goal should I be shooting for?
[attachimg=#]
I forgot to mention, that this test was done on a single zero wheel
A progression skews everything. Flat betting is the true test of a winning system. If it wins more than it loses in the long run. I can never see the difference in a neg progression or a pos one. Eventually they will both show a loss.
Quote from: Spike on July 01, 2009, 08:56:49 PM
A progression skews everything. Flat betting is the true test of a winning system. If it wins more than it loses in the long run. I can never see the difference in a neg progression or a pos one. Eventually they will both show a loss.
You missed the point of the pictures Spike. The summary screen shows that it has won more of the bets than it placed, I just don't know how common the results of the "summary" screen shot really are. That is really what I was getting at. Forget it's profit or loss. Had I been flat betting it still would have made a profit.
@ Spike >> After re-reading my post, I realized I was a bit vague with what my question actually was.
What kind of win to loss percentage ratio is actually a good or great result? You can see what it is in the above picture.
I am now at 3,800 bets placed on red/black and my win to loss ratio is 2.02%
I'm wondering if that kind of result is good, or just very common. I ask because while it has won more bets than it lost, it is still an awfully low percentage.
So, am I wasting my time? Is that kind of result normal?
I'm asking these questions because that ratio is one I've always ignored, as I was only ever shooting for a system to make a profit after x amount of spins. I never bothered taking into account it's win to loss ration (which was always drastically low in the systems I've created, as they used up as you lose progressions)
Alright, here is another Summary screenshot, as the BR Balance Trend is irrelevant. It is in profit, however the largest drawdown is enormous....but again, that isn't the point. It has won more bets than it lost betting only on red and black. The progression can be changed or even eliminated, that's irrelevant.
So again, is this common? Am I wasting my time thinking this is successful when it actually isn't?
[attachimg=#]
Oh, and here is the betting history for anybody interested. Any input from anybody?
2% is very very low, well within the variance range. Test it on 100K bets and you'll see.
@ Spike >> Thank you, that is exactly what I was looking for. I wasn't sure if what I had was actually a good result, or the kind of thing you just mentioned. I actually just let it go to 13,000 bets and the win to loss ratio was at a measly 1.87% so I just stopped it...oh well...
Hi Rob :)
I don't know whether what I am about to write is going to help, but anyway,
provided we are talking flat betting :
- I don't see the problem with losing more often than winning if
- 1) you can establish some sort of regularity (can roulette give you that ?) and
- 2) if the win is significantly big so it does more than wiping the losses, and
- 3) if you have established beforehand by a series of tests a correct ratio of winning bets. Let me try to explain further. For example, if you have a system that loses two or three times 6 units whereas it gets you 29 units in case of a win, you are happy. Say you have designed something that triggers the bet within a margin error of 3-4 spins, and you decided to stop betting after 4 losses in a row. Sometimes you hit at first try, sometimes not, but your system is such that you know that on average you will have x % of winning bets every y spins. Which bet, x % and y are the big questions, of course ::). But then, you can perfectly be happy with a ratio of 1 win for 2 or 3 losses, because 29 > (3 x 6) and 29 > (4 x 6)
- How frequent is the situation of winning more often than losing that you show with your statistics ? I wish I could program RX, and I would run my system n million times to get the answer
My concern with the system shown is clearly the huge draw-down. I would think, test, rethink, etc to find ways to keep the system profitable as well as keeping the negative oscillations under control. Easier said than done in addition of being only indirectly related to your question.
On another side note, I have changed my mind regarding outside bets because of this problem of "being forced" to win more often than losing. For example, although I like the initial "Lw" strategy with dozens or columns, I have a problem with betting 2 units for a profit of one while betting for a hit within the last two. In case of a loss, it would take two consecutive wins just to break even.
Ok, enough with digressing...
Cheers :)
Stepydan.