Premier League Statistics from different angles

Introduction

For each of the last 12 seasons, I have kept records of  EPL results. 

One extra piece of data that I have maintained is the odds available at kick-off. 

To keep these as meaningful and consistent as possible, only a few of the leading bookmakers' odds are used.

The data has then been split both by results and odds in order to see whether any 'unusual' trends develop. 

These are the findings with 2018/19 included. 

Circumstances have led to time being on the short side, but it is hoped that more narrative can be added in the lead up to next season.

As ever, any comments or suggestions added to the blog will be answered.

For those readers interested in data, stats, and football form analysis please check out the betting school article - how to assess football form.

RESULT OF THE MATCH Based on HOME DRAW AWAY

The above table shows the returns to a 1 pt stake for every match by season. In general, the returns for home wins are far more consistent than those for draws and away wins.

The average return by season has been HOME 385.90 DRAW 365.56 AWAY 352.72.

RESULT OF THE MATCH Based on FAVOURITE DRAW UNDERDOG

The above table shows the returns to a 1 pt stake for every match in each season. Again, the favourite returns are more consistent. 

It is possibly no coincidence that the season with worst returns for favourites, best for underdogs, and only just short of best for draws was when Leicester unexpectedly won the title with several fancied teams under-performing.

The average return by season has been FAVOURITE 378.21 DRAW 365.56 UNDERDOG 360.41.

ODDS ON favourites

This survey commenced life as an attempt to see how closely odds related to goal difference (supremacy) and while that continues, obviously the data for any subsection is rather small. However, it has thrown up the unexpected. Backing all ODDS ON favs would have shown a profit, and this can be explored further. As this was so unexpected, there is currently no split between home and away, or specific ranges.

I did not anticipate this section would show any sort of profit and it has yet to be split into the results of home favourites and away favourites. I may also be able to find if there is a range of odds hat consistently returns a profit. 

Searching for DRAWS

The shortage of draws this season (71) can’t entirely be put down to the gulf in class between the top few teams and the rest. However, only two of the previous eleven seasons saw fewer than 90 draws, 78 in 2013/14 and 84 in 2016/17.

Any clues to CORRECT SCORES

I now have statistics for the previous 12 seasons, so the calculation may be done covering 4560 matches, a reasonable sample. These figures show the number of draws by frequency of home and away favourites, along with the respective annual season %.

That does not look significant at first glance, but on closer inspection, there is a slight bias in favour of score draws with an away favourite and a massive difference in favour of 0-0 draws with home favourites. Other scores are detailed later.

Have statistics / trends changed much over 12 seasons?

With a template in place for each season's statistics, I took the opportunity to check this season’s figures against those for the previous 11 seasons. These are my findings.
This blog now covers 380 matches per season over 12 seasons, so 4560 in all.
To call it a blog may be slightly distorting the situation though. The object of this particular exercise remains to offer as much information in as small a space as possible. It can then be an easy reference point without a load of clutter.
On the downside, the onus will be more on any reader to interpret the statistics relevant to themselves.

BOKO (Best Odds at Kick Off), These are compiled using just a few of the leading bookmakers.

After 5 seasons where the % overrounds had steadily shrunk, they were constant for 2 seasons, before each of the last 4 have shown an increase on the previous year.

UNDER / OVER 2.5 Goals

The number of overs grew steadily and reached a peak in 12. They have dropped since but generally remain at a higher level than several years ago.

The overall figure is 51.73% over 2.5 goals for 12 seasons.

BTTS

This has been completely random as far as I can see. The two lowest figures were found in the season with fewest goals and a season with one of the highest totals.
For the record, these were the average figures for the 12 seasons.

This season:

GOALS in a SEASON

Overall number of goals:

GOALS in a MATCH

These are the averages per season since 2007/08:

The totals for 6+ goals in a match over the 12 years have been 22 12 26 26 23 25 24 16 23 23 16 23.

Exact Goals in a Match (2007/08 to 2010/11):

Exact Goals in a Match (2011/12 to 2014/15):

Exact Goals in a Match (2015/16 to 2018/19):

Here's a breakdown of Over 1.5, 2.5 & 3.5 goals per match over the past 12 seasons:

LEVEL STAKES returns

Returns for HOME DRAW AWAY

Level stakes of 1 pt on all Home teams would have returned an average of 385.90 but this was due to a large degree on a very high figure in one year. The last 3 seasons have produced 3 of the top 5 returns.
Level stakes on the Away team varied enormously from 272 to 413 leaving an average of 352.72.
Level stakes on the draw returned an average of 365.56. This was a particularly bad season, returning just 282.83 pts. The previous lowest was 344.57.

Returns for FAVOURITE DRAW UNDERDOG

Blindly backing all favourites would have returned 378.21, the draw 365.56 and the underdog faring worst of all at 360.41.

ODDS ON FAVOURITES

As stated earlier, I do not currently have a split for Home v Away favs. Part of the reason for that is I am still in shock about the profit shown at level stakes over 12 years!
An average per season of 203.25 bets with a return of 206.63: virtually the same as when recorded after 8 seasons. This included a 25.40 pts loss in one season that somewhat put a dent in the overall figure.

Matches with Home favourites

These consistently outnumber the matches with Away favourites roughly in the ratio of Home 70% Away 30%
The seasonal average is 264/380 (69.47%) and the return for backing Home is 263.79 Draw 259.65 Away 238.30.

Matches with Away favourites

The seasonal average is 116/380 (30.53%) and the returns being Home 122.10 Draw 105.91 Away 114.42.

BACKING the DRAW

While backing draws blindly would have produced an average loss of about 11 pts per season from 380 bets, backing the draw where there was a big priced underdog in the hope that the outsider could manage a draw would have been significantly worse.
Draws are rather unsurprisingly more common when the sides are more evenly matched but of course, this is reflected in shorter odds for the draws.
To reduce the number of bets, the relative home and away odds can be used instead of the draw odds.

#1 Backing the draw when both sides are priced at evens or higher would have yielded an average annual profit of 6.48 pts from 176.42 bets @ ROI 3.70%
#2 Backing the draw when both sides are priced between 2.50 and 3.00 would have yielded an average annual profit of 2.01 pts from 37.60 bets @ ROI 5.34%.
#3 Backing the draw when both sides are priced between 2.50 and 3.50 would have yielded an average annual profit of 3.87 pts from 60.80 bets @ ROI 6.38%.

In the table these show as:-

#1 Narrow #2 Tight #3 Odds Ag

All three ranges had an LSP loss this season and in 4 of the 12 overall. #2 and #3 had 6 consecutive winning seasons prior to this.

CORRECT SCORES

The 12 most common scores, along with a home/away favourite split are as follows. The numbers represent the average occurrence per season over the whole period under review: all other scores have less than 10.

Perhaps the stand out figure is the 2-0 score which is so rare with an away favourite.

PERFORMANCE by RETURNS

For teams that have played in the Premier League since the 2007/08 season, here are their average returns per season:

Certain teams seem better value to back blind than others. Blackpool actually lead the way with a return of more than 52 pts over 38 matches but I am going to concentrate on teams that have played at least 5 of the last 12 seasons at this level.

HOME

12 teams are in profit: from the top.

Burnley (av 25.99), Stoke, Leicester, Manchester City, Newcastle, Everton, West Ham, Manchester United, Fulham, Tottenham, Swansea, and Arsenal (19.35)

AWAY

7 teams are in profit: from the top.

Crystal Palace (av 29.78), Swansea, West Brom, Aston Villa, Wigan, Watford, and Leicester (19.89).

COMBINED

12 teams are in profit: from the top.

Crystal Palace (av 47.58), Swansea, Leicester, Burnley, West Brom, Manchester City, Tottenham, Manchester United, Wigan, Newcastle, Stoke, and West Ham (38.44),

PERFORMANCE by POSITION

Trends suggest that the top teams represent little value due to the number of matches they must win to make a profit.

The 2018/19 season shows the following leading the way in each section. As you will see, they are mainly mid-table sides.

HOME Newcastle, Wolves, Leicester, Bournemouth, Burnley, Arsenal.

AWAY Crystal Palace, Leicester, Cardiff, West Ham, Wolves, Tottenham.

COMBINED Crystal Palace, Leicester, Newcastle, Wolves, West Ham, Cardiff.

Only Chelsea who topped the list won the title in 2016/17 but did not top either the home or away list. No other team has topped the list and finished in the top 5.