MLS Possession with Purpose: The best and worst of Week 5

And so it goes; another week completed where you got the heart-pounding excitement from the Cascadia Cup clash, a defensive struggle from two of the best teams in MLS, and lopsided victories for two others. Before digging in, a couple of links to consider: if you missed the match between Portland and Seattle here are two articles you may want to read if the opportunity presents itself. This one was offered up by MLS, and then here's mine offered up on my home site here in Portland with the Columbian Newspaper.

Also, if PWP is new for you there may be value in reading what that's about through this link, an introduction to PWP and some explanations in case this approach is new to you.

With that out of the way it's time for some grist... who was the most effective and efficient team in my PWP Composite Index this week, and who was the worst?

To set the table here's my standard diagram for Week 5 only. Later this week I will publish the Cumulative Index - when I do I'll pop that link here.

Observations:

There were two games this past weekend with lopsided scores (3-nil LA over Chivas) and (4-1 FC Dallas over Houston).

Up until the completion of the LA Galaxy 3-nil thrashing of Chivas USA, it looked like FC Dallas would be top of the heap for Week 5 - and rightly so given they put 3 past Houston plus they got the Orange team to give them an own goal as well.

But as you can see, LA were tops this week; more later on why, and it may surprise you.

The tough part about the FC Dallas game, for Houston, was the straight red and sending off of David Horst on what the Referee considered was a rash foul down Houston's right sideline. I'm not so sure about that but as we have seen so far this year, the Referee's are stamping their authority with no reservation whatsoever.

Spilt milk and, as it goes. Dallas scored three goals within 15 minutes of that Red Card to take three points. If you run a team defense in Fantasy football and have Houston in that role you got hosed - I do and I got hosed in the back-four; still got 52 points though!

But back to the LA Galaxy match on Sunday

If anyone wasn't sure about how Landon Donovan could operate in a Diamond 4-4-2, be advised that he can - he has - and he will; when surrounded by other strong players, he's tough to stop.

For me, though, this game wasn't about just scoring goals. It was also about defense, and it's that defensive mindset that put LA at the overall top this week - clean sheets matter!

As for the bottom side of the Index...

If it's LA shutting down Chivas that garners the top offensive spot, then it's reasonable the flip side is the complete lack of structure and focus from Chivas that sees them at the bottom.

Carlos Bocanegra has great mental awareness, but he can't stop an aggressive Galaxy attack on his own - and in the short glimpses I had of their video it certainly looked to me like Baptiste and others were simply outmatched.

As for Houston - enuf said - the Red Card to Horst directly influenced the outcome of that game.

As for the middle of the middle of the pack... if you read my recommendations above about the Timbers-Sounders match, you know that game was all about possession with the intent to penetrate. There was absolutely no possession, that I saw, where the intent was to possess just for the sake of controlling possession.

There is no love lost between those two teams, and it seems every time they meet both just simply want to smash each other senseless. It makes for great entertainment, but there are times in my book where negative football has value, and securing three points (like it or not) is a time where negative football has value.

So on to the PWP Attacking Index; here's how they lined up head to head...

Observations:

For the first time this year the magical 3.00 barrier was broken in the attacking index. It comes on the heels of the USMNT also breaking the 3.00 barrier in the first half against Mexico. If you missed my thoughts on that game, you can review them here.

Onwards and upwards - for the first time this year Portland broke the top-five barrier in team attack for a given week, getting two stunners from Diego Chara plus another couple from the Argentina contingent of Diego Valeri and Maxi Urruti.

Not to be outdone was the final 20-minute performance of Clint Dempsey - aye - he got a goal early on, but for much of the game his influence and presence was pretty much unnoticed. Indeed, the chalkboard tells us that between about the 40-minute mark and the 70-minute mark, he had just 9 touches of any sort with no shots or key passes. It wasn't until the 70-minute mark where he started to directly influence and impact the game. After that point, Dempsey had no fewer than 18 touches in the run of play with one key pass, three shots on target, and a goal.

I'm all for highlighting his hat trick in that game, but he simply wasn't solid through the full 90+ minutes, and his team barely eked out a draw.

Understanding that and seeing the red-card tainted blowout of Houston by FC Dallas my PWP Attacking Player of Week 5 was...

Observations:

That may be a surprise to some on two fronts. One - Dempsey got a Hat Trick. And Two - what about Watson or Diaz?

Well, as already noted, Dempsey simply didn't play well for a full 90+ minutes and his presence and influence did not prevent Portland from having their best attacking performance of the year.

In addition, it's likely Seattle drops three points if Ben Zemanski doesn't do what he did in the box, and I'm simply not in favor of seeing someone getting an Attacking Player of the Week award when his team loses or draws. 3 points is the objective in this game - it's not all about just scoring goals.

And two - while Diaz is the spark that lights the Dallas attack this year and Watson donated a brace of goals this weekend, I feel and think Michel had more overall responsibility on the pitch; therefore - given his vast number of touches on both sides of the ball I give him the award.

And in case you missed it, I don't view this weekly award as going to someone who just lights up the front end without also considering how well they supported the back end.

Time now for Defending PWP Team of the week, where the LA Galaxy really made it count

Observations:

It was a close call between LA and Sporting KC this week when it came down to it; any team who can get a clean sheet against Real Salt Lake has really done their job.

But... alas... the Top PWP Defending team was LA. Why? Well it really came down to how poorly Chivas USA performed against LA, and not how poorly Real Salt Lake performed against Sporting; remember - this Weekly Index does not get influenced by previous performances on a week to week basis only the Cumulative Index does.

When checking out my Cumulative Index later this week, you may see a change in who the top defending team is overall - for now though - this is just Week 5.

In looking at the player statistics I had considered awarding the PWP Defender of Week 5 to Landon Donovan, and here's why: he had three key passes, five recoveries and an 85% passing accuracy with 2 assists. But the more compelling case fell to Juninho, given his combined efforts (like Michel) playing on both sides of the ball.

Here's the Diagram offering up his team effort on Sunday...

In closing

Week 5 saw some individual players step up and some team performances improve as compared to previous weeks. It's a long season, and it's likely the Cumulative Index will continue to take shape - especially after the (unexpected) Toronto victory over Columbus in Ohio.

You may have thought that game flew under the radar, but it hasn't, and Toronto will look the better for it in my Cumulative Index... all is not lost when a team gets a big victory without getting the headlines for that week.

If curious - here's a link to my Weekly PWP analysis on the Red Bulls of New York.

All for now,

Chris

USMNT - My thoughts after 2-2 Draw with Mexico

If you're like me you were pretty impressed with the first half Wednesday evening as Jurgen Klinsmann deployed a Diamond 4-4-2 in the truest sense - narrow and focused down the middle with the intent to manage the wings by channeling things to the middle. It worked really well in the first half. To give you a comparison on how well it went, here's a table on their Possession with Purpose (six steps in Attack) in the first half compared to that of the second half with the average for MLS Teams in 2013.

But before offering the here's a link to what PWP is all about in case you've missed it before.

Team	Possession Percentage	Passing Accuracy Percentage	Penetration Percentage	Creation of Shots Taken based upon Penetration Percentage	Shots on Goal compared to Shots Taken Percentage	Goals Scored compared to Shots on Goal Percentage
USMNT 1st Half	59%	85%	13%	14%	80%	50%
USMNT 2nd Half	41%	80%	18%	25%	14%	0%
Mexico 1st Half	41%	75%	21%	5%	0%	0%
Mexico 2nd Half	59%	80%	23%	35%	44%	25%
MLS 2013 Average for Comparison	50%	76%	22%	20%	34%	30%

Observations:

I won't offer up anything new here that I didn't already offer on twitter during the match but in case you missed some of those streaming thoughts here they are without limiting my words to the format of twitter.

Bradley and Beckerman needed to be the fulcrum between the defending side of the pitch and the attacking side of the pitch if that Diamond 4-4-2 is to be successful - I'd offer that most would agree they were (at least in the first half).

In considering I had never seen Michael Parkhurst in a left fullback position I opined that the way this team lined up some good chances would come down the right side with Beltran running overlaps or supporting Zusi in deep penetration on the wings.

I'd offer that was also the case in the first half - nothing better as an example than the goal Wondolowski had working from the Zusi cross that Bradley flicked on for Wondolowski to poke home.

What was surprising to me (a very welcome surprise) was how effective Michael Parkhurst was in the first half working the left side with his own mix of penetration combined with Davis --- I really did enjoy seeing that Wednesday evening and support like that from Michael reinforces his ball handlling skills - and - In my view makes him a very credible selection to start at Centerback along with Matt Besler.

If you didn't already know Michael Parkhurst was my PWP Defender of the Week #1 and here's that article supporting that analysis.

I'm not sure why I've never rated Omar Gonzalez highly but I don't - maybe it's his defensive positioning that makes me nervous but I'm a defensive minded guy in football and while there are good points in having a CB who can attack the box on set-pieces my view is that they are first and foremost on the pitch to (STOP) the opponent from scoring - all else is a bonus after that.

As for the goals against in the second half - other pundits have already offered up the Capt. Obvious here that Gonzalez was directly accountable for both goals scored by Mexico - so I ask (rhetorically) did he really add value to this squad in that game in his primary role and if not - who's better?

That's not a question for me to answer but I think it is a question Jurgen Klinsmann needs to ask himself and his new staff...

Like many things in life, I'm not particularly fond of folks who offer up a problem (be it real or perceived) without also coming up with a solution/recommendation to that problem.  So with that here are my options knowing that some players are simply not going to get selected that haven't already played under Jurgen recently.

Goodson - Not sure here either - I personally have not seen him enough to offer a view that (in my view) has merit - he does well for San Jose but he didn't get particularly good minutes overseas with what I feel and think is a top rated club.  More information needed.

Parkhurst - I have seen him probably as little as I have seen Goodsen but in those few short games (and his impressive showing to me on Wednesday evening) it is clear he has the pace to cope with the wings and also has the passing accuracy and understanding of a broader role in positional play to make an very effective CB (starting CB) provided he can handle the more physical side of the game when teams include a more traditional #9 who plays more with his back to goal than trying to run on to through balls.

Cameron - His time overseas is seeing the game as a right back for Stoke - is that mix the right mix to settle in alongside Besler - and how is that chemistry going to take shape?  He has positional awareness of how positional play works down the wings so that adds great value - just like that in seeing Parkhurst play the left side Wednesday evening.

For me Parkhurst is a first option to pair with Besler but my view is limited - call it a gut instinct - but do folks really expect a CB who has played as long as Gonzalez to say in passing he needs to be more dominant in his role as a CB in protecting the box?  Wow - I hope not.  That is something a CB should KNOW and understand from day 1...  oh my...

Perhaps a more compelling question is how long has this weakness (lack of being switched-on to the true purpose of a CB) been or not been recognized by the USMNT staff?

And then to throw a teammate under the bus - bollocks - it just reinforces my own views that Gonzalez is not the right choice to represent the USMNT as a starting CB in the World Cup.

A winning World Cup team must be linked in and switched-on to roles and responsibilities for 90+ minutes for at least 3 games in 8 days??? in order to advance - and then it just gets tougher and tougher... that speaks to having resilience in a squad and throwing a teammate under the bus is not an example of resilience - it represents a shirking of responsibility.

As for Green - as noted in my finishing tweets for the match - in my view Green is still green - that was worthy and notable of Klinsmann to put him in as a way to begin his trail of caps - but as an option going forward now?  Unless his attitude is so positive and infectious for others I just don't see him having any role of substance this World Cup - the hype is what it is - hype...

Tough question here for those who've been around footy for some time.  If a real stud, do you really think Green would miss an opportunity to play for Germany in a World Cup or European Championship in the very near term?

The pedigree with the German side is simply to strong to even think for a second that (if a starter there) he would ditch that opportunity to be a starter here).  Like it or not the USMNT's progress has not made it that far in being that good...  if you think they have your emotions are overwhelming your senses.  Bringing Green into the side is more about 2018 than 2014; and for that I tip my hat to Klinsmann...

In closing...

A welcome sight to see the USMNT open in a diamond 4-4-2 and the pieces to that puzzle looked pretty good when considering who started and who didn't.  I would offer that style of play speaks to some of the stronger styles we see in MLS - is that the intent of Klinsmann - to stamp a particular style of play that suits the stronger and more possession oriented sides in the MLS who are also known for closing down and giving the opponent very little space and time to work with?

I think so - and yes - width is critical to manage when working a narrow approach and the right pieces need to be there to do that.  Evidence of that was very clear Wednesday evening - as the second half opened and the subs began to rotate for the USMNT the Mexican side went from offering up just 8 crosses in the first half to a total of 23 crosses for the second half.

Clearly the change in players on both sides, to include complacency and fatigue of the USMNT, directly impacted and influenced the team attacking style of Mexico.

In looking towards the final selection Klinsmann has some issues to wrestle with - how does he balance the chemistry of the "MLS Players" playing together versus those guys who play abroad - how does Altidore fit into a Diamond 4-4-2? 

He's been laboring with Sunderland this year and I'm not familiar enough with that team to know what system they operate - but given their position in the League Table it would appear they are not very good in scoring goals - which for me tends to indicate their midfield isn't that strong.  To paraphrase Harrison on this one - Jozy doesn't have the right complimentary pieces to go with his skill set...

And with a trend of American players returning stateside might we see Jozy make a transfer move similar to Bradley and Dempsey this summer?  Hard to say now but if the USMNT chemistry continues to mature, using a majority of players from MLS, it just might mean the most effective move for him is a return to America.

Before signing off I have one final postulate for consideration.  In seeing how the game went Wednesday evening - has anyone considered that - given it was a friendly - the intent of the second half might have also included studying how Mexico may adjust, in pitch activity, to the Diamond 4-4-2, in order for Klinsmann to gather data on how other opponents might adjust (real time) in the World Cup?

This also provides Klinsmann some real data to evaluate on ways he might counter the opponents counter...  I wouldn't put it past him - especially since the game wasn't a real win or lose game of consequence...

That's all for me for now... More to follow on twitter as I join the crowd at Providence Park for the Cascadia clash between Seattle and Portland.

Montreal and Philadelphia Swap Young Strikers

Okay, I'm sure by now that, given you follow our site, you've also probably been made aware of the fact that the Philadelphia Union (an underrated team in my opinion) traded their young 20-year old striker Jack McInerney to the Montreal Impact for their young 22-year old striker Andrew Wenger. The trade has a very Matt Garza for Delmon Young feel to it, leaving me with an odd taste in my mouth. Are the Montreal Impact selling low on Andrew Wenger? It's, at the very least, presumable that they know something that we don't about him and his nature. The question becomes, then, is that assessment accurate? Obviously the idea of a poacher is one that is met with a bit of contention,  in the sense of how do you measure being in the "right place at the right time" for an individual? However assessing the 86 shots taken by 'JackMac' from the 2013 season, we can know that no fewer than 57 of them came from inside the 18 yard box, courtesy of digging around on the MLS Chalkboards. It's obvious that he's a player that can get the ball in advantageous locations. Already on the season he's put together 12 shots and 11 of them have come inside the 18-yard box with 6 coming directly in front of goal. He's been appropriately tagged on twitter as a "fox in the box"---hold the sexual innuendos---and I think the term poacher probably comes naturally with that association. Unfortunately, that term may harbor and imply the idea that he's more lucky than good. I'm not sure I entirely buy that approach.

Meanwhile with everyone's attention directly focused on McInerney--audaciously stamped as 'The American Chicharito'--having already being called in the USMNT Camp for training during the Gold Cup, people are forgetting about Wenger and his potential that once made him a #1 overall MLS draft pick. Back in 2012, Wenger was painted as a potent and rising talent in MLS, named to MLSSoccer.com's 24 under 24 roster, coming in 7th overall. Just one year later McInerney jumped onto the list himself, rocketing to 4th overall, while Wenger was left off. The perpetual "what have you done for me lately?" seemed to come out in these rankings.

Wenger--despite all his talent--has run into a slew of various injury-related setbacks the last two seasons; it's so much failing to perform. The talent is still there, and I fully expect John Hackworth to tinker in an effort to get as much out of him as possible. The easy narrative here might just be the returning home to "revitalize his career" or something like that. Instead I think Philadelphia possibly got an undervalued piece in this move.

Looking at the last two years and a total of 31 shots Wenger has taken, 24 of those came from inside the 18-yard box, a higher percentage than that of JacMac. With that you can see above with xGpSH (expected goals per shot) that Wenger's average shot has been more likely to become a goal than that of his counterpart. Now, understand that this all comes with the requisite small sample sizes admission. Wenger has played less than half the amount of time as McInerney and has less than half the amount of shots. However, estimations based upon their current performances with creating shots has them near the same level as that of Eddie Johnson, Will Bruin and Chris Rolfe in years past.

Creating shots isn't everything. Creating shots in important positions is something. As we attempt to analyze the value of certain events on the pitch and how certain players are responsible for those events, we'll see some things and maybe understand how to assess performances. It's easy to overact to certain things that come with doing this type of analysis--- Such as McInerney, Wenger, Bruin and Rolfe all averaging about 4.0 shots created per game individually. That seems rather important, but there is additional data that is missing. How much was each shot that they created worth? What other attributes do they bring to the match? This is just an simple break down between two players and comparing how they've impacted their respective clubs.

Personally, looking at all of this data, I'm of the mindset that Montreal got the better player. However, it's extremely close and that isn't taking into account the rosters in which they are joining or how they might be utilized on the pitch with their new teams (4-3-3 concerns vs. 4-4-2 placement). I would say at this time the difference between the two is that one is younger and has more experience. That might be a bit simplistic approach but honestly both create shots the same way in the same space. McInerney does so at a higher rate but Wenger has made up for taking less shots with taking advantage of his more experienced partner, Marco Di Vaio, and feeding him opportunities.

This may be one of the more interesting trades in recent memory. I'm fascinated to watch what happens next and how each of these two players develop. Their career arcs will go a long way in providing the narrative for this trade and I'm not so certain that this is as one-sided as some people might think. Referencing baseball again, the Tampa Bay (then, Devil) Rays were largely regarded as having "sold low" on Delmon Young. We can now see, looking over the past decade,  that he never managed to put together all those tools that we once believed he had. The lesson being: don't be too quick to judge Philadelphia. This isn't necessarily going to be something as easily evaluated by just a single season, and time will reveal the significance of this day.

How It Happened: Week Four

Week four was a great week for MLS fans. It seemed like nearly every game ended with a stoppage time goal to rescue a draw or clinch a victory, and there were plenty of great goals and saves to go around. On a personal level, week four was less great: between a car breakdown, being super busy at my real job and the USA-Mexico friendly happening at 11 PM EST, I come to you for my weekly column a day late and a game short. Sincerest apologies to my loyal readers (both of you).  

Philadelphia Union 1 - 1 Montreal Impact

Stat that told the story for Montreal: Long passes and counter-attacks

So this isn't necessarily a stat, but watching Montreal is a clear lesson in direct attacking with long passes. The image above shows all the completed passes by Montreal in the middle third of the field, and you can tell that they tend to be pretty long. And these are only the completed passes - long passes have a higher tendency to be incomplete, so in fact Montreal attempted way more long passes than are in the picture. On the bright side, Montreal has some guys who are pretty darn good at those long passes (and another who's pretty darn good on the end of them: check out this beauty that Mapp hits to Marco Di Vaio for the Impact's lone goal. This isn't new for Montreal: it's exactly how they played last year when they rolled to a hot start and then struggled mightily down the stretch. The hot start hasn't exactly happened this year; will the rest of the season play out any better?

Stat that told the story for Philadelphia: 3 blocked shots by Amobi Okugo & Aaron Wheeler

What hurts for Philadelphia fans is that if the centerback pairing of Okugo and Wheeler had managed to block a fourth shot, the Union probably would've snatched three points instead of just one. From watching Philly a couple times now this year, it seems like their backline, particularly Okugo, blocks a ton of shots. MLS fantasy stats say Okugo is averaging 9.5 CBIs (Clearances, Blocks & Interceptions) per game, and he tallied 9 against Montreal. But only two were blocked shots, and if he had closed Marco Di Vaio down just one step earlier on that goal.... Still, MDV is a class player even at his advanced age, and Okugo has saved enough goals this year that one slip up is hardly enough to pile on the guy.

Seattle Sounders 1 - 2 Columbus Crew

Stat that told the story for Columbus: Location of Federico Higuain's touches

If you haven't watched any Columbus Crew games yet this season, you're missing out on the most entertaining team in the league. This tweet from Armchair Analyst Matt Doyle sums it up better than I ever could: the Crew is so fluid and so good at creating space in possession that they create a ton of chances. The straw that stirs the drink for all this is their talisman, Federico Higuain. In the past, Higuain has floated all over the attacking half of the field to get on the ball, but this year he's extended his meanderings to the entire field (see exhibit A: heat map above).

Another part of what's making Columbus so successful this year is how well the rest of the team plays off his movement. For example, when Higuain slides onto Bernardo Anor's flank, sometimes Anor or Dom Oduro or Tony Tchani makes a run off the movement to create an attacking chance. This kind of interplay is awesome to watch, and if Gregg Berhalter can keep his team's creative spark alive then Columbus could not only make the postseason, but make some noise once there.

Stat that told the story for Seattle: 20 recoveries + interceptions in attacking half

After writing an opus to Columbus' early play, it's time to bring them back down to Earth a little bit. If it weren't for the red card issued to Djimi Traore, the Crew's perfect start would've been seriously in jeopardy. Sigi Schmid came out with a good tactical plan to counteract Berhalter's attacking possession style, pressing high up the field and trying to win the ball off Columbus' defenders.

Before going a man down, the pressing led to a number of balls won in the attacking half which led directly to dangerous Seattle counterattacks. Some of this high pressing could also be attributed to the 4-3-3 that the Sounders employed in the absence of Clint Dempsey. I might be alone in this opinion, but I wouldn't mind seeing Dempsey in that formation alongside Obafemi Martins and one of Lamar Neagle/Kenny Cooper. I know Sounders fans want Dempsey to be the focal point of this team, but he might just be more effective as a complimentary piece in a balanced formation.

Agree with my assessments? Think I'm an idiot? I always appreciate feedback. @MLSAtheist

Predictive strength of Expected Goals 2.0

By Matthias Kullowatz (@MattyAnselmo)

It is my opinion that a statistic capable of predicting itself---and perhaps more importantly predicting future success---is a superior statistic to one that only correlates to "simultaneous success." For example, a team's actual goal differential correlates strongly to its current position in the table, but does not predict the team's future goal differential or future points earned nearly as well. I created the expected goals metrics to be predictive at the team level, so without further ado, let's see how the 2.0 version did in 2013.

Mid-season Split

In predicting future goals scored and allowed, the baseline is to use past goals scored and allowed. In this case, expected goals beats actual goals in its predictive ability by quite a bit.*

Predictor	Response	R2	P.value
xGD (by gamestate)	GD (last 17)	0.805	0.000
xGD(first 17)	GD (last 17)	0.800	0.000
xGA (first 17)	GA (last 17)	0.604	0.000
GD (first 17)	GD (last 17)	0.487	0.000
xGF (first 17)	GF (last 17)	0.409	0.004
GA (first 17)	GA (last 17)	0.239	0.024
GF (first 17)	GF (last 17)	0.155	0.099

Whether you're interested in offense, defense, or differential, Expected Goals 2.0 outperformed actual goals in its ability to predict the future (the future in terms of goal scoring, that is). That 0.800 R-squared figure for xGD 2.0 even beats xGD 1.0, calculated at 0.624 by one Steve Fenn. One interesting note is that by segregating expected goals into even gamestates and non-even gamestates, very little predictive ability was gained (R-squared = 0.805).

Early-season Split

Most  of those statistics above showed some predictive ability in 17 games, but what about in fewer games? How early do these goal scoring statistics become stable predictors of future goal scoring? I reduced the games played for my predictor variables down to four games---the point of season we are currently at for most teams---and here are those results.

Predictor	Response	R2	P.value
xGD (by gamestate)	GD (last 30)	0.247	0.104**
xGA (first 4)	GA (last 30)	0.236	0.033
xGD(first 4)	GD (last 30)	0.227	0.028
xGF (first 4)	GF (last 30)	0.140	0.093
GF (first 4)	GF (last 30)	0.022	0.538
GD (first 4)	GD (last 30)	0.015	0.616
GA (first 4)	GA (last 30)	0.003	0.835

Some information early on is just noise, but we see statistically significant correlations from expected goals on defense (xGA) and in differential (xGD) after only four games! Again, we don't see much improvement, if any at all, in separating out xGD for even and non-even gamestates. If we were to look at points in the tables as a response variable, or perhaps include information on minutes spent in each gamestate, we might see something different there, but that's for another week!

Check out the updated 2014 Expected Goals 2.0 tables, which now just might be meaningful in predicting team success for the rest of the season.

*A "home-games-played" variable was used as a control variable to account for those teams who's early schedule are weighted toward one extreme. R-squared values and p-values were derived from a sequential sum of squares, thus reducing the effects of home games played on the p-value. 

**Though the R-squared value was higher, splitting up xGD into even and non-even game states seemed to muddle the p-values. The regression was unsure as to where to apportion credit for the explanation, essentially. 

D.C. United: Shooters, Providers and What?

As you might have seen from our twitter stream, I kind of wrote an article on DC United last night. Then I scrapped it. Then, Alex Olshansky dropped this brilliant mess concerning Michael Bradley, and I was like "that's basically what I was doing... on a team level!" So it kind of nudged me to at least put forth an effort to finish it...only not really. What I did was basically compiled stats for four "core" attacking players on three different clubs. Two of those clubs (Sporting KC and Houston Dynamo) have shown consistent success the last two years, while D.C. United...well, you know, they have kind of stunk the place up.

The rest I submit to you without further inane commentary.

D.C. United

SH=shots, KP=Key Passes SH/KP = Shots/key passes ratio ShCPG =Shots created per 90 minutes played %ofTeam= the total percentage of the teams shots that the individual created

Houston Dynamo

SH=shots, KP=Key Passes SH/KP = Shots/key passes ratio ShCPG =Shots created per 90 minutes played %ofTeam= the total percentage of the teams shots that the individual created

Sporting Kansas City

SH=shots, KP=Key Passes SH/KP = Shots/key passes ratio ShCPG =Shots created per 90 minutes played %ofTeam= the total percentage of the teams shots that the individual created

Location Adjusted Total Shots Ratio

Millionaire Malcolm Forbes was famous for his quote, "He who dies with the most toys wins." And while that might not be the most moral mantra for life, sports fans have a hard time arguing with the logic. After all, a game is about runs, points or goals, and after enough of those it's about shiny trophy cases. But in the world of sports analysis there is no such victory in the absolute. Analysts need to explain how those runs, points or goals came about. In the world of soccer especially, there is never a complete answer. Goals are exceedingly rare, so explaining how they grace us with their presence mathematically is difficult, to say the least. We're happy with higher R-squareds and other such geeky descriptive metrics. Have you ever seen a trophy case filled with strong correlations? Nope, all we get is a little blog post, and if we're lucky, some twitter praise. Still, we search.... One of the more popular explanations for winning in soccer is Total Shots Ratio, which calculates the percentage of shots taken by a team in games played by that team. A 60% TSR means that a given team took 60% of the total shots fired in the games they played. The logic isn't all that difficult to wrap your head around. If you can take more shots than your opponent you are likely to score more goals. For the English Premier League, TSR explains 68% of the variance in the point table, which is impressive for one statistic. TSR happens to be less important in MLS.

The variance prediction is just 37% and this is likely due to the lower finishing rates in MLS compared to the EPL, rendering shots less effective. But there are probably a number of other reasons why TSR is less predictive of points in MLS. There are a larger percentage of teams employing counterattack strategies which have significant impacts on finishing rates, which would in turn alter the effectiveness of TSR. But what if the shots were weighted to account for the location of the shots? It would be logical to assume that better teams take better shots and make it more difficult on the opposing shooters. But does that logic actually manifest itself when predicting points? ASA's Expected Goals 1.0 worked pretty well, so a TSR adjusted for shot locations ought to work better than the original TSR.

The first thing required would be a fair weighting of shots by location. To do that I took the ratio of the finishing rate by location and divided by the average finishing rate. Here is the resulting table for adjusting the value of shots.

Location	Weighting
1	3.14
2	1.79
3	0.72
4	0.54
5	0.24

For the sake of simplicity I have collapsed zones 5 & 6 into a fifth zone. This table illustrates that a shot from zone 1--inside the 6-yard box--is actually worth 3.14 average shots. And a shot from zone 5 is worth just .24 average shots. Adjusting all of the shots in MLS in 2013 yields the following result when attempting to predict table points.

You can tell from just eyeballing the dispersion of the data points that the location adjusted TSR better aligns with points and the Rsquared agrees. There is a 17-percent increase in R-squared. Not just pure volume of shots, but the location of those shots is vital to predicting points in MLS. It would be interesting to see if location is equally important in the EPL, where TSR is already such a strong predictor.

For the curious, the New York Red Bulls were the team that was best at getting better shots than their opponent. Their TSR improved from 47% to 52% when adjusting for shot location. Real Salt Lake actually took the biggest hit. Their TSR was 53% and their location-adjusted TSR dropped to 48%.

It's only one season's worth of data, but with such an impressive increase in the ability to explain the variance in point totals, it confirms that location does matter, and that teams are rewarded by taking better shots themselves while pushing their opponents -out farther from goal. And perhaps soccer analysts have another statistical toy to add to the toy box---Location-Adjusted Total Shot Ratio.

MLS PWP: Team Performance Index through Week 4

With just four weeks of play in the books, it's unlikely that the bottom feeders by the end of this year are those at the bottom right now. But this analysis should be an indicator on who needs to get better in attacking (and in defending) when it comes to Possession, Passing Accuracy, Penetration, Creation of Goal Scoring Opportunities and converting those opportunities into goals. As a reminder here is the initial PWP article offering up an introduction and explanation to the PWP Strategic Index, if this is a new approach for you to consider.

For the future - please get used to the abbreviation and hashtag #PWP-TPI. I will refer to my cumulative analyses on MLS team performance this way.

And in case you missed it here is a link to the Week 4 PWP Analysis.

Here's how they stand after Week 4:

Observations: (click to enlarge)

For me it's no surprise that Columbus remain atop considering they have 9 points after three games and have done pretty well in all phases of PWP. They rank 3rd in Possession percentage at ~57%, 1st in Passing Accuracy at ~81%, nearly half of all their shots taken have been on goal (41.48%), and they've converted 62.86% of those targeted shots into goals. Only FC Dallas has a better rate in converting targeted shots on goal to goals scored (71.25%)... Conversely, DC United have only converted 7.41% of their shots on goal into goals scored.

What is somewhat surprising is how far down Toronto have dropped after their 3-nil loss to Real Salt Lake. As noted in my Week 4 Analysis Toronto did well in getting results their first two games, but when going up against a traditional powerhouse in MLS their tactics and strategies were simply dominated.

Again, to drive home some points about PWP - Passing Accuracy and Possession Percentage will influence the bottom line, and in the first three games Toronto have simply been very poor in passing accuracy (67%, 59%, and 54%), and their highest amount of possession percentage came against RSL at 46%, while they had 38% against DC United and 32% against Seattle.

We already know that DC United has not started with a bang--and given Seattle's loss to Columbus--it isn't quite so surprising after all.  (Perhaps?) the league table position for Toronto is more a reflection of luck and good fortune than a comprehensive approach to attacking and defending with purpose?

I'm not alone in ranking Toronto a bit further down the scale...  Whoscored.com currently ranks Toronto as 11th. I also checked Squawka.com, but their data currently only go out to week 2 for team comparisons.

So enough about Toronto. Dallas, Salt Lake, Houston and Vancouver round out the top 5. In considering those other four, the new kid on the block for me here is Vancouver. Why?

Well last year they had issues in defending while clearly having a very strong attack. Early indications are that the defending side of the equation has been fixed... does that hold true as the grind begins in April and the heat of summer sets in?

As for Dallas, Salt Lake and Houston - Pareja is no stranger to fielding a top team in MLS when it comes to Possession with Purpose (Colorado was in the top ten last year), and Houston has done what it needed to do (at least so far) to get tighter in defense and offer up better balls for Barnes and Bruin to score.

The boring team in the top 5 goes to Real Salt Lake. Simply said, they just keep doing what they need to do - polish their Diamond 4-4-2 and let it shine, regardless of who plays up top - be it Garcia, Plata or Saborio. Then there's the ever present and dominating defensive central midfielder, Kyle Beckerman, who controls the back.

Speaking of the Diamond 4-4-2 - as the year continues, I'll be able to offer up additional analysis on what teams run what basic formations with the intent to really peel them back to see if specific team performance indicators increase or decrease based upon that simple filter.

In considering the bottom feeders outside of Toronto so far...

DC United - Aye - great weekend in that game against Chicago, and there remains no question DC United like a possession-based attack. The difference this week was a highly engaged Johnson and Espindola in attack that also included a steady stream of pretty good crossing and wing penetration.

How well that holds up is hard to tell. Head coach Frank Yallop has been known to cede possession, given his days in San Jose, but the direct attack for the Fire is more ground-based this year given the types of strikers. Has everybody by now realized that Yallop was sacked because he wanted to change the attacking approach in San Jose, and Watson--along with the front office--didn't?

To be sure - look no further than what team sits just above DC United - it's San Jose. For now, that's not an indicator that the Earthquakes are a bad team... no... I'd offer that it's more of an indicator that their approach in attack needs refinement as does their back four being a bit unlucky with that own goal against New England.

Not last and not least are Portland and Montreal... For me, seeing Montreal on the low end is not surprising - head coach Frank Klopas ran a pretty weak defensive team in Chicago last year, and it seems to have translated over to Montreal. It's true that Di Vaio was missing in games 1 and 2, so this will be a team to watch as well.

As for Portland,  here's a possession-based team that simply hasn't clicked yet. And in all that there remain holes in the defending side. One might say that there are distinct instances of distinction where they have instinctively distinguished themselves as lacking instinct in where to defend. Or - in other words - the back four have been better at ball watching than defending the ball...

Mid-table: Sporting, Colorado, LA, New York, Philadelphia and New England

It's early days and given some top activities in defending and attacking for Sporting it is likely their run of games inside and outside of the MLS Regular season will now give them a breather to prepare for Month 2; the same can probably be said for LA as well as the Earthquakes (who a lower at this time).

As for New York - I'll touch more on them in another article but for now I'm still not sure Petke has found the right mix. Philadelphia, from what I have seen, has some dangerous players in the Midfield (Edu and Maindana come to mind first) - so how they fare in April will be interesting to see in contrast with New England.

With all those observations I'll simply offer up these two diagrams to give you an idea on where each team stands in the PWP Strategic Attacking Index and Strategic Defending Index.

As a reminder here are the individual players who have been highlighted as the PWP Attacking Players for March - they might be reasonable targets for your MLS Fantasy team if funds are available...

1. Federico Higuain 2. Jaoa Plata 3. Bernardo Anor 4. Graham Zusi

On to the Strategic Defending PWP and the PWP Defenders of March.

Top Defenders for the Month of March:

1. Michael Parkhurst 2. Corey Ashe 3. Ike Opara 4. Kyle Beckerman

In closing...

Remember that the season is still young - but in about 4 weeks time it won't be as young, and in 16 weeks time this Index should begin to settle in and hopefully, like last year, paint an early picture on who's up for the Playoffs and who might be making some summer transfers to bolster chances for a late season push.

Teams to watch in this Cumulative Index are numerous.

In the bottom end let's see about Toronto, Portland, San Jose, and DC United - will the rest of MLs figure out the way to beat the Toronto approach of counter attacking? Will the return of Donovan Ricketts after the Seattle match spell a recovery in the back four for Portland? Can DC United really make better use of the accuracy and possession based approach? And finally, can Watson continue to make use of his aerial attack in getting penetration through the air as other teams seem to build greater strength in the midfield?

On the top side can Columbus continue their early run and does the sleeping giant residing in Houston awaken even more to clinch dominance across the East? Can FC Dallas hold it together this year under the new guidance of Pareja?

How about the other money bag teams like New York, Seattle and LA Galaxy - have other teams in the mix figured a way to bypass the top-flight DP approach used by those guys or are they real challengers for the Cup and the Shield?

Last but not least - the Champions from last year - Sporting continues to show well in defending, and Zusi has lifted his game on both sides of the pitch. Will the World Cup really be his time to shine? For the sake of the USMNT I hope so.

As always my thanks to OPTA and the MLS for continuing to provide free information in order to conduct this analysis.

If interested here's a link to my latest article on New Sports Hub about the Red Bulls of New York, including their PWP Attacking and Defending Players of the Week 4.

You can follow me on twitter here: Each week I look to offer up twitter comments for the MLS nationally-televised games as well as those for the Timbers.

All the best, Chris

MLS Possession with Purpose Week 4: The best (and worst) performances

The beauty and maybe the curse of Possession with Purpose (the weekly update) is that it completely ignores past performance and only gives you a view on how things went this weekend.

For me, I like that, as it helps point out how well a team performs against their counterparts as a snapshot in time. The value in seeing that weekly effort then has more grist when viewing those positive and negatives in a cumulative effort week to week.

So without further ado here's how the teams stood toe to toe in Week 4:

Observations: Considering the complete blowout that Real Salt Lake had against a supposed "elite" Toronto team, it's no wonder those guys headed the composite Index this week.

Like many things, there is a tendency to over-embellish on occasion when things look really good for a team early in the season, and those reminders on how far FC Dallas fell last year - from first to eighth in about 3 months - you just can never tell how good a team is until they go up against the likes of a team like Real Salt Lake.

Last year we saw Salt Lake lose out on both the Open Cup and the MLS Championship Cup - they have been consistent in their consistency for a consistently long time; Toronto may have done well in the first couple of games but as noted in an editorial of mine the other day, the word 'elite' really doesn't belong to a team like this just yet.

In taking a further look at this Index, note again how far down the rung Portland is; it indeed was a close match and there were many chances for the Timbers, but an Index like this points out that the weaknesses in defending far outweigh their strengths in attack. If the Timbers are to turn this around (quickly) then a run of clean sheets is in order. No better way to test that than in their next game when they host the Sounders at Providence Park. If their blood doesn't boil up for this game, it may be a very long season indeed for Portland.

So after seeing the Composite winner, who actually took top honors in the overall team attacking department? You may be surprised that it wasn't Real Salt Lake---lest it's forgotten, the Index above points out the overall difference between Attacking PWP and Defending PWP.

Here's the PWP Attacking Index:

Observations:

First out of the blocks this week is Sporting Kansas City - and yes; they did pull a Lazurus this week by scoring the late winner in stoppage time. When they scored their goals doesn't matter to this Index, as it is context neutral in that respect. The fact that they came from behind to defeat a worthy Rapids team speaks more to the overall outcome and all the compelling work that went with it.

And yes - their defense gave up two goals. So, no - you won't find them in the top ten in the PWP Defending Index.

Other teams doing well in attack this week were FC Dallas, Columbus (also with a late winner) and Chivas USA (who lost 2 points by giving up a late equalizer to New York).

By the way - that's a pretty good run so far for Dallas and Columbus; Higuain should probably be on your fantasy team, as should Diaz; as a Timbers supporter it was hard to acknowledge that it was a blindingly good 'turn and strike' Diaz had against Portland yesterday.  It's like Diaz turned on a dime and struck it home - unfortunately it appeared that dime was in Pa Modou Kah's pocket.

As for Real Salt Lake - no surprise they are in the top three given their 3-nil win against Toronto.

Perhaps another surprise for some is DC United being in the top 5 - if you watched the game, like me, there were moments where DC really looked good in attack. A 2-2 draw was worthy and should give Olsen some breathing room for a wee bit longer; note - having a passing accuracy of 93.50% in that game 'will' influence the Index just a wee bit.

On to the internal process data for Sporting and the PWP Attacking Player of the Week:

Observations:

Critical data in the PWP analysis hinges on (but doesn't solely rely upon) Possession percentage, passing accuracy and the ability to leverage those two primary statistics in 'creating and scoring goals'; Sporting were 4th best in accuracy, 2nd best in possession, and 2nd best in scoring goals based upon shots on goal.

Those outputs had strong influence in them being top of the Index this week.

The individual team leader in Possession percentage this week was Columbus (shaded somewhat given the red card to Traore) - Passing accuracy was tops with DC United at a whopping 93.50% while Chivas USA made the best of their chances by scoring one goal from one shot on goal while the other five were blocked...

Graham Zusi is my top PWP Attacking Player of the Week #4:

Observations:

If you had Graham Zusi in your Fantasy team you should have gotten some pretty good points from him - I don't, so I have no clue how well he really did.

It's not hard to see how outputs like that correlated to a big three points in Colorado.

And even more compelling was his motor - he had two interceptions, one clearance and two recoveries in the defending half - some good evidence that he played on both sides of the pitch today.  Saborio will probably get all the notoriety this week - so be it. Sometimes things work out that way...

On to the defending part of this game. Recall RSL were top in the overall Composite Index - here's why:

Observations:

A clean sheet is a clean sheet is a clean sheet - and when you can go a whole game without conceding one foul in your own defending third, it's no wonder Toronto were held scoreless. Even more so when considering how many PK's have already been awarded this year (16) {is that a record already?}.

Bottom line here is the bottom line: Toronto, a team not known for liking to possess the ball, did not really possess the ball, and indeed had just 3 shots on goal with 15 shots taken. Playing counter against Real simply doesn't work - just ask the Timbers of 2013.

Here's how Real Salt Lake scored in the PWP Defending Process:

Observations:

Pretty compelling - but what stands out most to me is not the Real Salt Lake defense as a whole, but rather that Toronto have yet to break the 70% Passing accuracy percentage total this year - they had 67% accuracy in Seattle, 59% accuracy against DC United and 54% accuracy against RSL. If anything, perhaps those first two wins for Toronto were more of an oddity than a trend of things to come? Guess we'll have to wait and see about that.

One other, more devious consideration, is that Toronto are simply refusing to make attempts at engaging a possession-based approach. Instead one might offer that, if the conditions aren't right for a counter-attack, they simply just push it back into the opposing teams' half any way they can, and then 'wait' for the right moment to engage an attack when the opponent is a bit more disorganized.

I'll have to watch for that a bit more closely - and given Nelson has been in Europe for a while, he may have picked up a few things that are a wee bit different, tactical wise, than how things have been played here in the States.

The PWP Defending Player of the Week was Kyle Beckerman - and here's why:

Observations:

Gotta love that picture - anyhow - some very compelling information here with Kyle on both sides of the pitch - he's in the central defending midfield role with perhaps the most responsibility of anyone on RSL, and here he is with 2 key passes, 80% accuracy in passing, with 6 interceptions and 14 recoveries.

Bottom line this week is that Kyle Beckerman basically ate Toronto for lunch.

All that said, there is always next week, Toronto; for now this game and others already played this year by Beckerman do him well in his opportunities to continue to play for the USMNT.

In closing...

In about a day or two I'll be offering up the Cumulative PWP Composite Index and all that goes with it.

All the best, Chris

ASA Podcast XLII: Expected Goals 2.0

I apologize for putting this up so late. I intended it to be up yesterday, and things just happened this way. The audio quality for my mic the first half was not good, but that was fine because Matty filled us in about Expected Goals 2.0 or something like that. Then, Drew and I do a rundown of Week Three's games, packed with some of our typical witty commentary. We hope you enjoy it!  

[audio http://americansocceranalysis.files.wordpress.com/2014/03/asa-podcast-xlii-xgd-2.mp3]