Scorecasting Economists

Premier League 26 and 27 December 2020

Premier League 26 December and 27 December

Scorecasting returns today after Christmas Day and gives West Brom almost no chance at all against Liverpool.

Our model has West Brom at 1.6% to win with Liverpool at over 90%. Liverpool are showing at 3.99 (so pretty much 4) expected goals, the highest our model has seen this season. Will West Brom play with ten men behind the ball for 90 minutes or has Allardyce already written this match off?

Bookmakers have almost exactly the same odds for Liverpool v West Brom and Man City v Newcastle, both with an 85% chance of win for the home team.

Elsewhere, apart from Aston Villa v Crystal Palace at 66.9% – 14.0%, matches appear much closer.

One big difference this week in our model and bookmakers is the Sheffield Utd v Everton match, where we favour Sheffield Utd (41.%) and bookmakers favour Everton at 53.1%.

 

 

 

Premier League 19/12/20 – 21/12/20

Premier League 19/12/20 – 21/12/20

The Premier League gets Friday night off but there are matches from Saturday through to Monday this weekend.

West Brom v Aston Villa

Big difference here between our model and bookmakers Mean Odds. Bookmakers have Aston Villa strongly favour Aston Villa (48.4% chance to win), whereas we actually have West Brom just ahead at 38.1% to 34.1% for Villa. This will of course be Sam Allardyce’s first match as West Brom manager so plenty of people will looking to see if he can repeat another relegation rescue. We’ve predicted a draw overall as the most likely result.

Elsewhere a number of matches look pretty uneven in terms of odds, Chelsea v West Ham, Palace v Liverpool, Southampton v Man City.

Manchester United v Leeds might be a bit closer than the predicted outcomes are showing (67.9% chance for a Man Utd win) with bookmakers at 56.2%; Tottenham v Leicester also has a 56.9% chance for Tottenham and bookmakers favouring Tottenham at 46.3%.

Everton v Arsenal

Everton v Arsenal should prove to be close though. We’ve predicted a 1-1 draw, favouring Arsenal at 41.2% and Everton at 31.8% to win, bookmakers slightly favour Everton (40.9% to 31.7%).

 

Championship 18/12/20 – 19/12/20

Championship 18/12/20 – 19/12/20

Across the games, there isn’t a lot of difference this weekend between what our model predicts as likely outcomes and bookmakers odds.

Wycombe v QPR

This is the one match that stands out with a big difference in odds with bookmakers favouring QPR, whilst our model favours Wycombe.

Wycombe are ahead on expected goals at 1.23 to QPR at 1.09.

We favour Wycombe at a 39.3% chance to win with QPR on 32.6%.

Overall the most likely score is 1-1 draw (13.2% chance of that score happening).

Bookmakers however favour QPR at 43.4% to Wycombe’s 29% (this is still far better than the 9% chance to win they gave them against Bournemouth earlier in the week). Neither team has won in the last 5 matches.

 

Brentford v Reading 

This match has the highest likely outcome in the division at 56.6% for a Brentford home win (see Outcome Probs). Bookmakers agree with this as they have Brentford at 55.3% for a home win.

Brentford are unbeaten in their last 5 matches, Reading have lost 2 of their last 3. Also Brentford have only lost three times this season, whereas Reading have lost 6 overall.

Our model has predicted a 1-0 win for Brentford (this score has a 12.6% chance of occurring).

On expected goals, Brentford have 1.71 to Reading’s 0.9.

The other match worth mentioning is Sheffield Wednesday v Coventry, where we marginally favour the away team, whilst bookmakers favour Sheffield Wednesday, our model still predicts a draw though.

 

Lower Leagues, December 15

More matches already? It’s hard enough work keeping up as a forecaster, let alone as a player. While Jurgen and Pep are moaning in the Premier League, it really is the lower leagues where playing squads aren’t so deep that the intense Covid-influenced schedule bites most.

So, another (almost) full round of fixtures. Lee Johnson’s Sunderland entertain AFC Wimbledon in what looks like League One’s biggest mismatch of the week.

In League Two, Oldham slayed their third (Bradford) manager since the start of 2018 on the weekend to move a little closer to the ultimate manager slayers Gillingham, and tonight face Walsall, in what should be a very exacting fixture given Oldham’s relatively poor home form and the resurgence in form of the Saddlers, who have dispensed with Tranmere and Bolton in the last week. Our model fancies Oldham (48% to win), whereas the bookies see Walsall as ever so slight favourites at 36%). I’d say it’s bound to be 1-1. But Oldham matches don’t tend to be that low scoring any more…

Championship midweek 15/12/20 – 16/12/20

Championship midweek 15/12/20 – 16/12/20

Reading v Norwich

Fifth placed Reading v second placed Norwich is one of this week’s biggest games in the Championship.

From our model it looks good for Reading, with Reading leading on expected goals 1.75 to Norwich 1.14.

The model has predicted a draw as the most likely result (see Score Picks, Most). A 1-1 draw is 11.1% likely as an outcome (see Draws in the table below).

However, if that is conditional on the most likely result occurring (usually a home win) then Reading are more likely to win 1-0 (see Score Picks, Cond). In terms of likely outcomes we favour Reading at 51.5% for a home win and Norwich at only 24.9% for an away win.

There is a ‘but’ though. Bookmakers disagree with us favouring Norwich at 41.1% with Mean Odds for an Away Win with Reading only at 30.9%.

Interestingly past results, courtesty of www.11v11.com

https://www.11v11.com/teams/reading/tab/opposingTeams/opposition/Norwich%20City/

show Reading and Norwich each having won 29 times (although over half these matches were in the old Division Three South!).

Bournemouth v Wycombe

Bad news for Wycombe this week, up against Bournemouth who are the highest scoring team in the league.

Bookmakers give Wycombe only a 9% chance of winning. Our model is slightly nicer (and this has nothing to do with a certain Wycombe fan in the Politics department at Reading) with Wycombe at 19.5%.

The good news for Wycombe fans though is that our model has actually given 1-1 as the most likely score.

Derby v Swansea

A mixed set of predictions here as our model favours Derby over Swansea (45.6% v 27.9% on expected outcomes), whereas bookmakers favour Swansea at 39.8% to 29.8% for Derby.

Swansea are third in the league though with Derby unbeaten in 5 but only one win in that period.

As with all bar two matches in the division, the model has predicted a 1-1 draw (the 2 exceptions being Cardiff v Birmingham City and Coventry City v Huddersfield, where home wins are predicted).

 

Premier League 15 12 20 -17 12 20

Premier League mid week 15 December – 17 December 

Liverpool v Tottenham

The top of the table clash clearly favours Liverpool in terms of odds.

We have a 1-0 Liverpool win predicted (see Score Picks, Most in the table below) – a 1-0 has a 13% chance of happening (see home wins, 1-0).

Liverpool lead on expected goals at 1.77 to Tottenham’s 0.84.

Our model favours Liverpool at 59.5% and Tottenham at 17.4% with bookmakers more or less agreeing at 53.6% to 21.6%

If this had been last season I’d agree with all this. However…. this season, it is clear that Tottenham are set up to deal with attacking sides and Mourinho will fancy his chances. Should be really interesting to see how they cope with Liverpool’s attacking force.

Fulham v Brighton

Near the other end of the table, this match is interesting because our model is predicting a clear Fulham win, whereas bookmakers slightly favour Brighton.

I think our model has this just about right as Fulham have started to look more organised in recent weeks. We have a 1-0 home win for Fulham as the most likely outcome (1-0 itself is at 10.5% but a Fulham win is at 58% overall (see outcome probs).

Man City v West Brom

Bookmakers give almost no chance of a win for poor West Brom (3.7%!) and sadly this is probably about right. We give them a marginally greater chance (7.4%!) but not much, with this being one of two matches we’ve predicted a 2-0 win for the home team (the other is Leeds v Newcastle).

Arsenal v Southampton

If i had to pick one match where I thought both the model and bookmakers were most likely to be wrong, this week it would be this match.

Both our model and bookmakers have Arsenal as favourites (51.5%, our model and 42.9%, bookmakers for Arsenal), our model even predicts a 1-0 win – and I suspect this is largely because in other seasons this would be right, as historically Arsenal beat Southampton.

However – Southampton are in 4th, something that actually makes the league look interesting, in addition to Leicester in 3rd and West Ham in 6th adds to this early season intrigue.

At the other end of the table, Arsenal are 15th and haven’t won in 5 matches (that win against Man Utd is starting to look like an anomaly). Also if you speak to Arsenal fans, the consensus seems to be, they think their own team looks rubbish at the moment!

 

Championship December 12 2020

Championship December 12 2020

Cardiff v Swansea

The two Welsh teams in the division meet today, with both teams in the top half of the table. Both teams have the same goal difference (+9) but Swansea are 4th, with Cardiff 9th.

Cardiff however have won their last 4 games and our model favours them over Swansea (1.37 – 0.91 on expected goals) and 47.6% (Cardiff) against 24.9% (Swansea). Bookmakers have the match much closer at 38.8% (Cardiff) against 31.7% (Swansea)

QPR v Reading

We have this as a draw, with bookmakers appearing to come to the same conclusion (almost nothing between home and away odds with bookmakers (35.9% for a win for QPR; 35.6% for a win for Reading).

Our model slightly favours QPR in terms of outcome probabilities and expected goals (39.4/34.6 for who is the most likely winner; 1.38/1.28 for expected goals). That said, Reading have often confounded our model and bookmakers predictions and are 5th in the league.

Wycombe v Coventry

In terms of odds, in our model, this  game looks to have quite a similar outcome to the QPR v Reading game.

It is slightly lower on expected goals and bookmakers more clearly favour Coventry (they are at 40.6% to win with bookmakers) but we have it as a draw and Wycombe at 34.1% to win and Coventry 37.1%.

 

Premier League Weekend of 11 November

Premier League 11 November 

Arsenal v Burnley

Both our model and bookmakers odds are not giving 18th placed Burnley much chance of winning against 15th placed Arsenal.

Despite Arsenal only scoring 10 goals in the league so far this season, Burnley have scored even less (five!) so this looks like one of the easiest games Arsenal are likely to get. We have a 72% chance for an Arsenal win and bookmakers have them at 61.7%. There is a higher chance of a 2-0 win for Arsenal (12.9% – see ‘Home Wins, 2-0’) than a win at all for Burnley (10.5%).

Manchester United v Manchester City

Somewhat strangely, the Manchester derby is actually 6th v 7th in the league  (although both have a game in hand). City have the 2nd joint meanest defence in the league (with Chelsea) with only 11 goals against. Man Utd, despite speculation about the sack for Ole Gunnar Solskjaer, have actually won their last 4 league games in a row.

We have a draw predicted as the most likely outcome but along with bookmakers heavily favour Manchester City (we have a 48% chance of City winning as the likely outcome – see ‘Outcome probs’ in the table below and bookmakers have them on 54.7% for a win).

Wolves v Aston Villa

Tomorrow lunchtime’s game looks like a close enough affair. Wolves are in 10th, with newly promoted Villa in 12th. Villa have only played 9 games though to most teams 11 and have scored 20 goals to Wolves 11. As with the Utd v City game, we have a draw as the most likely outcome but heavily favour one of the teams. In this case our model has Wolves a home win at 52.2% and bookmakers have them at 41.9%.

The next full set of games starts on Tuesday and ends on Thursday.