Monday, 26 August 2013

Spurs 1 - Swans 0 Match Stats and Chalkboards

Two games in, pointless and bottom of the league doesn't sound too great, but to put it into a better context a point against West Brom next week would mean we'd have taken the same number of points in total (1) when compared with the three corresponding fixtures last season so there's an awful long way to go before needing to panic.

There's no doubt the better team won, but in games like this the margins between courageous draw and disappointing defeat are pretty small, especially when defeat comes as a result of a soft penalty.  Spurs had a few good chances, but a lot of their efforts were from long range that did little to trouble Vorm (who I thought had an excellent game).

The starting line-up was a bit of a strange one to some extent, not so much in not starting with Bony: when you're playing away at one of the top sides, it's perfectly reasonable to set up more defensively.  What I didn't quite get though was having both Shelvey and De Guzman: in a game where you're maybe setting up for a point then I'd arguably have Britton (or Ki) taking one of their places.

Overall, Swansea's passing didn't seem as fluid as it usually is but this has to be taken in the context of playing away at one of the top sides, you're not going be able to just turn up and spray the ball around.

Spurs' midfield were impressive in allowing Swansea little time on the ball around the centre circle, as can be see by the two sides passing charts, Swansea had plenty of the ball (even if a disproportionate amount of this was after the goal) but relatively little of it made its way into more dangerous areas.
Total Passes by each Team.  The key things of note here are the distinct lack of blue in the attacking third in central areas for Swansea compared to Spurs as well highlighting what attacking play Swansea did have was heavily down the left hand side
This can also be seen from the attempted forward passes from Shelvey and De Guzman with neither having huge success, as the passes were often rushed or played under pressure:
Forward passes by Shelvey/De Guzman
Pozuelo was impressive when he came on after the goal, but how much of this was a combination of fresh legs and the advantage of coming on when Swansea were forced to be more attacking is difficult to tell, but he certainly added a dynamism and willingness to take people on.  Before the goal, Swansea had attempted only 4 take-ons (attempted dribble past an opposition player) compared to 12 in the final 35 minutes or so after the goal.
Swansea attempted take-ons pre (left) and post (right) Spurs' goal.  Pozuelo had most successful take-ons for Swansea with 3.
The other thing that's noticeable from looking at the take-on's, is how much play went down Spurs' right hand side with Andros Townsend in particular involved in a lot of their attacking play:
Total take-ons by team, Spurs with plenty of take-ons in the middle and down the right
The level of involvement from Rangel is often a good barometer of the overall level of play from the team. As can be seen from where he received passes (and how many) vs. Spurs compared to the game against Man Utd, Rangel's role vs. Spurs was mainly a defensive one.
Rangel's Passes received vs. Man Utd (left) and Spurs (right), Rangel figured little in attack against Spurs, partly due to the difficulty in maintaining possession long enough to enable him to be included further up-field. 
Writing a regular blog about stats in football can sometimes feel like trying to suck the joy out of something beautiful.  Rational decision making based on solid quantitative analysis is arguably the way to get the most out of a team but cannot capture how much I love Chico Flores.

In terms of stats Chico won all 3 of his attempted tackles, won 6 of his 7 aerial duels and had the most clearances (14) of any player - next highest was Ash with 9.  This of course ignores his incredible volley, his dribbling skills around the Spurs area and picking fights with Jermain Defoe.  Chico plays like he's having a bloody great time being there and I have a great time watching him play.

Other Posts: Ki's shooting from last season reviewed

Twitter: @we_r_pl
Match Stats and Chalkboards: Created using Statszone

Monday, 12 August 2013

Don't You (Forget About Ki)

It was good to see that most of the main members of the first team squad got a start in one of the two Europa League games against Malmo, but the most noticeable person who didn't was Ki Sung Yueng.

Despite all the fuss around Shelvey (I thought the original scepticism when he signed was too harsh and the subsequent euphoria once he's started playing a bit premature) and also the signing of Canas and the renewal of De Guzman's loan, I still think Ki has plenty to offer the Swans.

Ki's versatility is a big plus in a weaker squad but now the squad has been strengthened (especially in midfield) it will be tougher for him to get regular action.  With Michu and Bony expected to start the majority of games, with Michu filling the attacking midfield role, that leaves two midfield places for Britton, Canas, Shelvey, De Guzman and Ki to fight over.

Arguably Ki isn't as good at breaking up play as Britton or Canas or not as much as an attacking threat as Shelvey or De Guzman but there could well be another 50+ games to come this season so he will undoubtedly get some opportunities.

Ki had a couple of stand out stats last season, he had the highest overall pass completion rate (92.7%) but also at one stage had the most number of shots (38) without scoring a goal.  This was overtaken by Jobi McAnuff who played in all 38 of Reading's league games, had 43 attempts at goal but failed to score.

Overall for Ki's 38 goal attempts in the Premier League last season, 13 were On Target, 14 Off Target and 11 Blocked.

Looking at when Ki's shots happened, the chart below show's how the lions share comes fairly early on in the season when he played the full 90 minutes in 7 successive games, beyond that shots were relatively scarce and Ki didn't feature in the last 4 league games (due to injury).

As mentioned previously Ki is a pretty versatile player and played in all 3 (defensive/central/attacking) midfield positions at various points in games across the season so this change in shot volume is not necessarily a tailing off in performance and it should also be noted that after playing the 7 full games in a row, around half of his subsequent games were not full appearances.

That said, there did seem to appear times later in the season when he was hesitating to shoot but that could easily just be because it was something in particular I was looking out for rather than any real change in attitude.
Ki's Shots by Game (DNP= Did Not Play), number in brackets indicates minutes played (otherwise full 90 minutes played)

Going beyond shot volume, I've had a look at where these shots were taken and it helps to add to the understanding of why Ki failed to score last season.  The figures below are rough estimates cobbled together using Statszone but a number of Ki's efforts were so far out they didn't fully appear on the Statszone chart.

Location of Ki's 38 attempts at goal in the league last season

 From the chart you can see that the vast majority of them are 'hit and hope', the kind you might score around 3-4% of the time (if that), there's only a couple of attempts that could be classified as 'prime' positions and even then there's likely to have been plenty of bodies in the way.

I've produced an interactive version of the chart above where you can highlight any of the attempts to show which game they came from (I'm hoping to do more interesting versions of this as the season goes on).  The interactive version is located here.

Getting back to Ki, I think there's still plenty to see from him this season: as well as the volume of games, I can see Shelvey getting the odd red (as will Michu if he tackles like he did in the 2nd leg against Malmo) and there will also be games where only one of Bony/Michu starts and an extra man in midfield may be needed or where Ki's ability is required to close games out where Swansea are ahead with 20-30 minutes left.

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

Thursday, 1 August 2013

Vorm and Tremmel Stats Compared

The fact that there's any sort of debate about who will start in goal against Malmo is testament to how well Gerhard Tremmel played when called on last season, both in the league and also as part of the League Cup winning side.

All through Swansea's first season in the Premier League, from the bombardment that came in the opening match against Man City where Vorm's acrobatics kept the score respectable through a series of great performances and clean sheets, Vorm was undoubtedly the player of the 2011/12 season for Swansea.

It got to the stage where he was maybe playing too well and might get picked off by a big club leaving us with the relatively unknown (at the time) Tremmel who had let in 2 goals from 2 shots in his only league appearance that season although Gerhard did of course have plenty of experience from 'The Continent' so would hopefully be a safe pair of hands both literally and figuratively.

Injuries to Vorm last season meant that Tremmel started 12 League games and came on as a sub in another two, with the run starting due to Vorm's groin injury stretching to reach Tevez's goal at the Etihad meaning Tremmel started 8 league games in a row during Nov/Dec 2012.

In terms of stats comparing the performance of the two in the league, I've compared them using the relative chance types they've faced (more details on that here, but the main thing to note is that the chance types are mutually exclusive so Penalties are not included within Clear Cut Chances which are not included in Inside Box Shots) but excluded the two games where they've both played as won't have the full chance breakdown for each player in those games.

If you just looked at the figures in the bottom right, you'd see that Vorm conceded a goal from 10.6% of shots faced where Tremmel only conceded on from 7.1% of shots faced.  This however ignores the fact that comparatively speaking Vorm had a much tougher time of things with the chances he faced being much better quality.

When looking by chance type faced the stats for the two of them are pretty similar, Vorm saved 4 of the 5 penalties he faced which is pretty much par for the course for penalties faced in the Premier League last season.

In terms of errors leading to shots/goals, Vorm made 7 errors, 5 of which led to goals and Tremmel made 2 errors, 1 of which resulted in a goal which is slightly in Tremmel's favour but not a huge difference.

In this analysis I've looked at all attempts faced, you could argue about excluding shots blocked by defenders as the keeper has no role in them and I'd agree but that split isn't possible on data publicly available with regards to clear cut chances, inside and outside box attempts.

I would however argue that just because a shot isn't on target the keeper hasn't done his work.  Closing down the attacker quickly or good positioning are things that could result in a striker missing the target but could be down to the goalkeeper's ability as much as any failing from the attacker.

Vorm for me is still the number 1 keeper, but we have the situation now where Tremmel's name being announced on the team sheet will be met with an "OK" rather than a "Oh Shit".

With regards to the Malmo game, Laudrup has said that there isn't a First XI as such this season, with possibly 55+ games to play there's going to be a fair bit of rotation (there is of course the risk that we're out of Europe before the domestic season starts and have an abundance of midfielders and not enough games to play them in).

I'm going to stick my neck out and think that Laudrup will start with Bony on the bench, partly as he's new and partly do be a bit more defensive minded given that Malmo are likely to be more prepared as they are already half way through their season and Laudrup has stressed the importance of not conceding.

I think the midfield three will be Canas, Britton and Shelvey with plenty of options on the bench according to how the game pans out.

Twitter: @we_r_pl

Match Stats Created using: EPL Index 

Note: After almost finishing this piece I realised that freekicks faced were not included in the analysis.  I have factored in the free kick conceded by Tremmel v. Norwich but not included freekicks saved by either keeper.