Football Thoughts and Figures After Dec 1st
Making More Grumpy Liverpool Fans
The last time I wrote about Liverpool, I expressed some concerns about their fundamental numbers despite good results so far. That was after game 5, when Liverpool were 3rd in the table. Now we’ve just completed match 13 and Liverpool are in 4th place.
That’s the good news.
The bad news is that the numbers are a lot more solid now, and there are legitimate, clear problems. Here are the important baseline stats that I track for all teams.
(The first two numbers are for the offense. The second two are defensive.)
Offensively, Liverpool’s total shots have dropped, but chance quality has likely gone up and the effective shots are basically in line or better than last season. That’s fine. They are going to miss Sturridge, but the offense should survive.
The problem comes on the defensive end. Last year, Liverpool were excellent at limiting opposition shots (4th in the league), and as you can see, this year they are in the bottom half of the table. Not only that, but out of the 2.3 extra shots per game they are allowing, 1.6 of those per game are going straight through to Mignolet.
That’s a huge warning sign.
There is a positive here in that I checked with fellow ‘Bomber Colin Trainor and he says that Liverpool are doing fine in terms of location. Though they do concede a lot of shots, the ones they give up are generally bad shots (they have a fairly low probability of being a goal), so the shots the opposition takes are generally less valuable than what you’d see teams get against Manchester United.
The negative is that based on probability alone, shots conceded at this rate will start to bite a team in the ass fairly regularly, even if they are mostly “bad.”
I checked with Ben’s game state data (information that looks at what the score was when shots are being taken), and it indicates that Liverpool probably have a poor defensive shell, which is something we talked about back in September. They are now getting good chances on the counter coming out of it (Sturridge and Suarez are amazing counterattackers), but they aren’t nearly good enough at disrupting opponent attacks when employing it and because of this, you can expect them to give up too many goals.
To compare within the Premier League, we know that Chelsea and Arsenal will employ a defensive shell fairly regularly. Chelsea gives up 3.23 SOT per game, Arsenal give up 4.15. Liverpool’s 5.23 is dire.
As in, the teams around them in that stat are all relegation candidates.
Curious to do a bit more research on it, I looked at all the Champions’ League teams from the big leagues for last season. The most shots on target any team gave up and still managed to qualify were Sociedad’s 4.45 followed by Chelsea’s 4.21. Liverpool are definitely in the red zone.
It was clear early in the season that Rodgers switched his system from a possession-oriented one to one where they get ahead and then attack out of a defensive shell. That shell isn’t right though, and it will cost the team points and goals as the season goes on. It would be best if Rodgers could figure out some acceptable hybrid like Arsenal or Tata’s Barcelona now employ where they switch from possession to counter at need while still being able to limit opposing shots.
I think BR is doing good work, but this team and their tactics are still very much a work in progress.
Just Drawful or Welcome to Moyesland
Manchester United drew 2-2 at White Hart Lane this weekend, giving them their fourth draw of the season. No big deal, right? Well…
Over his last five seasons at Everton, David Moyes’s team drew 34.7% of the time, an average of 13.2 per season. That compares with a 26.7% draw average for the league as a whole during that time. Moyes’s percentage isn’t normal - it’s actually really high.
During the same five seasons, Sir Alex Ferguson’s United teams drew 16.3% of the time or an average of 6.2 draws per season (brought up much higher by one season where United drew 11 times). Obviously the quality of the teams were different and really great teams will generally draw less than teams in the middle of the table, but again, Moyes’s draw percentage lead the league consistently.
Now draws against superior opposition are mostly okay, but draws in general are to be avoided if teams want to maximize their chances to pick up points. So far under Moyes, United have drawn 4 league matches out of 13, or about 30%, which is above league average and would give them 11 draws across a whole season. To be fair, Fergie did win the league with 11 draws in 10-11, but only had 4 losses all year. Additionally, only 4 of 20 teams during that period had that many draws and still made the Champions’ League.
Draws for top teams are bad.
Moyes needs to get more comfortable with letting Rooney and the boys off the leash if they are to challenge for top spots in the league, including a Champions League place.
“Arsenal Are Not a Good Defensive Team, Despite Their Record”
We all say dumb things from time to time, but most of the dumb things Stewart Robson says seem to be about Arsenal. And this statement is waaaay dumb.
Fact: Arsenal were second in goals allowed across 38 games last season, conceding 37 goals in that time (City were a hair better with 34 against).
Fact: This season Arsenal are tied for 1st in the league, conceding only 10 in their first 13 matches.
Combine the numbers and Arsenal have allowed 47 goals in 51 matches, which is one behind City’s output of 46 in that time.
Arsenal are second in goals against across 51! games. I don’t care if you don’t like how they defend or can’t fully understand what they are doing. At some point you have to admit that based on a fairly large sample of results, this team is actually really good on the defensive side of the ball.
They also lead their very tricky Champions League group with only three goals conceded in five matches.
Arsenal aren’t just good defensively, they are outstanding.
Full credit to Steve Bould for helping to make that happen.
Goodbye Martin Jol
Normally I am fairly conservative about suggesting managers should be sacked. After all, if you are going to bother firing one, they should probably be really bad and you need to be able to replace them with someone better.
On the other hand, I have publicly battered The Guardian’s writers on Twitter for writing that Fulham were too good to go down.
Martin Jol should have been fired months ago.
As of October 1st, Fulham were giving up 20 shots per game while only taking 8 themselves. This is very, very bad. Through this weekend’s dire performance, where they lost to one of the worst offensive teams in the league by a 4-0 score line Fulham are…
Giving up 20 shots a game, while only taking 8.
Even Pescara, who most models considered to be the worst team in the big European leagues by far last season took 10 shots while giving up about 18 a game. Fulham were worse than that.
(Yes, I know our predictive model liked them more than I do, and I know why, and all I can say is that it was horribly wrong here. Welcome to modelling.)
The important question is: Can they be saved?
I honestly don’t know. I think Riether and Hangeland are pretty good in their back line, and I think Berbatov and Ruiz are very good up front. Everyone else on that team is either a question mark, actively bad, or truly horrific. The midfield simply doesn’t work with how Jol wanted to play, and the axis of Sidwell and Parker seems to do absolutely nothing to help either end of the pitch. Meulensteen has his work cut out for him, but the next few months should teach us a lot about whether a bad team with pockets of talent can be saved by a new manager.
[Note: it did not work last year for QPR.]
They need to unearth players in January to stabilize the center of the pitch, or they are goners.
- Chelsea beat Southampton pretty handily this weekend (I still think Ba should play more), but their midfield hole looked worse with Essien in it, not better. Injuries have taken too much of a toll on the former great midfielder, and they’ll definitely buy in January.
- I expected Southampton to get more than 1 shot on target against Chelsea. I have yet to see Osvaldo look like he’s worth anything near what Southampton paid for him.
- Remember when everyone was slagging off Gareth Bale at the beginning of the season? Now he’s in shape and producing even better than he was last season (probably because he has better talent surrounding him). His ability to get off quick shots and passes is absolutely tremendous.
- Delofeu had a Bale-eque game this weekend against Stoke with a goal, an assist, and four of his five shots on target. The kid is lightning in a bottle and starting to get his head up and look for teammates as well. He’s also only 19. Everton fans need to hope he rubs off a bit on Ross Barkley.
- Barcelona lost! This is good for La Liga. They need to pretend the league is competitive beyond the top 3 from time to time.
- Juventus won! This is bad. Despite Roma’s hot start, Juve are almost certainly going to run away with Serie A again. Too deep, too good, and a manager who almost never makes mistakes. Speaking of, Roma still have only given up 4 goals in 14 matches. Can they average under .5 goals conceded a game for the season? Doubtful, but hey, no one expected them to be anywhere near this good.
- Teams the model thinks are definitely getting relegated at the end of the season: