Investigating Total Goals Lines (Part 2)

In the first part of this piece I looked at the impact of goals across various leagues to see how they affected Total Goals lines and the Asian Handicap market. The aim was to use StrataData to see if there is any edge to be gained from the patterns that we identify.

To expand the investigation further I now want to take a look at how chance creation affects these lines, if at all.

For starters, let me quickly explain the concept of the data we collect to refresh your memory…

At Stratagem we put chances into six buckets, labelled “Superb”, “Great”, “Very Good”, “Good”, “Fairly Good” and “Poor”. Each of these buckets has its own conversion rate, which is linear across all 22 competitions that we cover and is proven to hold significant edge over purely shot-based expected goals models. For reference, a fuller explanation can be found here.

figure1

The above table shows the average conversion rates for each of our six chance types. Using these figures we can estimate an expected goals total for each team in every game they play. This is the very top level of what can be done and utilises a simplistic method, as things such as game state are not taken into account for ease of use in the rest of this post.

An example of how this can be used is shown below:

figure2b

When summed, these chances equal 1.96 to Manchester City and 0.69 to Arsenal. As the game finished 2-1 this can be considered a good example of chances mirroring the outcome of the match. However, it should be noted that these are average conversion rates and that there will often be games where a team scores three Poor Chances and others where they create a host of Greats and miss them.

Going further, the combined total of chances created in this game equals 2.65, which makes it reasonable to assume that the Total Goals line for this game could have been set at either 2.5 or 2.75. It was actually 2.75 in this case, which proved to be extremely accurate.

Looking in broader strokes we can see how often the chance data is close to the Total Goals line and if it can be used as a predictive measure.

To do this I looked at how often the Total Goals lines were within a fixed margin of both the actual goals and the expected goals. This allows us to see how accurate StrataData is at determining how many goals there would be in a game if average conversion rates held true for all teams in all games.

Over the course of a number of games it can be useful for seeing which teams are over or underperforming consistently, against both the goal line and their expected goals scored and conceded. In theory this would allow smart traders to get ahead of the curve if the markets are slow to react.

Beginning with a 0.25 goal margin either side of the Total Goals Line I looked at how often the actual number of goals and the StrataData expected goals fell within this range, before increasing this margin to 0.5, 0.75 and 1:

figure3

The above graph shows that StrataData is better at being within the specified range of the Total Goals Line than actual goals is. This may seem counter-intuitive, but as both are reactive measures known after the game has been played it is key to see how teams could perform on a repeated basis rather than what they actually did.

To give a clearer example I will take two games and break them down completely:

The first is West Ham United vs. Watford from the 10th September 2016 in the Premier League. This game finished 4-2 to the away side and comfortably beat the pre-match Total Goals Line of 2.5.

So does this mean the line was set wrongly and it should have been higher to start with?

Not necessarily, because StrataData tells us a different story. The game was actually much tighter than the total goals scored suggests, with poor quality chances being converted at a much better than expected level to give an actual xG over the whole game of 1.47, which was well under the natural goals line.

So, then, is it safe to assume that Watford could consistently outscore and beat the Total Goals Line?

No, not at all… Take their home game against Hull City on the 22nd October in the Premier League, which had a Total Goals Line of 2.25 pre-match and was expected to be low scoring.

This time it was, as it finished 1-0 to Watford thanks to a late own goal. However, once again the score did not tell the whole story of the game. Indeed, StrataData suggests that had average conversion rates applied then the total xG would be 3.82, which comes out well above the Total Goals Line.

Of course it is easy to pick out individual games and fit a narrative to them, but crucially what StrataData can do is better estimate teams that are consistently over- and under-performing against the Total Goals Line. This in turn can be used in a variety of models as a better predictor of future performance than simply using actual goals, or just use as an indicator for which side of a Total Goals trade to take for bettors who rely more on feel.

Ultimately, any edge that traders can get over the market can be significant in terms of long-term gains and that is what StrataData aims to provide. So if you are a StrataBet customer and want more information on how to use the data collected in our Reviews to improve your trading, please feel free to get in touch at info@stratabet.com for some tips.

Dave Willoughby (@donceno)

Tactical Substitutions: Making a Difference from the Bench

Substitutes, those brave footballing soldiers thrust into the thick of battle in the hope of either turning or stemming the tide of momentum in a game. Some are intended to bring about positive change, while some are simply utilised to kill time at the end of a game.

As football supporters, we will all have our own opinions on our respective clubs and their substitution policy. However, just how effective are teams at employing their bench players over the course of 90 minutes?

Observing purely tactical substitutions made in the Premier League up to and including Round 16 back on the 13/14th December, I have burrowed into the approaches of different teams and the impact of changes made, with particular attention paid to impacts in the Total Goals market.

To begin with a very general base layer of information, I wanted to see the average time of the first tactical substitutions made in the Premier League. Looking at all 20 teams, the average is just under 66 minutes for the first tactical alteration (65.66 to be exact), giving justification and substance to the oft-stated thought that managers are generally prone to waiting until an hour of play has elapsed before tinkering with their system. However, does this leave coaches with enough time to affect the outcome of a game?

What we can immediately see from the table below is that there is quite a big difference of 11 minutes between AFC Bournemouth, the most proactive team, and Chelsea, unsurprisingly the least proactive team to make a first tactical change, but interestingly the remaining 18 teams all fit within that window. There also appears to be no discernible correlation between league position and average time of substitutions, with managers happy to stick with their preferred method regardless of the standings:

Figure1.png

So why do teams wait to make changes? And is there a reason why Bournemouth are making tactical changes before everyone else?

The short answer is “yes”. In their 16 games played (at the time of writing), Bournemouth have either trailed or been level 11 times when adjusting personnel. In this light, it’s hardly surprising that of the 43 tactical changes made by Eddie Howe, the vast majority have been positive, with just two of those bench appearances representing a defensive move.

So we know that Howe is willing to be the first to blink and try to change the course of the game, but how has it worked out for him?

What is noticeable is that they have scored 10 times after making a first change, which equates to one goal scored for every 4.3 tactical switches. However, on the other side of the coin, they have also conceded nine, though this will inevitably be because they have needed to chase a result and have opened up in search of an equaliser or a winner.

Does this first change make Bournemouth a more attractive option for in-play trading when considering the OU Markets?

Ultimately, not really, as on the eight occasions their games have been over the natural goal line (the line closest to 50/50 odds at kick-off), six of those matches have already passed the threshold by the time they’ve made a change, meaning that the odds for the next goal tend to be pretty short (goals breed goals and all that jazz).

In contrast, Chelsea are the last team to make their opening tactical substitution, almost by a full two minutes to Middlesbrough, who ranked second. The interesting finding with Chelsea is that though they are often last to make changes, they have actually made more tactical alterations than anybody else. This gives suggestion that they are simply looking to take time off the clock and become more conservative while leading.

Indeed, in their recent astonishing run of ten consecutive wins (at the time of writing), they have been leading in every single game at the point they make their first substitution (72 minutes, on average). Attention in the final quarter of matches therefore turns not towards extending their lead, but to consolidating their position. They have proven far happier to bring on a less attacking player such as Chalobah or Ivanovic, who have both made five appearances off the bench each, though Fabregas has also proven himself to be a useful resource in terms of keeping possession and creating chances, stepping off the bench four times with two assists.

Hazard and Pedro have been two of the most substituted players in the league this season, suggesting Conte either doesn’t trust their defensive work rate late in games, or treasures security above maintaining a threat going forward. Perhaps the most impressive thing about Chelsea is their ability to close out games, particularly those that are low scoring. They’ve not conceded a single goal after making their first tactical change, being the only team in the league to have such a statistic.

In general they have shown that they are more than comfortable seeing out games that fall under the natural goal line, as when they do rack up bigger score lines, the damage is often done much earlier as they blitz opponents with quick goals. This makes them hard to gauge from an in-play perspective, with there likely being very little value in backing unders, unless they are nursing a small lead or sit level at half time. You can often take advantage of an overreaction at 1-0 in a game involving a heavy favourite, if you know them to be the type of favourite who are just happy to win, rather than going out looking to win big.

In terms of the volume of goals scored and/or conceded after the opening tactical substitution there are a couple of other teams who jump out, ranging from Crystal Palace at the top of the chart with 27 goals, down to Middlesbrough with a paltry 6:

Figure2.png

Starting with Palace, it’s been a roller coaster of a season for Alan Pardew and his men, who have found results very hard to come by. Their attacking nature has yielded a lot of goals, but has also seen a lot going in at the other end, particularly later in matches and after they make their first tactical change around the 63 minute mark.

On average their matches witness 1.6 goals in either direction after this mark, making them a very good candidate for backing the overs both deadball and in-play, particularly if the match is finely poised. They simply don’t have the mentality or ability to keep things tight, which is something that has proven very costly with a string of late defeats. Up to and including Round 16, Crystal Palace’s last four losses have come as a direct result of a goal conceded inside the final 10 minutes of play.

Games that go over the natural line have been the way for Palace this season, beating it in 59% of matches played. With their current defensive woes showing no real sign of letting up, it could be a trend that continues, even though they are without Connor Wickham for the rest of the season, who had been making an impact from the bench with two goals so far.

Concurrently, Middlesbrough are very limited in their output after their first tactical substitution, though that could come down to leaving themselves just 20 minutes to influence the game, with Aitor Karanka reluctant to make changes in personnel. He’s made just 37 tactical switches all season, not even using his allotted three options on average.

Goals have been hard to come by at the best of times for Boro, who have set their stall out to be a defensively organised unit first and foremost, which has reaped big rewards with points gained in difficult away matches to both Arsenal and Manchester City. However, what is perhaps even more intriguing is that aside from a pair of defeats to Liverpool and Everton and a win over Bournemouth, all of their games played up until Round 16 have been decided by a one-goal margin when there has been a victor.

In addition to this, half of their games have fallen under the natural goal line, while contributions from their substitutes have been minimal, with nobody stepping off the bench in a tactical change producing a goal or an assist. With substitutes being made late and often of a like-for-like nature, Middlesbrough are generally happy to stick with their game plan instead of throwing a bit more caution to the wind, which goes a long way to explaining their lack of goals so far. Karanka’s emphasis seems to be more on not losing games, rather than actively pursuing wins.

At a slightly more granular level, it’s not a major surprise to see that Arsenal, the team with the most goals scored after their first tactical substitution, are also seeing big contributions from those stepping off the bench. Olivier Giroud stands out as the Premier League’s super-sub this season, having chipped in with three goals and two assists to change the course of matches, though Man City’s Nolito has contributed a goal more. His goals alone have secured four points for City, providing an invaluable contribution. However, at Arsenal it’s not just the Frenchman doing the business when coming on, with Iwobi and Xhaka also having found the net, while Ramsey, Mohamed El Nenny and Oxlade-Chamberlain have weighed in with an assist apiece.

Arsenal themselves sit around mid-table in terms of the average time of their first tactical substitution and quantity of tactical changes, but the impact those alterations are having is yielding a big return, highlighting the value of a strong bench to call upon when needed. 53% of Arsenal games have finished over the natural goal line, while their ability to finish matches well is also a beneficial factor, with the team not letting up, which can also be of help to their Asian Handicap marks, which they have beaten 41% of the time in 2016/17 to date. There seems to be more benefit in sticking with Arsenal for the full 90 minutes if backing overs, even if the game is still in the balance or fairly low scoring heading into the final phase of play.

Looking through a wider lens, it’s interesting that just four players have managed more than one goal from the bench this season, making it a fairly exclusive club, though in total the net has been found 38 times by those introduced for tactical reasons:

Figure3.png

In terms of creativity, even fewer players have made multiple assists, with Liverpool’s Daniel Sturridge something of a surprise inclusion, taking some weight out of the suggestion that he isn’t a particularly strong team player:

figure4

Other notable discoveries from this piece are how indifferent Man Utd are after making a change, scoring just four times, while Hull seem to let the floodgates open when pressing to get themselves back into matches, shipping close to a goal a game in the minutes following their first substitution. It is easy to argue a case for why Hull are struggling at the foot of the table with statistics like those, with their squad depth being poor after a summer of late and rushed recruitment, as well as an extensive injury list that has been present almost all season, making it difficult for Mike Phelan to have a big influence with his touchline decisions. Man Utd on the other hand are a bit more intriguing, with their struggle to kill games off being evidenced by the lack of goals coming towards the end of matches.

Overall it seems that there are a few valuable pieces of insight with regards to the Total Goals market and substitutions being made within the Premier League this season, with it appearing teams near the lower reaches of the table are seeing more goals and perhaps providing more opportunities for in-play trading, with their need to play more openly to claw back deficits making for more expansive matches.

Arsenal and Man City both sit highly in the tables in terms of goal output, however, with strong attacking threat coming from the bench making a big impact in the latter stages of matches. There is a notable correlation with the more proactive teams and the number of goals scored, though in theory the more minutes players have to make an impact the greater their chance is of doing so. Chelsea and Middlesbrough, the two latest teams to make changes, unsurprisingly concede fewer goals in the final quarter, but their defences stand up over 90 minutes generally so there is no real need for them to be reactive and push forward, though in Middlesbrough’s goal-shy case it certainly wouldn’t help!

In summary if looking to improve your in-play goals trading it would be prudent to consider substitutes as a potential trigger with some teams to get in/out of trades quickly, where value could be had before the alterations have the chance to really take effect.

Alec Payne (@Payney3)

Weekend Review – 20/12/2016

Highlights

Italian Serie A
Juventus 1-0 Roma (Saturday 7.45pm)
The biggest game of the weekend also turned out to be our biggest position, as we backed Juventus to maintain their formidable home record against the second best side in Serie A. We wobbled a little bit when the starting elevens were announced and both Barzagli and Dybala were only named among the substitutes, though solace was taken from the fact that Salah was in the same position for Roma. In the end the 1-0 result was fair, with Juve creating the only two Great Chances in the contest and ending up with a final Expected Goals sum of 1.32, according to our model.

Roma gave the hosts a little scare in the middle portion of the second half but were never really able to take control of the game, with the momentum line staying firmly with Juventus as they maintained a slight edge from Higuain’s superb opener until full time. We shied away from doing anything in the Total Goals market on this one, but it was very much an Under 2.5 game, so that was a missed opportunity. However, the -0.5 paid off handsomely and Juve look a shoe-in to retain their title for the sixth consecutive season.

Recommendation: Juventus -0.5 @ 1.80
Actual Outcome: +1 (1-0)
Fair Outcome: +1 (1-0)

English Premier League
Manchester City 2-1 Arsenal (Sunday 4pm)
There was a definite sense of déjà vu for Arsenal here, as they surrendered an early 0-1 lead on the road for the second time in a week to lose 2-1. This time around Wenger’s side did not start especially well, but were able to grab the lead against the run of play and looked threatening against City’s shaky back line almost every time they got the ball upfield. Then, for whatever reason, they ceded control and the below-strength hosts took full advantage.

There was no way our tentative recommendation of Arsenal +0.5 deserved to pay out, but the half-win on the Total Goals trade came in quite comfortably despite the total sum of xGoals only being 2.65. Looking longer-term this was a huge win for City, while Arsenal are wobbling a little earlier than normal this season. Before 2016/17 kicked off we took a speculative position laying them for Top Four and with Tottenham finding some consistency and Man Utd starting to motor it’s beginning to move towards green after a long time red.

Recommendation: Arsenal +0.5 @ 1.85 and Over 2.75 @ 1.94
Actual Outcome: -0.5 (2-1)
Fair Outcome: -0.5 (2-1)

French Ligue 1
Monaco 1-3 Olympique Lyonnais (Sunday 7.45pm)
As sure as the sun rises and sets, Monaco’s home games will go over the natural goal line. Apparently this is even the case if the game itself should only have been 1-1, as according to our review. Our analyst rightly noted that although the total sum of xGoals was 3.45, Lyon’s best pair of chances came late and with the game “dead” at 1-3, while Monaco were denied a clear penalty and Lyon were awarded one incorrectly.

This gives a good illustration of the complexity that goes into generating an Analyst Fair Score, although whether it should have been 1-3, 1-2, 1-1 or even 2-2 we cannot shy away from the fact that every outcome would have led to a loss for our Asian Handicap recommendation. Put simply, Lyon turned up in this ill-tempered affair, which hasn’t been the case too often this season. They seem to be back.

Recommendation: Monaco -0.5 @ 1.95 and Over 2.75 @ 1.90
Actual Outcome: 0 (1-3)
Fair Outcome: -2 (1-1)

Dutch Eredivisie
Ajax 1-1 PSV (Sunday 3.45pm)
This was a rare game where our recommendation of Under 2.5 paid off in actuality, despite the Analyst Fair Score pegging it as a 2-2 due to the surprisingly open nature of the play. Taking subjectivity out of it the teams each had two Great Chances, three Very Good Chances and four Good Chances between them, which would easily point to a contest that should have ended up over the allocated Total Goals line. From our perspective this was very much a “big game” bet and one that followed recent trends, but it provides a good lesson in how historical patterns can occasionally mean little when there is so much motivation and relative quality on show. We got away with this one in the end.

Recommendation: Under 2.5 @ 1.95
Actual Outcome: +1 (1-1)
Fair Outcome: -1 (2-2)

English Premier League
Everton 0-1 Liverpool (Monday 8pm)
“Merseyside Monday” produced a cagey affair that never looked like seeing our overs recommendation come off, although the xGoals sum was 2.64 thanks to things opening up more towards the end when Liverpool sought the winner that they deservedly got. This was one of those games that would have likely exploded with an earlier goal, but Mane’s strike arrived deep into added time and left no room for a response. It would have been nice to see the Blues having a go at a Liverpool defence minus Matip, but we have to chalk this one down as a bad tip.

Recommendation: Over 2.5 @ 1.79
Actual Outcome: -1 (0-1)
Fair Outcome: -1 (0-1)

Ones to Watch

German Bundesliga 1
Hoffenheim vs. Borussia Dortmund (Friday 7.30pm)
Hoffenheim remain the only unbeaten side in the Bundesliga after a deserved 2-2 draw with Dortmund, where the hosts led twice even before Reus’ controversial red just before half time. There was great value to be had at +0.5 here, with that line a winner whichever way you slice it.

Recommendation: Hoffenheim +0.5 @ 2.05
Actual Outcome: +1 (2-2)
Fair Outcome: +1 (2-2)

German Bundesliga 1
RB Leipzig vs. Hertha BSC (Saturday 2.30pm)
Pre-game we wondered whether Leipzig could bounce back from their first Bundesliga defeat of 2016/17 against similar surprise package Hertha, and the answer was a resounding “yes”. We shied away from their line of -0.75 but were incorrect to, with the home side outshooting their visitors by 1.65 on xG.

Recommendation: N/A
Actual Outcome: N/A (2-0)
Fair Outcome: N/A (2-0)

Spanish Primera Division
Barcelona 4-1 Espanyol (Sunday 7.45pm)
An enforced goalkeeping substitute for Espanyol was largely to blame for their failure to cover a huge handicap in the Catalan Derby, with the away side last until 67’ before things started to disintegrate. However, the performances of Messi and Co. were well worth a three-goal margin by the end.

Recommendation: Espanyol +2.75 @ 1.83
Actual Outcome: -0.5 (4-1)
Fair Outcome: -0.5 (3-0)

Portuguese Primeira Liga
Sporting CP 0-1 Sporting Braga (Sunday 8.15pm)
Braga were without a permanent manager, but made a mockery of their +1.25 Asian Handicap with an almost deserved victory. Sporting were again made to pay for poor finishing, but in truth they could not argue for anything more than a 1-1, which would still have been a losing pick.

Recommendation: Sporting CP -1 @ 1.70
Actual Outcome: -1 (0-1)
Fair Outcome: -1 (1-1)

Australian A-League
Melbourne City 1-2 Melbourne Victory (Saturday 8.50am)
Cahill did end up playing and scoring to help us on our way to a half-win on this goals trade, but his side were second best for almost the entire contest after he found the net early on. Three goals was fair enough, with 3.33 being the total xG sum.

Recommendation: Over 2.75 @ 1.95
Actual Outcome: +0.5 (1-2)
Fair Outcome: +0.5 (1-2)

11 Recommendations
Total Actual Outcome: +0.5
Total Fair Outcome: -2

Weekend Preview – 16/12/16

Highlights

Italian Serie A
Juventus vs. Roma (Saturday 7.45pm)
For the second consecutive week the game of biggest interest comes from Serie A and involves Roma, with Luciano Spalletti’s side heading to Turin to face Juventus. Last time out they overcame Milan 1-0 at home to create daylight between the two teams in second and third, but this time they are looking to reduce the deficit to the only team above them. Juventus have a four point lead and have surprisingly been over the natural goals line in four out of five recent league games, while their team news is improving all of the time with Barzagli back to bolster the defence and Dybala ready to start up front.

Roma are still without Peres and Florenzi, but Salah could make the bench and both Paredes and Totti should be back to add to the options available to the manager. The market currently has the lines set at -0.75 on Juventus and 2.5 goals, which is hard to argue with, and Juventus have been backed in steadily since the open. We would take the hosts on at -0.5 based on their superior team news and because they’ve won their last 24 home league games, but we’ll dodge the total goals.

Recommendation: Juventus -0.5 @ 1.80

English Premier League
Manchester City vs. Arsenal (Sunday 4pm)
Pep Guardiola and his players went some way to quelling the questions around his methods with a 2-0 win over Watford in midweek, but despite a fairly strong performance in terms of territory and chances the result was not completely convincing. City are still without Aguero, Fernandinho and Kompany, while Gundogan has joined them on the sidelines with a knee injury that looks set to keep him out until next season. Arsenal have no Cazorla and continue to miss his creative influence in the middle of the pitch, while Mustafi and Ramsey are both absent again too.

This is not an ideal time for Guardiola’s defence to come up against Wenger’s attack, which makes a goals line of 2.75 look attractive, but we find it difficult to trust either team enough to support them in the Asian Handicap market. Still, if pushed we would back Arsenal for the simple fact that they can be had at +0.5 and have beaten their natural handicap lines 44% of the time this season, compared to City’s measly 31%.

Recommendation: Arsenal +0.5 @ 1.85 and Over 2.75 @ 1.94

French Ligue 1
Monaco vs. Olympique Lyonnais (Sunday 7.45pm)
There is not usually much in Ligue 1 to get the pulse racing, but Monaco continue to surprise and delight this season. They are now averaging 3.12 goals per game following their 0-4 win at Bordeaux, which came off a tiny -0.25 handicap. This was then followed up by a 7-0 demolition of Rennes in the cup on Wednesday, while visitors Lyon were labouring to a penalties defeat to Guingamp at the same time.

Given the respective performance levels of these two sides in 2016/17 it’s a bit of a shock to see Monaco as short as -0.5, but Lyon are in their best run of the season and have a completely healthy squad to choose from. In addition to this the Total Goals line of 2.75 looks a bit off, with basic averages of the two teams closer to 3.5. Unsurprisingly we like the hosts and the overs here.

Recommendation: Monaco -0.5 @ 1.95 and Over 2.75 @ 1.90

Dutch Eredivisie
Ajax vs. PSV (Sunday 3.45pm)
Two of the Netherlands’ three dominant teams meet here, with a draw the outcome that neither want due to Feyenoord hosting Vitesse and being expected to pick up yet another three points. PSV have the momentum in terms of their actual and fair form, which means that an Asian Handicap of -0.25 on Ajax owes everything to their home advantage. The big team news question mark is over Guardado, who has been out since early November but has a chance of making the bench, while for Ajax only Viergever is missing due to suspension. Both sides have come in Under 2.5 a lot more than usual recently, which explains why the line looks low on first sight, and we will back that trend to continue.

Recommendation: Under 2.5 @ 1.95

English Premier League
Everton vs. Liverpool (Monday 8pm)
The first Merseyside Derby of the season sees Everton on a high after their thrilling comeback win over Arsenal, while Liverpool enter off a routine 3-0 victory at Middlesbrough. A natural handicap of -0.5 on the visitors makes sense, given that they are second and Everton eighth, but we like the look of the goals here, which can be bought for close to 1.80. It’s hard to imagine that either side will fail to hit the net, especially if Matip fails to pass fit for Liverpool, and we’ll bank on this game having a decisive outcome instead of ending 1-1.

Recommendation: Over 2.5 @ 1.79

Ones to Watch

German Bundesliga 1
Hoffenheim vs. Borussia Dortmund (Friday 7.30pm)
Dortmund continue to flounder this season, sitting in sixth ahead of their trip to the surprisingly high-flying Hoffenheim. The market has set this one at +0.5 on the hosts and with a goals line of 3.25, and we feel like Hoffenheim on the double chance represents solid value on form.

Recommendation: Hoffenheim +0.5 @ 2.05

German Bundesliga 1
Leipzig vs. Hertha BSC (Saturday 2.30pm)
The other game of interest in the Bundesliga revolves around whether Leipzig can bounce back from their first defeat of the season by beating Hertha, who are right behind them in the table. With Compper back we think this is probable, but are not confident enough off a -0.75 mark.

Recommendation: N/A

Spanish Primera Division
Barcelona vs. Espanyol (Sunday 7.45pm)
Both teams are near enough full strength for the Catalan Derby, with Neymar back from suspension for the hosts and Leo Baptistao expected to pass fit for the visitors. Perez could miss out with a knock, but even if he does we still like Espanyol to beat the +2.75 handicap.

Recommendation: Espanyol +2.75 @ 1.83

Portuguese Primeira Liga
Sporting CP vs. Sporting Braga (Sunday 8.15pm)
The home side unluckily lost their big game with league leaders Benfica last weekend and should be fully motivated to make amends against a manager-less Braga side, who are occupying their typical spot of fourth. At first glance the -1.25 looks wide, but we’ll gladly take on the -1 instead.

Recommendation: Sporting CP -1 @ 1.70

Australian A-League
Melbourne City vs. Melbourne Victory (Saturday 8.50am)
The second Melbourne Derby of the season unsurprisingly has the home side on a -0.25 handicap, with a typical A-League goals line of 2.75. City’s Cahill being a doubt is the biggest worry for this recommendation, but Kamau is set to return and Rojas has passed fit for the visitors.

Recommendation: Over 2.75 @ 1.95

*Odds accurate at time of writing

Weekend Review – 13/12/16

Highlights

Italian Serie A
Roma 1-0 Milan (Monday 8pm)
These two teams were joint-second in Serie A before kick-off, but a narrow and slightly unmerited win for Roma means that they now have three points of daylight back to I Rossoneri. Ahead of this one we thought that the -1 Asian Handicap on the hosts looked wide, but that the Total Goals line of 3 was fairly pitched based on the teams recent performances when home and away, respectively. In truth the first observation proved to be more accurate than the latter, though the goals line could have been under threat had Niang scored from the spot in the 26th minute.

Szczesny denied him after giving away the penalty in the first place, but even factoring this in the final xGoals values for both teams by full time were just 0.74 versus 0.72. Roma’s goal came from a poor chance and owed something to sloppy goalkeeping by Donnarumma, but the home side held out well after taking the lead and gave next to nothing away in the latter stages. Ultimately it is tough to justify a 1-0, with 0-0 or 1-1 more Fair Scores based on the very balanced nature of the play and the number of chances created.

Recommendation: N/A
Actual Outcome: N/A (1-0)
Fair Outcome: N/A (0-0)

English Premier League
Manchester United 1-0 Tottenham Hotspur (Sunday 2.15pm)
Jose Mourinho clawed back three points on Tottenham in what felt like a pivotal win for the Portuguese manager at Old Trafford. He again saw his side create many more chances than they were able to score, but this time the opposition had a below-par afternoon and did not punish his players for their profligacy. Tottenham were surprisingly open given that their first choice back four was intact again, perhaps pointing to Dier’s importance as a screen in front of the them.

In the end the xGoals values for each team were 1.46 to 0.80, showing that United were deserved victors. This means that our recommendation of Spurs +0.25 was way off in actuality, though we were a little hard done by with the game not at least reaching two goals for a half-loss on the Over 2.25 call. We even got the relatively early one that we were looking for, but Tottenham’s best and only significant chance came from a set piece when Wanyama made a mess of a headed opportunity from Eriksen’s free-kick.

Recommendation: Tottenham Hotspur +0.25 and Over 2.25
Actual Outcome: -2 (1-0)
Fair Outcome: -1.5 (2-0)

French Ligue 1
PSG 2-2 Nice (Sunday 7.45pm)
A performance of supreme attacking efficiency saw Nice grab an unlikely 2-2 draw at PSG, comfortably beating the market line of +1.5 and giving us a solid win on our recommendation. However, the analyst covering this contest gave it a Fair Score of 3-1 because PSG created chances in every segment of the match, ending up with an xGoals number of 2.89 versus Nice’s 0.45. This is one that went our way and it’s rare to see such a swing on the Fair Score.

Edinson Cavani was front and centre for the hosts again here, scoring a brace but being guilty of numerous poor misses that could have given his team the comfortable victory that they deserved. It’s becoming cliché to criticise the Uruguayan’s finishing at this stage, but in such a strong team playing in such a weak league he should be much, much more decisive than he is.

Recommendation: Nice +1.5
Actual Outcome: +1 (2-2)
Fair Outcome: -1 (3-1)

Portuguese Primeira Liga
Benfica 2-1 Sporting CP (Sunday 6pm)
If we got away with the Nice recommendation then we had just been punished immediately before that one, with Benfica beating Sporting 2-1 despite the visitors having the better chances throughout the game. Indeed, Benfica’s xGoals was just 1.68 compared to Sporting’s 3.30, though the away side were trailing for over 65 minutes and so would naturally be expected to be pressurising and creating. The analyst adjusted for this with a Fair Score of 2-2, which would have given us a half-win instead of a full loss, which is a hefty enough swing. Benfica have now strengthened their grip on top spot and look formidable, especially with key striker Jonas due back imminently having played just 89 minutes of league football so far this season.

Recommendation: Sporting CP +0.25
Actual Outcome: -1 (2-1)
Fair Outcome: +0.5 (2-2)

English Premier League
Leicester City 4-2 Manchester City (Saturday 5.30pm)
Jamie Vardy and company reminded everyone of their explosive counter-attacking power in this one, ripping through a porous Man City backline to take a 3-0 lead by the 21st minute and never looking back from that point on. The visitors showed a lack of respect to the Champions in the circumstances, where they were significantly weakened at both ends due to team news, and were punished. We just ended up going with the Over 2.75 line in the end, which was never much in doubt, but we did not have enough conviction in our lean towards the hosts at +1.

Recommendation: Over 2.75
Actual Outcome: +1 (4-2)
Fair Outcome: +1 (3-2)

Ones to Watch

Italian Serie A
Torino 1-3 Juventus (Sunday 2pm)
Juventus survived a wobble to continue their imperious record in the Turin Derby, though the final score flattered them ever so slightly. There is no doubt that the Old Lady deserved to win after reacting well to going behind, but on total chances 1-2 was a more reasonable outcome here.

Recommendation: Over 2.25
Actual Outcome: +1 (1-3)
Fair Outcome: +1 (1-2)

German Bundesliga 2
Stuttgart 1-2 Hannover 96 (Monday 7.15pm)
An excellent contest between the two best sides in Bundesliga 2 ended with a justifiable narrow win for Hannover, who comfortably beat the handicap of +0.5 to provide a solid win for our recommendation. After this, we remain steadfast in our belief that they will be crowned champions come May.

Recommendation: Hannover 96 +0.5
Actual Outcome: +1 (1-2)
Fair Outcome: +1 (1-2)

English Championship
Brighton & Hove Albion 2-0 Leeds United (Friday 7.45pm)
It took two penalties to earn us a push here, but this was a comfortably fair win for Brighton. Leeds suffered badly from the early red card and penalty decision, creating just a solitary chance thereafter, while Brighton were professional enough to see out the job at the other end.

Recommendation: Over 2
Actual Outcome: 0 (2-0)
Fair Outcome: 0 (2-0)

Spanish Primera Division
Villarreal 3-0 Atletico Madrid (Monday 7.45pm)

This was a clear loss for our DNB recommendation, with Atletico suffering from some very uncharacteristic defensive errors and then struggling to attack cohesively enough to make a comeback. 3-0 was a bit of an unfair picture in the end, but there was no doubt that Villarreal deserved the win.

Recommendation: Atletico Madrid DNB
Actual Outcome: -1 (3-0)
Fair Outcome: -1 (2-1)

Austrian Bundesliga
Sturm Graz 0-1 Salzburg (Sunday 3.30pm)
A stop-start contest with a lot of bad blood ended up with a deserved victory for Salzburg, despite them playing the last 20 minutes with ten men. Our Over 2.5 pick did not come in on the actual score, but the game had enough chances to have beaten the line.

Recommendation: Over 2.5
Actual Outcome: -1 (0-1)
Fair Outcome: +1 (1-2)

10 Recommendations
Total Actual Outcome: -1
Total Fair Outcome: +1

Weekend Preview – 09/12/16

Highlights

Italian Serie A
Roma vs. Milan (Monday 8pm)
The biggest game of the weekend comes from Italy, where Roma are fresh off a hugely impressive derby win over Lazio and hosting a revitalised Milan, who are joint-second with them in Serie A. The home side rounded off their Europa League group with a 0-0 draw in Romania on Thursday evening, while the visitors have had eight days to prepare since their late win over Crotone. Team news favours Milan, with Montolivo and Kucka the only absences, while Roma are without Salah, Florenzi and potentially Perotti as well.

A handicap of -1 on the home team seems wide, though it is difficult to argue with a Total Goals line of 3 given the fact that both sides have averaged over this amount in 2016/17 so far. Roma have won 3/5 recent league games, with Milan triumphing in 4/5, though our Fair Form actually has the hosts one point better off during the same period. It is also key to note that Roma have won every home league game this season and have never failed to score in front of their own fans. We are so close to backing the visitors, but even +1 smacks of playing for a push.

Recommendation: N/A

English Premier League
Manchester United vs. Tottenham Hotspur (Sunday 2.15pm)
Tottenham joined United in the Europa League thanks to their professional 3-1 win over CSKA Moscow in the Champions League on Wednesday, while Jose Mourinho’s men were in Russia on Thursday ensuring their progression from Group A as runners-up. This extra day of rest and no travel is a bonus for Mauricio Pochettino’s side, who also welcome back the pivotal Alderweireld to have arguably their strongest eleven available (with some apologies to Lamela). Things also look good for the hosts in terms of team news, with Rooney back from a ban and Smalling the only player currently certain to miss out.

This all adds up to mean a very high quality contest should be on the agenda, but with the line at -0.25 and edging towards -0.5 we feel Tottenham are being harshly judged by the market. In addition to this a Total Goals line of 2.25 is understandable on the averages, but a little too low in the context of this type of fixture. Any sort of early goal should open things up significantly.

Recommendation: Tottenham Hotspur +0.25 and Over 2.25

French Ligue 1
PSG vs. Nice (Sunday 7.45pm)
Having slipped up against Ludogorets on Tuesday to lose top spot in their Champions League group to Arsenal, PSG need to refocus when league leaders Nice visit Paris. Following the dismal 3-0 loss at Montpellier last time out the Champions are having a strange week, with questions remaining over their collective attitude and the ability of Emery to motivate his players. Still, they had won 4/5 before that defeat and were beginning to make ground on the likes of Nice, who continue to move along nicely despite the fact that they have been without Balotelli for weeks.

The mercurial striker is back for this one, however, with centre-back Baysse now their only certain absentee. With Rabiot the only first teamer missing for PSG it seems both sides will be strong, but if 3 goals is pitched about right then the -1.5 handicap on the hosts is eye watering at present.

Recommendation: Nice +1.5

Portuguese Primeira Liga
Benfica vs. Sporting CP (Sunday 6pm)
Neither side enjoyed their Champions League commitments in midweek, with Benfica being beaten to top spot in Group B due to their home loss to Napoli and Sporting being eliminated from Europe altogether when losing to Legia in Poland. Sporting have the better team news in terms of availability, but have had a day less to prepare, while Benfica were also at home and did not face any travel. In a game of this magnitude between the two title favourites our preference is usually to side with the slight underdog, and while the risk feels elevated due to the aforementioned factors we will do the same again here. Jorge Jesus has enjoyed coming back to his old stomping ground and would happily take a draw.

Recommendation: Sporting CP +0.25

English Premier League
Leicester City vs. Manchester City (Saturday 5.30pm)
The Champions are without Schmeichel and Drinkwater again, which is obviously causing them some distress, but the visitors have even bigger team news issues with Aguero, Fernandinho and Otamendi suspended. As Kompany is still injured the only natural centre-back remaining is Stones, which could force Guardiola to use three in defence once more. Taking everything into account we have a lean towards Leicester on a +1 handicap, but the Over 2.75 has more momentum with the team news situation on both sides and with Claudio Ranieri’s side coming in at over the natural Total Goals line in 4/5 recent matches.

Recommendation: Over 2.75

Ones to Watch

Italian Serie A
Torino vs. Juventus (Sunday 2pm)
The Turin Derby is another huge game in Italy this weekend, especially with Torino on the uptick this season and currently sitting seventh. They have only beaten their big rivals once since 1994/95 but might feel emboldened given the Old Lady’s patchy away form, which makes us like the overs.

Recommendation: Over 2.25

German Bundesliga 2
Stuttgart vs. Hannover 96 (Monday 7.15pm)
In pre-season our belief was that Hannover would win Bundesliga 2 with Stuttgart right behind them, which still looks very much on the cards. This is a matchup between the two best sides in the division and naturally the line is -0.25, though you can get the visitors at +0.5.

Recommendation: Hannover 96 +0.5

English Championship
Brighton & Hove Albion vs. Leeds United (Friday 7.45pm)
Leeds are resurgent under Garry Monk, beating the handicap in 4/5 recent games and climbing to fourth in the table. Brighton have the chance to go top with a win, however, which explains a wide-looking -0.75 line on the home side. More safety can be found in the goals market.

Recommendation: Over 2

Spanish Primera Division
Villarreal vs. Atletico Madrid (Monday 7.45pm)
Both come into this on the back of disappointing 0-0 draws last weekend and having fulfilled their European group stage obligations in midweek, though Villarreal had much more at stake than Atletico. That’s enough for us to edge the visitors at DNB, as this should be their type of game.

Recommendation: Atletico Madrid DNB

Austrian Bundesliga
Sturm Graz vs. Salzburg (Sunday 3.30pm)
It is difficult to oppose Salzburg at the moment, as they are clearly the strongest side in Austria by far and no longer have the distraction of European competition to contend with. Still, 7/10 recent games for these two combined have ended Over 2.5, which looks the most likely outcome.

Recommendation: Over 2.5

Going In-Play (Part 2): Goals vs. xGoals per Segment

Thanks for joining me on the second part of my in-play goals journey!

I hope you found last week’s blog useful for your weekend trading, as I was certainly happy with how the majority of the games panned out versus what I expected to see.

I’m going one step further this time around and will analyse each team’s average goals per 15 minute segment relative to their expected goals. My hope here is that you’ll be able to get a broader understanding of not just when teams score and concede goals and chances, but how frequently they take the opportunities they create, and when they are converting them better or worse than expected.

For this week’s blog, I’ve picked three games that will be broadcast live on British television this weekend. Some of the teams featured may be familiar from last week, but seeing as it’s a new gameweek, we’ll be looking at the inverse of the patterns we looked at last time out. So for instance, last week we were treated to a goals-fest when Liverpool played Bournemouth away, but this week, Liverpool will be playing at home, so I’ll be making sure to give you a novel view on Goals so far this season.

As mentioned, I’m going to be investigating an enduring hot topic in the analysis community in “expected goals” (defined by us and many as “xGoals”). xGoals is, as the name suggests, the number of goals a team is expected to score based on the number and quality of chances they create. I’ve previously discussed how we at StrataBet calculate this, but for a quick recap it’s important to remember that the granularity of StrataData allows us to combine chances not scored with those that are scored to calculate what a team is expected to score given the overall number of opportunities, weighted by the quality of each chance.

My colleague Dave Willoughby has previously written a great blog post on how we rate our chances, which I’d again encourage you to read, but briefly, I’ve calculated xGoals using the average conversion rates of our six chance categories, labelled Superb (~75% conversion), Great (~40%), Very Good (~25%), Good (~15%), Fairly Good (~8%) and Poor (~2%), summing a team’s total chances per game by conversion rate rather than by category. This has allowed me to weight each team’s chances and goals by the quality of each opportunity, in order to provide a firm xGoals number.

I’ll be using xGoals to identify when teams are over-performing and under-performing, by creating tables where you’ll see have some numbers coloured either red or green. These numbers are the difference between xGoals and actual Goals (xGoals minus Goals). The difference can be positive or negative of course, but the interpretation is different depending on whether the team is scoring or conceding. For instance, if a team is scoring (“goals for”), then a negative number is good, as it means the team is scoring more than they are expected to (i.e. a team is finishing the chances they are presented within that time period). However, when looking at conceding goals (“goals against”), negative numbers are bad, because this means a team is conceding more than they are expected. I’ve highlighted these metrics in each of the graphics below to help you identify these moments.

So let’s jump straight into the first fixture and put this into practice…

LEI MAN.png

Leicester City vs. Manchester City (Saturday 5:30pm, BT Sport 1)

Champions Leicester City welcome Manchester City to the King Power on Saturday and both teams could do with a good result considering their recent stutters. With Man City sitting on a -0.5 handicap, there doesn’t seem to be much in the market to separate the two after the suspensions for Aguero and Fernandinho, while Schmeichel and Drinkwater remain sidelines for the host. To begin, we can take a look at the game play data to help us predict what the outcome will be here.

Focusing the on the top left graphic, this season Leicester appear to be slow starters at home, having yet to score inside the first 30 minutes of a home match (Jamie Vardy has come closest to this mark in gameweek 2, when he scored in the 32nd minute). Ironically, Man City are leakiest away from home in the first 15 minutes, conceding just over 0.4 goals in the first 15 minutes of away games this season. If Leicester are to capitalise on this, they’ll have to capture some of their form from last season where they averaged 0.2 and 0.3 goals in the first and second 15 minute segments respectively!

Despite the expected slow start, Leicester can be positive after the first 30 minutes. This period is marked by a strong scoring period for the hosts and a bad conceding period for the visitors, so much so for Man City that they concede 0.15 goals more than expected (table on top far right). If Leicester can’t convert some of their chances (top right figure) in this period, they can still hope that their better than expected performance in first half added time can reward them with a goal, where they are over-performing (i.e. converting more goal scoring opportunities) by 0.08 goals (table on top far right, segment 45+).

On the flip side and completely unsurprisingly Manchester City appear to be more dangerous than the league average when away from home (bottom left graphic), as well as over-performing more than expected in 4/8 time segments (see green numbers in table in the bottom right graphic). Of these segments, the spell between 76 and 90 minutes could define this game. Whilst Man City are scoring more than expected, Leicester are conceding 0.1 more than expected, making this time segment the leakiest in their home games this season. So if the scores are level after 75 minutes, it might be a good idea to look at the price of Pep Guardiola’s side!

ManTot.png

Manchester United vs. Tottenham Hotspur (Sunday 2:15pm, Sky Sports 1)

Sunday afternoon sees Tottenham travel north to take on Jose Mourinho’s Manchester United side. With a handicap of -0.25, there isn’t much at all to separate these two other than home advantage, so hopefully the data will help us uncover some useful trends. Indeed, I’ve picked out three key moments in this match that you should certainly be looking out for.

For starters the opening 15 minutes look interesting, but mainly from Tottenham’s point of view. Focusing on the bottom left graphic, it appears that Spurs start strong away from home, a fact that can be coupled with Man Utd conceding the same number of goals at home as Tottenham score away (0.3). This is further evidenced by both team’s expected performance, because while Tottenham are scoring 0.17 goals more than expected, United are conceding 0.04 more (table on the bottom far right, 1-15). So don’t discount an early away goal here!

Manchester United generally peak between 31 and 45 minutes, scoring most of their home goals this season in that period, with an average just shy of 0.8 goals. This is Manchester United’s best spell when playing at home and they are over-performing by 0.22 goals (table on the top far right, 31-45). Mourinho can be proud of this, as not only does this stat way surpass the league average, but it also surpasses his predecessor (Louis Van Gaal), who only averaged 0.16 home goals last season in this same time period! This coupled with Tottenham’s propensity to concede most of their away goals in this segment (under performing the league average by 0.17 goals), bodes well for United.

Importantly we can expect some more twist and turns going into the second half, regardless of game state. This time however, Tottenham maybe able to impact heavily upon the Total Goals line, as at the beginning of the second half (bottom left graphic, 46-60), Tottenham peak with an average just shy of 0.6 goals. This appears to be a key period for Mauricio Pochettino’s men when they are away from home, as they’re over-performing by 0.25 goals (bottom right graphic and bottom right table, 46-60). If Tottenham are able to convert any of the chances they create in this period, then we may see the total goals line of 2.25 surpassed before the 60 minute mark!

LIVWHU.png

Liverpool vs. West Ham United (Sunday 4:30pm, Sky Sports 1)

This fixture looks set to be very volatile, particularly when looking at Liverpool’s attacking explosiveness at home (top row graphics), and with the total goals line set at a meaty 3.25, it would still not be a huge surprise to see this one surpass that. I’ve picked out a few key moments from this contest that I think may be instrumental in determining its outcome.

Focusing firstly on the top left graphic, Liverpool are naturally strong at home, generally surpassing the league average goals mark over the course of the 90 minutes. In the first 15, Liverpool peak in terms of their average number of goals scored, scoring 0.83 between 16-30 minutes at home so far this season. This is further evidenced by the fact that they are scoring 0.09 goals more than expected. Interestingly, however, it is in the period after this (31-45) where Liverpool are most efficient. During this segment, they over-perform by 0.39 goals, and while West Ham are consistently conceding 0.3 goals in the first 45 minutes away from home, this 15 minute period may also provide some first half goals.

I’ve also picked out two key moments in the second half that could sway the goals market. At the start of the second half (46-60 mins), Liverpool’s scoring and West Ham’s conceding rate are almost identical, with both hovering around 0.6 goals in that period. West Ham have done well so far this season to not concede more, particularly given the peaks in their xGoals graphic (top right graphic), but the data highlights that this is a period that Liverpool could exploit.

I’m also keen to see how the end of this fixture plays out. Looking at the data, we may be in for a late Liverpool goal. Between 76 and 90 minutes, both teams are expected to score (Liverpool) and concede (West Ham) more (top right graphic, 76-90 minutes). However, having said that, a Total Goals bet could even still be alive come injury time, where Liverpool are over-performing and West Ham are under-performing. This is evidenced by the table on the top far right, where at 90+ mins, Liverpool are scoring 0.26 goals more than expected, whilst West Ham are conceding 0.12 goals more than expected. With Liverpool handicapped at -1.75 and expected to win, the final 15+ minutes of the game may be instrumental in whether they beat the handicap or not.

West Ham on the other hand can take solace in their away performances between 61 and 75 minutes. They generally peak during this period, scoring just over 0.4 goals so far this season, but importantly they are over performing between 61 and 90 minutes for scoring when away from home. If West Ham are able to continue their form during this spell, we may be in for an exciting last 15 minutes to go with an exciting 75 before that!

Conclusions

Once again my hope for this post is that not only will it provide some interesting data driven observations for those of you who enjoy making in-play goal trades, but also highlight where teams are over-performing and under-performing relative to the chances they make and concede. This may be the difference between a team being exploited or exploiting another.

Last week we discussed how Jurgen Klopp’s Liverpool would need to be wary of Bournemouth’s attacking talent throughout the game, as Liverpool had already been conceding many more goals than he would like away from home. However this week, Liverpool may be in for an easier ride. Our analysis shows Liverpool are consistently outscoring the league at home, and importantly, creating opportunities to score – opportunities that marry up with West Ham giving teams plenty of chances to score against them.

Manchester City are naturally strong away from home and we may well see another early away goal from them in this fixture, though Aguero being absent will dampen this somewhat. Although Leicester are generally strong defensively in the first half, the 76th may be the turning point, as Man City’s scoring power peaks at this stage, while Leicester’s defence crumbles at this stage more than any other. In the other fixture focused on Manchester, the Reds will need to be prepared for Tottenham’s attacking threat, particularly at the beginning of each half.

With the inclusion of xGoals, I hope here you were able to see even greater detail about when a team is more dangerous or vulnerable at a particular stage of a game. I especially hope that this information will help you with your trading again this weekend. As mentioned last time out, stay tuned for some more in play analyses, investigating the relationship between Goals and xGoals depending on the game state, as well a journey through corner production and concession.

Until next time!

Sagar Jilka (@DrSagarJilka)

Weekend Review – 06/12/16

Highlights

Spanish Primera Division
Barcelona 1-1 Real Madrid (Saturday 3.15pm)
The first Clasico of the season ended up all square, with our recommendation of Real Madrid +0.75 always looking a strong one despite Barcelona taking the lead early in the second half. Before Suarez’ goal the visitors had actually been the better side, creating the greater quality chances in the first half and deploying a game plan that did a good job of restricting the space for Barca’s attacking weapons. Even without Casemiro, who started the game on the bench as we thought he might, Zidane’s side kept compact and always maintained a threat on the counter.

Madrid had to show supreme resolve to fight back from 1-0 down, though they did also have a few wobbles as they left gaps by going in search of the equaliser. Neymar in particular missed a huge opening, while Messi’s finishing also let him down after a superb through ball by substitute Iniesta. In the end we should have had more conviction in the Under 3.5 goals line, regardless of Casemiro’s inclusion, as it very rarely looked like a game that could explode into life. Real remain the team for the big occasion under Zizou and maintain a six point lead over their rivals.

Recommendation: Real Madrid +0.75
Actual Outcome: +1 (1-1)
Fair Outcome: +1 (1-1)

English Premier League
Manchester City 1-3 Chelsea (Saturday 12.30pm)
We thought this looked like the best Premier League game of the season to date and thankfully it didn’t disappoint, with a superb tactical battle ending in a thrilling win for the visitors after Man City spurned an enormous opportunity to go 2-0 ahead and put the contest to bed. It was refreshing to see two of the big teams able to attack each other with full strength line-ups and after De Bruyne’s huge miss it was Diego Costa who came to the fore, though Chelsea’s right side again deserves some kudos as Willian slotted in brilliantly for the injured Pedro and struck up an instant rapport with Moses.

All-in-all Man City can feel very hard done by with the loss, as was shown by the late red cards for Aguero and Fernandinho that have made Chelsea even stronger title favourites than the result alone would have. Our recommendation of Over 2.5 paid off handsomely, though the full win on the +0.25 handicap was much more questionable and probably should have resulted in a half.

Recommendation: Chelsea +0.25 and Over 2.5
Actual Outcome: +2 (1-3)
Fair Outcome: +1.5 (2-2)

Italian Serie A
Lazio 0-2 Roma (Sunday 2pm)
A surprisingly tight Rome Derby never looked like pulling us in a win on the goals line of 2.75, as Roma’s switch to a 3-5-2 nullified the hosts and meant they never created anything better than a Good Chance (~15% conversion). The visitors scored from a defensive mistake and a hopeful shot from distance, meaning that the Fair Score looked much more like a 0-1 based on the overall opportunity counts.

Before this 64% of Roma’s games had been over the natural goal line, but the preparation time each coach had for such a big fixture meant that the intensity was nullified and it led to a deserved Unders. This is always something worth considering in such a setting, and while we were at pains to note it in the preview perhaps we should have allocated more weight to it ourselves. Still, Lazio being so unproductive was a surprise regardless.

Recommendation: Lazio +0.25 and Over 2.75
Actual Outcome: -2 (0-2)
Fair Outcome: -2 (0-1)

Italian Serie A
Napoli 3-0 Internazionale (Friday 7.45pm)
Our intuition that the handicap on Napoli looked wide proved to be off in actuality, though with a Fair Score of 2-1 we do feel hard done by to come out of this with a negative. Whichever way you slice it a game with this many chances (3 Greats and 5 Very Goods) should have always beaten the goals line of 2.75, though the big surprise was that Inter failed to find the target. A mixture of excellent goalkeeping by Reina and poor finishing ultimately denied them, though it has to be noted that they were forced into chasing after falling 2-0 down within just six minutes. Pioli’s men continue to be far too open at the back, but it’s making for some entertaining games.

Recommendation: Internazionale +0.75 and Over 2.75
Actual Outcome: -0.5 (3-0)
Fair Outcome: 0 (2-1)

Turkish Super Lig
Fenerbahce 0-0 Besiktas (Saturday 4pm)
This was one of those big games that just did not live up to the hype, as Besiktas came to stonewall their hosts and preserve their unbeaten record. They achieved this admirably, but were somewhat fortunate to come away with a 0-0 after Fenerbahce created the more meaningful chances and played with much greater intent. The visitors arguably could have had a couple of red cards in the second half too, though this was very much a contest that verged between 0-0 and 1-0 on the Fair Score. In essence it was very much a -0.5 handicap on the hosts.

Recommendation: Fenerbahce -0.5
Actual Outcome: -1 (0-0)
Fair Outcome: +1 (1-0)

Ones to Watch

Portuguese Primeira Liga
Porto 1-0 Sporting Braga (Saturday 8.30pm)
Things were going smoothly for Braga until Artur Jorge’s 35th minute red card, awarded for the concession of a penalty. However Marafona made a huge save and then the ten man visitors withstood a barrage for well over 60 minutes of play, until deservedly losing 1-0 deep into added time.

Recommendation: Sporting Braga +1.25
Actual Outcome: +0.5 (1-0)
Fair Outcome: -1 (3-0)

English Premier League
Everton 1-1 Manchester United (Sunday 4pm)
Our suggestion of Over 2.25 goals never looked like coming off in a relatively sterile encounter between two unconvincing sides. There was never anything better than a Very Good Chance created and it took a goalkeeper error and a penalty to get us to a half loss in the end.

Recommendation: Over 2.25
Actual Outcome: -0.5 (1-1)
Fair Outcome: -0.5 (1-1)

Scottish Premier League
Rangers 2-1 Aberdeen (Saturday 3pm)
For a long time it looked like this game could go either way, with both sides playing in an open, attacking style that made for a very good watch. In the end the victory went to the team that carved out the better chances and 2-1 looked the Fair Score.

Recommendation: Aberdeen +0.25
Actual Outcome: -1 (2-1)
Fair Outcome: -1 (2-1)

English Championship
Leeds United 2-0 Aston Villa (Saturday 5.30pm)
We thought that the -0.25 handicap on Leeds and a goals line of 2.25 were nicely pitched by the market, so opted out of this one. In truth it proved to be a solid decision, as a case could be made for numerous Fair Scores – though all have Leeds winning.

Recommendation: N/A
Actual Outcome: N/A (2-0)
Fair Outcome: N/A (1-0)

Austrian Bundesliga
Salzburg 4-1 Rheindorf Altach (Saturday 5.30pm)
Salzburg sent out a timely reminder of their quality by sweeping aside league leaders Altach, crushing the -1.5 handicap. As soon as they took the lead just two minutes in this recommendation was in real danger, and our fears of them being “in the mood” proved to be well founded.

Recommendation: Rheindorf Altach +1.5
Actual Outcome: -1 (4-1)
Fair Outcome: -1 (4-1)

12 Recommendations
Total Actual Outcome: -2.5
Total Fair Outcome: -2

Use Your Head! Assessing conversion rates of headed chances

Assessing chance quality has been a hot topic in the football analytics community lately and, though there are some differing opinions, one thing that most people seem to agree on is that it is fundamentally harder to score when using your head compared to using your feet. Anyone who’s watched or played the game will likely agree with this and so do I, but what I’ve come to realise lately is that headed chances are often rated unfairly, mostly due to a lack of more granular data.

The main inspiration for this insight came from my colleague Dave Willoughby (@donceno) a while back when he, as is his custom, challenged my rating of a certain headed goal, causing me to realise how my work with Expected Goals models had made me biased towards headers. He encouraged me to explore the true value of a headed chance using the detailed data we collect at Stratagem, and so here we are.

Before you continue reading, please be sure to check out Dave’s great blog post on how we currently rate all chances here.

As he stated in the above post: context is key. Indeed, when assessing headers this tends to be especially true, but is often overlooked due to the lack of more detailed data. So what type of data does Stratagem capture in order to better understand the context of a chance, besides the usual variables like chance location?

1. Defensive Players

How many defending players (including the goalkeeper) were there between the ball and the goal? This shouldn’t be thought of as a straight line between the ball and goal, but as more of a funnel shape moving out from the ball to the frames of the goal, incorporating every defender who could influence the strike with a block or by exerting positional pressure.

2. Attacking Players

How many attacking players were there between the ball and goal? Again, this would include every attacking player who could influence the strike, like causing a deflection on the original shot, moving into the goalkeeper’s line of sight or blocking him from making his dive.

3. Defensive Pressure

How much defensive pressure was the player taking the shot under? This is rated on a scale of 0 (no pressure at all) to 5 (significant physical pressure) and is often a key thing missing from most models. It is important because it takes the defenders’ position and their ability to hinder the strike from hitting target into account.

These variables are factored in together with more qualitative ones like the quality of the pass for example, to come up with the chance ratings Dave described in his post. I’m sure you all have a good feeling of the general effect these variables have on the quality of a headed chance – but let’s quantify them to explore how they all interact.

Using a basic machine learning logistic regression algorithm, I’ve created a simple Expected Goals model just for headers based on the above variables together with chance location. This is represented by whether or not the chance was inside the “prime” goal scoring area just in front of goal, what we at Stratagem call the “C1 Zone”. Data investigated is from all Stratagem league seasons starting in 2016, up until the 21st of November.

graphic_pitchgrid_v2

graphic1

There are few surprises here, as most of you certainly would have guessed that Goal Expectancy decreases with more Defensive Pressure and more Defensive Players. In addition to this, the fact that the number of Attacking Players between the ball and goal at the point of strike has a positive effect makes sense as well, because they usually block the goalkeeper’s view and, unlike defenders, won’t try to hinder the strike. It’s also clear that headers are very sensitive to chance location, with the prime C1 zone enjoying a far higher Goal Expectancy throughout.

So it seems that the model and the above variables’ impact on Headed Goal Expectancy makes “football sense”. Nice. However, the main point to take away from this simple model is how much variation there is between the different scenarios, even from the same location. Take the most common scenario inside the C1 zone for example: Defensive Pressure 1, Defensive Players 1 and Attacking Players 0.

As can be seen from the green line in the lower-left plot, this chance would have quite a high Goal Expectancy, with the player heading the ball under very little pressure and with only 1 player between him and goal (normally the goalkeeper), meaning the model would give him about a 43% chance to score. However, if it would instead be 5 Defensive Players blocking the goal, the chance to score would drop to about 16%. Keeping the Defensive Players at 1 but increasing the Defensive Pressure to the maximum of 5 would also see the chance of a goal drop, this time to about 29%.

In the above plot with all variables ranging from 0 to 5, the model gives a Goal Expectancy range of 0.09 to 0.74 inside the C1 zone, while outside it would range from 0.04 to 0.50. This is significant variation that a model without these variables simply wouldn’t be able to pick up on.

Like I’ve already mentioned though, when Stratagem analysts rate chances they take not only the above variables into account, but evaluate the whole context of the situation. As a result, using these chance ratings should provide us with a better model. From now on, whenever I mention Goal Expectancy I will refer to a new model, based only on the chance ratings mentioned in Dave’s blog.

So who is creating the best headed chances? To limit the scope of the analysis, I’ll have a closer look at the two leagues I’ve worked the most with: Sweden’s Allsvenskan and Norway’s Tippeligaen. Before looking at Headed Goal Expectancy though, we have to take into account the number of headed chances created by each side. Also, as stronger teams tend to create more chances overall and this should reflect in the number of headed chances as well, regardless of playing style, we’ll have to normalise the number of headed chances created. So I’ve created a metric I’ll call Header Focus, which is simply the number of headed chances created divided by the total number of open play chances created (defined here as any chance that is not a penalty kick or a direct free-kick), as this would give an insight to which teams focus more on headers as a part of their attack.

graphic2

Looking at the Header Focus for both leagues in the 2016 season, I’d say it passes the eye test. The teams at the top are all known to cross a lot and use target players up front, while the teams at the bottom end tend to keep the ball on the ground more. Häcken really stands out here with only about 7% of their open play chances coming from headers, which makes sense given Peter Gerhardsson’s preferred style of play and the type of players they have in the squad.

graphic3

When looking at the mean Goal Expectancy per headed chance, a clear pattern becomes visible: Norwegian teams tend to create better headed chances, with only three Swedish teams inside the top half. Of course, it could also be looked at the other way, in that Norwegian teams are worse at defending headed chances.

The top teams in Norway Odd and Rosenborg are among the best, but surprisingly Start, who finished rock bottom, sit in sixth here. It’s also a bit shocking to see Djurgården so far down given their high Header Focus, indicating that they’ve relied more on quantity than quality when it comes to headed chances. Aalesund and Haugesund seem to have done the opposite with low Header Focus but high Goal Expectancy chances. Finally, Häcken, Östersunds and Sundsvall don’t seem to care about headers at all, combining low Header Focus with poor mean Goal Expectancy.

graphic4

Given how the leagues share the same structure, we can also easily compare raw total numbers. The Norwegians top the chart here as well, with the same top three as with mean Headed Goal Expectancy. However, there are a few Swedish teams in the top of this ranking, and Norrköping and Djurgården’s high Header Focus clearly saw them able to rack up high total Headed Goal Expectancy numbers.

It’s maybe a bit startling to see bottom teams like Tromsø, Start and Gefle well inside the top half here, as they should be expected to struggle offensively, but it does make some sense because these poorer teams could look to use more long balls, crosses and headed chances to drive their attack. Logic would suggest that these are easier to create chances from for teams such as this, rather than playing their way through the opposition defence to set up high expectancy chances on the ground.

graphic5

Putting all three metrics into one scatterplot really drives the point home in terms of how the two leagues differ when it comes to headed chances. The teams in Sweden’s Allsvenskan seems to be worse at creating high quality headed chances (note again that the inverse could also be true and that they could simply be better at defending), while the Header Focus varies a lot from the lowest (Häcken) to the highest (Norrköping).

In Norway’s Tippeligaen, however, teams’ Header Focus varies less and the numbers are generally higher, while mean Goal Expectancy per headed chance varies more, with most of the teams able to create high quality chances. So generally speaking, Norwegian teams seem to have a more deliberate focus on headed chances, as they both produce higher quantities and better overall quality.

The size of the dots in this plot represent the total Headed Goal Expectancy for the teams and demonstrate how a team needs to perform well in at least one of the two underlying metrics to produce high numbers overall. Out of all the teams, only Odd combined a high Header Focus with high Mean Goal Expectancy, indicating a deliberate tactical plan to use headers as an important part of their attack.

graphic6

Finally, looking at individual players’ Headed Goal Expectancy per 90 minutes, we see how the top ten is entirely made up of strikers, which make sense as they are usually positioned in front of the goal at the receiving end of crosses. The Norwegians occupy the top three spots here as well, but what really stands out is just how dominating Odd’s target man Olivier Occéan has been with a value of 0.3 Headed Goal Expectancy per 90. In fact, his total Headed Goal Expectancy of about 8 made up over 70% of the team’s total production.

That’s it for now, but I will certainly dig deeper into headed chances now that there are more questions to answer than before I started this piece. To be specific, I wish to investigate how much Headed Goal Expectancy each team concedes, who is best at converting headed chances and finally look into the big issue I uncovered: just why are the Norwegian teams so much better at producing high quality headed chances? Hopefully we’ll find out!

Alexander Tanskanen (@zorba138)

Weekend Preview – 02/12/16

Highlights

Spanish Primera Division
Barcelona vs. Real Madrid (Saturday 3.15pm)
There is really only one place to start this weekend, as Barcelona host Real in the first Clasico of 2016/17 with a six point gap to make up in La Liga. The home side are boosted by being able to name a full strength team due to Iniesta’s return, but Madrid are without Bale and Kroos, while Casemiro is in question due to fitness and would be a huge loss in a fixture of this type. However, the playing field is somewhat levelled by form, with Real on a run of fully deserved wins and Barca coming in off the back of their worst showing of the season at Sociedad, where they were fortunate to escape with a 1-1 draw.

Swinging the pendulum back the home side’s way is the fact that they only have a meaningless Champions League game coming up in midweek, while Real host Dortmund in a battle to top their group. If you’re a purist you should probably just opt out and enjoy the spectacle, but we’ll put our neck on the line by going with Real +0.75 pending Casemiro’s inclusion and also give a nod towards the Under 3.5 if the Brazilian makes it too.  

Recommendation: Real Madrid +0.75

English Premier League
Manchester City vs. Chelsea (Saturday 12.30pm)
As a prelude to Spain’s big one, we have what looks like the best fixture in the Premier League to date as Manchester City host Chelsea at just the right time. The visitors are on a seven game winning streak since changing to 3-4-3, while City have been unconvincing in recent times but keep churning out the wins. This sets the contest up perfectly and there is no major team news to negatively impact either side, with just Kompany and Terry certain to miss out.

We have the typical big game set up of a -0.25 handicap on the hosts and 2.5 goals, with the fact that City have only beaten their line 36% of the time compared to 46% for Chelsea meaning we’ll back the visitors on the plus. We also like the idea of making a double with the Overs, as both teams have been over their natural lines 46% of the time this season. There should be good intent to take maximum points, but the visitors would be happier with a draw.

Recommendation: Chelsea +0.25 and Over 2.5

Italian Serie A
Lazio vs. Roma (Sunday 2pm)
There are a couple of great matches in Italy this weekend, with the Rome Derby taking top billing early. Lazio look likely to have a full strength squad, while Roma have concerns over Manolas and Salah, both of whom would nudge the line a little if they were absent. That line currently pitches the visitors as quarter-goal favourites, which looks questionable based on recent form and the fact that they struggle to keep clean sheets.

In total, an enormous 64% of their games have been over the natural goal line, which leads us towards liking the 2.75 in this very intense setting. However, it is important to note that the coaches have had plenty of time to prepare due to no midweek games, which could lead to a slightly cagier affair. Still, we’ll bank on the fact that neither can lie down if it goes to a +1/-1 game state.

Recommendation: Lazio +0.25 and Over 2.75

Italian Serie A
Napoli vs. Internazionale (Friday 7.45pm)
The standout game on Friday sees a rejuvenated Inter heading to stuttering Napoli on a surprisingly wide +0.75 handicap. The hosts have Albiol back but are still lacking threat without Milik and Mertens, while Inter are again without Medel but have no fresh injuries after the win over Fiorentina. In addition to this there is the fact that Napoli have a crucial UCL game on Tuesday night to prepare for, while Inter are already out of the UEL and so have no such concerns. Pioli has seen some very open games since arriving, but this is first away fixture and so he could be a little more cautious. Still, we’ll back him to get at least a draw here and to oversee another exciting match.

Recommendation: Internazionale +0.75 and Over 2.75

Turkish Super Lig
Fenerbahce vs. Besiktas (Saturday 4pm)
It seems like there is a huge game every week in Turkey at the moment, and they don’t come much bigger than Fener versus Besiktas. Both are in superb form, though the hosts have been more impressive and do not have the distraction of a big Champions League game on the horizon. A -0.5 Asian Handicap on Fenerbahce looks fair enough for those reasons, but we still have a liking for them to beat it based on how professional they were in the recent 2-0 win over Galatasaray. Lens is their only small doubt, while Besiktas are without two starters.

Recommendation: Fenerbahce -0.5

Ones to Watch

Portuguese Primeira Liga
Porto vs. Sporting Braga (Saturday 8.30pm)
Third travel to fourth on Saturday evening and going by the Fair Form, neither of these teams have deserved to lose any of their last five matches. This makes an Asian line of Porto -1.25 look on the steep side, especially with a crucial Champions League match on the horizon.

Recommendation: Sporting Braga +1.25

English Premier League
Everton vs. Manchester United (Sunday 4pm)
Seventh host sixth with both teams performing below expectations in recent matches, though United’s performances have generally been better than Everton’s according to the Fair Form. This goes some way to explaining a -0.25 line on the visitors, but we’re more interested in a low-looking Total Goals line of 2.25.

Recommendation: Over 2.25

Scottish Premier League
Rangers vs. Aberdeen (Saturday 3pm)
This is always a game with needle in it and the two teams are only separated by home advantage, with the market assigning Rangers a -0.25 Asian line. The Ibrox club have consistently been overestimated this season, losing against the handicap 60% of the time, so we’ll take Aberdeen here.

Recommendation: Aberdeen +0.25

English Championship
Leeds United vs. Aston Villa (Saturday 5.30pm)
Two historic clubs meet in the second tier, both having found form under new managers. One edge for Villa is that Leeds played a cup game with Liverpool in midweek, meaning that the line is DNB instead of -0.25 for the hosts. This and 2.25 goals look fair to us.

Recommendation: N/A

Austrian Bundesliga
Salzburg vs. Rheindorf Altach (Saturday 5.30pm)
Salzburg have been very inconsistent this season but remain the best team in Austria, while Altach are one of the group who have taken advantage of this by climbing to the top of the table. It all depends on the mood of the hosts, but we like the handicap here.

Recommendation: Rheindorf Altach +1.5