Written by Alex Vella. Originally, this was going to be a 200-word post on our Facebook Group but I got a little excited. Apologies in advance to Man Utd, Chelsea and Liverpool fans! Feel free to highlight my mistakes.

I find Matchday 1 of any competition the greatest metaphor for sports betting. The amount of narratives that form from a sample size of one game is so interesting. Manchester United are the real deal after their 5-1 victory over Leeds, even though they lost to Villarreal on penalties in their last competitive match. Chelsea and Liverpool eased past Palace and Norwich, whilst Man City now lack a ruthless streak after winning the league by 12 points last season and winning the EPL three of the past four seasons. Don’t get me wrong, all of this could be true but I’d much rather base my opinions on more than just one game.

Even if you dive deep into one game and look to create a narrative, do it evenly. Don’t just read or listen to one article, podcast or pundit’s view, look at a variety of sources. Look at both quantitative and qualitative data. If you read just one article on each of those games I mentioned above, your opinion will be shaped by one opinion. So it might be something like this; "Man Utd, Liverpool and Chelsea are brilliant and Manchester City will struggle to capture another title this season." So here’s a way I like to quantify narratives, especially short-term ones; xG! (Images courtesy of Understat).

Manchester United scored fives goals against Leeds, which was unexpected to most, especially when you take into consideration how many they were expected to score from the shots they took (1.67). Greenwood’s goal and Bruno’s second goal were respectively 0.08 & 0.07 xG chances. In other words, those goals only go in 7 or 8% of the time! Now I know Manchester United are most likely going to outperform their xG over the course of a season, due to having better quality players than your average football side, but scoring five goals from 1.67 xG every game is not maintainable over 38 games.

Chelsea, similar to Man Utd, had a resounding victory over Palace but really rode the good side of variance. Alonso’s free kick was a 6% chance and Chalobah’s wonder goal was a 2% chance, with Chelsea producing a total of 1.19 xG over the 90 minutes. Don’t get me wrong, they beat Palace pretty comfortably, but a fair result would have been a one or two-nil victory considering the chances they had.

You probably get the idea of where I’m going with some of the Matchday 1 results. Liverpool defeated Norwich 3-0, but only beat Norwich by 0.46 xG. Leicester defeated Wolves 1-0, but lost to them comfortably on xG. And finally, Manchester City are in dire need of a ruthless streak (according to some media outlets), although they doubled Spurs’ xG and dominated everyone last season.

Think Long-term

As I said at the start, this could all be true. But just like sports betting, think long term. Last weekend’s results are full of randomness, just like every singular sporting game! I implore you to look past wins, losses and wonder goals that happen a couple times a season. Instead, look at big sample sizes and whether your bet beats the closing line of the sharpest bookmakers in the world consistently. Judge your team or any other team over a month or more of fixtures, short-term-based opinions expire really quickly.

I’ll finish up with this:

Manchester City were in 8th spot on Christmas Day last season (small sample) and they won the league by 12 points (bigger sample).