SMRTILYTICS is a bit of a hobby of mine. For a full explanation of this project, check out the main SMRTILYTICS page. -OG
Week 19 – the Mid-Season Special
Hello and welcome to the Mid-Year Special of SMRTILYTICS!
I have not been as active as I’d like to be this year at posting updates but hopefully this special will make everyone forget about that – creating good spirits and joy along the way.
Before I get started, let me present a brief caveat to the data presented in this update. This data is up to and including Game Week 19 which may look a little wonky since we already know what happened in Game Week 20. It was very difficult to separate those weeks this year as there was actually only 1 day separating the two. I could have used Game Week 20 data but I wanted to be able to compare to last year’s Mid-Year Special so I have left it as is.
Right, got your calculators out and your thinking caps on? Here goes.
Let’s start by looking at the Big Chart:
Click this graphic to see it in full resolution
Team Performances
For the Mid-Season Special, I’ve divided up the individual team stats into three separate sections: the Top 6, Relegation Fodder, and Wild Cards. For the Top and Bottom teams, not only will I give you their current numbers, but also, I have gone back in time to display where they were just 12 months ago after Week 19 of the 2015-16 season (where possible). If you miss something, don’t worry, the full Rundown is in the next section so you can still view other teams I’m not highlighting and draw conclusions on your own.
The Top 6
Chelsea have been hotter than a pistol since their early season blowout to Arsenal – winning 13 on the trot to propel themselves to the summit of the table. Their numbers have been steadfastly improving to the point of being astonishing. Whatever Conte is putting in the water at Stamford Bridge is working. Their league leading Goals Against (13) and Total Shot Ratio on target (0.69) means that this is not an anomaly – they are in first deservedly (let’s skip the financial doping conversation). What makes it even more astonishing is the comparison to last year’s numbers for the same period.
2015/16 numbers after 19 games:
- Position: 16th
- Points: 20
- Goals For: 23
- Goals Against: 29
- TSR*: 0.48
- Combined Metric: 0.79
Ouch. Granted they have changed formation and manager but they have not experienced much turnover in terms of playing staff. What a turnaround!
Liverpool have been extremely interesting to watch under Jürgen Klopp since he took the helm last season. They really were a non-entity in any of the stats tracked under Brendan Rodgers – not showing in the top 4 for any stat tracked but they have steadily improved, particularly in Total Shot Ratio. This shouldn’t be too surprising since the pressing system Liverpool uses emphasizing winning the ball high up the pitch and applying pressure immediately. In fact, their offensive output has steadily increased while their respectable Goals Against has stayed relatively constant at just over 1 goal conceded per game. Don’t believe me? Take a look at their mid-term numbers from last season below and compare how many goals scored to this year.
2015/16 numbers after 19 games:
- Position: 7th
- Points: 30
- Goals For: 22
- Goals Against: 22
- TSR*: 0.61
- Combined Metric: 0.46
Arsenal sigh…where do we want to start? The team is much better this year, we’re scoring more goals (Alexis Sánchez baby!), and we have more points than last year. Our Goals Against is almost exactly the same and yet we find ourselves in third and not first. Additionally, our Total Shot Ratio is lower than last year (and not near the other Top 6 which is a worry – see the figure below).
Spot the Arsenal (hint, our Total Shot Ratio on Target is at the bottom of the Golden ‘V’ – compare that to our competition).
So what exactly is the issue? When I peak under the hood at our Total Shot Ratio, I discovered that we are allowing more shots on target than our counterparts. Remember that Total Shot Ratio is a ratio of shots conceded on target to shots taken on target and League leading Chelsea on average allow 2.53 shots on their own goal per game versus Arsenal’s 3.58. That isn’t a small difference over the course of a season.
Looking ahead to the rest of the season, tightening up defensively will be the most important factor to improving on last year’s second place finish – we are scoring the goals to be in it. And as if to prove my point, I can see one game into the future of the League and a 3-3 draw with Bournemouth awaits – heh.
2015/16 numbers after 19 games:
- Position: 1st
- Points: 39
- Goals For: 33
- Goals Against: 18
- TSR*: 0.61
- Combined Metric: 0.37
Tottenham are enjoying another good season in terms of numbers. Last year we saw beastly numbers from our neighbours in North London, only for them to spectacularly capitulate in the final weeks of the League. Make no mistake about it, they had numbers that should have won the League last year which makes their final day fuck up that much more awesome. Pochettino has been working wonders with their squad and has improved this year (note that this year’s numbers would be good enough to be first this time last year). If I’m Daniel Levy, I’d be doing everything to keep their gaffer from bigger and richer clubs this summer because once he goes, they’re screwed. For the record, I think Pochettino will be poached sooner rather than later.
2015/16 numbers after 19 games:
- Position: 4th
- Points: 35
- Goals For: 33
- Goals Against: 15
- TSR*: 0.66
- Combined Metric: 0.27
Manchester City continue to be a strange team to watch progress. Like last season, they started out like they were going to walk the League this term – only for the wheels to slowly fall off the bus. They still could win the League but they are just not posting the same elite level numbers as the teams above them. In fact, looking at their average position with our stats, they are exactly where they should be in 5th.
To be honest, I was expecting more from Pep Guardiola but perhaps that is expecting too much given he has a massive job updating an aging squad to be truly competitive again. The reality is that their numbers are better comparatively than last year (as are most in the Top 6), but it doesn’t look like enough of an improvement at this stage of the season.
2015/16 numbers after 19 games:
- Position: 3rd
- Points: 36
- Goals For: 37
- Goals Against: 20
- TSR*: 0.67
- Combined Metric: 0.35
I honestly thought Manchester United were shatting the bed this season under Mourinho and this brought me great joy – especially after the financial outlay they made this summer for Pogba, Zlatan and Henrikh Mkhitaryan. They, like others, are showing much higher numbers in what must surely be a more competitive league. I’d like them to hover mid-table for a while before firing Mourinho at the end of the season – let’s stretch their misery out; but they have shown a new found resilience lately with a series of wins that have propelled them up the table. There numbers are actually quite decent (Goals Against) and they should probably be a place or two higher than they actually are.
2015/16 numbers after 19 games:
- Position: 6th
- Points: 30
- Goals For: 22
- Goals Against: 16
- TSR*: 0.51
- Combined Metric: 0.41
Relegation Fodder
How is David Moyes still employed at Sunderland? When I started drafting my notes for this section, I noted that all three relegation candidates had fired their manager and then when I double checked, I realized that was a mistake (Palace sacking Pardew was awesome though). Sunderland just survived last year and seeing their numbers have only marginally improved this year, in a League that is vastly better stats wise, must make them a favourite for relegation.
2015/16 numbers after 19 games:
- Position: 19th
- Points: 12
- Goals For: 19
- Goals Against: 38
- TSR*: 0.36
- Combined Metric: 1.29
Hull are in trouble. They’ve sacked Mike Phelan and their brightest spark this term, Robert Snodgrass, has been linked with a move away to West Ham (West Ham are screwed too) and Celtic this month. When I look at the stats for Hull, it’s hard to see a promise of something better hiding beneath the surface of some truly rough performances.
There are no stats for last year for comparison.
Swansea find themselves in the basement of the League at the midway point. They’ve just ditched their second manager of the term and really the only good thing is that they haven’t lost touch with the teams just ahead of them. In fact their Goals For and TSR are maybe just promising enough to think if their new manager can get them a little more solid defensively, they likely will be safe (that average position of 16th is the giveaway). It could be a big ask for the Swans and I’ve picked them for the drop come the end of the season.
2015/16 numbers after 19 games:
- Position: 17th
- Points: 19
- Goals For: 16
- Goals Against: 24
- TSR*: 0.48
- Combined Metric: 0.66
Wild Cards
In this section, I just wanted to highlight a few oddities emerging this season. I won’t do full breakdowns on each team and I would encourage you to take a look at their numbers below in the Rundown section.
Crystal Palace find themselves in 17th despite an Average Position prediction of 12th. Their League position is really down to Goals Against as they’ve scored a fair amount of goals and their Total Shot Ratio on target is significantly better than the teams around them.
Burnley on the other hand should be far lower in the League than their 11 spot. Their Total Shot Ratio is the second worst in the League yet they are starting to get some performances in to assure their safety.
Leicester find themselves in 15th after 19 rounds this year and statistically, this is pretty close to where they are mapping to. Still worth pointing out since they are still the reigning League Champions.
Two Big Observations
Observation #1: The numbers are better this year
There were many occasions last season where Leicester’s successes were cited as results of a weaker league than normal. The likes of United and Chelsea fell away early and were relative non-entities when it came down to deciding the League. Additionally City, who started off so well, definitely dropped off after the strange mid-season announcement of Pep Guardiola taking over. As Arsenal fans, we couldn’t help but think it was an opportunity lost to win a League in a decidedly “off year.”
Statistically speaking, it certainly looks like we have proof that it was indeed a weaker league last term. Take a look at the comparisons from last year’s league leader and last place team after 19 games:
League Leader: 2016/17 Chelsea | 2015/16 Arsenal
- Points: 49 | 39
- Goals For: 42 | 33
- Goals Against: 13 | 18
- TSR*: 0.69 | 0.61
- Combined Metric: 0.21 | 0.37
Last Place: 2016/17 Swansea | 2015/16 Aston Villa
- Points: 12 | 8
- Goals For: 21 | 15
- Goals Against: 44 | 34
- TSR*: 0.45 | 0.38
- Combined Metric: 1.27 | 1.12
Observation #2: Good management trumps all
Often when I look at the numbers I crunch for SMRTILYTICS, I ask myself – “great, now what?” Don’t get me wrong, I really enjoy doing this – it absolutely gives me new insight into the Premier League and forces me to look at teams and their respective results in all sorts of new and different ways.
One of my initial goals was to use the lessons learned here to predict results, an experiment which yielded some interesting results but was far too much work for what it was worth.
Now I tend to look at teams and try to figure out not only why their stats are the way they are, but also how they might improve a specific area. I often start by thinking a team that allows a lot of shots on target should buy a new defender next transfer window or a team not scoring enough goals, should get their striker off the treatment table. While I think there is merit in those strategies, after doing this for a couple of seasons, I’ve realized that it is much more complicated than the individual players.
Don’t get me wrong, having the best players is a great start to being successful – but it doesn’t guarantee success in itself. The few and far between elite world class players may be the exception to this thought.
Consider the change in fates of League leaders Chelsea this season. Yes, they did sign N’Golo Kanté and Marcos Alonso but ultimately those two simply cannot be the sole reason for their success this year. The change in system that has brought them a number of wins so far is invariably is down to Antonio Conte and his management team.
As another example, look at the results Mauricio Pochettino is wringing out of a slightly better than average squad in Middlesex.
I’ve written about this in the past on how managers are getting better and better at getting the most out of their squads and I firmly believe this to be true based on the various number crunching I do throughout the season – a change to the numbers is far better served by a change of system than a change to personnel.
The Rundown
New for this season and introduced in Week 10 is the Rundown. If you noticed the Big Chart above, you likely noticed I’ve gonged the Weekly Top and Bottom Threes as some of it is already covered in the Predictions section below. Additionally, I’m introducing this new feature which shows every team based on the current table and includes League Position, Points, Goals For, Goals Against, Total Shot Ratio on Target, Combined Metric and something new(ish) I’m calling Average Position.
Average Position (AVG POSITION) was used in the 2015-16 year end spectacular and has proven to be quite popular in terms of quickly (and somewhat unscientifically) looking at if a team is performing above or below their actual League Position based on the 4 metrics listed (GF, GA, TSR*, Combined). It’s calculated by ranking each stat against the rest of the League and then averaging that position. For example, if Arsenal has the best Goals For and Against, the 4th best TSR* and 6th best Combined Metric, then we would calculate that as [(1+1+4+6)/4] which equals 3 or third. A quick check of their actual League Position will tell you if they are over-performing (first or second), under-performing (fourth or lower), or bang on where their stats say they should be (third in this example – duh).
The AVG POSITION metric isn’t statistically bulletproof but it is handy in looking at how teams are performing relative to the metrics tracked and each other.
Predictions
Finally we move onto predictions! I’m pretty much out of gas at this point so I’ll let them speak for themselves. At this point, you know the drill anyways.
Right – that’s all I got for now. I hope you’ve enjoyed and see you back here in a few weeks.
Youknowit!
OG