Premier League 2021/22: Expected points preview

After a challenging 2020/21 Premier League season, thanks to the majority of matches being played behind closed doors, the compacted schedule and the lack of time available for pre-season preparations, the forthcoming campaign should hopefully see things return far closer to what we would consider normal.

But as much as the clubs who will be battling it out at the top and bottom of the table appear reasonably obvious, where does the value lie in the betting market? Using expected goals and expected points, plus a look at how teams have traditionally performed from one season to the next, we can identify some interesting bets.

Premier League 2021/22 predictions: What do the odds say?

Let’s begin by looking at Pinnacle’s predicted Premier League table for 2021/22, which is based on their forecasts in the Total Points market, and see how they expect teams to perform in comparison to their figure from last season.

Team	2021/22 Projected Points	2020/21 Points	Difference
Manchester City	86.5	86	0.5
Chelsea	76.5	67	9.5
Liverpool	76.5	69	7.5
Manchester United	73.5	74	-0.5
Arsenal	61.5	61	0.5
Tottenham Hotspur	59.5	62	-2.5
Leicester City	58.5	66	-7.5
Everton	53.5	59	-5.5
Leeds United	53.5	59	-5.5
Aston Villa	49.5	55	-5.5
West Ham United	49.5	65	-15.5
Brighton and Hove Albion	45.5	41	4.5
Southampton	43.5	43	0.5
Wolverhampton	43.5	45	-1.5
Newcastle	39.5	45	-5.5
Burnley	38.5	39	-0.5
Brentford	36.5	N/A	N/A
Crystal Palace	36.5	44	-7.5
Norwich City	36.5	N/A	N/A
Watford	33.5	N/A	N/A

There is obviously no point of reference for the three promoted teams, and seven others are not expected to change their performance by more than three points in one direction or the other. For the remaining half of the division, it’s interesting to note that three are predicted to improve (by 4.5 points or more) while seven are expected to drop off by between 5.5 and 15.5 points.

The leading team for improvement is the newly crowned champions of Europe, Chelsea (at +9.5), which is understandable as they were better under Thomas Tuchel – taking 38 points from his 19 games – than they were in the first half of the season under Frank Lampard (29 from 19). Liverpool are expected to be 7.5 points better than in 2020/21, in their case as their terrible misfortune with injuries will surely have to turn for the better this season.

At the opposite end of the spectrum, there are six teams looking at a drop of between 5.5 and 7.5 points, which leaves West Ham United as the furthest outlier. David Moyes’ side are predicted by Pinnacle to get 15.5 fewer points in 2021/22 than they did last season. They face a Europa League campaign, which they’re not accustomed to and has proved draining for teams in the past, but does that explain their expected performance slump?

We need to look at the teams’ underlying statistics to see if there is any explanation within the data for these discrepancies, and if this can provide us with any worthwhile pointers for our bets.

Premier League: 2020/21 expected points vs. 2021/22 projections

There are different ways of obtaining or collecting expected points (xPts). The Understat website carries figures for xPts, which are based on their expected goal model.

However, while it is fine to use, their method doesn’t line up with reality. Take Manchester City’s 5-0 victory over Everton on the final day of last season, for which they assigned the champions 2.52 xPts and 0.33 for the Blues. Those figures don’t add up to the maximum of three which are available, and they’re somewhat irrelevant when a team can earn exactly three, one or zero points from a fixture.

An alternate method is to take the xG data from FiveThirtyEight and calculate xPts yourself by assigning a win if a side has over 0.4 expected goals more than their opponents.

Using that figure ensures that you get a realistic total of draws, and 0.4 xG is also approximately the value of an Opta-defined clear-cut chance so feels like a reasonable margin for declaring a victor in a match. Here’s how Pinnacle’s 2021/22 predictions compare to the xPts from last season on this method.

Team	2021/22 Projected Points	2020/21 Expected Points	Difference
Manchester City	86.5	97	-10.5
Chelsea	76.5	88	-11.5
Liverpool	76.5	83	-6.5
Manchester United	73.5	75	-1.5
Arsenal	61.5	57	4.5
Tottenham Hotspur	59.5	56	3.5
Leicester City	58.5	59	-0.5
Everton	53.5	46	7.5
Leeds United	53.5	53	0.5
Aston Villa	49.5	55	-5.5
West Ham United	49.5	59	-9.5
Brighton and Hove Albion	45.5	65	-19.5
Southampton	43.5	37	6.5
Wolverhampton	43.5	49	-5.5
Newcastle	39.5	45	-5.5
Burnley	38.5	31	7.5
Brentford	36.5	N/A	N/A
Crystal Palace	36.5	32	4.5
Norwich City	36.5	N/A	N/A
Watford	33.5	N/A	N/A

Comparing this with the first table in the article reveals some interesting discrepancies. Pinnacle expect both Chelsea and Liverpool to improve, though the margin by which they are anticipated to do so is not as wide as their underlying performances last season merit. While West Ham are still facing a points drop, the gap is smaller as they overachieved against their xPts by six last season.

The most interesting team in this section is undoubtedly Brighton, whose expected goal statistics were streets ahead of what they achieved in 2020/21. According to the FiveThirtyEight data they should have scored 55 goals yet they only found the net 40 times and this wastefulness cost them plenty of points. It will be fascinating to see if they can match their xG stats this season.

Outright Premier League betting: Who will win the Premier League title?

Manchester City are the 1.666* favourites here and deservedly so. As well as obviously being the defending champions, Pep Guardiola’s side scored the most goals in the division in 2020/21, conceded the fewest, and had the best underlying performance data too.

Manchester City are predicted to get 86.5 points this season, having won the league with 86 last season.

The likelihood of City retaining the title will only increase if they sign Jack Grealish and Harry Kane, as has been rumoured this summer, but those moves have not been made at the time of writing.

Transfers can have a huge impact on your bets though, particularly this season when three rounds of matches will have been played by the time the window closes. It’s rare for a team to win the league when not in the top four after six games have been played, so clubs who complete their transfer business late could be at a distinct disadvantage in the 2021/22 title race.

City make more of an intriguing prospect in the Total Points market in which they are predicted to get 86.5 points, half a point more than last season. Between 1999/00 and 2019/20 there were 21 instances of a team accumulating at least 86 points and only two of them (Manchester United in 2007/08 and Liverpool 11 years later) improved their points tally the following season.

This isn’t to say that City can’t become the third, but with their ongoing quest to secure a first Champions League title bound to occupy their thoughts in the second half of the campaign, an increase in league points can’t be taken for granted either.

Their crosstown rivals will hope the addition of Jadon Sancho (and possibly also Raphael Varane) can help them to win the Premier League for the first time in nine years. But as our expected points table illustrates, United were only the fourth best team in the league based on their underlying performance last season. If City were to get 86 points as predicted, then United would need to improve by 12 points to match them, something only four teams this century have done after getting at least 74 points the season before.

This line of thinking would likely exclude Chelsea and Liverpool from the title conversation too, though unlike United they did at least have the xPts in 2020/21 (88 and 83 respectively) to suggest they might have the wherewithal to reel in the reigning champions.

The latter will lose Mohamed Salah and Sadio Mane, their top two scorers from last season, for the Africa Cup of Nations but as it coincides with the early rounds of the FA Cup, it shouldn’t affect their league campaign too badly. When placing your bet for this season, it is certainly worth bearing in mind that your team of choice may lose players to international football for around in month early in 2022 though.

It’s hard to look past City for the title regardless, and it would be surprising if the same teams as last season didn’t finish in the top four alongside them, as the odds and last season’s xPts reflect.

Making Premier League 2021/22 season predictions

The three promoted sides are the leading trio in the relegation market, and all are predicted to get no more than 36.5 points (or 33.5 in Watford’s case). The odds imply that Brentford and Norwich are expected to fail to hit that target too. However, you must go all the way back to 2004/05 to find the last season in which all three clubs freshly up from the Championship got fewer than 37 points. Indeed, in each of the last four campaigns a newly promoted team has gone on to finish in the top half of the Premier League table.

Everton are 1.869* to finsh with under 53.5 points.

If we repeat our xPts method on the second tier, we find that the three promoted clubs were the three best teams in the division last season. This contrasts with 2019/20, where Leeds led the way but West Bromwich Albion were third and Fulham were down in ninth. Second in that season’s standings, and top in 2020/21, were Brentford. Despite finishing third and going up via the play-offs, Thomas Frank’s side have been playing well over a long period and look good value at 2.150* to get over 36.5 points.

While they face three of the big six clubs in their opening eight fixtures, those matches are at home and their away games are relatively benign until they go to Tottenham in early December. If they can make a good start and retain their best players, the Bees should stay up comfortably.

As alluded to in the expected points section, Brighton have been performing at a much higher standard in the last two seasons under Graham Potter than a quick glance at the league table would suggest. The Seagulls need the likes of Neal Maupay to regress to the mean – as he scored five goals fewer than expected in 2020/21, per FBRef – but with Pinnacle’s odds implying they are expected to finish with under 45.5 points, the over looks to be a sensible bet at 2.190*.

At the opposite end of the ‘xPts vs. Actual Points’ scale last season were Everton. On one hand they were unfortunate to only finish 10^th with 59 points, as the five previous teams to accrue that tally finished between fifth and seventh. But with one eye on what is to come we should note that their performances were only worth 46 xPts. The 2020/21 Toffees are a good case study in why you must look beyond expected goal difference, as their figure of -1 matched their standard goal difference, and focus on the specifics of individual matches.

They won more than their fair share of close games, especially on the road. Everton had nine away games where the expected goal difference was no more than 0.4 in either direction (and in five of them it was 0.1 or even), yet they won six, drew two and lost one of them. While not as prevalent at Goodison Park, they took seven points from the four matches which were similarly tight on the underlying data.

When making bets at this time of year it’s crucial to consider that managerial changes between seasons can impact upon the validity of looking at the stats from the previous campaign. A team’s tactics will probably change and the new man in charge may favour different members of the squad than the previous incumbent. With an unpopular manager having been installed this summer and the club’s transfer business looking uninspiring at present, it’s easy to see Everton finishing with under 53.5 points. They have done so in four of the last seven seasons and are priced at 1.869* to do so once again.