Imagine what the league table would look like if every match this season had been suddenly ended at the same specific minute. For example, if every game had finished on 85 minutes, would your team be sitting higher in the table having not conceded all those last-minute equalisers and winners?
One of the things I love about the Internet is that once you find the right community, you can generate and develop ideas substantially quicker and more effectively. As part of some really helpful feedback he gave on my previous post, @neilallison of the Sky Blues Blog showed me this excellent graph he’d developed to understand how Coventry’s points tally would look if their matches had finished at different times.
The challenge he outlined was whether it was possible to see how the table would look if you applied this approach to every team, and I just so happened to have the required data to hand.
This post marks my branching out into video. Well actually that’s already happened, and by comparison this barely qualifies, but I wanted to try something a bit more experimental than usual. The following animation, which I strongly recommend you view in HD and full-screen due to the amount of data therein, steps through what the Championship table would have looked like on 16th March if every match to date had ended at a specific minute, starting with the 1st and going all the way to the 90th.
Burnley certainly start well, don’t they? It’s also interesting how many crucial goals Birmingham score late on. You’ll notice that I’ve split out home and away records so that you can see where each team’s goals are going in, and also that I haven’t taken Portsmouth’s points deduction into account (given that it isn’t a reflection of their on-field performance).
I’m aware that I could have done a better job of the video, so just in case it looks horrible I have prepared a backup plan. The following graphic (which clicking will enlarge) shows how the table would look after each 5 minute interval:
As always, feedback is welcome – I’d be very keen to understand whether animations like this are the way to go, or whether I should just stick to the ill-disciplined visualisations. If people like this then I’ll refine it and produce versions for Leagues 1 & 2 as soon as I can.
Timing of goals scored & conceded: Championship
Background
You may remember some monstrous graphs from last season which showed what proportion of goals scored and conceded occurred early and late in a match. These were fine but painted an incomplete picture of the distribution and required a lot of scrolling, so I’ve been looking at ways to both broaden and compact this analysis. This post is the result of a seemingly bright idea for doing so.
Approach
I’ve divided each match into 6 equal slices of 15 minutes which seems the neatest compromise between granularity and simplicity. The technical term for these is sextiles, which was what swung it for me (not really – well, as far as you know). I’ve then added up all of the goals scored and conceded (up to March 16th) in each of these intervals, both for each club and for the division overall, and compared them in the table below. Clicking the table will bring up a full-sized version in a separate tab:
How this works
- Each coloured cell shows the difference between the proportion of goals scored or conceded for that team, in that sextile against the divisional average. The greener the cell, the larger the positive variance (i.e. a higher proportion of goals scored on the left, a lower proportion of goals conceded on the right), with redder cells denoting greater negative variances (less scored or more conceded).
- The grey row along the bottom shows the average for the division, e.g. 16% of all goals scored in the Championship before yesterday’s matches came in the first 15 minutes of the second half. As you’d expect, this is the same for both scored and conceded, as we’re counting the same goals in each half of the table.
- Taking the very first cell as an example: Barnsley score 10% of their goals in the first 15 minutes of matches – the divisional average is 11%, minus the 1% in that cell. Moving across to the ’1-15′ column of the ‘conceded’ section shows that they concede 4% more of their goals than the average in this interval.
- This isn’t a particularly dramatic example – looking at the darker green and red cells shows that there are some significant outliers here, and I’ll talk a bit more about these in the ‘Observations’ section below.
Observations
The first 15 minutes
Crystal Palace like to score early – 22% of all their goals occur in the first 15 minutes of matches. Blackpool start much more slowly, netting just 3% of their goals in this interval. At the other end, Portsmouth often get games off to a bad start, shipping 22% of their goals in the first quarter of an hour, while Burnley (with 20%) are also prone to an early lapse. Brighton, Hull and Millwall are relatively good at keeping things tight early on – hardly any of the goals these sides concede are at the start of matches.
Approaching half time
Watford are especially fond of scoring just before half time, with 28% of all their goals coming in the last 15 minutes of the first half. Hull find it difficult to net before the break, with just 5% of their goals coming in this sextile, so their fans will seldom be punished for leaving their seats early to beat the half time rush. Unlike Peterborough, who ship almost a quarter of their conceded goals in the approach to half time, Reading seldom concede in the run-up to the break and they’re also one of the best sides in the division at keeping things tight in the final minutes of the second half, which suggests a well-drilled and physically fit defence.
After the break
If you’re at a Hull match then make sure you’re back in your seat promptly after half time, as 30% of the Tigers’ goals and a whopping 36% of those they concede come in the 15 minutes after the break. Those who enjoy watching Coventry score are likely to find this interval to their liking also, while Nottingham Forest fans can make their way back to their seats at a more relaxed pace.
The last 15 minutes
There are 5 sides who enjoy scoring late in a match and all reside in the current top 10: Blackpool, Reading, Birmingham, Brighton and Leeds all score a large proportion of their goals in the final 15 minutes. It’s not often that Derby or Doncaster will score this late in the game, but to the Rams’ credit they also keep things relatively tight at the back as the clock ticks down. Cardiff and Millwall both concede over a third of their goals here, which could point to a stamina or concentration issue.
Overall
It’s worthwhile just looking at the colours to see when a particular team tends to score and concede, particularly clusters of red or green cells. For example, West Ham definitely seem to front-load their scoring compared to other teams, being green (i.e. above average) in the first half and red (below average) in the second.
Important things to remember
- This compares the proportion of each team’s goals scored in each sextile, rather than the number of goals scored. For example, Bristol City and Ipswich have both scored 12% of their goals in the first 15 minutes, but this doesn’t mean the same amount of goals – this represents 4 of Bristol City’s 32 and 7 of Ipswich’s 56.
- The numbers in each cell show the difference between that team’s proportion and the average, so adding this number to the average in the grey row at the bottom gives you the actual proportion of goals scored or conceded for that team in that sextile.
Next steps
As is often the case on this blog, this is a scruffy prototype version which I’ll tidy up and produce for the other 2 divisions and entire Football League after I’ve determined that it’s worth following up and incorporated any feedback (which is always welcome). There’s obviously a lot more analysis which can be done with goal timings, which I fully intend to investigate in due course.
Football League attack & defence monster chart
As I’ve already analysed the divisions individually in the 3 previous posts this week, I’ll just leave the combined diagrams I promised here for your enjoyment. A quick explanation in case you haven’t seen these before:
- The first shows the average number of goal attempts created per match versus the average number of attempts required for a goal to be scored.
- The second shows the average number of goal attempts faced per match versus the average number of attempts it takes for the defence to be breached.
- Both axes are centred on the average values for the Football League, which allows the teams to be grouped roughly into quadrants summarising how they compare to the rest.
- Comparing a team’s position on both charts gives an (admittedly not definitive but nonetheless informative) view of their current attacking and defensive performance.
- Both charts can be clicked to bring up enlarged versions in separate tabs, which I’d strongly recommend.
A few quick examples to give you the idea:
- Bristol City have the most wasteful attack in the Football League, needing 14 shots on average to score a goal, while the likes of MK Dons and Port Vale only need around 5 and half.
- Brighton and Exeter create the fewest chances, but the Seagulls are much better at putting away the ones they do create.
- Crawley and Burnley create the most, but neither are particularly efficient. Eddie Howe’s men in particular are quite wasteful, while Hull‘s profligacy is also worth a mention.
Defence
Again a few quick observations:
- Swindon are the best side in the Football League at restricting their opponents’ chances, facing virtually half as many attempts at their goal as Peterborough.
- Torquay, Morecambe and Derby all have very busy but equally able defences, but they could all learn from Charlton, who are not only more efficient but seem to have the knack of stopping their opponents from shooting in the first place.
- Northampton and Nottingham Forest have the least resilient defences, although the latter are much better at closing off the other side’s supply lines.
Here’s a list of links to the detailed posts on individual divisions, which include more detailed commentary and also trendlines showing movement since 2011:
A very quick request…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had over 23,000 views so far, which will hopefully convince you that it’s worth watching. I’d love for it to be seen as widely as possible, so any effort you’re willing to make towards this will be warmly appreciated. As soon as you click play it counts as a view, so that’d be enough for me.
Many thanks!
Attack and defence deconstructed: League 2
Completing the trilogy, it’s League 2′s turn to be graphically analysed. As per my previous Championship & League 1 posts, the aim is to take a step back from analysing efficiency alone and instead compare it with how many goal attempts are being created at each end to give a more reliable measure of performance. I’m continuing the approach I piloted with the League 1 version and providing a comment on each team rather than just cherry-picking observations, as this seemed to work better.
In this post there are two charts, each showing the following for all League 2 teams:
- The first shows the average number of goal attempts created per match versus the average number of attempts required for a goal to be scored.
- The second shows the average number of goal attempts faced per match versus the average number of attempts it takes for the defence to be breached.
Both axes are centred on the average values for the division, which allows the teams to be grouped roughly into quadrants summarising how they compare to the rest. Comparing a team’s position on both charts shows their current strengths and weaknesses at both ends, with the dashed lines showing how their averages have moved since the end of 2011.
Firstly, the attacking graph (click to embiggen):
In summary the bottom right is the place to be, with teams there creating plenty of chances and proving clinical at taking them, although anywhere but the top left has some merit. It looks as though most teams are clustered in a band from top left to bottom right with relatively few outliers, suggesting that chance creation and conversion tend to go hand-in-hand in League 2. This is a different picture to League 1 where we often found teams good at one or the other, but not both.
I’ll outline some conclusions after we’ve looked at both charts, but first let’s look at the defensive graph (again, enlarge by clicking):
In this chart the top left is the best quadrant, where teams allow few shots and excel at keeping them out, whereas the bottom right signifies the reverse and therefore spells problems. There appears to be a stratification here, with the majority of teams near the bottom of the chart having similarly weak defences while a more thinly-spread minority enjoy a much greater degree of resilience.
Let’s see what we can extrapolate by reviewing these two charts together, taking each club in order of their current position in the table:
- Swindon are unsurprisingly in the best quadrants on both charts given that they’re the league leaders and in irresistible form. Interestingly there are sides who create more chances, but nobody is better at restricting those of their opponents.
- Cheltenham are another side in the top quadrants for each chart after an unexpectedly fantastic season so far, although there are some slight signs of decline.
- Shrewsbury are more efficient than effective, particularly in defence where they are the most efficient side in the division and soak up a lot of chances.
- Southend have gone off the boil lately after an impressive first half of the season, particularly in defence where things are noticeably deteriorating. They still have one of the division’s most lethal forward lines though, so shouldn’t be written off just yet.
- Crawley create significantly more chances than anyone else, but selling their two most accomplished strikers has undermined the accuracy with which they can convert these. While they’re adept at preventing the opposition from having many attempts at goal, they’re not all that difficult to score against if you can find a way through.
- As a Torquay fan I can vouch for both our frustrating inability to convert the chances we create and the number of shots our robust defence (and in particular our outstanding goalkeeper) can repel. It’s nice to see that we’re starting to clamp down on the volume of the latter.
- Oxford share Torquay’s frustrations going forward but boast one of the division’s best defences, having worked out how to keep a lid on their opponents’ forward raids as well as denying the majority of them.
- You have to feel for fans of Port Vale at the moment, whose team have been on a great run despite the off-pitch chaos. They boast the division’s most efficient attack but seldom create enough chances to turn this into an advantage – perhaps their strikers only shoot when they know they’re in a great position?
- Gillingham boast the second most efficient attack but it’s at the other end where they’re having problems, evidenced by the number of goals they’ve shipped lately: 3 times in 2012 so far they’ve conceded 4 in a match.
- Crewe were my outside bet for the playoffs a few weeks back, assuming they continued to improve. They seem to be tightening things up at the back, where they were allowing even more shots than the Gulls, and getting more accurate up front, although they still don’t create enough chances to make their case.
- Rotherham are treading water on the edge of the playoff picture at the moment, and are arguably deteriorating slightly at both ends.
- Morecambe have an attack worthy of respect but currently own the division’s most overworked defence, which is hampering their progress after an excellent start to the campaign.
- Aldershot are on a great run at the moment and are starting to overcome the goal-shyness that has held them back for so long while simultaneously becoming more disciplined defensively, but they’ll need to sustain this if they’re to advance much further up the table.
- Accrington aren’t moving in an exciting direction lately: the chances are drying up while the defence is springing leaks.
- Burton are another side who started the season well and have since faded badly. They’re currently on a 10-match winless run with their once-prolific strike force now struggling.
- Bristol Rovers seem to be stabilising their wobbly defence since Paul Buckle’s departure, but their energetic forward line have gone backwards slightly and could do with being a lot sharper.
- It’s at the back where Wimbledon have struggled to enjoy their debut season in the Football League, on account of operating one of the least effective defences. Up front things are a little brighter, but not enough to compensate.
- Bradford are the nearest thing to an ‘average’ team at the moment and aren’t making noteworthy headway in either direction. They’re starting to do more with less up front and have a much better goal difference than their league position would suggest, so don’t rule out a late surge up the table.
- Barnet have gone from winning 4 on the bounce to the same sequence of defeats, which has negated a lot of the progress they made here. They’re the highest-placed side to reside in both of the worst quadrants, which should have them looking nervously over their shoulders if it continues.
- Macclesfield share the Brewers’ 10-match winless run, which has seen the efficiency of their defence plummet and their attack become the division’s worst by the measure used here. These are worrying times at Moss Rose.
- After a horrific start to their season, Plymouth continue to move in the right direction at both ends of the pitch despite still occupying both of the ‘bad’ quadrants. You’d have to back an improving side to stay up though, given the chaos around them.
- Hereford are also improving, albeit less so. No side has improved its attacking efficiency more, although the Bulls’ was pretty lousy before, and their defence has started to restrict opponents’ chances. However the paucity of recent wins would suggest that a lot more needs to be done.
- Dagenham & Redbridge are enduring a hellish return to League 2 and are struggling to turn things around. Their attack is the least efficient in the division, which makes the recent reduction in chances created all the more worrying.
- Northampton are another side whose improvements have so far proved insufficient to move them out of danger – they still have the division’s least efficient defence, and the meagre improvement in efficiency up front is being offset by fewer chances being created.
I’d be very interested in the views of League 2 fans as to whether this has revealed or confirmed anything interesting, in addition to any feedback on this analysis and the methodology used. I’ll follow this post up with the one everyone likes, where I basically cram all 72 Football League clubs into one chart and then stand back to admire the chaotic mess.
A very quick request…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had over 23,000 views so far, which will hopefully convince you that it’s worth watching. I’d love for it to be seen as widely as possible, so any effort you’re willing to make towards this will be warmly appreciated. As soon as you click play it counts as a view, so that’d be enough for me.
Many thanks!
Attack and defence deconstructed: League 1
This time it’s League 1′s turn to be analysed. As per my previous Championship post, the aim is to take a step back from analysing efficiency alone and instead compare it with how many goal attempts are being created at each end to give a more reliable measure of performance. I’ve taken a slightly different approach to the Championship one, providing a comment on each team rather than just cherry-picking, as I’m still testing and refining the way this blog works.
In this post there are two charts, each showing the following for all League 1 teams:
- The first shows the average number of goal attempts created per match versus the average number of attempts required for a goal to be scored.
- The second shows the average number of goal attempts faced per match versus the average number of attempts it takes for the defence to be breached.
As usual, the chart axes are centred on the average values for the division. This allows the teams to be grouped roughly into quadrants summarising how they compare to the rest. Comparing a team’s position on both charts shows their current strengths and weaknesses at both ends, with the dashed lines showing how their averages have moved since the end of 2011.
Firstly, the attacking graph (click to embiggen):
In summary the bottom right is the place to be, with teams there creating plenty of chances and proving clinical at taking them, although anywhere but the top left has some merit. It looks as though most teams fall into one of two clusters in the top right or bottom left quadrant, suggesting they most can either create lots of chances or convert them efficiently, but few manage or fail at both.
I’ll outline some conclusions after we’ve looked at both charts, but first let’s look at the defensive graph (again, enlarge by clicking):
In this chart the top left is the best quadrant, where teams allow few shots and excel at keeping them out, whereas the bottom right signifies the reverse and therefore spells problems. This distribution is a bit strange and almost linear, with most teams crowded into a narrow band and therefore quite a few crowded into the bottom right. I even had to shift Rochdale’s label to the opposite end of their trendline to make everything fit, while Walsall are all on their own in the top right.
Let’s see what we can deduce from these, taking each club in order of their current position in the table:
- Charlton are in the top quadrants for both attack and defence, as you’d expect from the table toppers. While they don’t create an obscene amount of chances, they’ve got one of the most clinical strike forces around and their defence is virtually impregnable.
- Sheffield United look if anything more dangerous than the Addicks going forward, although their defence is nowhere near as ruthless despite some recent improvements in cutting their opponents’ supply lines.
- Sheffield Wednesday‘s decision to dispense with Gary Megson looks less mysterious here. Given that they tend to allow opponents more shots in a match than it takes to breach their defence on average, their deteriorating back line doesn’t look capable of standing up to the sort of sustained probing that a fellow promotion contender would dish out.
- Huddersfield are nestled just within each top quadrant and haven’t really moved much, which chimes with them having not been outside the top 5 since late September.
- MK Dons‘ once lethal attack has seen chances dry up of late. While still clinical in front of goal, it’s possible that opponents have worked out how to stifle their midfield – if this continues then they could start to struggle for goals and see their promotion challenge falter, although you’d surely back them for a play-off place.
- Stevenage don’t seem to be missing Graham Westley yet, with their attack continuing to improve at just the right time. They may have seemed to stutter a bit lately, but they’ve had some tough matches and an F.A. Cup distraction to contend with.
- Notts County aren’t moving too much in either chart despite having won 5 of their last 6. Their attacking prowess has made up for some occasionally lacklustre defending this season, although the latter is starting to improve at a crucial time.
- Carlisle‘s defence has been a real problem this season so it’s just as well that they have such a dangerous front line, which is increasing in efficiency.
- Brentford have been busy at both ends this season but seem to be getting sharper in attack – it remains to be seen whether this improvement will allow them to break into the play-off picture.
- Bournemouth have moved up a gear going forward, creating a lot more chances but without converting enough of them. Perhaps this explains their willingness to acquire a proven goalscorer in Matt Tubbs at significant cost, as they’ll realise that a play-off spot isn’t beyond their reach despite a poor start to the season.
- Hartlepool have gotten a lot better at the back lately and are starting to create more at the other end too, although their home record may cost them a tilt at the play-offs that few would have been expecting this season.
- Colchester are yet another side who’ve seen their relatively prolific attacking undermined by a porous (albeit slightly improving) defence, particularly at home.
- Preston are among the biggest movers since the start of the year, with a once-fearsome attack being starved of chances while their leaky defence continues to sustain a battering. This looks to be a season of consolidation for them now.
- Bury‘s defence has gone downhill lately (reflected by just 1 win in their last 9 games) in a season which has seen relatively little action for them at either end.
- Oldham are having trouble up front lately and their season hasn’t been joyful at the back either: despite some limited success at restricting opponents’ chances they’re still proving relatively straightforward to score against.
- Leyton Orient‘s fans probably thought they’d recovered from their dire start to the season, but they’ve started to stumble slightly and performances have deteriorated at both ends of the pitch.
- Tranmere‘s all-action approach to attacking hasn’t resulted in that many goals – they have one of the division’s most wasteful forward lines, which partly explains their worrying winless streak.
- Yeovil look to be turning a corner lately, with both attack and defence moving away from the worrying zone. One of the shyest attacks in the division is finally starting to pester defences and, given the narrow margin of safety at the bottom, it could just be enough to preserve their League 1 status.
- Exeter currently reside in both ‘naughty’ quadrants, but while their defence is improving their attack still looks blunt. The re-signing of veteran striker Jamie Cureton is a necessary gamble given their perilous position.
- Walsall‘s defence are the only occupiers of the ‘good but overworked’ quadrant, able to soak up a lot of attacks. However their own strikers are seeing most of their efforts end unprofitably.
- Scunthorpe still create more chances than any other League 1 team despite having reined it in lately. Their defence is also edging upwards in the efficiency stakes, but they’ll need to maintain this to avoid the ignominy of two successive relegations.
- Wycombe‘s defence have started to capitulate with worrying regularity of late, and is currently the least resilient in the division.
- Rochdale have a different problem to the Chairboys – they look like they’ve forgotten how to score at the moment.
- Chesterfield are moving rapidly in the wrong direction on both charts, creating less chances while allowing more at the other end.
Well done if you made it all the way to the end, as I appreciate this is more than I usually write but wanted to try a different approach. I’d be very interested in the views of League 1 fans as to whether this has revealed or confirmed anything interesting, in addition to any feedback on this analysis and the methodology used.
I’ll produce a League 2 version as soon as I’m able, followed briskly by the one everyone likes where I basically cram all 72 Football League clubs into one chart and then stand back to admire the chaotic mess.
A very quick request…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had over 23,000 views so far, which will hopefully convince you that it’s worth watching. I’d love for it to be seen as widely as possible, so any effort you’re willing to make towards this will be warmly appreciated. As soon as you click play it counts as a view, so that’d be enough for me.
Many thanks!
Attack and defence deconstructed: Championship
Following on from my previous posts where I plotted attacking efficiency against defensive efficiency, I couldn’t help wondering if I’d missed a step. After all, efficiency is one thing but, as I mentioned at the time, it needs to be considered alongside how many goal attempts are being created to give a more reliable measure of performance.
In this post there are two charts, the first showing average number of goal attempts created per match versus the average number required to produce a goal for each Championship team, and the second showing average number of goal attempts faced versus the average number required for the defence to be breached.
Again the axes are centred on the average values to allow the teams to be grouped into quadrants. Comparing a team’s position on both graphs shows their current strengths and weaknesses at both ends, and as before the dashed lines show how these averages have moved since the end of 2011.
Firstly, the attacking graph (click to embiggen):
In summary the bottom right is the place to be, with teams there creating plenty of chances and proving clinical at taking them, although anywhere but the top left has some merit. I’ll outline some conclusions after we’ve looked at both graphs, but I’ll make a few quick observations here:
- Barnsley and Watford are having fewer attempts on goal but are making these count, suggesting that their attacks are now more patient and / or focused.
- Nottingham Forest are still pretty poor up front, but they’re improving significantly – could this be the start their dash towards mid-table safety?
- Like Birmingham, Brighton‘s attack is getting a lot more clinical, although unlike the Blues they still create few chances – less than any other Championship team.
- Burnley create by far the most chances of any side, but they seem to be focusing on quantity over quality.
Now the defensive graph (again, enlarge by clicking):
This time the top left is the best quadrant, where teams allow few shots and excel at keeping them out, whereas the bottom right signifies the reverse and therefore spells problems. Again I’ll make a few quick observations before moving on to compare both charts below:
- Middlesbrough‘s once proud defence has sprung leaks lately and is now looking decidedly average.
- It’s a busy time in the Peterborough goal as the Posh continue to invite more chances, while Derby‘s equally punch-drunk defence look to be tightening things up slightly.
Comparing the two
- Hull were the example used during my greatly appreciated namedrop on the excellent @WAGUpodcast a few weeks ago, and their situation is easily diagnosable here. They have a fantastic defence, arguably the division’s best at the moment, but they struggle to score despite creating a host of chances. Finishing in the playoffs could well depend on Nick Barmby being able to sharpen his attack without destabilising their excellent back line.
- Peterborough and Ipswich have attacks worthy of at least a play-off spot but are let down by their defensive vulnerabilities.
- Brighton’s recent good form seems to have stemmed from their already-efficient defence clamping down on the number of chances they allow their opponents. Their front line can also take some credit – while still creating few goal attempts, they’re getting increasingly better at converting them.
A quick favour…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had 20,000 hits in its first week, which is evidence that it’s actually quite good and I’d love for it to be seen as widely as possible. As soon as you click play it counts as a view, so that’d be enough for me.
Many thanks!
Foul intelligence
Here’s something a bit more experimental than usual. I was interested in which teams committed the most and least fouls, but also wanted to understand whether some teams were more skilled or intelligent at fouling, i.e. stopping their opponent while escaping punishment.
On the chart below I’ve plotted the average number of fouls each Football League team commits in a match on the horizontal axis, versus the average number of cards they receive in a match on the vertical. For the purpose of aggregating yellow and red cards, I’ve arbitrarily decided that a red is worth 2 yellows, to enable a single ‘cards per game’ number to be calculated. There’s probably a more sophisticated way to do this but I’m tired.
Rather than identify every team, which would be messy on a chart with this much clustering, I’ve chosen to highlight some of the outliers: those who are fouling significantly more or less than others, as well as those who are being disproportionately over or under-punished for the fouls they commit. As usual, click the image to bring up an enlarged version in a separate tab:
The axes are centred on the averages, and the diagonal dashed line shows all points where the ratio of fouls to cards is the same as the average. The purpose of this line is to show where punishment would be consistent with the average, i.e. a team anywhere on this line would receive the average number of cards per foul irrespective of how many fouls they commit.
This line also allows us to measure how far from the average a team is. The greater the perpendicular distance a team is from the dashed line, the more disproportionate the number of cards they receive. Teams in the red area above the line collect more cards per foul than average, which could be due to naive defending, clumsy tackling, general ill-discipline, or a reluctance to slide in except where the attacker is bearing down on goal. Conversely, teams in the lower green area receive less cards per foul, which could indicate cool heads under pressure, deft tackling or a sinister ability to con referees.
It’s perfectly logical to commit fouls for tactical reasons – whether it’s to stop or ‘dissuade’ a particularly dangerous player, to disrupt your opponent’s rhythm or to produce a break in play for your team to regroup – but it takes a lot of practice to use fouls in this way without collecting a lot of cards and therefore suspensions.
These statistics alone can’t tell us which of these possibilities is the case for each club, and a lack of more granular information such as number of tackles attempted per match prevents us from drawing firmer conclusions, but the spread of data is still quite interesting and can direct us in digging deeper. I’d appreciate thoughts on how to explain the position of any of the clubs I’ve highlighted here.
Some observations
- Crewe are by far the cleanest team in the league, just as they were last season. It’s a credit to them that they consistently approach the game so cleanly, although you wonder if they’re too timid to assert themselves. For reference, they’re fouled an average amount: about 10 times per match.
- Despite committing an additional 3 fouls per game on average, Chesterfield are punished significantly less than the Railwaymen and are the only side to average less than one yellow card per match. This could be due to the greater experience of the Spireites’ side making them more effective tacklers than Crewe’s youthful squad.
- Brighton and MK Dons receive a lot more cards per foul than the average, despite both sides doing well in their respective divisions and the latter actually committing fewer fouls than most teams. I’m at something of a loss to explain this. Seeing as the Seagulls collect cards all over the pitch, perhaps their passionate manager Gus Poyet gets his charges a little too fired up before games.
- Rotherham commit almost exactly the same number of fouls as Brighton but receive less than half as many cards for their trouble. You may remember from this recent mega-chart that the Millers have one of the league’s less resilient defences while Brighton’s is actually pretty solid, so restraint in the tackle doesn’t always pay dividends.
- While hardly conclusive, some credence is lent to the commonly-levelled accusation that Stevenage employ spoiling tactics. No side commits more fouls but they receive less cards per foul than average, suggesting that they know how to disrupt their opponents’ rhythm without giving the referee cause to caution them.
- Yeovil are just as fond of halting opponents in their tracks, but lack Boro’s subtlety and are consequently racking up a surely unsustainable number of cards.
- League 1 is quite a violent place this season – it currently houses the 4 teams who commit the most fouls per game.
A quick favour…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had 20,000 hits in its first week, which is evidence that it’s actually quite good and I’d love for it to be seen as widely as possible. As soon as you click play it counts as a view, so that’d be enough for me! Apologies to regular readers of the blog who will have already seen this plea a few times – I will stop haranguing you soon, I promise.
League 2 attacking and defensive efficiency in more detail
To complete the recent trilogy of enhancements to the attacking and defensive efficiency graph (see earlier posts for a more detailed explanation), here are some enhanced charts for League 2. First of all, a look at how each club’s average efficiencies have changed since the start of the year (click to enlarge):
You can see straight away that there have been some significant changes since December, most notably:
- Macclesfield are deteriorating at both ends – although they still boast one of the division’s most resilient defences, their attack is now among the most wasteful.
- Torquay and Shrewsbury have significantly tightened their already-impressive defences, which now both take need an average of 12 shots to breach.
- Barnet and Swindon are improving their efficiency at both ends with some purposeful recent performance.
Some observations:
- Shrewsbury‘s defence can withstand a staggeringly high average of 20 shots for each goal it concedes, which goes some way towards explaining their unbeaten home record.
- Aldershot are incredibly wasteful up front, taking 6 shots more than the divisional average to record each home goal – this sort of profligacy could well drag them into a relegation battle. Cheltenham‘s promotion challenge could also come unstuck if they don’t get sharper in front of goal on their own turf.
- Matches at Port Vale were relatively devoid of goals last season, but this time around they’re flying in at both ends!
Again a few things I’ve spotted:
- The foundation of Torquay‘s ability to grind out as succession of away wins seems to be an incredibly resilient back line. Their inefficient attack away from home is what’s making these wins quite narrow.
- Morecambe are the division’s most effective away side; tough to break down at the back and able to take their chances up front.
- Dagenham & Redbridge haven’t enjoyed their return to life in League 2, and it’s on the road where their troubles are sown – it takes more than twice as many shots for them to score as it does for them to concede.
A quick favour…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had 20,000 hits in its first week, which is evidence that it’s actually quite good and I’d love for it to be seen as widely as possible. As soon as you click play it counts as a view, so that’d be enough for me! Apologies to regular readers of the blog who will have already seen this plea a few times – I will stop haranguing you soon, I promise.
League 1 attacking and defensive efficiency in more detail
Building on my latest set of posts, I’m currently enhancing my attacking and defensive efficiency graphics one division at a time. Tonight it’s League 1 which is getting a polish, starting with a look at how each team’s attacking and defensive efficiency (calculated from the ratio of shots taken / faced to goals scored / conceded) has changed since the start of the year.
As usual you can click these to open larger versions in a separate tab:
As usual, the further left a team is, the fewer shots it takes them to score, and the further up they are, the more shots it takes to score past them. Put simply: top left = good, bottom right = bad, everything else = a mixed bag.
The small dot at the other end of each line shows where that club was on 31st December 2011. A few observations to give you the idea:
- Charlton are sharp at both ends of the pitch and have been getting even better defensively since the start of the year.
- Rochdale‘s defence is quite poor at repelling shots, needing an average of only 6 to breach it, while their already-inefficient attack is getting worse.
- Chesterfield are becoming more efficient in both attack and defence, which could see them escape the drop zone if they can build on that improvement.
It’s interesting how this seems to follow a slightly upward-sloping line apart from a few outliers at the top.
- Tranmere are one such outlier – their attack is extraordinarily wasteful at home, but only leaders Charlton have a more efficient home defence. This quirk is probably what’s keeping them out of the relegation zone.
- MK Dons are in the opposite corner and are very interesting. What you can’t see from this chart is that they are phenomenally good at restricting their opponents’ shooting chances at stadium:mk. What you can see is that this isn’t necessarily helping as their defence takes the fewest shots to breach, which suggests that they might get found out against opponents who excel at creating chances. It also somewhat undermines the benefits from their parallel achievement of having the most efficient home attack in the division, and potentially explaining why they’ve drawn as many as they’ve won at home in spite of this.
Here we can see a slightly downward-sloping line of sorts, again with a few stubborn outliers at the top.
- Stevenage are astonishingly resilient at the back away from home, able to soak up over 15 shots on average before conceding. That’s nearly 3 times more than Sheffield Wednesday can on their travels, which could well cost the Owls their promotion bid.
- Things may not all be going Huddersfield‘s way at the moment, but Lee Clark did manage to get them focused on scoring away – it took them less than 5 shots on average to score each away goal, which is less than half as many as some of the division’s strugglers have needed.
A quick favour…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had 20,000 hits in its first week, which is evidence that it’s actually quite good and I’d love for it to be seen as widely as possible. As soon as you click play it counts as a view, so that’d be enough for me! Apologies to regular readers of the blog who will have already seen this plea a few times – I will stop haranguing you soon, I promise.
Championship attacking and defensive efficiency in more detail
Hello again. Thanks to everyone who viewed, shared and gave feedback on the efficiency graphics I produced over the weekend. Based on the experience I’ve made some experimental tweaks I’d like to share.
Firstly, here’s the Championship chart updated for Tuesday’s fixtures and in addition showing how the clubs have moved since 31st December 2011, so you can see how their recent performances are reflected in their efficiency:
As usual, I won’t analyse this to death for you, but a couple of illustrative examples never hurt anyone:
- Birmingham are the most improved side, having increased both their attacking and defensive efficiency significantly since the New Year
- West Ham, Middlesbrough and Millwall are moving in the opposite direction, becoming significantly less efficient at both ends of the pitch
- Hull have tightened up considerably at the back but this seems to have come at the expense of their attacking sharpness
In response to a request in the comments section, I’ve also separated out the home and away records. Let’s have a look at the home chart first:
Again, some brief observations:
- It looks like Birmingham are virtually impenetrable at St. Andrews and their pretty dangerous at the other end too. Blackpool is also a tough place to outscore the hosts.
- A trip to either Nottingham Forest or Bristol City isn’t likely to inspire fear at the moment – Forest’s defence and City’s attack are respectively the least efficient in the league on their own turf.
- Middlesbrough and Crystal Palace can keep it tight at home, but struggle to turn their home advantage into goals
And finally let’s look at the away records:
What does this tell us? Well:
- Coventry are really struggling to create decent chances away from home.
- Hull are defensively much stronger away from the KC Stadium – I’m not sure if this is down to them setting themselves up differently away from home.
- Ipswich have the least efficient away defence by quite some distance, but they’re doing something right at the other end – only 4 sides are better at converting the chances they create.
A quick favour…
If you could click play on the video below that’d be great, and if you find yourself watching it all the way to the end and decide to share it with others that’d be absolutely brilliant.
It’s a short and very well-made film about my using Twitter to connect fans of Torquay United all around the world, including lots of inspirational footage from an evening match at Plainmoor. It’s had 20,000 hits in its first week, which is evidence that it’s actually quite good and I’d love for it to be seen as widely as possible. As soon as you click play it counts as a view, so that’d be enough for me! Apologies to regular readers of the blog who will have already seen this plea a few times – I will stop haranguing you soon, I promise.
