Since I started my Exploratory Data Analysis series, a potentially profitable pattern that stood out for me was betting on draws over the course of a season. For both the Bundesliga and EPL, betting on draws blindly over the season would have made you a small profit. On the other hand, this wouldn't have been the case in La Liga. Lesson learnt: betting on draws doesn't work in all leagues.
So clearly we want to bet on draws in leagues where the frequency of draws is higher. So how do we identify these leagues? As great as football-data.co.uk is, it doesn't have the numbers for the minor leagues. Thankfully, soccervista does: http://www.soccervista.com/soccer_leagues_ordered_by_number_of_draws.php
Italy's Serie B is listed in Group 1 - a league where the frequency of draws is high. We also have data from football-data.co.uk to check if we're on the right track. 429 matches have been played so far this season, with 149 ending in a Draw (34.7%).
If we were to bet a dollar on Draw for each of the 429 matches, our return would have been $480. That's a pretty decent 11.9% return by just betting on an outcome.
Anyone reading this must surely be wondering, "don't the bookie know about this already and if so, surely it has already been discounted?" That's a good point. My own guess is that surely they've done their homework and do know that a draw is more likely in leagues such as Serie B. What they're hedging against are the bets of the punters and people tend not to bet on draws.
