Going In-Play: Breaking down Premier League goals by time

If you’re an in-play trader with specific focus on the English Premier League, then my hope is that you will find this post extremely useful ahead of this weekend’s features.

The idea behind it is to provide some captivating visualisations that plot each team’s propensity to score in the 15-minute time segments of a 90+ minute contest. I thought I’d take a longitudinal approach here, because knowing when a team typically scores and concedes their goals can be very useful in numerous markets. In addition to this I’ve added some further context by plotting the league average against a team’s performance both at home and away, as we know behaviours can differ drastically depending on where they play.

To calculate goals per time segment, I used StrataData to extract a total of two datasets. The first comes from the current season, which includes all Premier League games from the 2nd of August until the 29th of November, while the second comes from the entirety of the 2015/16 campaign. The reason for using last season’s data is to look for trends and patterns that have continued on a team-by-team basis into this one, as our dataset for 2016/17 is still small when splitting by home and away.

As mentioned, I have split the 90+ minutes of games into 15-minute segments and separated out additional time altogether, which gives us a total of eight in all:

  1. 1 – 15
  2. 16 – 30
  3. 31 – 45
  4. 45+
  5. 46 – 60
  6. 61 – 75
  7. 76 – 90
  8. 90+

To provide you with a trading edge that you can apply immediately, I’ve focused on a sample of this weekend’s fixtures, where we’ll look at a home team’s goals for (i.e. goals scored at home) versus an away team’s goals against (i.e. goals conceded away) for this season. In addition to this we’ll also consider the inverse, so the home team’s goals against (i.e. goals conceded at home) versus the away team’s goals for (i.e. goals scored away from home). This gives us a direct comparison of the two teams that are playing each other and I’ve also plotted the league’s average goals scored for good measure:

averagegoalsperts_final4-03

Manchester City vs. Chelsea (Saturday 12.30pm)
There was only one place to start and what an absolute treat we have in store here! Chelsea’s sensational recent form has seen them rise to the summit of the Premier League and with Manchester City still title favourites and currently sitting on a -0.25 Asian Handicap, the market can barely separate the two. To try and help make up your mind, I’ve analysed their goal trends and found three key periods of the game to focus on.

Firstly, let’s look at the top-left of the graphic. This shows us that the opening 15 minutes of the game looks like it could be interesting, as Man City are typically strong starters at home. The data shows that they score most of their goals in this period (0.5 on average) and comfortably outscore the league average. This, coupled with Chelsea’s defensive frailties on the road in the same time period (0.32 goals conceded on average) may see City taking an early lead at The Etihad.

Another vital stage of the game looks likely to be the first 15 minutes of the second half, as City have averaged the same number of goals scored in the first 15 minutes of the second half as Chelsea have conceded (0.32). Again it points to the hosts being extremely dangerous after the whistle blows, while Chelsea look prone to lapses.

So, if Chelsea manage to survive those two periods unscathed then it is important to note that they are most clinical when it matters most, in the final 15 minutes of the game (76-90). They average 0.5 goals away from home in this spell, significantly outscoring the league average by over 0.4. So if it’s level heading into the final phase, it may be worth having a look at Chelsea’s price!

averagegoalsperts_final4-04

Tottenham Hotspur vs. Swansea City (Saturday 3pm)
Tottenham’s unbeaten run came to an end last weekend, but their good form in front of goal is highlighted in the data (top-left of the graphic), particularly in the second half of games. The first 15 minutes of the second could be decisive here if the contest is a close one, with Tottenham averaging 0.32 goals there and Swansea conceding an average of 0.5 in the same spell. So don’t be too surprised if the outcome of the match is determined early in the second half! If you are looking at trading the Total Goals in this game, there could be value in hoping against an early goal and buying in when the line drops to 2.5 (probably around the 20-30 minute mark) as it looks like it could explode into life after the interval.

Although Swansea’s second half away defensive performance this season (top-left again) appears to have improved slightly from last year (bottom-left), it seems the first 30 minutes of the second half is still their Achilles heel. However, there is reason to be positive, as the Swans can score late in an away game, averaging 0.32 goals between 76-90 minutes when on the road.

This bodes well for the Swans, but they will have to break down Tottenham’s defensive wall who have yet to concede after 76 minutes at home this season (top-right). Spurs can take pride from that fact, particularly given that they conceded an average of 0.3 goals between 76-90 minutes at home last season (bottom-right). There has been definite improvement there in 2016/17, so far at least!

averagegoalsperts_final4-05

West Ham United vs. Arsenal (Saturday 5.30pm)
Saturday evening sees Arsenal travelling to West Ham’s new home for the first time. Wenger’s men are probably banking on a strong start to the game, as they are yet to concede a first half goal this season away from The Emirates (top-left). On the other hand West Ham look to be consistent, scoring just under 0.2 goals in each segment of their home games. Arsenal should look at the 61-75 minute segment to ensure their defensive concentration doesn’t lapse, as they concede most of their away goals in this period (an average of just over 0.3).

The visitors have demonstrated a consistent attacking performance away from home this season and will be boosted by West Ham’s second half vulnerabilities at home (top-right). This season we see in the top-right that Arsenal are fast becoming dangerous at the death, scoring just shy of 0.7 goals in the last 15 minutes of games and just over 0.3 goals in added time. Although West Ham are yet to concede at home between 76-90’, it may be wise to not rule out a late Arsenal goal on Saturday, particularly if you’re looking for a late bet. All-in-all the away side at -0.75 looks good from the outset, but the braver amongst you might bank on a relatively slow start and then look to buy in as the line drops to -0.5 to get a full payout on a single goal victory.

Arsenal’s propensity to score late away goal bodes well with their ability to keep it tight at the very end of games on the road, again conceding 0 goals in added time (top-left). This is an improvement from last season (bottom-left) whereby Arsenal gave up 0.1 goals in added time. Finally, just a small note to say that West Ham have conceded more home goals than the league average this season, whereas last season they conceded less. For whatever reason, their home advantage has definitely diminished.

averagegoalsperts_final4-01

AFC Bournemouth vs. Liverpool (Sunday 1.30pm)
Liverpool travel south to Bournemouth and the data above highlights some very interesting trends. The erratic peaks and troughs in the graphic tell us that we could be in for a game that surpasses the total goals line of 2.75, and using StrataData I’ve picked out some key moments that could determine the final outcome for you.

Focusing on the top-left of the graphic (Bournemouth scoring at home versus Liverpool conceding away), both teams appear to be quite matched, which does not bode well for the visitors! The game looks like it could well start with an early home goal, as Bournemouth are averaging 0.32 goals in the first 15 minutes of games, whereas Liverpool are conceding just shy of 0.3 when away from home. In the 30 minutes after that, Liverpool are again conceding more than the league average, whilst Bournemouth are scoring more than the average.

This pattern also holds in the second half, as Liverpool appear to struggle between 61-75’, conceding an average of 0.59 goals in that period. If Bournemouth can continue their scoring form, don’t be surprised to see a home goal in this spell!

However, on the flip side Liverpool can change the course of this game with their attacking talent (top-right), even without Coutinho. Although they start slowly, averaging 0 away goals in the first half hour, they peak between 31 and 45 minutes, with an average of 0.5 goals scored. Bournemouth should be especially vigilant here, given that they concede over 0.4 goals in that timeframe. Another big period of volatility should then come between 61 and 75 minutes. Liverpool come to life here, with another peak average of 0.32 goals, and coupled with Bournemouth conceding almost 0.3, this spell may be worth keeping an eye on in relation to the Asian Handicap and Total Goals markets.

It’s interesting to note that Liverpool and Bournemouth’s goal trends from this season marry up with each other (bottom of graphic). In particular, Liverpool’s scoring pattern and Bournemouth’s conceding pattern (bottom-right) follow a similar trajectory, meaning that Liverpool are typically most dangerous in the moments when Bournemouth are most vulnerable. However, the same can be said about Liverpool’s vulnerability away from home (bottom-left), as they concede in the same junctures as Bournemouth score.

No doubt this one will finish 0-0 now, but it really looks to me like we should be in for a thrilling contest with plenty of goals!

averagegoalsperts_final4-02

Everton vs. Manchester United (Sunday 4pm)
From a game that should have plenty of goals to one that looks like it could be surprisingly tight…

Everton have been finding the back of the net consistently throughout home matches (top-left), often outscoring the league average. However, and quite ironically, in the period of a game where Manchester United concede most of their goals (0.67 in 61-75’), Everton have failed to find the back of their opponent’s net at home this season! Having compared United’s away goals conceded trajectory to last season (bottom-left), it’s interesting to note that their managerial change hasn’t drastically altered that trajectory, as they still concede most of their away goals in the middle of the second half.

On the other hand, after dispensing of Martinez it will be no surprise to hear that Everton have tightened up at the back this season (top-right) compared to last season (bottom-right), particularly between 61-90+ minutes. In fact Koeman’s side are yet to concede a home goal in this period (top-right), whereas last season they conceded most of their home goals (0.39) in that spell. United will be fully aware of Everton’s stronger defensive performance this season, and if they don’t take advantage when they average 0.31 goals in the second half (between 61-75’), then it probably doesn’t bode well for the rest of the match, as they average 0 away goals between 76 and 90 minutes (top-right)!

Having said that, Mourinho’s men are getting back into the habit of scoring added time goals. Last season under Van Gaal, they averaged 0 added time goals away from home (bottom-right), whereas so far this season they are averaging just shy of 0.2 (top-right), which is slightly better than the league average. With a reasonable Total Goals line set at 2.5 and an Asian Handicap of -0.5 to the visitors, this one looks likely to be pretty tight in the first 30 minutes especially, but there may be some action to be had from there if it remains 0-0.

Conclusions
Given the data, it looks like we could have some late goals on the cards this weekend. In particular there could be drama at the London Stadium if it stays tight, with Arsenal favoured to snatch it, while in the big contest Chelsea seem more capable of grabbing a late winner or equaliser depending on the game state.

Elsewhere Liverpool need to be wary of Bournemouth’s attacking talent throughout the game and an Over 2.75 trade looks good here. The Reds are already conceding too many goals for Klopp’s liking, being consistently equal to or greater than than the league average in the first half, with a spike in goals conceded between 61 and 75 minutes. Bournemouth’s scoring powers appear to peak in the segment after that (76-90mins), which means the rate of scoring should be consistently high for the entire contest.

Ultimately, Tottenham’s attacking talent throughout the game should see them through, but they will probably need to get into a +2 game state, as Swansea may be able to steal a late away goal based on past data. Finally, Everton look most likely to win it between 61 and 75 minutes, where they could take advantage of Man United’s weakest defensive period in a game.

So I hope that StrataData has helped you to see where some in-play opportunities might exist in this weekend’s Premier League fixtures. In a future post I plan to expand this analysis by incorporating game state behaviours, while also looking at when teams are most likely to win and concede corners, or create and give away chances.

Until next time!

Sagar Jilka (@DrSagarJilka)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s