TALLERES – A. SARANDI. KEY FACTS ABOUT THE MATCH
- Talleres spent his last meeting in the Argentine championship on July 17 against the Newells Old Boys and was defeated 2: 3 (Michael Santos scored in Talleres). Talleres has not won four matches in a row – he lost twice and drew twice.
- Arsenal Sarandi held their last meeting in the Argentine championship on July 18 against San Lorenzo. The winner could not be identified – a draw 1: 1. It was the second consecutive draw for Arsenal Sarandi. And he cannot win in three matches in a row – he lost once and drew twice. In each of the 4 previous matches with the participation of the team, the performance did not exceed two goals.
- The previous head-to-head match between Talleres and Arsenal Sarandi in the Argentine Championship was played on February 11, 2020 and ended in a 1: 1 draw.
- The home team is stronger offensively than defensively (defense rating is 949 and attack rating is 1113). The away team plays at about the same level in defense and attack (defense rating is 898, and attack rating is 872).
- The ratio of the home team’s attack rating and the away team’s defense rating is 1.24. With this ratio, the average expected number of goals in a match for the attacking team is 1.55.
- The ratio of the away team’s attack rating to the home team’s defense rating is 0.92. The average expected number of goals for the attacking team for this ratio is 1.26.
- In the last five matches of Talleres, bets on a draw (+ 34.0%) and on a loss for the team (+ 15.0%) have brought profit.
- In the last five meetings with the participation of Arsenal Sarandi, it was profitable to bet on the team’s victory (+ 91.0%), on total under 2.5 goals (+ 53.0%) and on a draw (+ 28.0%).
Match predictions were generated using machine learning: the predictive model was trained on a large sample of played matches and now, based on the data obtained, it generates predictions for upcoming matches, taking into account the incoming parameters. One of the key factors for forming a forecast is the ratio of the teams’ attack and defense ratings.
A score of more than 1 means the odds are underestimated and the bet is mathematically beneficial. If the score is less than 1, the coefficient is overvalued and the rate is mathematically disadvantageous. The predictions for the match are formed based on the estimates of the odds, but this is not a decisive factor. For this reason, sometimes a prediction can be made for the outcome whose coefficient score is below 1.
Note: Bookmaker Points – the number of points that the team should have scored based on the bookmaker’s odds in the last five matches (or less if the team does not have five matches with odds). Calculation according to the formula: 3 / k-t for a win + 1 / k-t for a draw. Actual points – the actual number of points that the team has scored in the last five matches (3 points for a win, 1 point for a draw).
TALLERES – A. SARANDI. PROJECTED SCORE AND GOALS
For each team, the machine learning model calculates the expected number of goals. Based on the values obtained and using the Poisson distribution, the probability of the exact number of goals scored (0 goals, 1 goal, 2 goals, and so on) is calculated. After that, the probability of all scores in the match is calculated. The calculation results are presented in the “Accurate Score and Goals” section. Then the probability of each team winning and the probability of a draw are calculated. Final forecasts are generated using these very calculations. Note: The Key Match Facts section lists the expected goals for the teams based on actual matches played, while the goal prediction for the Opponents Attack and Defense Assessment is generated by a machine learning model.