Developing a Sport Trading Strategy Part 1: Building The Framework

This is the first of a four part post looking into developing a sports trading strategy. For this strategy we will be trading in the football markets, however the concepts and pointers can be applied to any sports.

This plan will consist of multiple steps and go from conception, testing through to completion, I hope that viewers reading this will find it useful, but will also point out any flaws in the methodology. For the benefit of this post, we will be tweaking a common strategy that is well known amongst football traders; backing under2.5.

Research The Market, Discover Potential Value Points

The first step is to identify a market where there is potential value to be extracted. In our case we will be looking at the under/over 2.5 market, we know that the market behaviour is predictable in this market and this allows us to gather potential value based on our market data. But if you are looking to apply this to a potential strategy you are working on, you are looking for markets that correlate well with key events in your historical data. How often teams score in a particular league? how many clean sheets? In what timeframe are the most goals scored? What betting markets can this data be applied to? Does the expected value of the prices in these markets map to the data you have? (For example: In your data, does the probability of teams scoring (or a condition being met) correlate with the pricing of the market you are looking at)?

In addition to this, how does the market behave leading up to the entry point for your strategy? For example, for some markets the price drifts just before kick off and then corrects a few minutes in, other markets shorten before kick off and then correct a few minutes in. Not only can these predictable market variations be exploited, they can be used to maximise the potential profit for your strategy... So watch the market a few times before dipping your toes in, if you have Bet Angel you can record charted data to see the trend for multiple games, this can help you with identifying market behaviour leading up to and during a match.

A quick google shows that only 10% of games have a goal in the first 10 minutes, so we will look to exploit this statistic. (Really you need more data than this, but for this under 2.5 example it will be enough).

The data will have also shown that we need to go against the market to get any real value. 

Narrow Down the Time Frame

Its important to narrow down the timeframe using market data and analysis that we mentioned earlier. The data would have highlighted key points about the ideal time to enter the market and the ideal time to trade out. It should also be able to highlight trade out points if the market is going against us (stop loss points). Remember, the more time our money is in the market, the longer it is exposed to market volatility.

In our example of the under/overs market, we know that the market for under 2.5 starts high and shortens down until a goal is scored, with this in mind, we want to enter the market at about 3 minutes (because the market will drift for the first few minutes) then exit at around 15-20 minutes.

We also need to identify good value entry points, not all games will have good value pricing on the unders market. This is because if two teams are 'low scorers' then the market will price them accordingly. So in this example we will be looking at unders markets priced at 1.9 or above, you can of course tinker with this, I've just set this as an arbitrary figure.

The time frame and market conditions for our framework are:

  • Entry Point: 3 minutes after kick off
  • Team Must be at odds of 1.9 or higher.
  • Exit Point: 15 mins after kick off
So far, we've identified a market that has some value, we've identified the ideal time to enter and exit that market. Now we need to think about the discipline required to make this strategy work long term.

Discipline and Bank Management

Now we are on to the crucial part of the strategy. Without a well-defined set of rules, stop losses and bank management, we will struggle to make even the best strategies work. Failing to create a well thought out money management plan, or failing to stick to one, is one of the main reasons strategies fail. There are some psychological aspects at play here, but we can help ourselves by defining a well thought out (what I like to call) limitation process. So let's take a look.

All strategies need to have a stop loss, even if it's just the case of cashing out what is left of our trade. It's imperative that we limit our losses because they will occur. With football, if we are trading in play there may be some indicators that automatically result in a stop loss. For example, if a team ends up with a red card, if this negatively affects our strategy. then we should cash out when the market settles down.

Determine the amount of bank we are willing to stake per trade. When it comes to staking, its very easy to just chuck the whole bank balance in. This can be dangerous and is not something that should be done lightly. Not only will we not be able to track the win/loss ratio, we will not be able to recover well from a string of losses.

For our under 2.5 strategy, we will be looking to stake not more then 5% of the bank, more importantly we will be backing rather then laying. This gives us more more leeway in terms of the amount to stake thanks to no liability. So this is our bank management ruling:

  • Not more then 5% of initial bank roll per market trade
    • With stake reassessment every month
  • Only back to lay. Never Lay to Back

Stop Loss. Assuming a goal is scored between the 3rd minute and the 15th, then we will be executing our stop loss between 5 and 10 minutes after a goal has been scored, giving the markets time to settle. This process can also be applied to other situations that go against our strategy, for example if a team member gets a red card in the same time frame.

  • Apply stop loss 5-10 mins after a goal has been scored (Allowing market to settle)
    • In the case that another goal is scored in this time frame, repeat the process again
  • Apply stop loss 5-10 mins after a red card has been given.
    • In the case that another goal is scored in this time frame, repeat the process again
  • Always exit at 15 minutes regardless of profit loss (remember, money in the market is exposed).
Some people put a protective stop loss in, for example if a team has had 8 attempts on goal then that may be a trigger to exit the market (whereby the average goal is scored after 4 attempts). This is defensive so it may be optional and possible depend on the odds for the unders market. But lets chuck it in anyway, we can always tweak it during the testing phase.
  • Apply stop loss if more then 8 attempts have been made on goal (on target or otherwise, both teams included)

Some people like to put a global stop loss in, such that if they've lost a certain amount in a day then they cease trading for the day. There are pros and cons to this. Firstly a previous event will have no bearing on the next one, each market is unique and has its own set of variables that effect the probabilities for the out come of that event.** Translated - Placing a global stop loss is kind of pointless from a logical point of view. However, from a psychological point of view, it can make sense. A string of losses will cloud your ability to make informed and well reasoned choices when picking a market, you may be more inclined to over stake or place bets on markets that don't meet all of your strategies criteria.From this perspective, it can be a crucial help. This should be something you must consider doing if you feel you are susceptible to chasing losses or letting losses cloud your judgement.

**with football this is not always true, you should avoid games that are effected by other games  that are playing at the same time. (for example if a team needs to win and rely on another team to win in order to stay promoted/win the league. They have the potential to be exploited. But that's an edge case that is not covered by this strategy framework.


Right, so we have the framework for a strategy, what should we do next? Well, we need to do a few things with the framework, we essentially need to give it some meat.
We need some historical data of the league of your choice, in our case, the Premier League from which to build a better picture of the market.
We will be using the Poisson Distribution System to convert the seasons average statistics (like goals scored per game) into their probability. One thing is for sure, if you do have a new strategy that you're working on, you don't want to go in all guns blazing with it. You need time to test it. One of the most effective ways to do this is to back test it. Back testing involves taking historical data, and applying our strategy to it to see just how effective it is (or would have been?).

In the next post we will use some jiggery pokery (and some math) to extract variables needed to do back testing. We will be using the Poisson Distribution system to convert score averages into probabilities, and the BInominal Distribution system to determine over under probabilities from the goal scoring average. Working out how effective it is, what teams to avoid, what teams to stick to and how does the 'Force Majure' effect our profits (ie red cards and stuff). It is important to have good data, the more the better. we'll be using probabilities and statistics retrieved from this data to test the effectiveness of our strategy.

There are pros and cons to back testing in this sport. It doesn't take into account managers or players leaving/joining teams for this season. We can apply our tests to the results of the current season also to try and minimise this variation.

Part 2 Is now available


Popular Posts