Friday, December 2, 2011

There Are Other Arsenal Players

Jesus Christ, it's been a solid month! I knew going in that getting a new job would keep me from blogging as regularly, but not to this degree. I'm sorry to all of my fan. Anywho, a lot has happened in the past month, but I'd like to focus on what has been happening at Arsenal.

Since my latest post, Arsenal have played really well. They had a five game win streak snapped with last week's draw with Fulham. The reason for their success? Everybody in the media will tell you it is Robin Van Persie. The accusation that Arsenal are a one-man team has been reinforced by the spectacular failure of their other strikers to actually score, which is the most important part of any striker. However, I'd like to point out some significant holes in this idea.

First of all, I think people are forgetting that this is not the first year Robin Van Persie was good. He averaged a goal a game last year, albeit in less than half a season. His strike rate was better than anyone in the league on a per-game basis, proving that he is a world-class talent, if a little injury prone. I think he is one of the best players in the league and I'm not belittling his contribution.

Still, he is not the only person responsible for Arsenal's recent success. Let's look at some numbers shall we? Here is the efficiency table for Arsenal, the top numbers being over their last six games and the bottom for the entire season:

Off Eff Def Eff
FR Opp FR Possession
Last Six Weeks 130.35 73.46
38.10% 37.50% 59.67
Total 121.85 106.12
31.33% 36.00% 58.92

As you can see, Arsenal's Finish Rate has indeed gone up, meaning that more of their chances are being taken. That is probably attributable to Van Persie, true. But what jumps out at me is the Defensive Efficiency number of 73.46, meaning that Arsenal's opponents are having a lot of trouble creating chances even after normalizing possession. This defensive efficiency increase has led to Arsenal conceding .6 fewer goals per game over this stretch. Van Persie's finishing exploits over this period have garnered an increase in the Goals For column of the same amount and everyone is talking about it, but the defensive improvements have not merited comment.

Of course, the real story behind Arsenal's success (stop me if you've heard this before) is that they've played an easy schedule. Four home games and only one team in the top half of the table. Most people have been saying they would have been toast without Van Persie in these past few weeks; I counter those were the weeks they actually would have survived his absence.

Friday, November 4, 2011

Weekend Preview: What to Make of Newcastle?

Newcastle is 3rd in the table and one of two unbeaten teams in the league. This wouldn't have been a big surprise not that long ago, but given the disastrous tenure of owner Mike Ashley, this is a huge shock. Ten weeks in, the punditry are finally forced to accept they need to take a look at Newcastle. Some see a fluke, some a genuine contender. But which is it?

Unfortunately for Newcastle fans, it's a fluke. First off, I feel I need to make the somewhat obvious point that being unbeaten is meaningless. In this league, a win and a loss is better than two draws. Saying you're unbeaten sounds great, but it really doesn't make you a great team. Moreover, if we analyze the four drawn matches that Newcastle has had so far, only in one of them has Newcastle produced more shots on goal than the opponent, meaning they were the more likely loser in the other three. If they had lost one or two of these, I doubt Newcastle's start would be that big of a deal.

The more common argument for the unbelievers has been that Newcastle haven't played anybody good yet. Newcastle have played the easiest schedule out of any team in the league. Their opponents have average 1.04 points per game. Averaging 1.04 points per game for the whole season would put a team below the 40 required to avoid relegation. Moreover, Newcastle's toughest two games (Arsenal and Spurs) were both at home. Suffice to say Newcastle have been a little lucky on the scheduling front so far.

That wouldn't be as much of a concern if Newcastle were dominating their opposition, you can only beat the teams in front of you after all, but they're not. Though they are tied with Chelsea for third in goal differential, they are only sixth in SOG differential, and that is with an absurdly easy schedule. In fact, their average goal differential is higher than their average SOG differential, something that is usually untenable.

Just to be clear, Newcastle are not a bad team. I think they'll end up fighting for 7th with Everton and Sunderland. However, they are not better than any of the top six and their easy early schedule hasn't given us the best indication of their true talent level. We'll see what happens after their next four games of Everton, City, United, and Chelsea.

Tuesday, November 1, 2011

Away Teams Pay the Penalty

Chris Anderson has another excellent post at Soccer By The Numbers about the differences between home and road teams. He notes that not only do home teams accumulate more shots on goal, but also that they have a higher Finish Rate (i.e. they convert a higher percentage of their shots on goal) and wonders why that would be the case. Fortunately, I think I have an answer for him.

First of all, let’s ask the important question: why should being the away team make any difference in the outcome of a game? The field, the ball, the rules, the play; they’re all the same. What is it about playing away from home that makes any difference at all? There’s the comfort factor obviously. You are certainly more apt to be ready for the game if you wake up in your own bed and can follow your usual routine. Traveling can also fatigue the body, especially if it is a long trip. There may be certain details about the field of which you are unaware or can’t take advantage (ever notice how all of those ballboys at Stoke have towels to wipe off the ball for throw-ins just to help Rory Delap?). But the main difference is you don’t have the support of the fans.

What kind of impact can the fans have on a team’s ability to convert their chances? If you answered “negligible”, I think you’d probably be correct. However, fans do have a significant impact on the referee and how he makes decisions. In every sport, referees tend to give more calls to the home team and soccer is no exception. Fair enough, but what does this have to do with improving the home team’s Finish Rate? Simple. The refs not only award more free kicks for the home side, they also award more penalties.

As we know, penalties have the highest Finish Rate of any type of shot on goal; 80% of penalties that result in shots on goal go in the net. If a ref awards only a few more penalties to home teams relative to away teams, home teams will get a relatively massive boost to their Finish Rate because it is so easy to take those opportunities. It follows then that a team playing at home would post a higher Finish Rate than they would on the road. So far this year, the evidence has backed this hypothesis. Including penalties, teams have a Finish Rate of 31.14% at home, compared to 30.45% in all games. If we take out penalties, teams have a Finish Rate of 27.84% at home, below the 28.12% they average in all games. It seems that penalties are indeed what inflates the Finish Rate of home teams.

Though there may be differences in approach that make certain teams have a better Finish Rate at home or on the road, league-wide that doesn’t seem to be the case. Unfortunately, what seemed to be an intriguing difference between a team’s play on the road and at home is no more than a function of the referee.

Sunday, October 30, 2011

Is Petr Cech Good Anymore?

Arsenal's victory at Chelsea was certainly impressive, but what I took away from it was how bad Petr Cech has become seemingly overnight. He was beaten three times at his near post, all of which were stoppable. It is the latest in a disturbing trend for Chelsea this season: their Opponents' Finish Rate is 3rd-worst in the league at 41.18%, behind only Bolton and Blackburn.

It looks even worse when you take a harder look at the data. This early in the season, one might suppose that Chelsea's struggles in this department might be the result of a tough schedule. In other words, perhaps the teams Chelsea is playing are very good at taking their chances. However, their opponents have posted an average Finish Rate of 26.84% when not playing Chelsea, below the league average.

Perhaps this is a function of the team instead of Cech: Chelsea's bad defending leads to easy opportunities which their opponents duly dispatch. Chelsea may be liable give up more fast break opportunities because Ashley Cole and Jose Bosingwa push up so high and more fast break opportunities generally lead to a higher Finish Rate. Or they may be susceptible to free kicks and corners, which although they do not often result in SOG, typically go in at a higher rate when they are on target. I don't really have any data on that, but it's possible that it's not Cech but rather a function of the way Chelsea play. Arguing against that is that last year Chelsea posted a 23.74% Opponents' Finish Rate, second-best in the league. Other than Villas-Boas, the personnel is the same, but yet the results are completely different.

Cech is only 29, not old by football standards, and certainly not for a goaltender. I'm really unsure what has caused this dramatic turnaround, and this may prove to just be a rough patch. However, it's such a dramatic difference that I can't help but wonder if this may be the fatal flaw in Chelsea's title challenge.

Friday, October 28, 2011

Weekend Preview: Most Exciting Teams

Last season, I created a simple little measure to determine which teams were the most exciting to a neutral. I did so by finding the average number of scoring chances both created and allowed (using shots on goal as a proxy) by each team in a given match. Last year, the three most exciting teams were West Ham, Blackpool, and Chelsea and the three least exciting teams were Stoke, Blackburn, and Manchester City. Obviously, a lot has changed in the excitement table, as West Ham and Blackpool are unfortunately no longer in the league, while Man City have been much more entertaining this season. Here is the table in full, the number is the average number of SOG occurring in each team's games:

Tottenham 12.50
Man United 12.33
Man City 11.89
Bolton 10.89
Wolves 10.56
Arsenal 10.00
Norwich 9.89
Chelsea 9.67
Swansea 9.44
Fulham 9.22
Liverpool 9.22
Blackburn 8.67
Wigan 8.56
West Brom 8.56
Everton 8.50
Sunderland 8.22
Newcastle 8.00
QPR 7.89
Aston Villa 7.56
Stoke 6.78

As you can see, it is quite different. Manchester City have gone from being one of the most boring teams to one of the most exciting almost overnight. Wolves and Man U are also high up there, so it's no surprise that City's last game against both saw 7 goals between the sides. Stoke is still the worst value for your money, though Aston Villa and QPR are right behind. For TV viewers, I would say you should check out City vs. Wolves and Chelsea vs. Arsenal, and steer clear of Stoke vs. Newcastle.
Gambling Picks: Bolton over @Swansea 5/2.  I think Bolton are still underrated. They have played a very tough schedule and deserved to win last time out against Sunderland, but ultimately lost 2-0. I think they'll have a bounce back against a Swansea side that plays very open football which will be to their benefit.

Thursday, October 27, 2011

Six and the City

Sorry, couldn't resist throwing in one of the many puns out and about. I honestly am still having trouble believing that it actually happened in that fashion. As I said before the game, I expected City to win, but to win 6-1? Obviously, the red card had an impact on the result, but even if Evans hadn't fouled Balotelli, Balotelli would surely have scored and we would have been 2-0 up at the start of the second half. I'm sure we would have gone on to win from that position.

The game itself was amazing to watch. Some of the moves that City put together were mind-blowingly good. I loved Aguero's goal, the flick from Balotelli in the buildup was sumptuous. Silva's no-look 30-yard volley splitting two defenders to find Dzeko for the sixth was simply unbelievable. That game will live on in the memory for quite some time.

However, it is way too early to say that this means anything definitive. City are the best team right now, I don't think there's any doubt about that, but it's not necessarily going to continue and we should all just...

Nah, screw it. City to win the league. Not because of this one game, but because through the first quarter of the season, they have been demonstrably better than everyone else and they have a large enough squad to cope with just about any injuries. City's goal differential is higher than their points total at +26, they have the largest SOG differential, the 3rd most possession, the highest Finish Rate, and the second-lowest Opponents' Finish Rate.

More importantly, the other challengers haven't looked nearly as good. United have the worst defensive efficiency in the league, meaning they give up more shots on goal per opponents' time of possession than anybody. Chelsea don't have a real weakness the way United does, but they have been behind City in all aspects of the game. It's hard at this point to see either outplaying City over the rest of the season.

Saturday, October 15, 2011

Contrary to Popular Belief, De Gea is Really Good

Throughout the season, there has been no summer signing subject to more scrutiny than United's new goalkeeper David De Gea. Somehow, every game people have been questioning his positioning, or his ability to deal with crosses, or his command of his penalty area. And yet, somehow, he's not only been good, he's been the best goalkeeper in the league through these first eight games.

Judging goalkeepers is somewhat easier than other players because they are solely limited to the defensive side of the ball. The best measure of a goalkeeper is what I term the Opponents' Finish Rate, or the percentage of shots on goal that lead to goals for the opposing team (G/SOG). In games that De Gea has started this season, he has posted a 13.5% Opp FR, best in the league among starters (though his understudy Lindegard has a 0% Opp FR thanks to his one clean sheet). The next few names on the list are Krul at Newcastle (mainly due to Newcastle's easy schedule thus far; I expect him to drop), Begovic at Stoke, Hart at Man City, and Given at Villa. Krul aside, those are some of the best keepers in the league.

To be sure, that one statistic doesn't tell us everything. What about his struggles with crosses? Actually, we would expect that to show up in the above statistic. Think about it. Shots on goal directly from crosses produce goals at a higher rate than other shots on goal, mainly because they tend to be taken closer to the net. If De Gea were struggling with crosses and allowing opponents chances that are easily convertible, we would expect his Opp FR to be high, but it's not.

A more legitimate concern would be the command of his penalty area. It's possible that if he can't direct his defense, opponents will get more quality opportunities. Since Opp FR is a rate stat not a counting stat, this would not show in the data. There may be some truth to this as United have allowed a lot more shots on goal this year compared to last. However, I don't think the blame should be entirely directed at De Gea: as I pointed out in a previous post, United's defense has actually been very poor this year. Even if it was true, I would expect it to improve as De Gea improves his English and gains experience in the EPL.

It's funny, but I believe United to be very overrated and their most commonly called-out player to be very underrated. I expect City to win next weekend, but I don't think De Gea will be the reason.

Wednesday, October 12, 2011

Meet the New Boss, Same As the Old Boss?

It's starting. Despite the season only being seven games in, speculation on which English Premier League manager will be fired first has begun in earnest. The favorite right now would have to be Steve Kean (1/2 odds on Skybet), whose Blackburn Rovers were recently demolished by Manchester City, but he is far from the only one in the hotseat. While it is only natural that fans tend to blame the manager for the team underperforming, I can't help but wonder if fired coaches are the victims of overreactions. After all, replacing a coach mid-season generally requires that you replace him with someone currently out of work, not usually the best pool of applicants. Also, despite the influence the manager has, it is the players who most determine success on the field. Do teams that fire their coaches mid-season tend to do better after the firing? To help answer the question, here's a look at the four teams last year who fired their manager during the season (Blackburn, Liverpool, Newcastle, and West Brom) and how they fared under their new manager.

Friday, September 30, 2011

Weekend Preview: City vs. Bayern Aftermath

The City-Bayern game was painful to watch. Not just because we lost, but the fact that the referees didn't give two clear penalties and missed Gomez being offside for Bayern's second. Mancini's substitutions were also bizarre. I understood the reasoning for bringing on De Jong, we were far too open at the back. What I don't understand was not playing him in the first half and taking him off in the second to add more to the attack. This was Mancini's modus operandi all of last season and I have no idea why he abandoned it in our biggest game so far this year.

Then there's the Tevez situation. I said before the season that if we acquired Aguero, we needed to get rid of Tevez. We didn't and now we are reaping the consequences. The thing is that if Tevez had no suitors in the summer, how are we going to offload him in the short window? And that's without considering that teams may be afraid to sign him because of dressing room issues now. A real mess without an easy answer.

  • Last week I picked Stoke to beat Man U. Unfortunately, they only drew. However, I think my point about United being weaker than they seem was somewhat vindicated. I think the Basel game showed why I was concerned about their defense as well.
  • Gambling Picks: Everton (19/10) over Liverpool and Newcastle (19/10) over Wolves. Liverpool haven't played well lately and I like Everton to win at home. Wolves just suck and Newcastle are actually pretty good. Just goes to show if you have a striker you can get 35+ million pounds for, SELL HIM.

Friday, September 23, 2011

Weekend Preview: Buy/Sell

So apparently, this is the favored new format for ranking systems: viewing teams as if they were stocks and then deciding to buy or sell depending on if you think their value will rise or fall. Without further ado, a few stock tips (in addition to selling on Stoke which I prophetically talked about last column) I think people should have.

Manchester United: SELL. Don't let the scorelines fool you. Chelsea outplayed them last game at home, West Brom outplayed them in the opener, and they have been really lucky in their Finish Rates. Right now, ManUre allows more SOG per time of possession than any team in the league, but those shots simply aren't hitting the back of the net. Either De Gea is brilliant, which fails the eye test, or they have been lucky. The same is true at the other end of the pitch where they have posted a Finish Rate of 46%. Half of their SOG going in the net? Not going to continue. This is not to say that they won't do very well this year, I still think they're the favorites for the title, but they won't run away with it like people seem to be thinking.

Chelsea: BUY. The game against Man United was just the latest good display by Chelsea. They have been consistently getting more SOG than their opponents and only failed to win at Stoke and against United. If AVB figures out his best starting lineup, they could be really scary. I really think they're closer to City and United than to the teams below them. By the way, I'm starting to think that AVB bought Mata just because he's his doppelganger.

Everton: SELL. I actually think Everton might win against Man City tomorrow, given that they have always won against City the past few years. That said, their excellent start has been fueled by 3 out of 4 games being at home against very poor opposition. I'd wait until tomorrow, and then get rid.

West Brom: BUY. West Brom have played an incredibly tough schedule, have played well, and have earned next to no points. I think Hodgson is an excellent manager who was unfairly blamed for Liverpool's poor start and I hope he isn't unfairly blamed for West Brom's start. West Brom have the worst Finish Rate differential of the league and a large part of that is luck. I think they get back on the winning track tomorrow against Fulham.

Aston Villa: SELL. I don't like this team. They don't keep the ball well and they don't have the efficiency on the break that carried the team in the past. Bent is good and Given has started very well, but I don't see their unbeaten start continuing, particularly since they have played an easy schedule so far.

Arsenal: HOLD. Arsenal's defense has been certifiably terrible, but I think once Vermaelen comes back it will improve. They've also been a bit unlucky in their Opponents' Finish Rate and their first two games had red cards, which limits how much we can take away from their performances. All in all, we haven't had enough time to judge how the new-look squad will shape up.



  • Gambling picks: @Stoke over Man U 5/1. I know, I know. I think both of these teams are overrated, but I like how the matchup plays out. United haven't yet played a team with a particularly good defense and I think Stoke will be a problem for them. United have also allowed too many chances on the break and I think Stoke can take advantage of that. This is the part where I urge you to not put too much stock in my advice.
  • I am officially christening Bryan Ruiz, who wears "Bryan" on the back of his jersey, William Jennings.
  • On a personal note, I have a new Sharp Aquos TV sitting in my living room. I can't wait to watch City-Everton in HD tomorrow on it. Good stuff.

Thursday, September 22, 2011

Fouling Is Not Cheating And Foulers Don't Prosper

I feel I need to respond to Chris Anderson, the Cornell professor who writes the blog Soccer by the Numbers. In his most recent post, he looks at whether a team fouling more is correlated with winning and finds that it is not (I also have to take issue with the title. Cheating is breaking the rules without being punished; fouling by definition is punished). He then looks at the same question on a team by team basis and finds that there are some teams with which fouling more is correlated with winning, including our beloved City. However, this analysis doesn't hold water if you take a deeper look at the statistics.

It's true that City fouling more has a slight correlation with earning City more points (emphasis on slight; the R-squared is .0214). However, the correlation is slightly stronger between City suffering more fouls and earning more points. It's even stronger between the total number of fouls in the match and earning more points. This suggests to me that City do not play better when they foul more, but rather the games in which they do well tend to have more fouls in them. Why would that be? One possible explanation could be refereeing differences. It's possible that City fare better when games are called rather tightly, protecting our creative players and allowing them to work. That solution doesn't make too much sense though, as we would expect the same to hold true for other talented offensive teams were that the case. A more compelling explanation is that City favor a game that is more fast-paced, i.e. a game in which possession changes hands more frequently. In such a game (although I have absolutely no evidence for this), I would posit that there are more fouls because the game is less settled. This would pass the sniff test, as City last season tended to labor when they had steady possession (one of the many reasons the recent comparisons between City and Barca, who are excellent at making possession count, are weak). They tended to do better in games with more fouls since they are indicative of a more fast-paced, counter-attacking game. Unfortunately, there's no way I can really test this hypothesis as possessions are not shown in the official statistics.

Friday, September 16, 2011

Weekend Preview: Not Stoked About Stoke

Currently, Stoke lie 5th in the table, earning eight points from four games played. That's pretty impressive, especially considering they've already played Liverpool and Chelsea. However, I think that Stoke will be lucky if they finish in the top ten this year. Their good start has been fueled almost entirely by luck and it is extremely unlikely that it will continue.

Take a look at their most recent Premier League game against Liverpool. Stoke produced zero, yes zero, shots on goal from open play and allowed seven. They won because they were awarded a penalty, an event that is extremely unlikely to be repeated weekly. They allowed seven shots on goal, none of which went in the net. It's hard to claim that Stoke deserved to win that game, or even that they defended stoutly. They were lucky, plain and simple.

This is not the only game where Stoke have ridden their luck. Against Chelsea, Stoke had one shot on goal, Chelsea had seven. The game was a 0-0 draw. Against West Brom, Stoke had one shot on goal, West Brom had four. Stoke won that game 1-0. All told, Stoke's opponents have averaged a Finish Rate (G/SOG) of 4%. The best mark last year was Man City with 23.4%. This is unsustainable.

Even the lone game they actually outshot the opposition is not that impressive. Stoke did have five shots on goal to Norwich's three in that game, but Norwich went down to ten men in the 63rd minute. Before that, the teams were equal in shots on goal.

At the moment, if we looked at Expected Goal Differential, which assumes teams convert chances at the same rate, Stoke is second-worst in the league. Now this doesn't account for the fact that Stoke have played a relatively tough schedule (Tottenham are at the bottom at the moment), but still it illustrates that Stoke do not belong in the top half of the table on the merit of their play so far. And the thing is, this is before Stoke's commitments in the Europa League come into play. If Stoke were a stock, I would urge everyone to sell high immediately.
Other things:

  • Gambling picks: Sunderland 7/5 over Stoke, West Brom 13/8 over @Norwich. Given that I just wrote about Stoke in detail, I don't think I need to go over the logic of that pick. West Brom have been criminally unlucky so far this season and should beat Swansea even on the road.
  • Two very interesting games this weekend in Tottenham-Liverpool and Man U-Chelsea. I see these two as toss-ups, but would probably go with the home sides if you held a gun to my head.
  • Man City drew with Napoli during the week. I was fascinated by Napoli's approach as a team, talk about intelligent counterattacking football. The 3-4-3 has been a favorite formation of mine since I started playing FIFA 05 (FC Porto with Anderson, Quaresma, and Lucho Gonzalez was pretty unbeatable) but I've never seen it used at the top level. But man, was it effective. It frustrated City's attacks through the middle and led to some great breakaway opportunities. Adam Johnson coming on earlier would have helped a lot. Also, Tevez really sucked. He looks fatter than Michael Johnson now.

Sunday, September 11, 2011

City vs. Wigan Recap

Another excellent win. I was surprised to see Tevez start, particularly as this might have been a chance to get Balotelli a game (he's suspended for the first few Champions League games). Tevez was probably the weak link in our lineup, he sat too deep and clogged the linkup play, and generally didn't play anywhere near the level Dzeko has over the past few games. And that's without mentioning his missed penalty.

It's interesting to contrast him with Aguero. In a lot of ways, they are similar. In particular, I love how they both have such quick releases on their shots. How many times do they get a shot between the defender's legs? The quick shot just after a dribble is a huge weapon for both of them. However, Tevez needs to be a focal point of the offense to be effective in a way that Aguero does not, hence Tevez's struggles today in this team.

Still, the team played well. Apart from the usual suspects (Silva, Aguero, Toure), I was very impressed with Clichy. When we signed him, I had heard concerns about his crossing, but it was excellent in this game. I much prefer him to Kolarov at this point.

At this point, it's very hard to see a weak link in City's play. Midfield has been solid, the defense has been excellent even without DeJong in front of them, and of course the attack has been overwhelming. The most important thing is that the rotation of the squad hasn't affected the level of performance.

Comparisons with United are inevitable of course, given that United are the only other team with maximum points. I still believe that City have played better though. The old adage "Form is temporary, class is permanent" springs to mind. United still have an incredibly high finish rate (44.7%) and an incredible low Opponents' Finish Rate (12%). It's the latter number that is clearly unsustainable and I expect United to start drawing some games once they play teams that sit back more. United's defense to date has been unimpressive, posting the league's second-worst Defensive Efficiency, or shots on goal allowed per time of possession. Bolton, despite only having the ball 42% of the time, had six shots on goal, none of which went in. That will not always happen. United are certainly in good form, but I don't think they have the class of City.

Wednesday, September 7, 2011

Swansea is a crazy team

Browsing through my data, I came upon a really interesting observation: Swansea City is second in the league in time of possession. The tiny team from Wales has had more possession in its first three games (61%) than any team other than Chelsea, more than last season's leaders Arsenal and the Manchester duopoly. Obviously, it's a small sample and can be partly attributed to scheduling (playing Sunderland and Wigan at home), but they did play City on the road and had the ball 57% of the time.

But here's the crazy bit. Swansea have also been extraordinarily bad at creating and allowing chances given their very high possession. I mean, second from bottom of the league in both categories bad. If we normalized possession, giving both Swansea and their opponents 50% possession, Swansea would have the worst SOG differential in the league. And it's not even close, the next lowest would be Arsenal and they're a full standard deviation better than Swansea.

So what does this tell us? One thing is for sure, Swansea have decided to continue playing possession football in the top flight. This team is not going to sit back and absorb pressure a la Stoke. It might seem that this gameplan doesn't suit them though, as they are poor at creating chances from their possession. However, a lot of the bad statistics are the result of being crushed by Man City in the first game. It would be foolish to read too much into that, particularly since they have played a little better in subsequent games. Still, I would be interested to see how they would fare sitting back and playing more of a counter-attacking style. Their match at Arsenal this weekend might well be instructive.

Monday, September 5, 2011

Goodbye Irene

Yes, I am still operating this blog. Unfortunately for me, Hurricane Irene saw fit to dump tons of rain in the creek next to my house, resulting in the loss of my garage, my driveway a third of my land, my basement, and (up to and including the present) power and internet. So one quick thought and I'll get back to enjoying cooking on a gas stove and using hurricane lamps.

It was certainly a great weekend in the Premier League. I did manage to catch the Blues' destruction of Tottenham in its entirety, but missed United's thrashing of Arsenal. And in some ways, I think it's good that I did because everybody is making way too much of this game. As of right now, United are not even close to City as a team if you take a deeper look at the stats.

The issue is that SOG are more informative than goals at this stage of the season. Because there are more SOG than goals per game and SOG are highly correlated with goals, looking at SOG gives a larger sample and thus tells us more about a team's true talent level at this point in the season. One way we can look at how good a team is at this point in the season then is to look at their goal differential given that the the team and their opponents convert SOG to goals at the same rate, what I call adjusted Expected Goal Differential (the assumed rate is .3 for both teams and the number is expressed per game). Here are the leaders so far this season.

Man City 2
Chelsea 1.1
Man United 0.9
Liverpool 0.5
Wigan 0.4
Wolves 0.3
Fulham 0.2
Newcastle 0.2
West Brom 0.1
Aston Villa 0
Blackburn -0.1
Everton -0.15
Sunderland -0.3
Bolton -0.4
Norwich -0.7
Stoke -0.7
QPR -0.7
Arsenal -0.8
Swansea -1
Tottenham -1.35

Now obviously, caveats apply due to scheduling. I'm still bullish on Tottenham, for example; they won't play Manchester teams every week (by the same token, I'm very down on Wigan, who got to play the three promoted teams in succession). However, City and United have played comparable schedules (United's was tougher, but they have played two home games to City's one) and City have been far superior. Consider that United have allowed 17 SOG in the past two games and only allowed two goals. That rate of allowing goals will not continue. Through this point of the season, City have been a goal a game better, a significant difference in quality.

Now this is not to say that City will be better over the course of the season. United's true strength is its depth, not its firepower, and I still expect them to win the title. However, people should not be making so much of United's start to the season. It is much less impressive than City's.

Friday, August 26, 2011

Weekend Preview

Couple of things to look for this weekend.

  • I'm very interested in the Villa-Wolves game. The interesting fact about this game is that both teams have played the same schedule so far, playing against Fulham and Blackburn. This means they are much more easily comparable than most other teams at this point in the season. Unfortunately for Villa, Wolves have played better against those teams than Villa so I'm predicting them to get an away win.
  • Man City-Tottenham will most likely be the most entertaining match of the round. Based on a formula for excitement I created last year (the total number of SOG in games featuring the team), they are ranked numbers one and two respectively.
  • Several people have mentioned the relative low scoring in the early part of this season. I don't believe that this will be a problem for the rest of the season. The chances being created are about equal, it's just that teams are finishing their chances 22% of the time rather than the usual 30% of the time. I predict scoring will increase in the coming weeks.

A Lost Week

Okay, finally back from NYC/Adirondacks/job search. Lots of things happening. Let's get to it, bullet-style.

  • Nasri signs for City. Color me unimpressed. A good player certainly, but one I don't think we need. It really all depends on how it impacts FFP. I'll be much happier if we manage to unload some of our unwanted players, the way we did with Adebayor.
  • Speaking of FFP, Malcolm Gladwell has a good article on Grantland where he talks about why sports franchises shouldn't make money. Replace franchises with clubs and he could easily be talking about soccer.
  • My first gambling pick certainly went pear-shaped. Note to self: never bet against the team you picked to win the title on the basis of one game. In my defense, I do have to note that the line moved toward Tottenham and I expected Modric to play. Oh well.
  • Champions League draw was certainly tough. Still, as a fan, it's more fun to see City play the best as opposed to some random team from Romania or Czech. The Bayern games in particular should be fantastic.

Sunday, August 21, 2011

Bolton vs. City Recap

Different formation, same result. Mancini started with a much more attacking formation, having Aguero and Dzeko paired together up top. The wrinkle I really liked was that he started Silva out wide right in a kind of 4-4-2, but Silva constantly moved into the middle behind the two strikers, dragging his defender with him. This made it easy for Richards to maraud forward on the right side with only Petrov defending him.

I liked this win a lot because we managed it with a fair bit of squad rotation. There was no De Jong in the lineup; the midfield consisted of Silva, Barry, Milner, and Yaya in De Jong's usual spot. Also missing from last weekend's win was AJ, replaced by Aguero, and Clichy, replaced by Kolarov. Good to know that we can change formations and players and still be successful.

Not too concerned about the goals we gave up because we were so potent offensively. I was a little disappointed with Richards defensively, he should know not to give Petrov that much space to cross. The free kick, well, Lescott challenged well, but Davies got a head to it first. Not sure exactly what you can do about that. Side note: I hate Kevin Davies. He constantly backs in to defenders, which while seeming innocuous enough, doing it when said defenders are in air it can really mess them up. The problem is that he never gets cards for doing it, so no matter how many times he gets whistled, he keeps on doing it.

Anyway, still top of the league. At this rate, can't see any reason why we shouldn't be there for a good long while.

Friday, August 19, 2011

Gambling Picks

Every Friday, I'm going to be making a few picks here, along with some reasoning. I usually only pick upsets, so I only pick a few games a week. This week I'm unsure how well I'll do, as there is not enough data to really tell what teams are for real yet. Nevertheless, I'm going to attempt it anyway. Follow my picks at your own risk.

Pick #1: Tottenham (5/1) at Man United
United would have been fortunate to get just a draw out of their last game, they were extremely lucky to pick up a win. They mustered a solitary SOG, yet scored two goals. Suffice it to say I don't think that is repeatable and they will have to be much better against Spurs to get three points. United are at home and they are the better team, but between injuries to key defenders, the shakiness of De Gea, and the form Spurs showed in midweek, I like Tottenham to win. I like them even more at 5/1 odds.

Pick #2: Annndddd that's it. Sorry, only one game I really like this week. More next week, I promise.

Introducing Advanced Statistics Part 1: Possession

This will be the first in a series of posts in which I attempt to explain some of the statistics I've created and why I think they are more useful than the traditional statistics.

One of the things I've tried to do with my statistics is separate out the different skills needed in soccer. As I see it, there are three skills needed on the offensive side of the ball: possession, creating chances to score, and turning those chances into goals. Similarly, on the defensive side of the ball the three skills are limiting opponents' possession, limiting opponents' scoring chances, and limiting the taking of those chances. The statistics I've created evaluate a team's skill in each of these areas. In this post, we'll focus on those evaluating possession. 

Possession is easy to calculate as it is provided in nearly every match report, with pundits often pointing to possession statistics as evidence of how a team played. In reality, there is no correlation whatsoever between possession and winning, but that doesn't mean possession is useless. Looking at possession can tell you how good a team is at keeping the ball and winning it back from their opponents.

Ideally, I'd like to look at the number of times a team had possession too, similar to what Hollinger does for basketball. Time per possession would tell us how good a team was at keeping the ball; time per opponents' possession would tell us how good a team was at winning the ball back. I think that would be much more useful analytically. However, the best measure of possession available to us is the Time of Possession statistic, or the percentage of time a team had the ball in the match. Since possession is zero-sum, we don't really need to calculate Opponents' Possession; in other words, we know from a team's possession how successful they were at limiting their opponents' possession. Last season, the league mean for possession was 50% (duh) and the range was from 60% (Arsenal) to 39% (Stoke).

What good is it to look at a team's possession if it is not correlated with winning? Well, for one thing it can tell us something about a team's approach. Take for example Stoke's game against Chelsea last weekend. Stoke had just 34% of possession. That tells us that Stoke was sitting back and absorbing pressure and that Chelsea was initiating most of the play. What it does NOT do, and what many commentators mistakenly think it does, is indicate that Chelsea was the better side. As stated above, last year Stoke averaged 39% possession and yet finished in 12th place. Possession is just one piece of the puzzle, the rest of which we will get to later.

Monday, August 15, 2011

City vs. Swansea Recap

Wow. That was better than I dared hope. I know it was only one game, but that was as complete a victory as you could ask for. Shots On Goal were 15 to 3 in favor of City, Joe Hart was truly tested just the once, and we took our chances well. Were it not for Swansea's new goalie Vorn, the game would have been even more of a slaughter.

But it was not just the goals, it was the manner of the performance. I don't think there was one player on the team who preformed poorly. Dzeko had his best game in a City shirt, Silva continued his ongoing audition for the role of Best City Player Ever, Johnson was threatening, the back line solid. And Aguero. Good God, what a debut! The first goal was good, his assist was fantastic, and his second goal sublime. his link-up play with Dzeko was perfect, it seemed as though they'd been playing together for years.

The only nitpick I have is that our defense did lose a fair amount when De Jong went out. Swansea's most threatening attacking play occurred in that period, having a couple of good counter-attacks. Obviously, Aguero's boost to the offense more than compensated for that, but it still seems we will have trouble being tight defensively in a 4-4-2 without De Jong. In fact, in Silva and De Jong, City have two players that seem to be irreplaceable. If we do get Nasri, he could maybe fill in for Silva, but De Jong has no backup.

However, that is a worry for the future. Right now, we are top of the league and hopefully we won't change our position in the table the rest of the season.

Wednesday, August 10, 2011

Preseason Predictions

This seems like the time to make them.

The promoted teams will struggle. As I have mentioned in previous posts, last year’s Premier League was both balanced (very low standard deviation of points) and strong (excellent UEFA coefficient). It would be tough for even very good Championship teams to break in and secure a place. However, it is not clear that the three promoted teams are even that good. Below is a table that shows the z-scores of promoted teams’ points in the Championship the previous season and whether or not they achieved survival. As you can see, things do not look very good for Norwich and Swansea, as only Hull City had a worse score and survived (by the skin of their teeth, I might add). QPR looks to be in a better position, and is certainly in a better financial position if they should have to spend to stay up. At this point, they are the only promoted team I’m backing to stay in the Premier League.

Year Team Z Points Survived?
2006 Reading 2.79 y
2010 Newcastle 2.70 y
2009 Wolves 2.17 y
2010 West Brom 1.96 y
2011 QPR 1.89 ?
2008 West Brom 1.84 n
2006 Sheffield United 1.78 n
2007 Sunderland 1.74 y
2008 Stoke 1.65 y
2009 Birmingham 1.63 y
2007 Birmingham 1.59 n
2011 Norwich 1.59 ?
2007 Derby 1.45 n
2011 Swansea 1.29 ?
2008 Hull 1.26 y
2006 Watford 1.21 n
2009 Burnley 1.08 n
2010 Blackpool 0.55 n

Fulham will struggle. Fulham were actually underrated last year, mainly because they had 16 draws, the most in the league. Their goal differential (+6) was the same as Everton’s, and close to Tottenham’s (+9). However, there are three sobering facts about Fulham:

1) They had the oldest average age of their starting eleven of any EPL team.
2) They had only 14 outfield players start 10 or more games.
3) They are playing in the Europa League this year.

This seems to me to be a perfect storm of having more games to play with a small and old squad.  It is bound to take a toll on their performance. People have questioned Mark Hughes for leaving, but I think he was quite smart to leave on a high. I don’t see them finishing in the top half again.

Liverpool will make the top four. Yeah, I know, I said this last year too. But unlike last year, when I was stepping out on a bit of a limb, this year everyone and their mother have Liverpool challenging for the title. That I don’t see. During the second half of last season, after Dalglish replaced Hodgson, Liverpool were undeniably a better team. In fact, they had the third-highest goal differential over that period. The problem is that only some of this resulted from Dalglish and his signings and some of it was luck. For example, the number of shots on goal per game allowed by Liverpool was exactly the same in each half of the season, but the percentage that were goals was higher in the first half. It’s hard not to see that as luck. Liverpool did improve offensively over the course of the season (they averaged about a shot per game more under Dalglish), but I don’t see any evidence that they are close to United at this point. I think they will finish third.

United will win the league. I know, for a City blog, I’ve been fawning on the scum far too much. That doesn’t change the fact that they still have the best squad. The null hypothesis is that the previous winners are the best team this season. I don’t see enough evidence to reject that hypothesis.

One of Wolves or Wigan will go down. They finished 16th and 17th last year, they were much closer to the relegated teams in the table than to the teams above them, and neither have improved their squads. If I were a betting man, and I am, my money would be on Wigan to go down. Not only did Wigan lose some of their best players, but they were incredibly lucky last year. Last year, the average shot distance of Wigan’s goals was far longer than anyone else. Scoring on long-range strikes, while impressive, is not exactly a recipe for sustained success because it is much harder to repeat. This is reflected in their abysmally low finish rate (goals divided by shots on goal), and without N’Zogbia and his penchant for scoring ridiculous long range goals, I see them struggling. Of course, I predicted that both of them would go down last year.

Chelsea will drop out of the top 4. I feel that the fourth spot will come down to Chelsea and Arsenal, and Chelsea will ultimately lose out. This is a tough one though, because we don’t have all the information about transfers. I’d feel more optimistic about Chelsea’s top four chances if they grabbed Modric, or if Arsenal sold Fabregas and Nasri. However, at the moment, Chelsea look like the team to make way. Arsenal didn’t perform as well as Chelsea last year, but they had a ton of injuries to key players. I think Arsenal’s first-choice starting eleven was stronger last year and I feel Chelsea may decline because they, like their neighbors Fulham, are a very old team. I also think Villas-Boas might have a hard time dealing with such an old squad, particularly since that was one of the strengths of Ancelotti.