Dr.Aroad Elo Prediction Method: The more points the home team has over the visiting team, the more likely it is to win

Dr.Aroad Elo Prediction Method: The more points the home team has over the visiting team, the more likely it is to win

On the basis of predicting and mastering the probability of the game, the gaming company formulates a business strategy for each game through the analysis of the probability, and finally expresses it in the form of odds. We summarize the methods of game analysis by bookmakers into static analysis and dynamic analysis. Among them, static analysis refers to a technical analysis mode based on the basic information of the competition team, and dynamic analysis is a comprehensive analysis mode based on market factors (which also includes insider information held by gaming companies).


微信图片_20221226143211.png


It should be noted that the difference between the two analysis methods is not only the difference in form, but the biggest difference is that static analysis is aimed at predicting the probability of the game, while dynamic analysis has entered the odds and index analysis program. Therefore, more strictly speaking, static analysis is a probability analysis method, and dynamic analysis is an odds (exponential) analysis method.


During the development of the gaming industry for more than 200 years, many game prediction methods based on probability and statistics have been produced. After countless practices and improvements, three more authoritative static analysis methods have finally been formed, namely: Prediction method, goal rate comparison prediction method, six-match record comparison prediction method. Practice has proved that almost all bookmakers in Europe have established mathematical models based on these three prediction methods to determine what odds the company will offer before the game.


Key points of this chapter


Point 1: A clear static mathematical model is the basis for analyzing the game.

Point 2: Static analysis is a model for demonstrating the rationality of probabilities.

Point 3: From the dealer's point of view, it is possible to make predictions about the probability of occurrence.

Point 4: Players above the intermediate level can try to use static mathematical models combined with other analysis methods to analyze the mode.


Section 1: The Errow Forecasting Method


The founder of Elo prediction method is Dr. Aroad Elo, Ph.D. in physics from the United States. The earliest application of this method is to use the method of scoring to compare the strength of chess players, so as to predict the result of the game. The method is described in detail in Arrow's book Chess: Then and Now. Erofa's scoring of chess players is based on the trend of the player's performance. When the player's performance rises steadily, the score will increase. If the player keeps losing chess, the score will decrease accordingly. The chess game is almost a purely technical game, and there is some scientific reason to predict the player's strength by scoring, but even so, some deviations will still occur in the actual operation of Erofa. Because the level and state of chess players sometimes fluctuate, and some unforeseen factors will also affect the performance of chess players or the result of the game. For example, a low-ranked player performed supernormally and won the chess champion, or a high-ranked player performed abnormally and lost chess, and so on. Nevertheless, in most games, Erofa can still predict the result of the game more accurately, so it has been promoted and passed down to this day.


Later, the Englishman Jaques Black improved and innovated the Arrow method through the research and demonstration of 1,500 Premier League (microblogging topic) matches, and it is widely used in today's football game predictions. Erofa's improved model is to predict the result of the game by analyzing the team's home and away points.


1. Distribution of integral difference


Jaques Black collected data from 1,500 games of the Premier League and UEFA Cup in the 1995-1996 season. The specific operation method is to subtract the points of the away team from the points of the home team, and divide the points difference into several grades. It turns out that about 700 of the 1500 games the point difference between the two teams was small, roughly in the range of +10 and -10, and outside of those 700 games the distribution was not so much Concentrated. Among them, there are only more than 50 games where the home team's points are more than 30 points higher than the away team's points, and there are only 50 games where the away team's points are more than 30 points higher than the home team's points.


2. The impact of points difference on the probability of home and away teams winning and losing


According to the basic principle of Erlofa, the more points the home team has than the away team, the greater the probability of the home team winning, and vice versa, the lower the probability of the home team winning. Finally, the following analytical equations are summarized:


Probability of the home team winning = 44.8% + (0.53% × point difference between the two teams)

Probability of the away team winning = 24.5% - (0.39% x point difference between the two teams)

According to the equation, we give an example to illustrate the method of its specific operation.


Example: The home team has 50 points and the away team has 40 points.


Step 1: Find that the integral difference between the two parties is 50-40=10;

Step 2: The probability of home team winning is 44.8%+(0.53%×10)=50.10%;

Step 3: The probability of the away team winning is 24.5%-(0.39%×10)=20.60%;


According to the winning probability of the home and away teams, if they are respectively subtracted by 1, the tie may be 1-0.50-0.206=29.4%. The final probability of winning the game is:


Probability of home win: 50.10%

Draw Probability: 29.40%

Away win probability: 20.60%


Example: The home team has 40 points and the away team has 50 points.


Step 1: get the integral difference between the two sides to be 40-50=-10

Step 2: The probability of home team winning is 44.8%+(0.53%×-10)=39.50%

Step 3: The probability of the away team winning is 24.5%-(0.39%×-10)=28.40%


According to the winning probability of the home and away teams, if they are subtracted by 1, the possibility of a tie can be calculated as 1-0.395-0.284=32.10%. The final probability of winning the game is:


Probability of home win: 39.50%

Draw Probability: 32.10%

Away win probability: 28.40%


The establishment of the above equation is completed through a relatively complicated calculation process, which is simplified here, as long as you master the formula and calculation procedures. The probability calculated by Errow's method is not only an analysis tool for dealers to predict the game, players can also use this method to make a preliminary analysis of the probability of the game. At the same time, after mastering the basic probability, you can also calculate the odds and handicap through the obtained probability.

The next article will bring you a static prediction method (2): the goal rate prediction method, so stay tuned.