How to predict the card count in the World Cup

How to predict the card count in the World Cup

By Mark Taylor Jun 11, 2014

Tweet

Read this Total Booking Points betting article to understand how researching referee nationality, historical trends and recent competitive form can help bettors predict the card count for World Cup games.

The abiding memory of the 2010 World Cup Final, won by Spain with a single goal four minutes before the end of extra time against the Netherlands, was the large number of cards issued by England’s Howard Webb. Nine yellows were shown in regulation time and five more, including a second yellow for John Heitinga in extra time.

Excessive, certainly, both in the context of the 2010 tournament and previous hosting’s of the competition and perhaps also surprising in view of a European referee officiating two European teams. But also, it was testament of the extremes that can occur over a single match.

Inevitably, Webb’s handling of the match received criticism, but a more measured assessment could hardly omit the part played in the large card makeup by both sets of players. Many considered de Jong’s chest high, first half tackle worthy of a straight red and Iniesta, was booked for removing his shirt to reveal a tee shirted tribute to a fellow professional immediately after scoring the only goal.

So in attempting to predict the card count in a World Cup match, we should, as a bare minimum consider the disciplinary record of the respective sides, along with the identity of the referee.

Theory behind bookings

Iniesta’s celebration is an uncommon contradiction of the more general truth that cards come about largely through an accumulation of fouls, which themselves arise through a need to make tackles. So while some sides may have a more aggressive approach that invites bookings, as a general rule, the more tackles and potential fouls a team is forced to make, the more likely they are to receive a card.

So invariably the weaker team in the match will tend to pick up a larger share of the cards and the general relationship using the universal 10 points for a yellow card and 25 for a red, derived from 10 seasons of EPL data is that:

Proportion of card makeup = 0.64-(0.28*(Proportion of total goals a team is expected to score)).

In a match where 2.6 goals are expected to be scored on average and Team A is expected to score, on average, 2 goals or 2/2.6 percent of the total, from the relationship above, that team would expect to receive 42% of the expected total card makeup in that game. The remaining 58% of the bookings makeup would go to their weaker opponents. (Read this article to understand – How bookings affect live soccer betting.)

To reinforce this relationship, fouls committed nearer to your own goal, usually by defenders or midfielders, tend to be treated more harshly by referees.

If you took all fouls given away on a match by match basis during the 2011/12 Premiership season by every participating player, you find that a defender who commits at least three fouls in a game is more likely than not to have also received a booking. For midfielders this tipping point is reached after four infractions and a striker can happily commit between five and six fouls before it is odds on that he has seen a card.

The 2010 final, therefore illustrates some of the general trends that contribute towards card totals. Spain had been exceptionally well behaved prior to the final match, picking up just three yellow cards in six previous matches. Indicative of the connection between tackling in your own half of the pitch, fouls and cards.

Spain is, of course a side that coverts possession and did the lion’s share of the attacking. However, when faced with an aggressive opponent and a hugely important game, even the best behaved of sides can flex their ill-disciplined muscles.

The identity of the tournament also determines card count. The Premiership, historically has a lower card total than La Liga despite both being European leagues. An EPL game averages around 35 points compared to over 60 in Spain.

World Cup finals matches lie between these two extremes, averaging around 50 points per game over the last five tournaments. Despite the different levels of average card count that exists, not only within the same confederation, but also across different regions. South American referees, in keeping with Spanish based referees are often shown to be more disciplinarians than their counterparts in England.

Nationality of the referee

These domestic differences largely disappear on the World Cup stage. Smaller sample sizes for the less well represented confederations can lead to occasional outlying results, but FIFA appear to dictate that a uniform approach is required for their show piece event to prevent a team unfairly receiving suspensions because of the national identity of the referees they have played under.

The makeup for each European, North and Central, South American and Asian referees, as a group has stayed within a point or two of the tournament average for the last four events, with the lesser represented regions, such as Africa and Oceania falling around half a dozen points below the norm.

Similarly, over the last six seasons, La Liga referees have officiated in their domestic league games with an average makeup of 63 points, yet when the best of their generation has taken charge of Euro and World Cup qualifying games in recent years, their games have resulted in an unremarkable average makeup of 38, very slightly below par for the tournament.

Both referees and to a degree, players appear to appreciate the need to tailor their disciplinary performance to the wishes of the organising committees, with FIFA able to eventually suspend unruly players and rest maverick officials.

Therefore, in attempting to predict card makeups for the World Cup we can expand on our rudimentary approach.

How to predict the number of cards in the World Cup

Firstly, assume that the average makeup per game will be close to the recent tournament average of 50 points. Spread quotes for the 2014 tournament as a whole appear to confirm this assertion.

Next, bettors should use historical statistics based around the whole game rather than the record of individual teams within the game, because of the inevitable human interaction that may drive on-field behaviour.

Then if bettors wish to estimate individual team totals, divide the expected match total, using pre game goal expectations to give more of the predicted total points to the weaker team which is likely to be involved in making more tackles closer to their own goal.

Draw data from the same tournaments, where possible and recognise that markets invariably refer to 90 minutes play. European teams have almost all played 10 competitive qualifying matches to reach the finals, South American qualifiers have played nearly twice that number and it would be wise to omit friendlies which are often refereed differently. So there is a pool of recent and competitive matches to draw data from.

For Premiership matches there is a reasonable correlation between a team’s average match makeup after ten games and their final makeup after 38 matches. But be prepared to rein apparent outliers closer to the likely competition average or perhaps bolster data from similar earlier competitive tournaments, such as the Euros.

Largely ignore raw figures, we are looking for trends within national sides to be involved in generally well behaved or relatively badly behaved matches compared to the overall makeup of the competition.

Turning to referees, again, ignore raw makeups for individual referees from their domestic leagues, but instead look for ones, who over a season may be either more lenient or disciplinarian compared to the average for their usual place of work.

In the four premier domestic European leagues, referees are most lenient in England, becoming increasingly strict in Germany, then Italy and finally Spain. So an average domestic makeup of 45 would be considered draconian for an EPL referee, but lenient for a Spanish La Liga official. The raw numbers can be misleading.

Natural Self-preservation should ensure that officials have FIFA’s required 50 makeup as a benchmark, but they may not be able to fully disregard their natural refereeing style. A lenient La Liga ref may also dip slightly below the required World Cup average.

Try not to be too influenced solely by the record of officials in the World Cup qualifying matches. Invariably sample size will be small and trends unreliable. For example, after six matches of the 2011/12 EPL season, Howard Webb’s average bookings makeup was 64. By the conclusion of the season, it was an unremarkable 34.

Finally, be aware of special circumstances within a game, although try not to over inflate the importance. Some sides are natural rivals outside of World Cups, analogous to derby matches in domestic football, where makeups increase and even at a World Cup some group games, especially the final group game, may carry more importance than others.

Also be aware that bookings quotes may be pitched slightly high as people tend to want to root for events to occur rather than hope that they don’t.

As an example, the heady cocktail of two sides used to participating in matches that produce above average makeups, overseen by a referee of similar leanings would reasonably expect, on average to produce a well above average makeup.

This simplest way to combine these three major factors on the card makeup is to multiply them together. For a less extreme example, a team whose historical record sees them play in matches that are 1.1 times the average for the competition from which the data has been drawn, facing a side that is 1.05 times average, refereed by a lenient official whose matches are 0.8 times average would score a combined 0.924 once these three rates are multiplied.

Again from historical data, this combined figure typically produces a bookings makeup that is 0.95 times the historical average for the current competition or 47.5 for a World Cup.

This final conversion factor is given by

Producing predictions is as much about avoiding plausible, yet erroneous assumptions and when the footballing family, with such great diversity, come together every four years, there are many such possible pitfalls. Not least in how the matches are likely to be officiated.

Mark Taylor is a freelance soccer and NFL writer who, along with producing expert content for Pinnacle Sports, also runs his own soccer analytics blog, the Power of Goals.

If you have feedback, comments or questions regarding this article, please email the author or send us a tweet on Twitter.