# How to predict draws in football

**Unlike sports with a lot of points, such as Rugby and American football, a football match has a significant chance to end in a draw. In this article we will discuss how to predict which matches are more likely to end in a draw, and which are not. This tool is rarely used of those who give free forecasts for football, however, quotes to draw, as a rule, always very attractive.**

About 26% of the matches of the English Premier League ended in a draw in 10 years from 2006 to 2016-th year. The most common for this period was 1:1 (42% of all draws), then a goalless draw (32%) and 2:2 (22%). So draws make up about a quarter of the selections in the Premier League during the season and are also an important component in the bet on the handicap. It is intuitively clear that a draw is more likely to occur between two teams with relatively equal opportunities, after the chances of both groups were incorporated in the calculation. In the matches between the obvious leader of the championship and team fighting for a place in the Premier League, usually you can see the quotes of the alleged draw at around 14%, whereas if there are two teams from the middle of the table, the probability will increase to 30%.

This can be confirmed by applying the widely used Poisson distribution for two nominally equal teams. Although a pure Poisson approach slightly underestimates the probability of a draw in football that requires adjustment, but in General, this technique is relatively simple.

Consider the match of the Premier League, where the average TB 2.5 expected very often between two equal teams. In this case, each team need to score an average of 1.25 goals per game against each other. According to the schedule of the Poisson distribution, we can assume that both teams have 29% chance not to score, so the probability that the game ends in a draw 0:0 is the product of these two probabilities.

0,29 * 0,29 = 0,08

Draw 1:1, in accordance with the actual data from the Premier League, is more likely to be almost 13% chance. Once these numbers are calculated for all possible outcomes, outlines the lines of probability on 0:0, 1:1, 2:2 and so on. They can be added together to obtain an overall picture of the chances for a typical Premier League game between two equal teams. In this example, without any adjustment, taking into account the small deviation of the curve from Poisson reality, the probability of a draw you get about 27%.

If the odds are equal strength teams to play a draw it is easy to determine the number of goals scored at the same time, the indicator which is often neglected in the prediction. The less likely to be broken total in the match, the greater the likelihood that each team will have a lower individual scores. The likelihood that teams will fail to score when the equity increases from 0.29 to 0.33. This, in turn, increases the probability of complete match 0:0 from 0.08 to 0.11 and the total is not adjusted, the probability of excluding a draw increases to 0.29.

So the choice of matches that are more likely to end in a tie inevitably leads us to teams of approximately equal class may have shown previously a tendency to a quality defense, and not a very convincing attack. However, a good strategy on these graphs, it is necessary to gather information. Sometimes teams play the season with the higher probability of a draw, sometimes less. Track is in the initial stages of the season is incredibly difficult, even tracking all training sparring teams in the offseason. The regression towards less extreme averages should always be considered when approaching the climax of the season when a draw is mutually beneficial for both teams, especially often it can be observed in Series A and Series B.