Monday, 29 April 2013

Chelsea 2 - Swansea 0 Stats and Chalkboards

Not too much to say about this one to be honest, this was a relatively comfortable win for Chelsea although who knows if things would have been different if it was still goalless (or even just 1 goal down) at half-time.

Couple of stats stood out for me from the game: Ki again has superb stats in terms of passing accuracy (62 successful from 64 passes).

The other was Aerial Duels where Luiz won all 9 of his Duels (Michu lost all 5 and Shechter all 3 of his) for Swansea.

If you were an optimist, you could argue that the last couple of months have seen the combination of tough fixtures and maybe something missing as mid-table safety has been secure for some time now.

A pessimist might say that over the last 32 league games (as at Monday afternoon) Swansea have 36 points and Aston Villa have 34 so the 6 points Swansea gained from the first two games have masked average form overall.

The reality is probably somewhere in between, the difference between mid-table and relegation dogfight isn't a huge one.  A couple of late winners, the odd injury of an influential player and things can turn.  Next seasons focus should again be around finishing 8th-12th but for me success is still anywhere from 17th upwards.

Sunday, 21 April 2013

Swansea 0 - Southampton 0 Stats and Chalkboards

This was a game that probably won't take up much space on the end of season DVD, there was some nice build up play but little bite in the final third, with the best chances arguably going to Southampton (as well as Southampton having a goal disallowed - probably unfairly).
Attempts by side, Swansea may have had more attempts but most carried little threat.  Both sides had two Clear Cut Chances each but failed to convert.  
The quality of Swansea's crossing was patchy at best during the game despite it being a tactic used far more than normal.  As a comparison, from open play, Swansea complete 3 of 25 crosses (12%) vs. Southampton compared with 5 of 13 (38%) against Newcastle a few weeks ago.

Biggest culprit against Southampton was Dyer who found a team-mate with only 1 of 8 crosses:
Crosses by Team - Including Corners.  Little success for either side
It was a pity about the end product from Dyer as I thought he had a pretty good game in parts and an awful lot of the ball was down the right with him and Rangel.

I don't know if it was the substitution of Moore at half time, but the first 10-15 minutes of the second half looked particularly disjointed with an awful lot of misplaced passes:
Forward passes during the first 15 minutes of the 2nd Half (left) and remainder of 2nd half (right).  Plenty of the ball down the right during the last half hour, but the number of incomplete passes through the centre is noticable.
Over the last few games I've been keeping track of the number of passes received by Michu when he plays up front compared to a deeper role.  Normally he receives the ball around 25-30 times when playing up front and around double that when in a deeper role.

The figures from the Southampton game were pretty surprising as Michu only received a pass from a team-mate 5 times in the second half.
Michu particularly quiet in the 2nd half, although the passes he did receive were certainly in more dangerous areas.

Twitter: @we_r_pl
Match Stats and Chalkboards: Created using Statszone / EPL Index

Saturday, 13 April 2013

Would you work for the Swans for free?

Over the last couple of days there's been a fair amount of attention on the practice of a number of Premier League clubs (including Swansea) using unpaid Graduate interns on long term (usually 11 month) positions.

As someone who is interested in performance analysis but not involved in the industry I've seen numerous jobs posted (The Video Analyst does a good job of collating them) and it's interesting to note that the publicly funded UK Sport website has taken a number of these roles down since this issue broke.

These roles follow similar lines where they are asking for someone ideally with a postgraduate qualification in Performance Analysis to work for nothing (to be fair to West Brom their job offers at least some sort of salary at £8k).

I'd imagine that working among the backroom staff for a football club, although being an incredible experience would pretty much take over your life with evenings and weekends taking a battering (not just because you often had to but because you wanted to).

The main roles that have attracted criticism are from Reading, Wigan and Swansea.  Of the 3, the Reading one is arguably the most ethically dubious as it involves working with the First Team so has the most bearing on the overall success of the club and where they are pretty much looking for an analyst who already knows what they are doing, the Wigan and Swansea roles are working at Academy level so there is less of the juxtaposition that has been presented by the media of unpaid serfs rubbing shoulders with multi-millionaires.

In relation to the Swansea role there was a ripple of discontent on Twitter about it being unpaid when the role was first advertised a few weeks ago but it took the highlighting of the Reading role to bring the issue mainstream.
Ad for 11mth Performance Analyst Intern position
If you put the legal and ethical considerations to one side, the opportunity to work at a Premier League club as a Performance Analyst is a fantastic opportunity and could save years of toiling about in lower leagues where they have a fraction of the facilities.  Put bluntly if the club were charging you £10k to do the work I'd still argue it would be a worthwhile investment.

This ultimately is the key point, it's a great job with great prospects but is only really applicable to someone with supportive parents or willing to get even further into debt.  It means you get the best person who can afford to work for free for a year, not necessarily the best person for the job.

Anyone who is interested enough in this field to have taken an MSc in it is not going to suddenly decide they want to work in Insurance or Marketing, they'll take anything they can to get a foot on the ladder and people can say 'It's Supply and Demand' or 'It's what everyone else does' but if you are getting someone to produce work that is of value over as long a period as 11 months without paying them then that has to be considered unfair.

Swansea are seen as the 'Mother Teresa' of football clubs with the Trust on the board and sensible financial planning and as a result the club tends to get more criticism for anything it does that would be considered standard behaviour elsewhere.  I don't agree with bringing up the issue of the recent dividend for the Directors, that for me was fair reward for the time and money those people have invested in the club.

Minimum Wage works at around £12k a year, if the person who takes the role isn't providing the equivalent value or more to the club then why have the you got the post?  If they are, then why aren't you paying them?

Twitter: @we_r_pl

Saturday, 6 April 2013

Norwich 2 - Swansea 2 Stats and Chalkboards

After spending most of last week putting a post together about big chance conversion and how Michu and Swansea were above average, it was sod's law that the game against Norwich saw Michu only take 1 of 4 'Big Chances'.
Big Chance conversion (figures as before Norwich game), with Michu having had 21 Big Chances prior to Norwich game, to have 4 in one match shows the level of opportunities he had
I've mentioned it before but I'm surprised at how relatively consistent the figures are, when Michu plays as the main attacker he receives the ball 25-30 times and around double that when playing as the attacking midfielder.
Passes received by Michu - Last 2 Games
Michu had the most attempts (9) of any player and he has repeatedly stated his preference for playing deeper. It's going to be an interesting summer to see the kind of striker that gets signed in the summer

Shots by Team - Swansea also had double the Big/Clear Cut Chances (6v3)
Overall this was the kind of game where any result was possible and you could make a valid case that we could have lost or won this game, in Laudrup's post match interview he mentioned the fact that in recent games we got punished for mistakes and this wasn't always the case here.  On the other hand Swansea had a number of great chances so this could easily have been another 4-3.

Twitter: @we_r_pl
Match Stats and Chalkboards: Created using Statszone / EPL Index

Friday, 5 April 2013

The Impact of Big Chance Conversion

At the end of last season I compared the top scorers in the Premier League using the MCFC Analytics data, one of the things missing however was goals scored via 'Big Chances' (defined as a chance where a player would reasonably be expected to score).

This season, using data from EPL Index, I've been able to look at the impact of Big Chances and have used that to look at the scoring behaviour of some of the Premier Leagues top scorers as well as discovering some huge differences in Big Chance conversion rates between teams.

As at 5th April, there were 22 players who had scored 10 or more Premier League goals:
Players with 10+ Premier League goals, some such as Berbatov and Lampard have a significant proportion of goals via penalties
The initial part of this analysis will look at the top 4 on the list (Suarez, van Persie, Bale and Michu):
In terms of Minutes Played, Suarez has played the equivalent of around 2 full games more than Michu and 3 more than RvP/Bale as well as having had far more shots
When breaking those shots down in to more detail, it can be seen that Suarez has a far higher proportion of shots that are blocked and Michu a higher proportion that are off-target:
Suarez and Michu have approx. a third of their shots hit the target compared to closer to half for RvP/Bale
Not all chances are the same which is where breaking shots down into Big/Other chances helps give more understanding as to how players are performing:
Top 4 Scorers and PL as a whole.  Goals figures exclude Own Goals, Big Chance and Other Attempts figures exclude penalties
From the figures it can be seen that these 4 players are the top scorers for different reasons:

  • Suarez - Excellent Big Chance conversion rate coupled with a few goals from a glut of shots
  • van Persie - Average Big Chance conversion rate but has had the most Big Chances of any player, good conversion of other chances
  • Bale - Excellent Big Chance conversion (but few chances), good conversion of other chances across a large number of attempts
  • Michu - Excellent Big Chance conversion with an OK conversion of other attempts
As a comparison Suarez only scored 25% of his Big Chances last season (7 of 28), with van Persie scoring 46% (21 of 46) and Bale 33% (3 from 9).

When looking at goals scored which are relatively rare compared to passes for example, there's always more room for differences in performance to be due to chance rather than a real difference. 7 of Bale's 17 goals have come from outside the area but you could argue he's done it often enough to show it's not a fluke:
Location of Goals from Bale (via Squawka)
As well as the top players there are few other players who have interesting stats around chance conversion:
Range of Players with stand out figures for a range of reasons (good and bad)
It's off a small sample size, but it was interesting to see the poor Big Chance conversion rate for Jelavic and Adebayor, both of whom did well last season (with Big Chance conversion rates of 43% and 41% respectively).

Giroud has very similar stats to Michu in terms of total shots/Big Chances with one very important difference:  Michu has converted 52% of his 21 Big Chances compared to just 19% from Giroud.  It's an almighty if, but if Giroud can convert his Big Chances at the same rate as Suarez/Michu have done this season then he'd probably get around 20 league goals given the number of chances Arsenal create.

Mata is arguably one of the most technically gifted players in the league and his conversion rate of non-Big Chances stands out at almost 15%.

I was under the impression that Pogrebnyak was having a disappointing season at Reading but looking at the stats, purely in terms of taking chances he's not got bad figures.  Looking at a team level however you can start to see why he hasn't scored too many.
Big Chances by Team
Reading have the lowest number of Big Chances created by some distance, but have converted the highest proportion of any team.

Man Utd have created the most Big Chances and also have one of the best conversion rates, in terms of total goals scored this explains most of the difference between them and Man City this season.

Where Reading don't make many chances but convert a good proportion of those they get, West Ham have the opposite problem, creating a decent number but with a terrible conversion rate.
Big Chances Created and Scored by Team - Those below the line have below average conversion rate
No single metric works in isolation from other figures but as almost half of all goals come as a result of Big Chances the value of a finisher is huge, the difficult part is differentiating between the flash-in-the-pan, the consistent goalscorer and the players who might come good again in time.

There's also other issues such as does a side have few clear cut chances because forwards are making bad runs or because the midfield are supplying poor balls.

Related Posts:
A Review of publicly available Opta Data Sources
Premier League Weekly Review
How Mark Gower became 'The most creative player in Europe'
2011/12 Top Scorers by club compared

Twitter: @we_r_pl
Stats and Chalkboards: Created using EPL Index / Squawka

Monday, 1 April 2013

Swansea 1 - Spurs 2 Stats and Chalkboards

Three league defeats in a row (and 4 defeats in last 5 league games) obviously doesn't make great reading but it should be remembered that all 4 have come to teams above us in the league, fluctuations in fortune combined with easier/harder fixtures mean streaks of a few good/bad results are always likely.

Although it could be argued that we 'won' the last 65 minutes against Spurs, the damage was already done by this point as a result of clinical finishing by Vertonghen and Bale, after which it's difficult to know how much of the change in pressure is due to a Swansea improvement or Spurs saving their energy for future games.  Swansea had 27 penalty area entries to Spurs' 9 but doesn't matter as much if you're two goals down for the majority of the game.

Overall Swansea had 15 attempts to Spurs' 10 (14 vs 6 when just looking at time after the Spurs goals) but Spurs had 7 attempts on goal to Swansea's 3 (as far as statistics are concerned Dyer's header against the crossbar is as off-target as a shot that goes out for a throw).  

As mentioned in the preview of the match, Adebayor's finishing has been pretty terrible this season (he's now scored 1 of 11 'Clear Cut Chances' this season) and his chance to make it 2-0 was taken by someone who certainly didn't expect to score.

It was interesting to see Ki come on for De Guzman, especially if it is the case that De Guzman is going back to Villareal.  It'd be daft though for Villareal to say anything other than they want him back, ultimately it'll be a question of whether Swansea can come up with a sufficient offer to tempt Villareal to sell, expect plenty of speculation over the next couple of months as both sides try and get maximum value.

Top passer by some distance on Saturday was Ashley Williams with the other notable passer being Leon Britton who had a 96% pass success rate.

From Williams' chalkboard always nice to see the long diagonal out to Dyer which helps to stretch play although the biggest pass combination in the match was Williams to Hernandez:

Other notable stats from the game were Dyer yet again being the most fouled player (4 times) and Routledge creating 5 chances, 2 more than any other player to continue what's been a superb season for him.

Twitter: @we_r_pl
Match Stats and Chalkboards: Created using Statszone / EPL Index