Scorecasting Economists

Lower League Forecasts, December 5-6

Fans are returning! In some parts of the country, that is. Is it going to affect outcomes on the field? The evidence suggests marginally, if at all. Covid doesn’t appear to have had much of a direct effect, either, really.

Here are our forecasts for this weekend’s matches. First, League One:

It’s hard to look beyond a Sunderland win at home against Wigan.

Here’s League Two:

A few unusual looking forecasts here, not least at Gillingham, Harrogate, Colchester and Barrow…

Premier League 5/12/20 – 7/12/20

Premier League 5/12/20 – 7/12/20

Man City v Fulham

Bookmakers are giving Fulham just a 4.4% chance of an away win against Man City later this afternoon (for Fulham fans, we have a marginally more positive chance of 5.8%!). Bookmaker mean odds are shown in the column on the right of the table below. The team we favour is shown in Outcome Probs in the 2nd column.

Chelsea v Leeds

Although Chelsea look very good this season, I can’t help feeling that Leeds have a slightly better chance than the 7.4% we give them of winning in our model. Bookmakers are inclined to agree as they give them a higher chance (16.6%)

Expect goals from the home teams in these games Man City’s is starting to edge towards 3 expected goals (2.52) and Chelsea are not far behind with 2.43.

Tottenham v Arsenal

The predictions for the North London Derby currently favour league leaders Tottenham. Arsenal are not helped by being 14th in the league whilst Tottenham currently have the meanest defence in the league (only letting in 9 goals so far). Arsenal have also only scored 10 goals so far this season so although we have predicted a draw as the most likely result (see Score Picks, Most in the 3rd Column below), both bookmakers and our model still favour Tottenham:

We have a 45.8% chance of a Tottenham win; Bookmaker mean odds have 48.6%

West Brom v Crystal Palace

Elsewhere in the division it looks like our model differs from bookmakers in the West Brom v Crystal Palace game where we favour West Brom, Crystal Palace are slightly edging ahead for bookmakers.

Burnley v Everton

In the first game today, Burnley v Everton, we favour Burnley whereas bookmakers are clearly favoring Everton (48.7% chance for an away win). Our model has predicted a draw though.

And of course Aston Villa v Newcastle last night was postponed.

 

Championship weekend of 4 December/5 December

Championship Scorecasts for weekend of 4 December 2020

This week we can see that every single home team in the Championship has at the very least a higher ‘Expected Goals’ score than the away team (See the Expected goals column in the table below)

In our model this is also reflected in the ‘Outcome Probabilities for Home teams’ (see table below, Outcome Probs). This means that in terms of percentage chance of a win, we favour the home team in each match.

Barnsley v Bournemouth

This is the only match where bookmakers favour the away team (Bournemouth) over the home team. They clearly favour second placed Bournemouth (44.9% to win) whereas we have a draw predicted as the most likely outcome (score picks, most) and only just favour Barnsley (by 0.5%) over Bournemouth.

Overall in the league though there are still lots of 1-1 draws being predicted, emphasising the closeness of the league.

Six 1-1 draws and six 1-0 home wins predicted as the most likely result from our model.

Relegation battle

The number of home wins predictions we have means likely bad news for all three of the teams in the relegation spots.

We have chances of an away win for Derby, Sheffield Wednesday and Wycombe ranging from 23.8% for Sheff.Wed. to 20% for Derby. Bookmakers largely agree with this – with their odds of an away win for these teams ranging from 24.2% for Derby to 21% for Wycombe.

Championship midweek, beginning of December 2020

Championship midweek, beginning of December 2020

A draw is predicted as the most likely result from our model for this week for 9 games in the Championship in this set of midweek fixtures (See Score Picks, Most).

Bournemouth v Preston

Exceptions however are Bournemouth v Preston – bookmakers agree with us (See Outcome Probs, for our likely outcome and Mean Odds, Home for average bookmaker odds) that top of the league Bournemouth against 18th placed Preston is a likely Bournemouth win. 2-0 is predicted with Bournemouth on 2.07 expected goals.

Cardiff v Huddersfield 

Cardiff v Huddersfield on the face of it sounds like much less likely to be a 2-0 victory for the home team if we look at where the teams are currently placed in the league (Huddersfield 13th and Cardiff 14th). However, in terms of the last few years, Huddersfield have not beaten Cardiff since 2003 – info on past results taken from:

https://www.11v11.com/teams/cardiff-city/tab/opposingTeams/opposition/Huddersfield%20Town/

which largely explains why the model has picked out a 2-0 victory for Cardiff. Bookmakers agree with us but there is a difference in how likely we think it is (62.4% in our model for a home win; 49.3% for bookmakers mean odds).

Elsewhere in the league our model predictions in Outcome Probs are pretty similar to the bookmakers Mean Odds with most games generally looking pretty close.

 

Lower League, December 1-2

After the FA Cup over the weekend, it’s back to the league for most teams. The exception basically is Bolton, Southend, Port Vale and Leyton Orient, who have already played their matches whilst everyone else was playing in the Cup.

The usual tables below, with expected goals (based on past matches), outcome probabilities (based on scoreline probabilities, which are further to the right in the tables). Then bookmaker odds furthest right, turned into probabilities by taking the reciprocal of the average decimal odds from Oddsportal for each match.

The most uneven match is at Sunderland as they host Burton. Sunderland are 58% likely to win according to our model, and 61% according to the bookies.

Focus on… Oldham and Tranmere

Hot on the heels of facing Bradford, Oldham are up against another of their traditional rivals – Tranmere Rovers.

Tranmere are Oldham’s third most common opponents, and tonight is the 89th clash. Oldham have won 35 of the previous 88, Tranmere 32, and there’s been 21 draws. Despite winning more, Oldham have scored fewer – 123 to Tranmere’s 124.

That’s in part because Oldham’s heaviest ever defeat came at the hands of Tranmere in only the second ever clash between the sides, on Boxing Day 1935. That day Tranmere finished 13-4 winners, and yet only a day after Oldham had won 4-1 against the men from the Wirral (yes, on Christmas Day). Tranmere have also won 6-2 (1969) and 4-0 (2005) over the subsequent years.

Those hectic days aside, it’s been a low scoring contest by and large, with 65 of the remaining 84 fixtures featuring each team scoring at most two goals. In the last 14 meetings since 2008 no teams has scored more than twice.

But both teams are in goalscoring form of late, with Tranmere dispatching Grimsby 5-0 and Port Vale 4-3, and Oldham winning 4-3 at Barrow.

That’s contained in the form in the table below:

Ignoring the fact Oldham are 17th not 16th, otherwise the stats slightly favour Tranmere, who are slight favourites to win tonight at 39% by our model, and 37% on average by the bookies.

But neither of these numbers factor in that Oldham have been doing well on Tuesdays nights in recent years. Since 2010, in general they’ve won just 32% of home matches, but when it comes to home matches on a Tuesday night, they’ve won 47%. That’s a significant difference. Who knows why?

It’s not obvious, which is probably why it may not be all that wise to bet on it.

Focus on Oldham Athletic… and Bradford City

Tomorrow, Oldham travel to Bradford. Not the for the first time, either. This will be the 60th time Oldham have made the 30 mile journey over the top to their Yorkshire rivals. And it’s going to be the 114th clash between the two sides, going back to 1907.

This may not be a Rangers and Celtic type rivalry, but it must rank as one of the most common fixtures amongst teams that have gone up and down the leagues. Looking into it, of course all the big Premier League teams have played each other over 150 times, and Rochdale, Hartlepool and Darlington have played each other a lot. But these are clashes that have almost exclusively taken place either in the top or bottom division of the Football League – in a time when relegation out of the Football League didn’t happen.

So, it’s one of the most common fixtures in English football, once you qualify it enough…

Below, you can see the two clubs and their Elo rating history since 1900. The blue line is Oldham, the burgundy line Bradford, and the black vertical lines at the bottom represent each fixture that’s taken place. So the fixtures have taken place in patches, but with a significant gap in the late 1960s, 1970s and 1980s, and with particularly many games in the last few years.

It’s a fixture with Cup pedigree. In 1950 almost to the day the teams competed in the FA Cup First Round, and on the exact day in 1950, November 28, that the clubs meet tomorrow, a replay took place at Boundary Park that Oldham won 2-1. Since then, the FA Cup history has really favoured the Yorkshire team, as they prevailed 3-1 in the First Round in 1955, and won 5-2 at Boundary Park in 1962 (also in the First Round), and then in 1987 hit Oldham for five again in a Third Round Replay.

So tomorrow will be the seventh FA Cup match. Even if the Cup history is a little rough, the overall recent history strongly favours Oldham. Oldham have won five of the last six matches, going back to December 2017. The summary of tomorrow’s game is given below:

Oldham are higher in the table, with four more points, and two more wins. Oldham have a worse goal difference though, but have scored and conceded more than Bradford.

The tables above also show recent clashes between the teams, and the recent matches played by both teams.

The bottom table is the likelihood of different match outcomes. As always, a 1-1 draw is most likely, at 10.7%. But that doesn’t mean it’s going to happen.

Bradford are slight favourites, but both our model and the bookmakers only put them in the region of 40-44% to win, with Oldham at 32-33% to win.

 

 

Championship 27/11/20 – 29/11/20

CHAMPIONSHIP END OF NOVEMBER

The English Championship continues to look closer each week than the Premier League (where some much higher odds for a win are now appearing weekly for the top teams).

Blackburn Rovers v Barnsley

I thought this one was interesting, firstly because Blackburn have so far scored more goals in the league than any other team (25 so far) but their expected goals are only 1.52 against Barnsley. The other thing that’s interesting is the bookmaker mean odds for Barnsley are identical to our model (24.4% of a win for Barnsley). There is only a 1.2% difference in the chances we give for a win (49.8%) to bookmakers (48.6%)

Birmingham City v Millwall and Reading v Bristol City

Both of these games have identical expected goals figures in our model for both home and away teams (1.17 each for the Birmingham v Millwall game and 1.23 each for the Reading v Bristol City game).

Both games are predicted to draw (See Score Picks, Most). There is almost nothing between Reading and Bristol City in our outcome probabilities either (Home 36.3% and Away 36.4%)

Derby v Wycombe

At the bottom of the table we have Derby beating Wycombe and bookmakers agree with us, with bookmakers having higher mean odds of a home win at 52.3% than our 45.6%.

Overall, across the games, our likely outcomes (Outcome Probs) are quite similar to bookmakers mean odds.

Premier League 27/11/20 – 30/11/20

PREMIER LEAGUE, END OF NOVEMBER WEEKEND

Man City v Burnley

Man City v Burnley is the first time this season I’ve seen the Mean Odds for a home win at over 80% (these are bookmaker mean odds taken from https://www.oddsportal.com). This seems all the more likely if Burnley goalkeeper, Pope is out with a head injury (not that our model can take this into account!).

We even have a 16.5% chance of a 2-0 win for Man City in our model (see Home wins, 2-0), higher than we have as Brighton’s overall odds of winning against Liverpool.

Chelsea v Tottenham 

The London derby of Chelsea v Tottenham should prove to be a close affair but both our model and bookmakers favour Chelsea. Chelsea lead on expected goals (1.61, Chelsea to 1.01 for Tottenham). The most likely score is a 1-1 draw according to our model.

Southampton v Man Utd

This match is interesting as the bookmaker odds are almost identical to ours in terms of outcomes (see Outcome Probs, for our model’s most likely outcome – 25.8% home win; and Mean Odds for bookmaker odds – 25.1% for a home win).

Elsewhere in the league bookmakers appear to agree with us that Leicester will win against Fulham, Arsenal are likely to beat Wolves (although their odds of that happening differ by 13.5%).

Overall though it likes another good weekend for goals. Leicester and Man City both have at least 2 for expected goals , Liverpool, Arsenal, Chelsea, Everton, West Ham and Man Utd are all on at least 1.5 for expected goals too.