Betting on Soccer Draws Provides the Best Value

Before you read any further, please note that these results are preliminary

---

I’ve always been curious as to whether consistent discrepancies exist in betting odds for particular events. I’ve previously heard of the favourite-longshot bias where betters tend to overvalue “long shots” and undervalue favourites. For example, in a horse race where one horse is given odds of 2.00 and another is given odds of 100.00, the true odds may be 1.50 and 300.00, respectively. However given that I bet frequently on soccer matches I wondered if any such discrepancies exist for three-outcome soccer betting.

Since the end of January I have been tracking virtually every football match offered by a particular bookmaker. For each match, I have logged the available odds at a random snapshot in time, along with the eventual outcome. Based on the data I have come to the conclusion that betting on draws provides better value than betting on a win or loss for the home team. Based on the results, betting on the home team provides the worst value of the three possible bets.

Why does this discrepancy exist? I believe it’s because draws are considered to be boring results, and people are more interested in betting on a win or a loss. This distorts the amounts placed on each result, forcing bookmakers to adjust the available odds accordingly. Because people systematically under bet on draws, bookmakers have to offer higher odds on them to get a more balanced level of betting across the possible outcomes.

Below is my summary data

Total matches: 1,097

Statistic	|     |	Home Team Win	|     |	Draw	|     |	Home Team Loss	|     |	Total
Occurances	|	458	|	331	|	308	|	1097
Frequency	|	41.75%	|	30.17%	|	28.08%	|	100.00%
Average Odds	|	2.32	|	3.37	|	4.09	|	3.26
Average Payout	|	$0.84	|	$0.99	|	$0.92	|	$0.92

 

If you had placed a bet on every home team you would have returned $0.84 for every dollar invested, while betting on a draw for every game would have returned $0.99 for every dollar placed.

After reading about the favourite-longshot bias, I decided to look at results for betting on the shortest odds, longest odds, and the middle odds. By that I mean if the odds were as follows for a match:

Portsmouth – 6.70
Draw – 3.75
Chelsea – 1.55

A Chelsea win represents the shortest odds, a draw represents the middle odds, and a Portsmouth win represents the longest odds. I found that betting on the middle odds provides the best value, and the shortest odds provides the least value. Perhaps people tend to overestimate the value of the “big” clubs that they’re more familiar with. Below are the summary statistics. Please note that I excluded any event where two or more outcomes shared the same odds, hence the smaller sample size.

Total matches: 1,035

Statistic	|     |     |	Short Odds  Wins	|     |     |	Middle Odds Wins	|     |     |	Long Odds Wins	|     |     |	Total
Occurances	|	474	|	320	|	241	|	1035
Frequency	|	45.80%	|	30.92%	|	23.29%	|	100.00%
Average Odds	|	1.97	|	3.28	|	4.61	|	3.29
Average Payout	|	$0.85	|	$0.98	|	$0.92	|	$0.92

 

These results are preliminary, and I will provide updates as my sample size grows. I will also take a look sometime at the value of bets with odds of 1.00 – 2.00, 2.00 – 3.00, 3.00 – 4.00, etc., to see if the favourite-longshot bias exists in football.

Disclaimer

Please note that none of my data has been double checked, nor have the summary statistics. Please don’t rely upon any of these results. This post is designed for your casual enjoyment only. As my sample size grows and I make a more thorough analysis of the data, I will be able to present these results with more confidence.

Suggestions?

If you have any requests or ideas for other analysis to run with the data feel free to let me know.