Hi , i know martingale is a strategy more used in roulette and it is a sure fire way to poorhouse. But i would like to offer an alternative view to it by using it on football.
The reason why martingale fails is due to the fact that current event has no memory of past history and every event is an independent event. In another words it is totally possible for a long losing streak in games where the above rule applies (such as roulette)
However in football, such rule does not apply, the preparation of this game is related to the result of previous game. So my hypothesis is using martingale in football betting would work.
I would suggest focusing on two markets, match odds and sent off. My idea is to lay the draw and lay the sent off - yes for every game pre off and hold it till full time. If a draw or a sent off happens in a game, lay the losses in the next game, and then if it happened again, lay the total losses in the subsequent game until you recoup the losses. The rational behind this is if a team keeps on drawing (which is unsustainable, a draw only gives 1 point, while a victory gives 3 points, it takes 1 victory to offset 3 draws), the manager will prepare his team to take more risks, and will result in either a win or a lost, so the chances of draw reduces by every game. The same logic works for red cards, if a team has accumulate a lot of red cards, surely the manager would cautioned his team to control their aggression, so the chances of sent off reduces game by game.
Checking back in history i have only seen 4 consecutive draws at most for the top 4 teams in major european leagues. And even that it is rare enough.
Obviously when a streak develops the red would look scary. In order to control risk and prevent a bank bust, i thought of two variations to this strategy : setting a loss cap where you no longer chase losses beyond it or only recouping 75 % of the losses in next match.
Thoughts?
martingale strategy in football
Check this link out..cyxstudio wrote:
Checking back in history i have only seen 4 consecutive draws at most for the top 4 teams in major european leagues. And even that it is rare enough.
https://en.wikipedia.org/wiki/Premier_L ... tics#Draws
THree teams on 7 consequtive draws..
So if you were laying at avr odds of say 3.5..(£100) you would be up around £60K liability by game 7 (just to break even).
Not saying its impossible as I once viewed the P&L of a profitable trader just on laying the draw (no greening), but he was very seective on which games he picked..and thats a whole new ball game (if you pardon the pun!)
When things are going well in a trade and you have a good green thats when you can start increasing amount of money used. i'v seen me willing to risk my whole green because my next trade looks almost certain to succeed, going from a decent green to £0 is sometimes worth the risk.
The problem i have with martingale/loss chasing stratergies is at there core they are always appiled in blanket form.
You should be looking for a stratergy to identify selections to use on a system rather than looking for a loss recovery stratergy that can be applied to a system as a whole.
Example system - Laying the field in horse racing
Concentrate on finding a stratergy to identify the best race types and circumstances in which to lay the field on.
Dont lay the field of every single race and waste time looking for a loss recovery strategies
You should be looking for a stratergy to identify selections to use on a system rather than looking for a loss recovery stratergy that can be applied to a system as a whole.
Example system - Laying the field in horse racing
Concentrate on finding a stratergy to identify the best race types and circumstances in which to lay the field on.
Dont lay the field of every single race and waste time looking for a loss recovery strategies
Some food for thought in the basic idea, but there are loads of potential problems.
Presumably you start laying the draw after 2 or 3 consecutive draws? Or do you propose just laying every single draw and hoping for the best?
If you're only applying it to the top 4 teams in the league, then the draw odds at home to weak teams will be quite high - not at all ideal for loss recovery. And who exactly are the top 4 teams? After 4 weeks of the season they could be almost any teams, and after 3 draws, a top 4 team might now be 7th. Do you still include them? If you're talking about the acknowledged 'big guns', then you have a problem where their next 2 fixtures are away to Man City and home to Arsenal. The manager might not be too unhappy with 2 draws, and I doubt he'd adopt a gung-ho approach in these games.
But there might be specific scenarios where the strategy could make sense.
Presumably you start laying the draw after 2 or 3 consecutive draws? Or do you propose just laying every single draw and hoping for the best?
If you're only applying it to the top 4 teams in the league, then the draw odds at home to weak teams will be quite high - not at all ideal for loss recovery. And who exactly are the top 4 teams? After 4 weeks of the season they could be almost any teams, and after 3 draws, a top 4 team might now be 7th. Do you still include them? If you're talking about the acknowledged 'big guns', then you have a problem where their next 2 fixtures are away to Man City and home to Arsenal. The manager might not be too unhappy with 2 draws, and I doubt he'd adopt a gung-ho approach in these games.
But there might be specific scenarios where the strategy could make sense.
I follow the logic in that long streaks should be less, however whilst roulette has no memory neither do it's odds whereas I would expect that punters are using the same aguments when deciding to back/lay and therefore odds will fall as the streak gets longer therefore negating the effect. Having said that, if the odds arn't changing in line with the memory then there may be an opportunity for some money to be made there with conventional lay/backing.
This is a question from a another site but it will help you calculate streaks
Q. I am working on a laying system based on favourites. I would like to know what the longest losing streak of favourites is.
________________________________________
A. There is a simple formula you can use to determine the longest winning / losing streak of an event. Here's the formula:
Log(RUNS) / -Log(PROB)
Where RUNS is the maximum number of runs in a sequence and PROB is the probability of the event happening.
e.g.
In the past 10 years on the Flat Turf there were 33,084 races so if we wanted to check the longest losing run of favourites over this period assume RUNS = 33084
The strike rate of favourites on the Flat Turf during this period was 31.7% so the probability of them winning is 0.317. But as we want to calculate the longest losing run we need the probability of them losing. This is simply
1 - prob of winning = 1 - 0.317 = 0.683
We therefore set PROB = 0.683
The formula is now:
Log(33084) / -Log(0.683)
You can work that out with a calculator or Excel. If you do you will see that the answer is
27.3
What this means is that in a sequence of 33,084 races (10 years worth of results) we can expect the longest losing streak of favourites to be around 27.
Real World Test
Running this report reveals the following results:
Total Runs 33,084
Total Correct Lays 22,609
Lay Strike Rate 68.3%
Max Winning Run 28
The real world test saw a longest winning run of 28 which is near enough what the formula above produced.
Q. I am working on a laying system based on favourites. I would like to know what the longest losing streak of favourites is.
________________________________________
A. There is a simple formula you can use to determine the longest winning / losing streak of an event. Here's the formula:
Log(RUNS) / -Log(PROB)
Where RUNS is the maximum number of runs in a sequence and PROB is the probability of the event happening.
e.g.
In the past 10 years on the Flat Turf there were 33,084 races so if we wanted to check the longest losing run of favourites over this period assume RUNS = 33084
The strike rate of favourites on the Flat Turf during this period was 31.7% so the probability of them winning is 0.317. But as we want to calculate the longest losing run we need the probability of them losing. This is simply
1 - prob of winning = 1 - 0.317 = 0.683
We therefore set PROB = 0.683
The formula is now:
Log(33084) / -Log(0.683)
You can work that out with a calculator or Excel. If you do you will see that the answer is
27.3
What this means is that in a sequence of 33,084 races (10 years worth of results) we can expect the longest losing streak of favourites to be around 27.
Real World Test
Running this report reveals the following results:
Total Runs 33,084
Total Correct Lays 22,609
Lay Strike Rate 68.3%
Max Winning Run 28
The real world test saw a longest winning run of 28 which is near enough what the formula above produced.
-
CLOWNSHOES
- Posts: 173
- Joined: Sun Dec 28, 2014 8:04 pm
Le Tiss would be as well creating an automated reply for these topics, every few days he has to tell someone that loss chasing is bad 
This is trueCLOWNSHOES wrote:Le Tiss would be as well creating an automated reply for these topics, every few days he has to tell someone that loss chasing is bad
Something I've realised over the years with trading though - psychology is quite often more important than graphs!
So many people struggle to accept a loss. This is why we have these Martingale threads, or systems that are similar. It's also why some struggling traders let bad trades go 'in play' for massive losses.
Lets's face it, if I put my cat in front of the keyboard......he'd have a 50/50 chance of picking a winning trade. This game is all about bank management, staking plans, entry & exit points, and learning from mistakes
It's really unfortunate that the word 'martingale' has even been mentioned on this thread, because it obscures what seems to me to have been a valid point by the OP that there might be some merit in betting against an existing sequence based on a very real incentive by at least one of the participants to avoid a continuation of the sequence.
Granted, no staking plan will, on its own, positively influence the ultimate financial outcome, but what about the merits or otherwise of the original proposition? It seems to hold at least some validity to me.
Granted, no staking plan will, on its own, positively influence the ultimate financial outcome, but what about the merits or otherwise of the original proposition? It seems to hold at least some validity to me.
-
CLOWNSHOES
- Posts: 173
- Joined: Sun Dec 28, 2014 8:04 pm
100%, Couldnt agree more regarding bank management and staking plans. Its just as bad to up your stakes too much during a winning run.
The proposition could only legitimately be that the market has not factored in the relevant information and hence the odds are too low and hence there is a value lay available. However, you should no more attempt to martingale this value lay than any other value lay. It might be value but nevertheless might go against you. Why risk more than 2% of bank on any trade?weemac wrote:It's really unfortunate that the word 'martingale' has even been mentioned on this thread, because it obscures what seems to me to have been a valid point by the OP that there might be some merit in betting against an existing sequence based on a very real incentive by at least one of the participants to avoid a continuation of the sequence.
Granted, no staking plan will, on its own, positively influence the ultimate financial outcome, but what about the merits or otherwise of the original proposition? It seems to hold at least some validity to me.
Even if you used Kelly criteria you wouldn't scale up in the way martingale requires and so using martingale simply allocates too much banks compared to the risk reward. Don't do it.