The last issue was "Static Prediction Method (1): Dr.Aroad Elo Prediction Method", and today is "Static Prediction Method (2): Goal Rate Prediction Method".
2.Goal rate prediction method
In the paper "Exponential Gambling in Games" co-authored by David Jackson and K.R. Mosheski, a research method for predicting the outcome of a game by using the past scoring rate of participating teams is proposed. The specific method is: Ra represents the past goal rate of participating team A, and Rb represents the past goal rate of team B; the total number of goals in this game is predicted by Ra+Rb; the outcome of the two teams can be predicted by Ra -Rb to predict. In the thesis, the 1990 World Cup match between England and Ireland was taken as an example to demonstrate their theory. England's average goal percentage in international competitions is 1.29 and Ireland's is 0.73. Calculated according to their method, England's winning probability is 1.29-0.73=0.56, and the total number of predicted goals is 1.29+0.73=2.02. Before the game, the bookmaker gave England a winning index of 0.85-1.10 for this game, and a total goals index of 2.10-2.14. Comparing the bookmaker's index with the predictions of the two scholars, it can be found that the bookmaker's England win index of 0.85 is much higher than the scholar's forecast of 56%, which means that the bookmaker's evaluation of this result is significantly higher than According to the positioning given by the scholars, the final game ended 1-1. If the punter sells England on the exchange according to the index of 0.85 and wins, then his income is 0.85× the transaction price, and the profit return is extremely considerable, while Buyers will suffer considerable losses. Judging from the prediction results of this game, the prediction methods of Jackson and Mosheski are very accurate.
The application of this method mainly grasps the following points:
1. The relationship between average goal percentage and team strength
First of all, it is necessary to clarify whether there is a functional relationship between the team's average goal rate and the team's strength. If this functional relationship cannot be established, this method cannot be used for prediction. The team's total league points represent the team's overall strength to a certain extent. According to the prediction theory of Jackson and Mosheski, the calculation method of the average goal rate is: average goal rate = total goals of a team / The total number of games played by the team. Take the Premier League (microblogging topic) and the Bundesliga 2008-09 season as examples.
Premier League, Bundesliga 2021-2022 Season Ranking and Average Goals
Team Rank Average Goals Team Rank Average Goals
1 Manchester City 2.60; 1 Bayern Munich 2.85
2 Liverpool 2.47; 2 Dortmund 2.5
3 Chelsea 2.00; 3 Leverkusen 2.35
4 Tottenham Hotspur 1.81; 4 Leipzig Red Bull 2.12
5 Arsenal 1.60; 5 Union Berlin 1.47
6 Manchester United 1.5; 6 Freiburg 1.70
7 West Ham United 1.63; 7 Cologne 1.52
8 Leicester City 1.10; 8 Mainz 1.47
9 Brighton 1.11; 9 Hoffenheim 1.70
10 Wolves 1.10; Monchengladbach 1.59
11 Newcastle United 1.16; 11 Frankfurt 1.32
12 Crystal Palace 1.32; 12 Wolfsburg 1.17
13 Brentford 1.25; 13 Bochum 1.12
14 Aston Villa 1.37; 14 Augsburg 1.15
15 Southampton 1.13; 15 Stuttgart 1.20
16 Everton 1.13; 16 Hertha Berlin 1.09
17 Leeds United 1.11; 17 Bielefeld 0.79
18 Burnley 0.89; 18 Tefelter 0.82
The average goal rate of each team in the above table is the result of dividing the team's total number of goals in the 2021-2022 season by the number of games played. For example, Chelsea have scored 76 goals this season and played 38 games, so their average goal rate is 76/38=2.00. We then conduct a correlation analysis on the relationship between the ranking of each team in the standings and the average number of goals scored. It can be found that the relationship coefficient between the rankings of the two major leagues and the goal rate is about 0.85 on average, and the relationship between the ranking and the average number of goals is more significant. Therefore, the average goal rate symbolizes the attacking ability of a team and can represent the overall strength of the team to a large extent. Therefore, we can conclude that the average goal rate can effectively predict the outcome of the game.
2. Matters needing attention
No method of forecasting is foolproof, and all we can do is try to be as accurate as possible. Using the average goal rate to predict the result of the game, three things should be paid attention to:
1. The strength of the team
The scoring rate reflects the team's attacking ability, and the higher the striker's scoring efficiency, the more guaranteed the team's victory. However, even the most aggressive team may not win every game, otherwise there would be no upset. For example, in the 2010 season, Atletico Madrid defeated Barcelona 2-1 at home (microblogging data). The average goal rate of the home team Atletico Madrid was only 1.32, while that of Barcelona was as high as 2.50. In theory, Barcelona should win, but in the end it was Atletico Madrid. Cold win. Instances like this are not uncommon. Just imagine, which super giant has no record of being upset? Therefore, the strength of the team is an important factor in determining the outcome of the game, but it is by no means the only factor.
2. Home and away factors
Just like the example of Barcelona's away game loss to Atletico Madrid, looking at the standings again, it can be found that Atletico Madrid's home goal rate reached 1.95, while Barcelona's away goal rate was 2.13. From this point of view, the strength gap between the two teams has been narrowed a lot. For any sports competition, the geographical advantage of the home field is very important. According to the analysis of the statistical results of the five major leagues, it can be found that the proportion of home wins in football matches is higher than that of away games. The real strength of the team has little to do with it. Almost all teams have a higher winning percentage at home than away.
3. Fighting spirit and status
The team's attitude towards the game and the state of preparation also have a profound impact on the outcome of the game. For example, some big-name clubs have to deal with two-line battles, injuries to main shooters, etc., which will have an impact on the outcome of the game.
3. Forecast definition
After actual combat testing, we can summarize the following operational definitions as the basic criteria for predicting games using the average goal rate. It should be noted that this definition is more suitable for competitions with higher league levels such as the five major leagues.
Definition 1: When the difference between the average goal rate of the two teams is more than 0.30 (excluding 0.30), the team with the highest average goal rate wins;
Definition 2: When the difference between the average goal rate of the two teams is more than 0.10 to 0.30 (including 0.30), if the average goal rate of the home team is higher, the home team wins;
Definition 3: When the difference between the average scoring rate of the two teams is 0.10 to 0.30 (including 0.30), if the average scoring rate of the home team is lower than the average scoring rate of the away team, the home team wins or draws.
Definition 4: When the difference between the average goal rate of the two teams is below 0.10 (including 0.10), the home team wins or draws.
The four criteria above cover team strength and home and away factors. In view of the fact that the average goal rate of each team can truly reflect the strength of the team after 10 games, the use of the average goal rate should start from the 11th round. It should also be pointed out that when using the average goal rate to predict a game, because the average goal rate of each team will change after each game, it is necessary to average the number of goals scored by the team based on the number of goals scored in the previous game. The adjusted goal rate can be used as the basis for prediction. For example, if Manchester United scored 20 goals in the previous 10 games, then before predicting the result of their 11th game, the average goal rate applied is 20/10=2. If Manchester United scored another goal in the 11th game, when predicting the 12th game, Manchester United's average goal rate should be adjusted to (20+1)/10=2.1.
The next article will bring you the static forecasting method (3): the six-field forecasting method, so stay tuned.