The basic idea for this function arose already in my very early youth. I wanted to get football results as realistic as possible, for playing fun reasons. One of my favourite games played was simulating German Fußball Bundesliga results—by very different means, one being table football –, build tables, compare it to real tables – and improve my own games to get even better results.
As soon as Computers began to take over – which happened to the end of my junior days, in 1978 – I got my first programmable pocket calculator, the HP 65CV. Still very much into playing, I produced first results with this simple tool, random numbers. Later at University I had access to one of the very early main frames. So I started to develop and improve a football programme already then, including one part with predictions and the first simulation.
This one wasnt great, but it was something to build on. In 1990 I felt ready – given the Hardware improvements – to develop a very much improved Programme on my Personal Computer. This was the basis for my entry into a pretty long professional gambling career, mainly focussing on football betting.
The idea was, to be able to predict football results on a probability basis. This is most likely also the main joy of watching football or playing football simulation games: you expect to get realistic results, which, on the other hand, should not be too predictable. You get surprises here or there and you get easy favourite wins. You get the top teams easily taking over and dominationg the League as well as a surprise drops that noone can really explain. That is responsible for the fascination of football – and lots of other sports as well–, in reality and in games.
A simulation should generate results that make you instantly believe that they could occur next Saturday on the pitch. You see Arsenal being beaten by Blackburn Rovers or Chelsea draw with Fulham, but you may also enjoy the 6-0 win from Man United against Burnley or Hull the other week, as well as Chelsea and Arsenal cruising to victories. You see Man City at the top of the table after 10 games, but you sense, that this could not be long lasting as they feel the hot breath of the chasing Big Four down their neck in the short but intensive days of winter.
My function provides all this. First of all, I had to reflect the strengths of the teams realistically. I did divide their abilities into an „offensive“ and a „defensive“ part, for a start. The offensive drives the ability to score goals, while the defense tells you, how well they can avoid goals. Given two teams, there is an easy mathematical way of getting a prediction for this certain game, which first of all takes these parameters, compares them and then compares them to the average, out come the goal expectancies for this certain game. One value tells you, how many goals team1 does expect to score (and thus team 2 to concede), the other one, how many goals team 2 are expected to socre (and thus team 1 to concede).
So far this explanation does not yet take the home advantage into account, which obviously exists in reality, whatever it is based on – be it „feeling“ better at home, be it the crowd behind one or be it even other teams having to travel or, last but not least, be it the referees, that may be affected by crowd reactions on certain foul, offside or handling occurrences. It does exist, no matter what. So there is another parameter, that reflects the homeadvantage, and, this one as well, as realistic as possible, making it even individual, because their may be teams making more use of it or others, that feel more comfortable on the road. This is accounted for and integrated in another simple mathematical method into these two values, called the goal exoectancies for a certain upcoming game, calculated individually and uniquely for this special occurrence.
As these two values are just expectancies, one needs now the simulation function, that produces in e.g. a thousand games on average exactly this amount of goals – and like that being as realistic as possible. So the one result is still pretty much random, but in the long run it would be, that the basic strengths of the teams would be well reflected. Doing this for every single game, you get realistic results throughout the season.
In the attached file one can see the expectancies in the first two values behind every game listed. The first 17 games of the simulated Bundesliga season are the real results, while the other 17 rounds have been purely simulated, producing the ensuing tables and the battle for the title, relegation or places in Europe can be looked up.
The simulation surely gives the opportunity to drive certain paramaters in any manner. There may be a couple of decent question to be asked – and answered. Here a few of them:
- Should the strengths of the teams be made purely Computer driven or could the User possibly take an influence?
Answer: Why not make it User accesable? It is necessary to have a certain understanding of it. But as soon as this is cared for, it could increase the play fun.
- Should the strenghts of the teams be updated during the season, based on their (simulated) results?
Answer: Well, this is almost a philosophical question. It is true, that results occur sometimes randomly (or they usually do), but this does not actually have an influence on the basic abilities of the players. Or does it? As there are such values like „self-confidence“ or „mental power“, that could have the decisive impact. And even, the Club Directors or Fans sometimes overreact on random but bad results. The answer should then be: Make it User optional. Either, he wants Inter Milan or Chelsea or Bayern always to be like they originally are, or he wants updates on random results.
- How quickly should the Computer react on results, if it does so?
Answer: My programme tries to reflect reality. So I have to try and find the correct values for all such parameters. In a game, this cold look different. Maybe, someone wants a quick reaction, someone wants a slow one, someone doesnt want any (see also: Question 2). It could be made User optional as well. What reaction time would you wish?
Just a short introduction into this main function of my footbal programme. I think, there would be many other areas, that I could be useful in order to improve the great Sports Simulation Games on the market nowadays. Feel free to get in touch with me.