PDA

View Full Version : modified martingale system



system1
27th February 2006, 17:13
hello
I have looked at and researched lots of systems. but i cant get this one out of my head, im convinced there is a way of making it work - you have probaly heard of it or seen it talked about in the forum, it is basically a stake system ie. you set a target ie 10.00 and back for example a 10/1. you would put a pound on to hit your target, should it lose you take the money you staked and add it to the target, if your next horse was 10/1 you would put 1.10 on to win your target and stake back. I know that this sytem always fails because of the long strings of losers. but there must be a way round this!, anybody with a half decent strike rate would make on this system if it wasnt for the losing streaks that come along from time to time. you obviously cant predict when they will come, but they always will. any thoughts?

presto
27th February 2006, 17:20
very, very risky system is the martingale. last year fav backers would have to stump up a £50+ million bet if they were using this system after a long run of fav's losing.

the most realistic martingale inspired staking plan i could think of would be 'set limits'. i,e, aim to win £10 on bet 1 - though if the losses add up to £100. then stop the loss chasing and re-start back to the £10 aim.

however i would not reccomend this system!

vegyjones
27th February 2006, 17:25
Although this system could possibly profitable...

there are systems that produce better results with less of the stress!

So for your own health and sanity, it is a system to avoid!

system1
27th February 2006, 17:30
Why do people always associate favs with the system?
Your quite right, the loss chasing is its downfall. I have thought about trying a variation where you would cut your target in half after 5 losses until you win. after that win the new target would be the half you cut earlier. this might sound stupid at first, but if you have ever researched it you find it works most of the time, and throws up clumps of winners, then a rare long losing streak. lasting a long streak and using the normal groups of winners to soak up your cut targets might work.

system1
27th February 2006, 17:31
I know what your saying veg, but like i said i cant get it out of my head. I know there is a way of making it work, and i,ll find it.

MarcusMel
27th February 2006, 19:03
The problem lies in the fact that at level stakes the system loses. If it made a profit them the staking plan could be implemented with possibly not betting into a long losing run. You would have to figure out at what losing length you stop and and wiat for a winner to restart. However no matter how you stake if the system loses money long term then ANY staking system will also lose. I belive this is mathematically certain where there is a limit to the amount you can bet. The only way to make money from gambling long term is to be betting on prices that are bigger than the strike rate.

Simple example of price being too big. You KNOW that on any spin of a coin you have a 50% chance of winning. If a rich mug bookie were to offer odds of 2-1 on heads instead of the true odd of evens the gambling public would poor money into the bet of heads. The percentage value of the bet is
1/(2+1)=1/3 or 33.33%. Therefore if you have a method of chosing a gamble and you know for certain that the true strike rate and you see the percentage of the bookies odds is less than your knowedge of the SR then you have a Value bet.

However after that things get a little more complicated as what you are really looking for is a set of average odds that overall give a better winning value.

system1
27th February 2006, 19:21
thanks. thats given me something to think about.:)

MarcusMel
27th February 2006, 19:33
Just edited my last post to make it a little more precise.

presto
27th February 2006, 19:52
if you are interested in 'value' then have a look at the 'kelly' staking plan - it's a long way from perfect, but there are many variables you can use,

system1
28th February 2006, 09:16
I have been experimenting with rules for selection. Up until now i have no limit as to how many horses are in the races i,m betting on. I was thinking of reducing the number to 8 or less runners, the logic being even random choices in 8 runner races will avoid long losing streaks better than say random picks from fields of 10 or more. any thoughts? for instance how long would you go without a winner if you randomly picked a horse from races that had three runners only? compared to picking from races with 12 runners for example.

wb
28th February 2006, 12:26
any thoughts? for instance how long would you go without a winner if you randomly picked a horse from races that had three runners only? compared to picking from races with 12 runners for example.

Thats very hard to answer, because that would be assuming that all horses have the same chance, which they don't.

With all respect, any system that randomly picks horses will almost certainly fail, otherwise, we would all be at it and there would be no need to study form, stats etc.

P.s. for your banks sake, make sure you are paper trading first :)

system1
28th February 2006, 15:27
I dont pick horses at random, i have a pretty good 5 star rating system to pick the winners. what i meant is any selection system will break down from time to time, which is what i want to limit, i was thinking that a limited field would just give me a very slight edge, because of the improved chances of randomly picking a winner, ie it has to be easier to randomly pick a winner from a 5 horse race over a 10 horse race. do you follow what i mean?

wb
28th February 2006, 15:29
oh, ok, I see what you mean, but don't forget that prices will be adjusted and shortened for a smaller field accordingly

system1
28th February 2006, 15:36
yes quite right. I am looking for a slight edge in this respect so that will come into the equation when i finalise the rule.