Soccer Game Prediction


 


1.  Soccer Game Database

You must find a way to "define" a soccer team, i.e., to use a set of number to describe a soccer team.
An example is given below.
 

1.Team - To recognise each team in the Divisions.
2.Rating - This is a rating applied to each team to help evaluate that
team's worth on merit,against other teams in their Division,at that point in
time.The higher the rating, the better the team.
3. League Position - The team's current position in the Division which
reflects its standing against all other teams in its Division.The higher the
better.
4. Average Goals scored at home - This is the number of goals scored by that
team,divided by the number of games played at home.The higher the better.
5. Average goals scored against at home - This is the number of goals scored
by the visitors against the home side,divided by the number of games played
at home.The lower the better.
6. Average goals scored for Away - This is the same as 4 above, but away
from the home ground.The higher the better.
7. Average goals scored against Away - This is the same as 5 above, but away
from the home ground. The lower the better.
8. Last 4 home results - say WLDD only included to help evaluate the
outcome.See other note.
9. Last 4 away results - say DLWL only included to help evaluate the
outcome.See other note.
10. Last home result - say D but this is included in the last row of 8
above.Only included to help evaluate.
11. Last Away result - say L but this included in the last row of 9
above.
12. % Home wins.
13. % Away wins.
For simplicity, in the following, we will choose the following variables to describe a team:

1. Team name
2. Team Rating.
3. Current League Position.
4. Avg goals for at home.
5. Avg goals for away.
6. % Home wins.
7. % Away wins

The team name is not a factor, for example, if a team changes name, it is still the same team. There are two teams in a game, the final model is:

Home team:

2. Team Rating.
3. Current League Position.
4. Avg goals for at home.
5. Avg goals for away.
6. % Home wins.
7. % Away wins
Away team:
2. Team Rating.
3. Current League Position.
4. Avg goals for at home.
5. Avg goals for away.
6. % Home wins.
7. % Away wins
Results:
            Result (10: home team wins, 0 home team loses, 5 draw)

This model has 13 variables. We will use the first several games to train the DecisionMaker, and we will use the game data to test.

The example only intends to show you how to use the DecisionMaker. You should establish your own numbers to descibe a game and collect your own data.