In today's exploratory data analysis, let's look at the stats from the English Premier League.
At first glance, it looked to me like the EPL was a lower scoring league than Spain's La Liga. The most frequent Total Goals in the EPL was 2, compared to 3 in Spain.
| Total Goals | Count of Match |
|---|---|
| 0 | 30 |
| 1 | 58 |
| 2 | 83 |
| 3 | 78 |
| 4 | 49 |
| 5 | 41 |
| 6 | 10 |
| 7 | 4 |
| 9 | 1 |
| Grand Total | 354 |
But digging a little deeper into the distribution, there are more games that finish over 2.5 goals in the Premiership compared to La Liga. There's not a lot in it, only a two percent difference but goes to show that the eyes do play tricks on you sometimes.
| Under 2.5 | Over 2.5 |
|---|---|
| 48% | 52% |
| Final Result | Count of Match | % |
|---|---|---|
| A | 101 | 29% |
| D | 95 | 27% |
| H | 158 | 45% |
| Grand Total | 354 | 100% |
Just as with La Liga, we observe a home field advantage, with the home team winning 45% of the matches. Comparing the final results between the two leagues, draws are more common in the Premier League, with 27% of matches resulting in a draw, compared to 23% in La Liga.
Again, let's define favorites as having odds to win at less than 1.8. Looking at Home games, there were 98 matches where the Home team was on odds less than 1.8. Here's how the matches ended:
| H | D | A |
|---|---|---|
| 73 | 18 | 7 |
Looking good for the Home teams so far, winning 73 out of 98 matches. But just as with the Spanish league, betting on the favorite for every match would have lost you money. Assuming a $1 bet each time, here's the actual return from 98 matches:
| H | D | A |
|---|---|---|
| $97.15 | $98.78 | $69.25 |
Betting on a Draw all the way would have netted you a small profit despite only 18 out of 98 matches being draws. In fact, if one had bet a dollar on draw in every match of the season, they'd have made a $16 profit (4.5%) over the 354 matches played.
Given that betting on draws are pretty darn profitable, can we predict them with greater accuracy than 27%? Here's something that has yielded decent results so far:
Unfortunately, i don't have the odds for a half-time draw but if you were to bet a dollar on a full time draw for all 97 matches, your return would have been $112.9. That's a 16% return.
My takeaway from analyzing both the premier league and la liga has been that betting on draws are a profitable strategy over the long run given the higher odds. If we could predict a draw with greater accuracy so as not to waste bankroll unnecessarily, then life would be perfect. 16% over a season is not too bad, but it isn't go out and buy a lambo kind of money either.
Betting on favorites seems to be a losing strategy both in the premier league and in la liga as well. Unless you have a decent filter, it isn't worth the risk.
All data from football-data.co.uk
