Over and Under bets are undoubtedly one of the most popular types of bets in the football field. In order to obtain value in such types of bets though, you need to prepare yourself with some detailed statistical analysis and, of course, the appropriate betting odds.
In this betting advice article, we will look to expand and reach out to introduce a slightly more complex approach that incorporates new thinking about over and under bets. It is an in-depth look at underlying data, statistical measures, and other factors that affect the number of goals scored in a match.
Hopefully, it is something that will make it easier to decide on specific bets.
In the following sections, we will open up new perspectives and show how complex over and under bets can be.
- Forecasting models and statistical measures
- Other factors for over and under bets
- Conclusion – Over and Under bets for advanced players
In giving out betting advice, it is our aim to give the reader some good juicy food for thought, as well as developing instincts to think more about betting and betting opportunities. With the right techniques and principles, you can positively influence your own success. In this article, we focus on over and under bets by looking at a variety of statistical approaches, as well as tactical considerations.
Forecast models and statistical measures
Statistical measures are the basis for forecast models to operate over and under bets. Statistics are hugely important, but you can’t just look at raw numbers, you also need to take into consideration specific information (for example, the playing style of the teams, the tactical orientation of the coach, injuries and occasion). But statistics are at least quantitative approaches, and certain odds for over and under bets are good in theory.
The goal difference
The goal difference is a first important yardstick for the style of play of a team. Compared with the other teams you can determine whether a team belongs to the “offensively strong” or “defensively stronger” category of teams. Say if Manchester City have produced the most goals in a league, that speaks well for the power and function of their offence. Here it is necessary to research which players score the most goals in the team and whether they are in form or even playing.
Defensive teams usually concede fewer goals than the other clubs in the league. Mostly, the average scored is under 1 goal, which means that it is not uncommon that a defensively strong team plays more often without finding the back of the net. Again, there is a need to analyze the squads of selected clubs. Are all the key defenders playing that day? What was the tactical orientation of the team in the most recent matches? Is there a series of games without conceding that they are on?
From the goal difference, defensive-oriented and offensive-oriented teams can be distinguished. Something important to note here is that this becomes clearer after several games have been played, as some distortions can occur at the beginning of the season because the results are weighted higher. From the tenth matchday, this levelling off and the “outliers” are collected statistically. To work with the goal difference, it is advisable to use some statistical measures.
The mean
The mean comparison is probably the easiest technique to make first observances about the number of goals. Determining the mean value (the average value) is very easy. You take the total number of goals and divide them by the total number of matches. If Liverpool has a goal difference of 24-18 after 15 match days, then the average score of the goals scored is calculated to be 24:15 (= 1.6) and the average score of the goals scored is 18:15 (= 1.3).
With both measures, you can then work based on the selected model. Overall, in Liverpool’s matches, the fictitious goal ratio fell 24 + 18 = 42 goals, so 2.9 goals on average (42:15 or 1.6 goals scored in the average + 1.3 conceded goals on average). From this, we can calculate even more complex models with such a simple calculation, which we will show in the next sections.
Frequency Distributions
Frequency distributions are very popular for over and under bets because they respond to specific goal events. Taking our example from Liverpool still, who scored 24 goals in 15 matches. It would be interesting to know how this number of goals were spread out over the 15 games. Were there multiple wins with three or four goals scored? Or did they have just one two high scoring games and struggled to score in the rest of the sequence?
If so, this would be followed by the reverse conclusion that there were also many matches in which the team remained goalless or only netted the one goal for example. This approach is suitable to determine the standard deviation and thus the variance of the average, plus at the same time it allows you to check whether scoring and goal-shy encounters are rather random or systematic.
Standard deviation and variance
The standard deviation is an important measure of descriptive statistics and aims to determine how robust a mean value is. The average of 24 goals is determined by the number of 15 events (number of goals in the 15 matches) and divided by the total number of games. So if Liverpool produced 2x 0 goals, 5x 1 goal, 5x 2 goals and 3×3 goals in the 15 matches, then the standard deviation is calculated as follows: the average is subtracted with each event and all terms are added together. This gives the variance.
In this example then, the computation would be (1.6-0) + (1.6-0) + (1.6-1) + (1.6-1) + …. + (1.6-3) + (1.6-3) + (1.6-3). The variance in this example would be 0.9067, the standard deviation would be the root of it. But this all sounds like a lot of time-consuming brain power, but you can use online calculators or set up a spreadsheet to work all these equations out automatically. Anyway, going back to the Liverpool example, the standard deviation is 0.9522.
Another goal structure would yield a different variance and standard deviation for the same number of goals in the same number of matches. Our second example: in the 15 matches scored by Arsenal this time is at 5x 0 goals, 3x 1 goal, 3x 2 goals, 2x 3 goals, 1x 4 goals, 1x 5 goals, then the bill would be (1,6-0) x5 + ( 1.6-1) x3 + (1,6-2) x3 + (1,6-3) x2 + (1,6-4) x1 + (1,6-5) x1 = 2,373 at variance and 1, 54 to standard deviation. In any case, the discrete events already differ greatly from the mean in tabular form.
What does this mean? The higher the standard deviation and the variance, the less robust the average is and the predictive power becomes in turn, less accurate. Betting experts with statistical models look especially at this measure.
The Poisson distribution
Advanced bet experts often use the Poisson distribution to calculate odds for over and under bets. According to the experts, this probability distribution comes closest to the most accurate predictions. The Poisson distribution can also be modelled using a standard Excel spreadsheet. The big advantage of this distribution (compared to the normal distribution) is that it respects discrete events. For this distribution, both the expected value and the variance are needed. We simply derive our expected value from the calculated mean value. It should be noted that an exact number of goals cannot be accurately predicted, but this distribution can be used to establish a relatively accurate probability “bandwidth” for a game.
Other factors for over and under bets
Still, even breaking down the type of team (defensive or offensive), and the goal distributions, how many goals are going to be scored in a match is not only dependent on just the factors above. Specific features also play a role in the forecast, not least a small residual share of luck or bad luck. Obtaining information such as the tactical set up of the coach, head to head form, injuries and even occasion (does one team need points to win the league while the other is in mid-table) are also crucial factors as they undoubtedly have an impact on the number of goals in a match.
The tactical alignment
The lineups of the teams are a first indicator of the style of play of a team. The classic lineups are clearly 4-4-2, 4-3-3 and 3-5-2. Of these three lineups, the 4-3-3 is a relatively offensive mindset. Despite a back four, three strikers offer far more punch than two spikes. The other two classic lineups are more likely to be ball possession football, while a 4-3-3 clearly makes counter-football possible. However, there are also coaches who vary their composition depending on the situation.
In their 2015/16 title-winning season in the Premier League, Leicester made a lot of adjustments to tactics according to the situation. Even if they were set up offensively in matches, when they would take the lead, they would tweak formation to go a little more defensive. So even though they got the title, they only won 23 matches and they made great candidates for under bets that season because they were happy to defend leads and were happy with their 1-0 and 2-0 results.
In the following season, Chelsea won the league title with 30 victories after sticking with their 3-5-2 formation and just staying the course with their foot on the gas. Why is this important? Well, there are leagues across the world which are little more defensive minded than say the English Premier League in general. Looking at the Egyptian and Argentine leagues, for example, you will tend to see a lot more conservative formations such as a 5-4-1 (or a 4-5-1) and that is just a matter of how certain things are done in certain leagues. It is important because naturally, this will have an effect on how many high scoring games are likely to be produced.
With the five-man defensive wall and just one striker up top, it tends to lead to more drawn games like (often with 0-0 or 1-1) and generally produces very low scoring fixtures. Although betting odds are lower for under-betting in such leagues, these bets are more likely to win. Specializing in such leagues and such types of bets can, therefore, be lucrative.
Injured players and scored players
Of course, one of the most important pieces of information besides the tactical orientation is also the setup. It is easy now, well ahead of games actually to look online at a list of injured players. If Harry Kane is out injured for Tottenham, with him being their main source of goals, the team’s chances of scoring well are going to be diminished. That increases the likelihood at the same time of them winning the game by a smaller margin.
The same thing would be if Liverpool, for example, were missing Virgil van Dijk from the heart of their defence, it would leave their defence more vulnerable. This, of course, favours the tendency for the opposition to score more goals. So if there are major disruptions in the lineups of teams like this, then it should naturally impact your decisions. A missing striker would lean towards more under bets for example.
So a look at the upcoming game formations and lineups can also be very valuable. With regard to an upcoming cup tie or a midweek league game against lower opposition, some top teams tend to give their more important players a breather. If Man City are facing Bradford in the FA Cup, then they are unlikely to send out their strongest starting eleven, more so if they have a derby against Manchester United in the league in midweek. So this also has an influence on the dominance and the own game, logically also on the number of goals.
Conclusion – Over and Under bets for advanced players
A perfectly accurate model that predicts the number of goals 100% of the time is something that doesn’t exist. Chance, luck and individual situations are influencing factors in a result and all the statistical data in the world can’t allow for things like that. But looking at under/over betting from a statistics-based point of view can help with the probabilities of making a higher return of accurate predictions.
So getting a good grasp of this, along with some common sense thrown in along with the identification of lucrative odds can make over and under bets a very profitable affair.
