Luck plays a huge part in sport. But there’s only so far good fortune will carry a football team over the long-run. In time, weak teams will be found out and strong teams will prevail.
There’s a simple way to determine just how lucky (or unlucky) a team has been over a period of time (e.g. a season) using Expected Goals (xG). This gives a good idea of what to expect from their results in the future.
Media Misrepresentation of Performance
Football is a game of emotion. And with that comes a lot of rash, unsubstantiated opinions about the game.
Fans and the media primarily care about winning — and much less so about performance. As a result football teams are often credited, and criticised, on their results irrespective of how they actually played or what they really deserved.
An underdog gets hammered by their superior opposition for an entire match, but defies all odds by keeping a clean sheet. The final score is 0-0. The favourite was still, by far, the better team.
The narrative for these kinds of results are usually skewed in favour of the underdog. Pundits and tabloids will typically label the performance as “resilient” and claim the players showed “true character” and “mental strength” to hold on. The usual clichés come out.
Meanwhile the favourite will often be condemned — regardless of how well they actually played. Minor imperfections in their game/formation will be highlighted, and dwelled upon. The manager may even be called into question.
The fact is, the entire result was a whisker away from being “yet another loss for a poor side”. It was just unlucky for the favourite.
An away team wins 0-2. The home team bossed the game and wasted great chances, while the away team was awarded with a dubious penalty and scored a late header (their one and only shot on goal). There’s no doubt who should’ve won the game.
The media will often make out that the losing team got what they deserved for being wasteful, and that the winning team had a plan all along. The winning manager is often credited for a “masterstroke”– completely disregarding the fact that his team displayed a lack of creativity, and that this type of performance is not sustainable and will result in losses.
In these situations the home team will often be unfairly criticised for displaying a “poor work ethic” and “lack of quality”. But the fact is, it was just a bad day at the office. Nothing more than that.
Games are inaccurately assessed because the football media always wants to attribute a reason for why something happened, or didn’t happen. Good and bad luck is often disregarded. And that’s a positive thing — because if it carries through to the sports betting markets then there will be opportunities to find value.
Performances Make Results
The way I see it is that deviations from expected results will always occur in football. Things can go for and against you. That’s sport.
The bottom line is, goals and points come from good performances. Therefore performance is most crucial to a team’s long-term success.
So you really ought to look further than results when analysing teams or measuring them up against one another. For this I recommend using xG, which accounts for the most significant moments in games: the goal scoring opportunities. That’s a great starting point for football analysis and prediction.
Premier League Expected Points Table
The following ‘Expected Points’ table (taken from Understat) shows the points every Premier League team was expected to have achieved in the 2017/18 season given the chances that occurred in their games.
I’ve sorted the table in order of ‘Expected Points’. The leftmost column lists the team’s real finishing position.
The table confirms that Manchester City, who won the title in this season, deserved it… by a long shot. However, they were expected to notch 91.09 points — which is 8.91 less than their glamorous 100. So they still over-achieved, based on Understat’s Expected Goal estimates.
I’ve highlighted all significant outliers in yellow. These teams obtained significantly more, or less, points than they were expected to according to xG. Using this simple technique you can identify both lucky (and unlucky) teams for any of the major top flight leagues.
The Luckiest Teams
For this particular season (2017/18) it’s evident that two teams in particular over-achieved.
1. Manchester United
- Expected: 62.33
- Achieved: 81
Achieving +18.67 points over Understat’s estimate is enormous.
Some Man United fans may have felt disappointed they didn’t close the gap on local rivals, City. Yet perhaps they should have been content with a 2nd place finish, considering the Expected Points table ranked them 6th.
I predicted: “Given that the other top clubs achieved points tallies so closely aligned with Expected Points, it’s reasonable to think that United need to improve if they’re aim is to replicate, or better, this season’s finish.”
- Expected: 41
- Achieved: 54
Burnley built a reputation for ‘organisation’ and ‘resilience’. But could it ever last?
No doubt, they should be commended for their finishing position. Given their resources, pound for pound, I think Burnley achieved the most of any Premier League club in the 2017-18 season.
I predicted: “despite their impressive season, I believe there’s tough times ahead for Burnley. They’re going to need to raise their game, and adapt, if they want to replicate a similar finish again. +13 points over the Expected Points is a significant deviation. I don’t believe Burnley will continue to over-achieve, to this extent, for another season.”
The Unluckiest Teams
Three teams in particular were unlucky in the 2017/18 season.
1. Crystal Palace
- Expected: 58.03
- Achieved: 44
It looked as if mid-table Palace were short-changed by the end of the season. Expected Points placed them 7th.
I predicted: “Expect a solid finish for Palace next season, without a serious relegation scare.”
- Expected: 48.72
- Achieved: 36
xG suggested Southampton didn’t deserve to be in a relegation battle.
I predicted: “They ought to remain fairly comfortable next season, with a view to finish mid-table.”
3. West Bromwich Albion
- Expected: 42.96
- Achieved: 31
Relegated West Brom should’ve been well clear of the bottom three according to their Expected Points tally. It placed them 13th.
I predicted: “We can’t use their Premier League performance to estimate West Brom’s chances in the Championship — it’s a different competition.”
*2019 Update*: How Did The Following Season Go?
The Expected Points system helped me to make some bold — but generally accurate — predictions for the following 2018-19 season.
In particular, the Burnley and Man United predictions proved to be extremely precise, and certainly wouldn’t have been obvious to the average football punter:
- Man United plummeted to 6th place, precisely where the Expected Points table placed them in the previous season.
- Burnley secured safety in the Premier League — but only managed a 15th place spot. Just one place lower than expected points placed them in the previous season. It was an impressive estimate.
- Crystal Palace finished 12th. Not quite in line with the 7th place finish Expected Points gave, but well clear from relegation, and comfortably mid table.
- Southampton finished 16th, just shy of relegation. The Expected Points placed them 9th the previous season. That prediction proved to be far too ambitious.
Can Expected Points Improve Your Betting?
Expected Points highlights that points on the table does not always tally with performance. So it can be used to identify under & over-achieving teams over the course of several games, which makes for better-informed predictions for upcoming games, or the following season.
However, it’s important to recognise that not all deviations from Expected Points will “even out” over time. After all, nothing remains static in football; clubs are constantly changing and adapting. Some teams will improve while others tail off. So there’s no guarantee that ‘lucky’ teams will get their comeuppance, or ‘unlucky’ teams will improve their win ratio going forward.
For Expected Points to benefit sports bettors, the available odds must be heavily based around results, disregarding performance to a large extent. That way the “luckiest” teams might be overrated in upcoming fixtures — meaning their opponents, or the draw outcome, has value. Or the most “unlucky” teams might be underrated with a great price available if you backed them.
Learn more on football prediction from my post on How Expected Goals (xG) Can Be Used for Football Betting.
You May Also Like…
- The Luck Index. These models incorporate various statistics (not only xG). The ESPN Luck Index (in collaboration with the University of Bath) concurs that Manchester United, followed by Burnley were the most fortune teams in 2017-18 campaign.
- Other football stats. I’ve created a list of recommended free football statistics sites. These will also enable you to quantify performance.
- Football betting advice. My Guide to Football Betting suggests factors bettors can use to analyse games.
- Football betting models. I’ve written a guide to Creating a Football Prediction Model. Those of you looking to take sports betting seriously should read this.
- Football tipsters. There’s a lot of dodgy Tipsters out there; most won’t earn you money. Always use a legitimate tipster site.
- Accumulator vs Bet Builder vs Request A Bet — What’s The Difference? - March 28, 2023
- Best Sites For Free Tennis Statistics | Top Tennis Stats Websites - March 28, 2023
- Variance In Sports Betting | Understanding Swings In Results - March 28, 2023
I see the Expected points as a measure of a team’s game quality.
Since they derive from the expected goals, they represent the ability to create better scoring opportunities and to concede the worst.
Consequently, the actual points indeed tell us if the team has overperformed or underperformed and it is equally true that they tell us about the quality of the players involved.
I mean, between two teams with the same expected points, there will be one with more or less points than the other because the quality of his players is higher or lower than the average.
Get lucky backing #ASteppingStoneClub