Smart football stats explained

Here at our team of alchemists have concocted 3 stats to help you improve your betting performance: attack force, defense strength and shape. We're not going to bore you with math so here's a short explanation of what each of these stats represent

Attack force

This number is calculated using the numbers of goals scored by each team. The graphs display the attack for each team depending on the type of game ("host" or "guest"). This means that for teams that are hosts the "attack force" that is displayed on the match page is calculated based on the number of goals scored while being hosts. This way you can take into account situations where a team's performance varies significantly when playing at home versus when playing away.

The values for this parameter varies from around 10 (weakest attack) to 100 (strongest attack).

Defense strength

This parameters is calculated based on the numbers conceded. Similarly to the attack force, it is calculated based on the type of game played (as a host or as a guest).

The values for this parameter varies from around 10 (weakest defence) to 100 (strongest defence).


This parameter is a bit more complex and might seem esoteric, but we think it is very useful. As we explained in a previous article, the bookmakers' predictions are, ON AVERAGE, usually pretty accurate. Thus, the bookmaker's predictions are indicative of the expected performance of a team. However, the bookmakers predictions are not 100% accurate so the actual performance of a team is different from the expectation.

The shape of a team represents how the team performed vis-a-vis of the expectations. A team may under-perform or over-perform. A super-team like Barcelona might be under-performing if it looses 2 games in a row after 8 wins and a weak team might be over-performing if it wins 2 games in a row after 8 loses. If you think this parameter is useless we advise you to read on the regression to the mean, a concept from statistics was the inspiration for this parameter ;-) Because some teams perform differently based on whether they play at home or away this parameter is also computed based on the game type played (as "host" or as "guest").

The values for this parameter varies from around 20 (least in shape team) to 100 (most in shape team).

Extra smart stats for machine learning

In order for someone to visually make sense of the data we are displaying only 3 parameters. However, if you are interested in playing around with our data you can subscribe to the "Legend" package which will give you the option to download CSV files with these parameters broken down into details. For example, with regard to the attack force for each team you will have access to the following parameters: "home games attack", "away games attack", "home games attack on 1st half", "home games attack on 2nd half", "away games attack on 1st half", "away games attack on 2nd half". Plug in this data into various machine learning algos and you might find a fortune :-)