2013

10 Points: Arsenal, Requirements, Set Pieces & Game State Effects

By Benjamin Pugsley | October 23, 2013
2013

10 Points: Arsenal, Requirements, Set Pieces & Game State Effects

 

1) Goal Of The Week

Kasaaaaami! This is what I said on twitter on Sunday morning:

@mixedknuts Unless van Basten, circa '88, scores today then Arsenal's first will be the nicest goal for a long time... — Ben Pugsley (@benjaminpugsley) October 20, 2013

And just like that the Pajtim Kasami scored a goal which had tiny hints of the peerless Marco Van Basten sprinkled over it.

2) Scoring

We've been through this before at 10 Points, but @richardwhittall brought it back up again!

Whittall: While it’s certainly possible that something was up early on and has now smoothed itself out, it’s equally possible that the slow opening was just a tick of the needle in the wrong direction, a quirk of random variation. Perhaps as the games pile up, that number may creep ever higher, or, as is more likely, plateau. We should keep this in mind before rushing off to search for meaning in seasonal goal-per-game trends. Historical trends are far more telling of the health of the sport.

Agreed. From the start of this debate I always maintained that shots on target were pretty stable, shots on target% was also pretty stable and at normal historic levels. Instead, the lack of goals was all due to the weirdness of scoring% (goals/shots on target).

The historical average for Scoring% is about 31%-ish. As we can see from the graph below the PL's first 4 weeks of the season saw pretty low levels of scoring% and, in week 3, a ridiculously low scoring% number. Reasons for that 4 week stretch to open the season? No idea, who the fuck cares, it's probably variance and luck and maybe some system effects in the early season.

As we can see from the chart below rolling scoring% is on the rise. REGRESSION! Scoring%, by the seasons end, will settle at ~31%.

Sc__wk_8_medium

3) Arsenal: Shots Influence

Arsenal are top of the league after 8 games and congratulations to them. They have recorded some excellent results so far, although slight negatives would be the weak quality of opposition that Arsenal have faced and Arsenal's non-stellar underlying numbers.

Anyway, I don't want to be a killjoy about a team who is playing some bloody nice football and should get better when some key players return from injury. Wenger must rotate his players in November and December though, gotta keep the players fresh.

Shots Influence

Shots influence looks at each players contribution to team shots when on the field of play. So, we count a players shots and shots assists and divide them by the team total. Giroud and Ozil are the two players who are driving Arsenal's offensive output thus far. For me, Ramsey is by far the most interesting: why is Ramsey's number rising on a game by game basis? is Ramsey's role changing, which allows the player to be more influential, or is it something else? Confidence or a self belief that he can be an important contributor on the offensive side of the ball? It'll be interesting to see how long Ramsey continues this improvement in terms of shots influence.

4) Set Piece Stuff

Set pieces are funny old things, eh. I can't decide just what the luck/skill split is for set pieces and this makes evaluating some of these numbers problematic. If pushed I'd say shots creation/prevention has far more skill than luck.

Still, here are the set-piece SHOTS numbers through the first 80 games. *Set piece shots= shots from corners, shots from free kicks, shots from direct free kicks. (I wish I could separate direct free kick shots but I can't)

 

Sp_shots_wk8_medium

Man City are doing some pretty wild things in terms of generating set-piece shots for (1st) and preventing set-piece shots against (1st). This doesn't really fit with Man City the weak set-piece team. City have conceded a low, low number of shots from set-pieces but what about goals?

Sp_goals_wk8_medium

So, City have conceded just 14 shots from set-pieces but those 14 shots have yielded 4 goals. City conceded% is28.7%, league average is 9.2%. In short, City won't carry on conceding goals at that rate.

A word on Fulham who have conceded the most shots and taken the least shots from set-pieces. Fulham have scored 5 set-piece goals from 16 set-piece shots. Unsustainable to say the least.

5) Week 8 = No Shocks

I had a mighty good week in terms of betting on the PL; if I believed in accumulators then the girlfriends shoe collection would look a whole lot better right now!

This was the week of the fav: Chelsea, Arsenal, City, Arsenal, Fulham, Everton and Swansea. Even the tricky picks weren't that tricky. Stoke were a good shout to draw and opposing Liverpool and Man United in last weeks games wasn't too difficult a decision.

Having no shocks in a round of games is a pretty rare occurrence.

6) Requirements: Man United

Man United: Currently have 11 points from 8 games, which is very un-United like. Would it be fair to say 83 points would be in the ballpark to win the 2013/14 PL? That means Man United need 72 points or 2.4 ppg from here till the end of the season.

2.4ppg is a 91 point pace over the course of a 38 game season. Absolutely nothing about this Man United team gives us the slightest hint that they can suddenly become a team capable of 2.4 ppg/91 point season over 38 games.

There is a glimmer of hope for United in that their schedule has been mighty tough so far this season and at some point (last 10 games?) United's schedule will cool down significantly. Still, United are facing a 30 game run of 23 wins, 3 draws and 4 losses to get to 83 points by the seasons end. Even 83 points may not be enough.

6) Requirements: Sunderland

Sunderland have 1 point from their first 8 games. 38 points is the line in the sand for PL survival. 37 points from Sunderland's remaining 30 games works out at 1.23 ppg, or, normalized over a full 38 game season, ~47 points. It seems to be very unlikely that Poyet can turn this Sunderland side into an upper mid table team in his remaining 30 games.

Obviously, we never want to deal in certainties, but it looks pretty clear from this writers point of view that Sunderland are in desperate trouble. If you need further evidence, just go look at Sunderland's SoTR and PDO numbers from this post.

To me, in my dark, attempting to quit smoking mood, this is what doom sounds like. This is Sunderland's sound:

8) Palace: How Low Will You Go?

Two points from 8 games and looking severely over-matched in most games they have played, what will become of Crystal Palace? Relegation seems to be the likely answer and, quite frankly, it was always on the cards what with minimal investment and a thin squad to begin with.

Palace are on course for 9.5 points. Obviously that is not going to happen, but is 20 points a fair line in the sand for this Palace team? Is that too low?

How about 25 points which is the only live line I can find right now. Do Crystal Palace make it to 25 points? I'd be amazed, quite frankly. They appear to be awfully weak right across all positions and without a huge investment in January (unlikely) this looks like the worst team to grace the PL since those glorious Sunderland and Derby County teams.

Obviously, when Palace survive on the last day with 40 points, I expect someone to dig this up and tell me how stupid I am!

9) Game State Effects On Shooting Accuracy And Prevention

Long title, a ton of work but by the end of the season I will have so much information on game state effects that it will make the effort worth it.

For now, I just want to show you how the game state and the tactical effects of the game state affect shooting accuracy% (shots on target/total shots)

Shooting_acc_game_state_medium

Now, the information at +2 game state/-2 game state will need some time to smooth out but there are some interesting things going at the Close Game States.

Tied game state shows us information about teams' behaviour without any score effects so things here look pretty normal.

Plus 1 Game State is mighty interesting through the first 80 games: The team leading by a goal (plus1) manages to get a higher% of it's shots on target than their opponents who are trailing by a goal (35.2% to 33.7%) and we know this is probably due to shelling effects.

Shelling effects? The team that leads by a goal tends to sit back in a tight defensive shell which restricts their opponents ability to get shots on target. Included in this shelling tactic is the counter attacking bonus: it's easier for a team leading by a goal to create better shooting opportunities as it's trailing opponent pushes hard for an equalizer and leaves a lot of space to be exploited. That space allows the team at Plus 1 to get a higher percentage of their shots on target.

Looking at Minus 2 game state, it's easy to see why the game is virtually over with. The team that is trailing by two goals finds it extremely hard to get their shots on target, whilst at the same time leaving their defense horribly exposed to efficient opposition attacks.

In short, if your team falls a goal behind it's more difficult to get your own shots on target and the opposition find it easier to get their shots on target. Falling a goal behind is a lose-lose situation.

10) Team Goal Of The Week

A thing of beauty. Controlled passing, speed of thought and feet, and a snap finish by Wilshere on his right foot foxes the 'keeper. Beautiful goal.