Bank Management Methods That You Should Not Use: Martingale, Fibonacci and Others

Martingale is one of the simplest and most famous staking plans there are. Many people still post articles and videos on the internet claiming that it works for some kinds of betting games or sporting events. But this article will show you why it is not a good way to manage your betting bank. And instead, we will tell you 2 much better staking plans to use.

What is the Martingale betting system?

First of all, let’s explain the concept in an easy way. Simply, if you bet on an event with even odds (2.00) and lose, then you double your next bet. You continue doing this until you win. You will then win all your losses back and also win the original profit that you wanted from the first bet. In theory it works, but in reality, it does not.

Where did people use the Martingale?

Originally the idea was used for the even-odds red/black bet in roulette. Players thought that if they bet on red, eventually it would have to come up red. Let’s look at what that would actually mean in terms of stakes and possible winnings.

wdt_ID Bet Stake Total Stakes Result P/L Notes
1 1 1 1 Lose -1 We stake 1 unit on the first bet. We lose.
2 2 2 3 Lose -4 We lost, so we stake 2 units on the 2nd bet which could win us 4 points if it comes through.
3 3 4 7 Lose -7 We lost, so we double the stake to 4, but we have now staked 7 points in total.
4 4 8 15 Win 1 The bet wins. We get 16 units. Minus the 15 points we had already staked, we get 1 point profit.

With a 3 game losing run, it doesn’t seem so bad. We needed 15 units to win 1 unit. But in reality, long losing runs in even odds events are much more common that you would expect. If you made just 400 bets on a European roulette table, it would be expected that a 9 bet losing run would happen. This is based on the Expected Losing Sequence equation (Log(n)/-log(1-SR)= E.L.S.) where we use the expected winning percentage and the number of bets to calculate the longest expected losing sequence. For a European roulette table with one zero on it, the expected chances of winning the red/black bet are 48.65%. In that case the Martingale betting would look like this.

wdt_ID Bet Stake Total Stakes Result P/L
1 1 1 1 Lose -1
2 2 2 3 Lose -4
3 3 4 7 Lose -7
4 4 8 15 Lose -15
5 5 16 31 Lose -31
wdt_ID Bet Stake Total Stakes Result P/L
1 6 32 63 Lose -63
2 7 64 127 Lose -127
3 8 128 255 Lose -255
4 9 256 511 Lose -511
5 10 512 1,023 Win 1

After 9 losing bets we would have had to use a betting bank of 1,023 units, and when we win, we only get 1 point profit. Using 1,023 units of capital to win 1 unit is not an efficient use of a betting bank. And in fact, the longer you use the system, and make more bets, the longer the losing sequences are likely to be. At 5,000 bets we can expect about 13 losing bets in a row. After 100,000 bets we can expect 17 losing bets.

wdt_ID No. of bets E. L. S.
1 100 6.90
2 200 7.90
3 300 8.60
4 400 9.00
5 500 9.30
6 750 9.90
wdt_ID No. of bets E. L. S.
1 1,000 10.40
2 2,000 11.40
3 5,000 12.80
4 10,000 13.80
5 50,000 16.20
6 100,000 17.30

Casinos have been running for a long time. The longest recorded losing run is said to be from 1943 when red came up 32 times in a row! Anybody doing the Martingale that day would have needed some big pockets. After 32 losses, it would have taken 8.5bn units to win that one unit of profit. Of course, even if you had the money, you would not be able to do this because casinos have table limits. These two factors put a stop to Martingale, the amount of cash needed to get one unit of profit is inefficient and also casinos will not let you keep doubling your bets to any level.

wdt_ID Bet Stake Total Stakes Result P/L
1 1 1 1 Lose -1
2 2 2 3 Lose -4
3 3 4 7 Lose -7
4 4 8 15 Lose -15
5 5 16 31 Lose -31
6 6 32 63 Lose -63
7 7 64 127 Lose -127
8 8 128 255 Lose -255
9 9 256 511 Lose -511
10 10 512 1,023 Lose -1,023
11 11 1,024 2,047 Lose -2,047
12 12 2,048 4,095 Lose -4,095
13 13 4,096 8,191 Lose -8,191
14 14 8,192 16,383 Lose -16,383
15 15 16,384 32,767 Lose -32,767
16 16 32,768 65,535 Lose -65,535
17 17 65,536 131,071 Lose -131,071
wdt_ID Bet Stake Total Stakes Result P/L
1 18 131,072 262,143 Lose -262,143
2 19 262,144 524,287 Lose -524,287
3 20 524,288 1,048,575 Lose -1,048,575
4 21 1,048,576 2,097,151 Lose -2,097,151
5 22 2,097,152 4,194,303 Lose -4,194,303
6 23 4,194,304 8,388,607 Lose -8,388,607
7 24 8,388,608 16,777,215 Lose -16,777,215
8 25 16,777,216 33,554,431 Lose -33,554,431
9 26 33,554,432 67,108,863 Lose -67,108,863
10 27 67,108,864 134,217,727 Lose -134,217,727
11 28 134,217,728 268,435,455 Lose -268,435,455
12 29 268,435,456 536,870,911 Lose -536,870,911
13 30 536,870,912 1,073,741,823 Lose -1,073,741,823
14 31 1,073,741,824 2,147,483,647 Lose -2,147,483,647
15 32 2,147,483,648 4,294,967,295 Lose -4,294,967,295
16 33 4,294,967,296 8,589,934,591 Win 1

The Fibonacci Staking Method

Another staking plan that people should not use is the Fibonacci method. This system increases the stake after each loser following the Fibonacci series 1-2-3-5-8-13-21-34. The next number in the Fibonacci series is the previous two added together. With this method, the stakes increase, but at a slower rate than Martingale. People using the Fibonacci series also often cap the series on the 8th bet and then go back to the beginning of the series if they have had 8 losses in a row. While a lot better than the Martingale system, it is still not a great system. Similar to the Martingale system, the Fibonacci method is an inefficient use of your betting money.

The 1326 Staking Method

One more system I want to advise against is the 1326 staking method. In this staking method, people go up one level in the betting series if they win, and go back to the start if they lose. They are basically betting more when they win. They bet 1 unit on their first bet. If they win, they bet 3 units on their next bet. If they win the second bet, they bet 2 units on their third bet. If they win their third bet, they bet 6 units on their fourth bet. The idea is to take advantage of hot streaks in your betting, as if somehow wins are like buses and 4 might come along at once. This is a strange system because it is suggesting that while you are more likely to have 2 wins in a row, it is not so likely a third win will come, so bet less. But if the third win does come, then there is a higher chance of fourth winner coming, so bet 6 units. The 1326 method is an illogical system.

So, what are two good bankroll management methods?

Instead of any of these three systems, I would like to suggest percentage staking and target profit staking as better bank management methods.

With percentage staking, you bet a certain percent of your bank on each bet. This is good because as your bank gets bigger, you bet more, and if your bank gets smaller, your stakes get smaller, too. It is very easy to use and it is comfortable to bet a small amount of your bank each time. In theory it should be very hard to go bankrupt following this method. One thing to be careful of here is sticking to the method too strictly at higher odds. If you have a 12 odds horse in a race, it is a lot of money to put 5% of your bank on it. So it is a good idea to decide on different percentages for different odds bands. For example 5% of the bank for lower odds 1.01-1.50, 4% for medium odds 1.51-2.50, 3% for higher odds of 2.51-5.00, 2% for very high odds 5.01-10.00 and 1% for long shots over 10.00 odds.

Target profit staking is also an excellent staking method. In this method, you change your stake to achieve a target profit depending on the odds of your bet. If your target is 5% of your bank, then a high odds game would need a low stake, but a low odds game would need a higher stake. The stakes reflect the risk in the bet. I am quite happy to stake 20 points on a 1.25 odds bet or 1 point on a 6.00 odds bet. With Martingale, Fibonacci and 1326, they all focus on the stake with no care for the bank balance or the odds level. Target profit staking has both. But I will give one word of caution, do not use target profit staking for odds under 1.25, because then you will be betting a large amount of your bank on one game, which is never a good idea. Some bettors put their limit at 5% or 10% of their bank. So in those cases, limit your stake at your personal comfort point. I would not bet over 10% on one game, even if the odds are 1.10 and it is Manchester City at home to a non-league team in the FA cup. Shocks do happen.

These two staking plans are good to use together. So for this reason I suggest Target Profit Staking with a target of 5% for odds of 1.25 or over and Percentage Staking 5% of your bank for odds of 1.25 or under.

Returning to Martingale once more, some people felt they could still use this system if they looked outside casinos for places to apply it. Some people turned to sports betting to look for events that are more connected and less independent like casino table games. Markets with even odds are often the over/under 2.5 goals bets, no. of corners bets, the Asian handicap bets or tennis bets between similarly able players. Some people feel that if they back over 2.5 goals and follow one specific team until they get over 2.5 goals, then they have less chance of hitting a long losing streak. Betaminic’s Betamin Builder Tool uses a huge database of soccer statistics and bookmaker odds to find value patterns in football games. Some of those strategies use the over 2.5 goals market and have many bets around the evens level. This makes some people think it is a good strategy to use the Martingale system with. One of those betting strategies is called “Over 2.5 Against the Trend”. (You can follow this strategy yourself by signing up for free on the Betaminic website.) The average odds of its bets are 2.29. It has made 139 points profit from 744 bets which gives it an amazing 18.79% yield and 52% win rate. That makes it look good in the eyes of some Martingale fans. But I will show how using the right kind of staking methods can be a much more efficient use of your betting bank.

Look at the results for these 5 staking plans and this actual Betaminic system. We can test the historical betting data in staking software and see what different settings would result in. We can also run Monte Carlo Simulations. This is where we take the data and shuffle the order of the bets. The order of the bets is decided by the luck of the fixture list. It is important to shuffle the data to see if you would get similar results if the fixture list had been different. The betting system’s edge or your betting skill can be constant, but it is the luck of the fixture list that can change how a staking system performs. A strong staking plan should not go bankrupt in any shuffled simulation.

wdt_ID Starting Bank Staking Plan Actual Profit 1000 Monte Carlo Simulations Average Profit % of Simulations Bankrupt 1000 Simulations Lowest Profit 1000 Simulations Highest Profit
1 100 Level Staking (1 point) 140 140 0 140 140
2 100 Target Profit (1%) 211 210 0 210 211
3 100 Fibonacci (1 point) 225 319 6.30% -64 436
4 100 Percentage Staking (1%) 286 286 0 285 286
5 100 1326 (1%) 312 287 0 191 406
6 100 Martingale (1 point) 493 491 47.90% 122 672
7 100 1326 (4%) 1,247 1,137 6.20% 518 1,588
8 100 Target Profit (5%) 15,569 15,565 0 15,540 15,597
9 100 Percentage Staking (5%) 31,866 31,866 0 31,820 31,910

We can see the Level Staking plan for this system made 140 points profit from 1 point stakes. All of these staking systems were tested with a starting bank of 100. Even on just 1 point stakes the Martingale strategy goes bankrupt in 47% of 1,000 simulations. Even if you did put more money into the Martingale system and kept betting, the final result would only be between 122 and 672 points, and an average of 493 points profit. The Fibonacci Staking plan also goes bankrupt in 6% of simulations with 1 point stakes and only got a profit of 286 points. The 1326 staking plan does not go bankrupt on staking 1% of the bank, but when we increase the staking size of the recommended Target Profit and Percentage Staking plans to 5%, their profit levels eclipse the 1326 system which quickly reaches bankrupt levels at just 4% of bank staking levels. The Target Profit and Percentage Staking plans do not go bankrupt in any of their 1,000 simulations and have far better results which prove their suitability.

A good staking plan is supposed to take a betting system that makes profit on level stakes, and then get even more profit out of it. Any staking plan that increases the risk of you losing money through bankruptcy is not a good one.

So, the most efficient way to use your betting capital is to follow a proper staking plan with a betting system that has been tested on its historical results and shown to have promising potential. Do not use the Martingale or the Fibonacci methods.

 

Sign Up for free to access the Betamin Builder here.

How Betamin Builder Helps You Make Money In Betting Exchanges Like Betfair

See the best betting systems ranked by profit, ROI and risk here.

Read more about the free Betaminic Public Strategies here.

Read more about opening and closing odds

Access Betlamp, the amazing free statistics tool here.

Comments 0

bigpunter

August 17, 2019

Hi, nice guide to understand not to go bankrupt.
IMO most of bettors know %stake is a safer method but it could be interesting to developp on how to bet in % stake. For example to comparate different simulations with a change of %stake bankroll every 10 bets or each Week or month. What to do when a DD arrive, change the %stake when ? waiting 10 losses and change or wait with the same stake until stop loss ?
Thanks for your job

Javier Paniagua

August 17, 2019

Hi bigpunter,

Thanks for your comments.
Regarding your question, my recommendation is that once you have decided on the management method you are going to use, you have discipline and follow it. You should only stop when you reach the stop loss you have set. In this way, you will protect your bank before possible contingencies, and you will not end up in bankruptcy.

Cheers!

John

August 31, 2022

In roulette if u stick with betting red at 47.4% win chance, the level staking wouldn’t work though right? Isn’t fabionacci the best for that combined with stop and reset threshold. Thanks