Rowdie: Mathematical football prediction and betting tips

Soccer stats are the key

Do you have any idea about how football fair odds at Rowdie are being calculated? Just like the title of this article suggests, soccer stats are the key. But things are not as simple. First thing you need is data. Up to 2 000 fixtures per day, then you have up to 4 000 line ups, up to 50 000 events etc.

The first step is choosing the best performing formula – actually a few of them. Calculations of fair odds are quite complex and complicated and you have to twist them and see what happens. The first question however is “What do I feed the formula with”? Do I care for last 10 or 15 fixtures? Do I care for head to head statistics only? What do I do when there are no head to head statistics? What does a new coach mean to the team? What stats are relevant and what have no impact?

If you made it up to here, you must have realized that coming up with accurate fair odds is not as simple. Actually it is really really hard. At the end your algorithm will spit out some numbers, but how accurate are they? You actually can not know… but you have to find out somehow. Especially if you are a bookmaker. If you fail to do the predictions accurately, the 7% market margin will not save you.

The way you verify your formula is to do the predictions retrospectively. This means, you feed the formula with numbers and try to predict the results of the last 5 rounds within the league. If your formula works, the deviation will be no more than 3%. Simple as that.

Let’s take the more difficult approach

Another option could be looking deeply at each player. Great examples are World cups or similar competitions where players play outside of their leagues. Teams many times consist of players who never played together. Let’s say the best players meet in one team. Is this a guarantee that the team will be the winner? The short answer is No. The long answer is “There is a big likelihood that the team will actually suck”.

We have seen that many times. A great example was the Russian ice hockey team in 2 000. A great blend of the best players in the World (at least at first sight) failed completely. The biggest problem when it comes to predictions on the player level is the ego. You have to have stars, but you also need players who support the stars. If the team consists of stars only, there will be nobody who supports them.

As you saw, there are many ways one can take and many ways one can fail. If you fail enough, there will be a reward – one or 2 formulas that work.

Soccer stats and data is the key to predict the results. But there are many ways to fail

Subscribe our newsletter

P