Chasing Abby or Catching Abby

By Benjamin Massey · December 1st, 2017 · No comments

Kyle Thomas/Canadian Soccer Association

For Canada’s women’s national team, 2017 was an intermission. No Olympics, no World Cup, no qualifying, 365 days of the calendar you got out of way to prepare for the next thousand. 2018, 2019, and 2020 are what count: Jessie Fleming, Deanne Rose, Rebecca Quinn, Ashley Lawrence, Kadeisha Buchanan, and hopefully players we didn’t expect developing into the most competitive lineup Canada has ever had. If we’re ever going to win some gold these are the women who’ll do it.

However, there’s another piece of history Canada is chasing for which 2017 mattered a great deal. In our last game against Norway Christine Sinclair scored the 169th goal of her international career and fourth of the year. She needs 16 more international goals to pass Abby Wambach for the most in the history of soccer.

16. Not that many, not with CONCACAF World Cup qualifying bringing games against Haiti and Puerto Rico and that lot next year. But Sinclair is 34, old for her position and her sport. It’s not certain she has enough time left.

The good news is we don’t need to put the old lady on the ice floe yet. Sinclair’s scoring rate over the past two seasons is better than Adriana Leon’s career rate and basically tied with Deanne Rose. Behind Nichelle Prince, but Prince has 1,013 senior minutes and hasn’t scored in the last 655. More importantly Sinclair’s been piling up the assists: seven in 2017, four in 2016, four in 2015, our leading playmaker each of the last two years. On a roster that despite Herdman’s improvements can get very direct, Sinclair and Janine Beckie are the only forwards moderately capable of holding up the ball and Beckie has better things to do. Sinclair would be in the eleven on merit if her name was Jane Smith and her idea of leadership was telling Sophie Schmidt “your she-man-bun looks very nice.”

You couldn’t say that about Abby Wambach at this stage of her career. When she retired at age 35 nobody, American or otherwise, argued it was too early. But then her team was a lot deeper. The American attack her last year included world player of the year Carli Lloyd, a not-very-broken-yet Alex Morgan, and Christen Press, with Crystal Dunn the coming woman. Running the attack through Wambach made no sense, using her off the bench rather than Dunn or the Amy Rodriguez/Sydney Leroux depth players of the future set seemed unwise and out of her character. If anything most of us seemed to think she’d gone a year or two late. But selfish old Abby was a much better goal-scorer than kind old Christine.

This graph compares Sinclair and Wambach’s goalscoring records in the seasons in which they turned a given age. For example, Christine Sinclair turned 34 this year, so 2017 was her age 34 season. Abby Wambach turned 34 in 2014, so 2014 was her age 34 season.

Sinclair and Wambach were both early June babies so we are truly comparing their performances at the same age. That’s only the first coincidence; it’s weird how closely they trend together. Right down to both slumping at 31 and bumping back in their age 32 seasons. Of course, when Wambach “slumped” in her age 31 season she started the year ice cold and ended it with four goals in six matches at a FIFA World Cup. Sinclair, with nothing much to play for in 2014, scored once.

My co-podcaster Carolyn reminded me that, as recently as 2009, Wambach and Sinclair were within a few goals of each other. Wambach made her century on July 19, 2009 at age 29 against, coincidentally, Canada. Sinclair scored her hundredth five months later, on February 20, 2010 against Poland despite being three years younger. Sinclair had 130 caps, Wambach had 129. It was only in old age that Wambach pulled away.

All through their thirties, Wambach outproduced Sinclair very heavily. Wambach’s worst 30+ season (2011 at age 31, 0.566 G/90) was essentially identical to Sinclair’s best 30+ season (2015 at age 32, 0.570 G/90). Wambach’s advantages included more home games, more easy qualifiers that Canada skipped for 2015, and teammates who deferred to her on scoring chances. Big goalscorers past their primes can make a lot of hay against CONCACAF minnows. Wambach’s age 30-34 seasons included an Olympic qualifying and a World Cup qualifying campaign, running wild both times. Sinclair only had 2016 Olympic qualifying, where she was mostly injured. But for the moment we aren’t asking who was the better player, but who put the ball in the net more. The good news is that Wambach’s production did not decline much from age 30 to age 35, and if Sinclair can do the same she might win on endurance.

In her last two years, Sinclair has scored 0.439 goals/90 minutes clip. That would put her 3,278 minutes away from goal #185 and immortality, or just about three years. Assuming she isn’t injured any more often, and scores at the same rate, and gets just as much playing time every game. As she goes from her mid to her late thirties none of those assumptions are safe. Healthy and productive 2018 Olympic and 2019 World Cup qualifying runs will help Sinclair; age, wear, and tear will hurt.

History’s best female strikers tend to retire around Sinclair’s age. Wambach was 35, Birgit Prinz, Tiffeny Milbrett, and Carolina Morace all 33, Mia Hamm only 32. The exceptions have either been multi-position stars like Kristine Lilly or Michelle Akers or, encouragingly, super-annuated veterans for second-rate countries who still had roles into old age. Scotland’s Julie Fleeting is 36 and still active. Italian Patrizia Panico hung on until age 39. The late 30s are not quite unexplored territory.

Canada’s second-best all-time goalscorer, Charmaine Hooper, not only kept going up until age 38 but provided value. Playing second and sometimes third fiddle behind Sinclair and Kara Lang, Hooper played 1,683 minutes over 24 games from age 36 and up, scoring 11 times, which for Sinclair would do nicely.

Alas, Hooper’s old age flatters to deceive. In her last eight caps Hooper hit one final vein of form, scoring eight times including a hat trick against Sweden to wind up with 71 senior goals in her career. But that was another age. When she last played in July 2006 (it would be inaccurate to say she “retired”) women’s soccer was still young. Today, give-or-take the inevitable blunders, decent teams can basically defend a savvy but unathletic striker. In 2006 players with sometimes questionable fitness could play 90 minutes every night and chip in multi-goal games. Melissa Tancredi got a miracle brace against Germany at the Olympics, but 2016 Melissa Tancredi was Alex Morgan in her prime compared to some of the forwards you saw ten years earlier.

On the other hand, Sinclair’s a lot better than Hooper was.

The odds may still, just, favour Sinclair. Barring injury or the unforeseen, the natural arc of Christine Sinclair’s career will close after the 2020 Summer Olympics in Tokyo, when Sinclair will be 37. It’s hardly likely she’ll hang on for another cycle into her forties, but it would also be surprising for her to retire after an Algarve Cup or something unless life makes her. Moreover, and I’m touching wood just typing this, Canada’s 2019–20 team might be highly competitive: the best Sinclair’s ever had, and worth hanging on for at the end of an endlessly frustrating international career. If Sinclair does play through 2020, and has a couple productive qualifying campaigns, that’s just enough time. She’ll never come close to Wambach on goal rate but she might show the fortitude to hang around and catch Abby with sheer guts and longevity, to say nothing of being popular enough that her teammates won’t freeze her out when they don’t need her anymore.

Which, come to think, would be pretty appropriate for a Canadian heroine.

Whatever It’s Called, the FIFA Women’s Player of the Year is Still a Joke

By Benjamin Massey · January 11th, 2017 · No comments

John Major/Canadian Soccer Association

In 2012, after the greatest season in her own and her country’s history, Christine Sinclair was nominated for FIFA’s women’s player of the year award. She probably deserved it: the leading scorer and most valuable player at the Olympics captained a perennial underachiever to its best-ever international finish, winning world-wide admiration for skill and guts. Instead she finished fifth, behind the winner Abby Wambach (a one-dimensional goalscorer with fewer goals), Marta (who was, honestly, terrible), Alex Morgan (also good, also outscored), and Homare Sawa (another forward on another good team without much to scream about individually).

Wambach wasn’t a bad winner but what hurt was how Sinclair lost. Among peers her value was more-or-less appreciated, but in FIFA a country with no real women’s soccer program gets just as many votes as Germany. It was the their uninformed votes that relegated Sinclair to fifth place while elevating the average Sawa and the famous-but-worse-than-Melissa-Tancredi Marta. It was proof that the FIFA women’s player of the year award was devoid of merit.

2016 was the franchise’s unoriginal sequel. Once again, Canada’s women beat France and won an Olympic bronze medal. Once again, captain Christine Sinclair played an important role for a team that beat expectations. Once again, Sinclair was nominated for international player of the year, now called “THE BEST” by semi-literate cretins, and once again her opponents included an American player with a probably-inferior season but a huge international reputation (Carli Lloyd), a notable Japanese player who didn’t do anything this year (Saki Kumagai, whose team failed to even qualify for the Olympics), and a fabulously-overrated Brazilian striker coming off a unremarkable season, presumably on the list because people have heard of her (Marta, again, harder to kill than Jason Voorhees).

The plot twist was that Sinclair didn’t deserve the award either. She had a good year for club and country, putting her in Camille Abily/Amandine Henry/Lotta Schelin nice-but-not-enough territory. She was not one of the three best players in world women’s soccer, nor one of the three best out of the ten nominees. The only hope was that Sinclair would scoop up reputation votes as a fabulous player approaching the end of her career who everybody should know has never gotten her due. It wasn’t enough and she finished eighth, ahead of Abily, behind Schelin, and pretty much tied with Henry.

Canada couldn’t complain, but the results were still stunning. Marta, once again, wound up in the final three for no obvious reason. Her 2016 was not nearly as appalling as her 2012, as she bagged a couple goals against Canada and one against France in friendly play, plus two against the Swedes at the Olympics. She was also effective with Scandinavian superpower FC Rosengard, though there’s something not-quite-World-Player-of-the-Year-flavoured about the phrase “joint-leading scorer with Ella Masar*.” The Germans had three nominees in tremendous all-round midfielder Sara Däbritz, retiring sentimental favourite and long-time talent Melanie Behringer, and young forward Dzenifer Marozsan, threatening to split the vote. France’s Henry and Abily were also present.

Maybe that’s why Carli Lloyd won. Carli Lloyd, who was the bona fide 2015 world player of the year after scoring about a billion goals in the World Cup and tearing Japan to bloody ribbons in the final. Carli Lloyd, still in her prime and now, technically, an award-winning author. Carli Lloyd, who in a jam-packed 2016 managed one and a half goals against real teams, knocking a rebound into an empty net against France and scoring a pretty good header against New Zealand (that’s the half). Carli Lloyd, who captained her American national team at the Olympics to, er, a fifth-place finish, the worst in their history. Carli Lloyd, who ran off from the Houston Dash[1] to chill and go on a book tour, while her coach literally did not know where she was. Carli Lloyd, who was outscored by Behringer at the Olympics despite Behringer being a traditional midfielder and Lloyd a very attack-oriented number 10. Behringer took penalties and played more games but she out-open-play-through-the-quarter-finals-scored Lloyd three to two.

Then there was the coach of the year award. Canada’s John Herdman was nominated, as in 2012. As in 2012 there was a very good argument that he deserved to be among the contenders, and as in 2012 he came up short, finishing fourth in a bewildering field. Of the ten coaches nominated three (Brazil’s Vadão, France’s Philippe Bergeroo, and South Africa’s Dutch boss Vera Pauw) had been sacked in disgrace before the finalists were announced. Another (the United States’s Jill Ellis) is unpopular with fans for the whole “leading the team to its worst major tournament finish” thing. The winner was Germany’s Silvia Neid, who absolutely deserved it and ran away with the voting. Sweden’s Pia Sundhage was third, and also a good pick. But Ellis, the catastrophic underachiever, was second, despite the fact that her team blew its brains out in Rio and—very sorry but I’m going to have to run wild with the formatting here—Sundhage outcoached her into the ground the only time they met.

At least Neid’s victory was justice at the end of a long and legendary career. But Lloyd was less deserving than the top Germans and the idea of Marta or Ellis in their respective top threes is laughable. To make matters worse, FIFA has introduced a fan voting component to their traditional format. The votes of the captains of each national team, the coach of each national team, selected media members from each national federation, and the overall fan ballot each account for one quarter of the result[2].

The fans did some damage. Their votes bumped Ellis ahead of Sundhage and lifted Marta ahead of both Behringer and Maroszan, who among captains, coaches, and media were still behind Lloyd but ahead of the Brazilian. As it was Marta beat the Germans handsomely[3], making one wonder what sorts of idiots vote for these things.

Yet the mathematically-gifted of you will have realized that if fans get 25% of the say, that means the “experts” get 75%. No amount of Marta-worship from ballot-stuffing Brazilians, no number of Tumblr campaigns for Carli 2017, would have mattered if the professionals had voted intelligently. And when you break down the voting you see that, just like in 2012, countries that don’t even really play women’s soccer dragged the whole award into the mud[4].

Below is a table showing each nominee, her final position in the actual player of the year award, the proportion of the vote she received from all the participating captains, coaches, and media (both raw and weighted one-third each), and the proportion of the vote she received from countries ranked in the top however-many of the most recent (December) FIFA women’s soccer rankings. Also shown are the votes for all countries with a FIFA ranking, which means any country that has played a single official match in the past eighteen months, and the votes for all countries without a FIFA ranking, which haven’t. There are a few countries that are not even “not ranked” by FIFA, but regardless sent in votes; they are lumped in with the “inactive.” And advance apologies to my mobile users.

Name # All Weighted Captains Coaches Media Top 10 Top 20 Top 30 Top 40 Active Inactive
Abily 10 5.39% 5.34% 5.26% 6.19% 4.55% 9.58% 9.60% 9.17% 8.31% 6.24% 2.81%
Behringer 3 14.18% 14.73% 10.19% 11.79% 22.22% 21.07% 20.15% 20.54% 20.02% 14.27% 13.92%
Däbritz 5 6.29% 6.44% 5.01% 5.85% 8.47% 11.11% 6.21% 4.65% 4.20% 5.80% 7.81%
Henry 9 5.60% 5.49% 6.02% 6.53% 3.92% 4.60% 6.03% 5.43% 5.21% 5.64% 5.49%
Kumagai 6 7.62% 7.48% 7.52% 9.41% 5.50% 0.77% 2.45% 2.33% 3.30% 5.64% 10.99%
Lloyd 1 21.92% 21.24% 26.48% 25.28% 11.96% 6.51% 16.95% 17.05% 18.12% 21.66% 22.71%
Marozsan 4 13.19% 13.19% 13.03% 13.32% 13.23% 21.46% 19.21% 20.54% 20.02% 15.75% 5.37%
Marta 2 13.16% 13.05% 15.71% 11.79% 11.64% 7.66% 6.03% 7.11% 8.41% 11.35% 18.68%
Schelin 7 7.08% 7.35% 6.27% 4.66% 11.11% 5.75% 6.40% 8.01% 6.71% 7.07% 7.08%
Sinclair 8 5.57% 5.70% 4.51% 5.17% 7.41% 11.49% 6.97% 5.17% 5.71% 5.72% 5.13%

This table raises questions. Questions like “when Marta dies is she going to get 10% of the vote, or 15%?” and “how did countries that don’t play woso develop such a girl-crush on Saki Kumagai?”

Though the top 10 isn’t a lot of countries, it amounts to 29 voters picking three winners for a total of 81 points in ballot strength (North Korea somehow neglected to appoint a media representative). 29 people is not an overwhelming sample but major awards in sports and entertainment have been decided by fewer. For Christine Sinclair to finish third among that elite 29, including her coach and her media rep, is a tribute from the very competitors Sinclair has been trying to lift Canada up to.

Carli Lloyd starts gaining ground early, driven by a strong coach’s vote in the top 20 and top 30. Of course Jill Ellis voted for Lloyd but the coaches of England (5), the Netherlands (12) China (13), Italy (16), Switzerland (17), South Korea (18), Iceland (20), Austria (24), Belgium (25), and Mexico (t-26) also put Lloyd first. Many of those countries played the Americans in US-based friendlies this year, and Lloyd scored on not a few, so thinking they were sunk by the player of the year must be a great consolation. Italy’s Antonio Cabrini completed his confusing ballot with Marta and Lotta Schelin, then for good measure listed Jill Ellis and Philippe Bergeroo first and third on his coach’s list, suggesting there may be a reason Italian woso has on the downturn lately. Pia Sundhage had Lloyd nowhere.

Maybe they just liked her book. Anyway, what counts is that, in the real woso world, Lloyd is in no danger of catching either Marozsan or Behringer and Marta is an also-ran. The top 30-ranked countries include everyone of even minor consequence in at the senior international level, save some token Africans. Had only the top 30 voted we would have finished with Marozsan and Behringer exactly tied with Lloyd a good step behind, and that would have been an excellent result. If you don’t hold her Houston Magical Mystery Tour against her it’s easy to defend Lloyd as the third-best player on this list.

However, that’s not how it works. Among both minnows ranked below 40th in the world and the teams that aren’t active at all, Lloyd had a decisive lead. Of the 3,321 points allocated in the player of the year ballot the top thirty countries disposed of 774. Inactive countries—national teams which literally do not exist—cast 819 points worth of votes. If you want this award you’re better off being a household name with a book deal.

You can follow Lloyd’s share of the vote rising as the calibre of the voters declines, and very satisfying it is. But Carli Lloyd is nothing next to Marta. Even as far as the top 40, as minnows fill the water, Marta was incapable of cracking 9%. But add in the true nobodies and Marta is on the podium: between them and the fan vote the Brazilian Ella Masar was anointed the second-best player in world women’s soccer for 2016.

If you have the endurance, this chart shows how players’ votes changes as we descend the rankings. Select a player to highlight her, and hover over a point to see which ranking that is. Each point is a player’s ballot position among voters within a set of ten ranking places (which usually doesn’t mean ten countries), with the last point being the not-ranked and not-even-not-ranked voters.

The coach of the year ballot, thank God, was simpler from both ends. At the top, except for us homer Canadians supporting John Herdman, Silvia Neid was a fairly obvious choice both on the basis of Germany’s gold medal and as an acknowledgement of one of the best coaching careers in women’s soccer history. You might chisel her out of first place, on the grounds that she did get beat by Melissa Tancredi in a game she didn’t really want to win and that Herdman or Pia Sundhage had done more with less, but leaving Neid out of your top three altogether would have been negligence. Sundhage was the obvious contender for best of the rest, with Herdman hanging around but probably impossible for a Canadian to neutrally rate.

On the other side of the vote were, well, the guys who’d been fired already. It’s a good bet you can’t be the best coach in the world if your employer decided they’d prefer anyone else. Except for French captain Wendie Renard, who loyally put Philippe Bergeroo third on her ballot, voting for France’s fall guy was a sign of mental illness. And he might still have been better than Pauw, obviously listed only as African representation, or Vadão, whose Brazilians beat nobody in particular and needed a win from the penalty spot just to reach a home bronze medal game in which Canada, a team he had met in two pre-tournament friendlies, destroyed him.

And the fired guys weren’t even the only randoms! Gérard Prêcheur, head coach of the Olympique Lyonnais women, winner of the last season’s Champions League and Division 1 Féminine as well as a favourite in both this year, would have been a excellent nominee if you could find anybody who prioritized European club play in an Olympic year, which you can’t. A notch below were Swiss boss Martina Voss-Teckleburg and Bayern Münich’s Thomas Wörle, both of whom are probably good coaches and neither of whom had much of a 2016. Switzerland dominated a European Championships qualifying group that had nobody in it and wasn’t at the Olympics. Wörle won the last Bundesliga but ain’t gonna win this one and went out of the 2015–16 Champions League to Twente, which is even worse than it sounds. It is, apparently, hard to find ten decent women’s soccer coaches in the world; Paul Riley must be throwing Heineken bottles at his television.

So, with such an obvious top four of Neid-Sundhage-Herdman-Prêcheur, how did Ellis get the silver medal? Oh boy here comes that big table again.

Name # All Weighted Captains Coaches Media Top 10 Top 20 Top 30 Top 40 Active Inactive
Bergeroo 10 3.54% 3.50% 3.85% 3.73% 2.94% 0.77% 2.82% 2.58% 2.60% 3.01% 5.19%
Ellis 2 16.62% 16.26% 19.74% 18.24% 10.80% 1.92% 7.72% 6.59% 8.01% 13.80% 25.31%
Herdman 4 7.27% 7.30% 7.26% 6.87% 7.76% 10.73% 9.04% 9.17% 8.91% 8.06% 4.81%
Neid 1 32.46% 32.86% 28.80% 30.87% 38.89% 42.15% 37.10% 38.11% 36.54% 34.26% 26.91%
Pauw 8 3.06% 2.99% 2.39% 4.58% 1.99% 1.15% 1.13% 2.33% 2.20% 3.29% 2.35%
Prêcheur 5 8.11% 8.23% 8.72% 6.02% 9.96% 11.11% 10.17% 9.69% 9.51% 8.26% 7.65%
Sundhage 3 17.89% 17.98% 15.98% 18.58% 19.39% 26.82% 24.67% 24.29% 22.92% 19.61% 16.16%
Vadão 6 4.97% 4.89% 6.15% 4.75% 3.77% 1.15% 0.94% 0.65% 1.10% 3.45% 9.63%
Voss-Tecklenburg 7 3.03% 2.97% 2.48% 4.24% 2.20% 3.45% 4.14% 4.01% 5.31% 3.37% 1.98%
Wörle 9 3.06% 3.01% 4.62% 2.12% 2.31% 0.77% 2.26% 2.58% 2.90% 2.89% 3.58%

Neid is never not winning, so the victor was the right one. But witness, friends, the Rise and Fall of Jill Ellis. From less than two percent of the vote among the top ten (one of whom was Ellis herself), she gets a boost as she rolls downhill from support that included the reliably-mental Italians and Swiss but was in no danger of bothering the top picks. Ellis is well above the three coaches who have actually been sacked, which is fair enough, as well as oddballs Voss-Teckleburg and Wörle. Then get down to the nowhere countries and all hell breaks loose. Among the minnows only does Ellis whip the superior Prêcheur and Herdman, pass Sundhage, and storm into the medals but, in the inactive countries, she very nearly catches Silvia Neid, which by itself proves they should have their votes taken away.

Yet, again, the American is not the only recipient of minnows’ largesse. At the bottom of the rankings Vadão outpolls not only the rest of the fired brigade but Herdman and Prêcheur! In the very last tiers Bergeroo passes Herdman as well; our Geordie John apparently doesn’t have great name recognition in Argentina. Prêcheur’s work at OL makes him a bit of an insider’s candidate, and he does very well all things considered among the elite, but his little rally doesn’t last long when the obscure countries get in. Wörle, Pauw, and Voss-Tecklenburg, lacking either big names or achievement, are basement dwellers all the way, though the minnows prefer the sacked Pauw to the useful Voss-Tecklenburg in only the least of their capricious whims.

Want to see it? Too bad; I just have another one of those crappy charts.

On the men’s side, where four billion people know what Claudio Ranieri did for Leicester City, these problems don’t arise to the same extent. Complete information is available to even the most sheltered voter. Women’s soccer is much more of a niche event, and huge chunks of voting power are handed to nobodies because they captain a team that, even if it bothered to get together for a game, wouldn’t win a decent Canadian metro league. You have to look for women’s soccer, you can’t just absorb it as with the men.

There’s no question that some captains, coaches, and media from the irrelevant nations took their duty seriously and came up with ballots at least as well-informed as a random Vancouver blogger’s, and from the US to Uzbekistan the media vote was “fair” at worst. But a statistically-obvious number voted for the people they’d heard of. It wasn’t a Canadian who got screwed this time; Herdman wasn’t going to win no matter how you divided it up and Sinclair wouldn’t have deserved to. But the essential truth has not changed in four years: the FIFA women’s awards are voted on in ignorance and therefore meaningless.

(notes and comments…)

Rowdies Can Rongen

By Benjamin Massey · August 21st, 2015 · No comments

Matt May/Tampa Bay Rowdies

Matt May/Tampa Bay Rowdies

Today, Thomas Rongen lost the NASL sack race. 261 days after being named Tampa Bay Rowdies head coach on December 3[1], the veteran has been fired along with general manager Farrukh Quraishi[2].

Whenever you fire your coach eighteen games into his first season while tied for third in the NASL combined standings you’re hitting the panic button. Was I Rowdies owner Bill Edwards I’d have given Rongen more time. Well, I tell a half-truth; was I Edwards I wouldn’t have hired Rongen at all. When I wrote my 10,000-word NASL preview in March I called Rongen, as politely as I could, “the sort of coach who always seems to get a new job somewhere.”[3] I’m sure he will, too. Rongen is affable, fun on social media, and wears a bowtie. He is a “character” with support in the soccer community and such people are never out of work indefinitely, though the litany of disappointment since the 1999 MLS Cup is now a bit longer.

It’s not like Rongen and Quraishi were serene themselves. The two ex-bosses tore down virtually the entire Rowdies roster over the 2014-15 offseason; they must take credit for the new team’s achievements and blame for its failures. Recently they made headlines by signing two high-profile busts, Freddy Adu and Omar Salgado. Adu, who is frequently injured, has played a total of fifteen minutes in one game. Salgado, who is even more frequently injured than that, spent some time waiting for paperwork, played twelve minutes in one game, and hasn’t been seen since. They joined a team of career underachievers like Corey Hertzog, Gale Agbossoumonde, Maicon Santos, and Rich Balchan, with only Hertzog inherited from 2014. There’s little material to suggest a championship run, and maybe a hot start raised expectations beyond reason. If so that would be cruel: the luck that made Rongen look good in the spring turned out to be really bad.

Tampa Bay had a fine spring season. They finished in second place, only a point behind New York (who nobody’s going to get) and were second in goal difference as well. If not for the Cosmos beating the Rowdies 2-0 in New York on April 18, Tampa Bay would already have booked a playoff spot thanks to the NASL’s split-season format. There’s no denying that firing the leadership of a team which came so close so quickly looks odd.

NASL TSRs and PDOs through August 21
Spring 2015 Fall 2015 Total 2015
Team TSR PDO TSR PDO TSR PDO
ATL 0.448 92.35 0.385 114.52 0.419 103.74
CAR 0.427 117.20 0.419 98.37 0.423 107.93
EDM 0.478 94.20 0.556 110.74 0.511 100.68
FTL 0.504 102.69 0.562 91.55 0.527 98.97
IND 0.535 103.77 0.558 90.68 0.546 97.62
JAX 0.503 80.00 0.509 103.57 0.506 102.04
MIN 0.528 99.74 0.503 106.60 0.518 102.04
NY 0.557 107.35 0.548 98.18 0.553 103.65
OTT 0.528 91.39 0.528 107.01 0.528 100.67
SA 0.500 94.28 0.422 95.00 0.469 94.54
TB 0.492 111.00 0.500 85.71 0.495 99.26

But the spring season is only ten games. Drawing conclusions from ten games’ worth of results is stupid. To the right are the Total Shots Ratio (TSR) and PDO of each team in the NASL so far this season. Fully explaining TSR and PDO is beyond the scope of this article but think of them as statistical smell tests. PDO is simply a team’s shooting percentage plus its save percentage, and TSR the proportion of total shots in a game taken by the team in question. So a team that’s generating more chances than its opposition will tend to have a higher TSR, and a team which is converting on a lot of its shots or getting lights-out goalkeeping a high PDO. This is important because history suggests that a team with an outrageous PDO, either high or low, is deceiving you: over the long haul a team’s PDO approaches 100.00, and if the PDO is way out from that said team is probably either better or worse than the standings show.

NASL statistics aren’t perfectly accurate, but people who minded their TSRs and PDOs would have guessed that the Cosmos would be killing it this year while the Scorpions and Railhawks slumped, and that FC Edmonton wasn’t nearly as good as their 2014 autumn. The point isn’t that comparing TSRs and PDOs allows you to glean soccer’s well-hidden secrets, but that it allows you to guess broadly who’s lucky and who isn’t.

Tampa Bay’s TSR is solidly middle of the road. In the spring they surged with the second-best PDO in the league and the best among anybody who should seriously be considering playoff spots. In the fall their luck’s run out and their spot in the standings has fallen with it. The result is that their current position is about right, or maybe a little flattering. They’re average. They have average tattooed all over their foreheads. That certainly isn’t enough to blow away your coach after eighteen games, but it’s not worth making a real fuss to defend the guy either.

Last year’s Rowdies, much-maligned with a long-time coach playing his kid in front of a crowd demanding a lot more, were pretty average as well until they were decimated by the new management. 2014 Tampa Bay ended the year with a perfectly respectable TSR of 0.509. Where they fell apart was in PDO: 85.85 was last place in the league by a lot. A 22.09% shooting percentage was bad, but their save percentage of 63.77% was the real killer, and so the Rowdies finished well out of the playoff running.

Break down Tampa Bay’s spring PDO and you’ll see how their save percentage improved and nearly brought them glory: 84.21%, first in the NASL by a long way. Quelle différence, but not a sustainable number. An anti-stats type might grumble that Rongen should get credit for finding talent in goal, but the problem is that Tampa Bay’s keepers have been Matt Pickens, one of the few returning 2014 Rowdies, and veteran Kamil Čontofalský, who despite obviously being a good player had an unlucky save percentage last year in Fort Lauderdale. Rongen did not find brilliant new talent capable of posting high percentages: two keepers coming off low percentages, one of whom Tampa already had, rebounded.

Moreover, the Rowdies are not a young team building for the future: as mentioned above, a lot of their players have been plucked from the busts of a higher level. Men like Adu, Hertzog, and Agbossoumonde will not develop into star players now that they’re in their mid-twenties with several professional seasons behind them. There’s been much talk of a “five-year plan” but, in Tampa Bay’s case, that didn’t mean developing home-grown talent this season at the cost of results. On the Rowdies roster Salgado, Darwin Espinal, Robert Hernández, and Jeff Michaud are under 22 years old, and only Michaud played his teenage years in Florida. Michaud and Hernández essentially never play, leaving Espinal as the only young-ish regular. Čontofalský is 37, Pickens is 33, and leading scorer Maicon Santos is 31. That’s a long way from a youth movement.

Again, there is nothing here to justify so large a change to the team so quickly. For Rongen and Quraishi’s firings to make sense on their own terms there has to be something else: maybe all those new players came at a high price their performances haven’t justified, maybe something went on behind the scenes, or maybe Bill Edwards is just sick of waiting and doesn’t see enough improvement on the horizon. So that’s my defense of Rongen and Quraishi. The other side is there’s no reason for Rowdies fans to stay up wishing they were still around, either.

(notes and comments…)

Those Matt Van Oekel Statistics, in Full

By Benjamin Massey · December 18th, 2014 · No comments

Today, FC Edmonton announced the signing of veteran second division goalkeeper Matt Van Oekel[1]. Virginia native Van Oekel had spent his entire seven-season professional career with various incarnations of Minnesota soccer clubs, starting with the Thunder in 2008, and has been the starter for the Minnesota Stars/United since 2012.

Though seldom classed among the NASL’s best Van Oekel’s had some good seasons behind stalwart defenses. He’s also one of the league’s most stylish players, his various haircuts being a bit of a running gag in NASL circles, and will fill the niche left by the departing Lance “Blue Steel” Parker. Probably more importantly, Van Oekel also brings experience to what is a pretty young goalkeeping corps: John Smits is the most experienced of the bunch with his three professional seasons.

We haven’t done one of these in ages! Here is Matt Van Oekel’s career to date. As always, regular season only, NCAA statistics are unreliable, NASL statistics are dodgy especially in 2013, and though he was the starter I haven’t got his 2007 Rutgers numbers at all[2]:

GP Strt MIN G A PKG Sh Sv GA Sv% GA/90 Yl Rd
2004 Longwood NCAA 17 17 1522 0 0 0 148 107 41 0.723 2.42 2 0
2005 Rutgers NCAA 10 9 829 0 0 0 64 52 12 0.714 1.09 0 0
2006 Rutgers NCAA 14 14 1279 0 0 0 64 52 12 0.813 0.84 0 0
2007 Rutgers NCAA statistics not available
2008 Minnesota USL-1 0 0 0 0 0 0 0 0 0 nan nan 0 0
2009 Minnesota USL-1 1 1 90 0 0 0 5 4 1 0.800 1.00 0 0
2010 Minnesota USSF D2 6 5 474 0 0 0 29 19 10 0.655 1.90 1 0
2011 Minnesota NASL 1 1 90 0 0 0 2 2.00 0 0
2012 Minnesota NASL 25 25 2250 0 0 0 107 77 30 0.720 1.20 0 0
2013 Minnesota NASL 18 18 1620 0 1 0 27 1.50 1 0
2014 Minnesota NASL 20 19 1720 0 0 0 81 62 19 0.765 0.99 0 0

Van Oekel’s college career began in 2004 at Longwood University. Those bold Lancers were taking their first step into NCAA Division I and were massacred like Russians at Sevastapol. But you mustn’t blame Van Oekel: the freshman started all seventeen games, got a tonne of work, and posted surprisingly reasonable numbers for a guy who conceded 41 times.

Sensibly, rather than get shell-shock as the college soccer equivalent to Ben Scrivens, Van Oekel promptly bailed to the Rutgers Scarlet Knights, a cavalry motif apparently suiting Van Oekel just fine. As a sophomore he fought Lubos Ancin for playing time and saw action October 8 at Louisville, a team that included his new Eddies teammate Frank Jonke. Jonke scored twice on Ancin so Van Oekel rolled out of the bullpen for his first action in seven matches; Jonke almost immediately beat Van Oekel to get his hat trick. But the Jonke show against Ancin established Van Oekel as Rutgers’s starter for the rest of the year. Truly, this was a partnership meant to happen.

Though it’s hard to tell because of Rutgers’s inability to post statistics for his senior year, and a pre-season ACL injury hurt[3], Van Oekel ended his college career in 2007 with a good record but few accolades. He was left off the MLS Combine lists and ignored in that year’s SuperDraft despite a weak year for goalkeepers. (It’s not like his program was overlooked: his teammate in 2005 and 2006 was Nick LaBrocca, while 2007 featured a young Dilly Duka.)

Patience is a virtue, even for Lancers and Scarlet Knights, and soon Van Oekel got his chance. Prior to the 2008 season Van Oekel went on trial with the USL First Division Minnesota Thunder and impressed enough to win a contract[4]. The Thunder had a decent veteran team but were weak in goal: Joe Warren had just retired and Nic Platter would see his first season as a pro starter. A great opportunity for Van Oekel, though Platter played every minute in 2008. Still, Van Oekel’s option was picked up for 2009, and on September 13 he made his professional debut at Stade Saputo, stopping four shots to help Minnesota earn a 1-1 draw with the Impact.

The 2010 season was a big one for American second division soccer and it was big for Matt Van Oekel. The United Soccer Leagues and North American Soccer League were having their acrimonious divorce, playing one last campaign together as the United States Soccer Federation Division 2 Pro League. The Minnesota Thunder were no more: the team dissolved, but their hosts at the National Sports Center created a new club known as the NSC Minnesota Stars stocked with former Thunder players. Van Oekel was not signed but stuck around on amateur terms, and while Platter showed interest in the new team[5] he soon went to Martin Rennie’s Carolina Railhawks. The Stars replaced Platter with two veterans: ex-DC United man and Liberian international Louis Crayton[6], already a seemingly washed-up wanderer at age 32, and, seemingly crazily, 35-year-old ex-Thunder goalkeeper Joe Warren, who had actually been retired for the past four years.

Crayton ended his professional soccer career 45 minutes into the Stars’ first game on a bonehead play at Swangard Stadium when he tried to fake out Dever Orgill for no obvious reason, collided with the young Jamaican forward, and blew out his ACL[7]. Warren came on in relief and began one of the more improbably successful second acts in American soccer history. With Crayton gone, Van Oekel was officially added to the roster, saw a few games, and did well enough to earn a contract for the inaugural 2011 NASL season as the badly under-financed Stars mounted a surprisingly decent run. For 2011 Warren remained the starter and Van Oekel played only one match, in (quite pleasingly) Edmonton on May 23, where the Stars lost 2-1 to a Kyle Porter brace. It was hard to blame coach Manny Lagos for sticking with Warren: after a dodgy regular season the Stars surprised everyone by taking the first NASL championship thanks in no small part to Warren’s heroics. The veteran goalkeeper retired for (presumably) the final time after 2011, though, and Van Oekel’s option was picked up.

In 2012 Van Oekel finally ascended to the starting job for his fifth season in Minnesota. With his only competition being rookie Mitch Hildebrandt, Van Oekel was assured the bulk of the minutes, and though Hildebrandt impressed when he played Van Oekel was Manny Lagos’s man. Minnesota, Van Oekel included, was inconsistent but (stop me if this sounds familiar) rode a mediocre regular season to a stirring playoff run that ended only with a defeat on penalties to Tampa Bay in the NASL final. Van Oekel also took part in a memorable US Open Cup giant-slaying when he and the Stars knocked off MLS title contenders Real Salt Lake 3-1 at Rio Tinto Stadium.

Van Oekel’s 2012 season was good enough to earn him an extended trial with MLS’s DC United[8]. It didn’t work out, and while Van Oekel signed a two-year contract with Minnesota the newly-rebranded United also grabbed veteran keeper Daryl Sattler, holder of the NASL Golden Gloves[9]. The good news for Van Oekel was that Sattler was injured midway through the spring season: the bad news was that both Van Oekel and Minnesota played poorly, conceding 14 goals in six games with a pretty lowly 0.611 save percentage. In the fall results improved, though we haven’t got the shooting data to say more: in any case Van Oekel played every minute between Sattler’s injury and the last two games of the year, when United was out of the running and Hildebrandt started.

Van Oekel hadn’t proven he was good enough to start for an elite team, but he hadn’t proven he wasn’t either. At the start of 2014 Mitch Hildebrandt had only played four NASL games and third-stringer Andrew Fontein, recently signed from Tampa Bay, was equally inexperienced. Manny Lagos stuck with Van Oekel through the first seven games , but an injury brought Hildebrandt in for the last two games of the spring. Van Oekel returned to the eighteen in the fall but Hildebrandt continued to play until August 9 when, in a game all Eddies fans will remember, he was sent off and Van Oekel took over. With Hildebrandt suspended Van Oekel played the next week against Indy and remained the starter for the rest of the season, apart from one game in Edmonton.

Now Van Oekel will face the first change of scenery of his professional career. For the second time he’s competing for a spot with the Golden Gloves winner: John Smits took the 2014 award for lowest goals-against average. Then again, goals-against average is not very meaningful as a statistic. Then again again, Van Oekel’s most successful seasons have come behind strong defenses. His 2012 was fairly good but nothing remarkable, his 2014 quite nice (in relatively limited minutes), but 2013 and his backup years showed little to get excited about. Van Oekel’s Minnesota teams have been consistently well-coached and solid under the tutelage of Manny Lagos; Edmonton is also quite a good defensive side but there’ll still be an adjustment there.

Van Oekel has struggled with consistency, which is probably the main reason Minnesota looked for other options while he was there. That said, the younger John Smits is no picture of consistency himself, and some of his mistakes have been high-profile ones. Certainly Van Oekel has the quality to fight for a starting job: equally certainly, neither he nor Smits will want to be the backup. Given FC Edmonton’s long and glorious history of serious goalkeeper injuries, getting veteran cover makes sense for Colin Miller. But the one team we thought would regularly start a Canadian goalkeeper in 2015 is now far from a sure thing.

(notes and comments…)

Comparing Independent and Reserve Attendances in Lower Divisions

By Benjamin Massey · December 2nd, 2014 · 1 comment

As you know the third division of American soccer, USL Pro, has become an affiliate league to Major League Soccer. While most teams remain independent, starting in 2014 USL Pro began admitting MLS reserve teams, and this system will massively expand for 2015 with several reserve teams in Canada and the United States.

Nobody runs their reserve team to make money, but many Major League Soccer front offices are marketing hard and hunting paying customers. Some, such as the Vancouver Whitecaps reserves, charge higher prices for tickets than the best reserve teams in the world. They’re making progress: how many times have we heard the reserve sides of Toronto FC, Montreal Impact, and Vancouver Whitecaps been called “new professional teams!!!” by the excitable, rather than an expansion of what already existed?

This model isn’t new. Several countries run reserve teams in the same league pyramid as independent clubs: Spain and Germany are the most famous but we see it all over the world, from Norway to Japan. Indeed, even in North America professional youth teams have operated alongside the independent semi-pros and amateurs of USL PDL for several years. So what does this mean for fans? Is a reserve team in a real league worth as much as a real team in the same league?

Inspired by an old Tyler Dellow post on mc79hockey.com, now removed from the Internet[1], I set out to compare the attendances of independent and reserve clubs in the same league.

Unfortunately, reliable attendance information for many such leagues, toiling in the lower divisions of non-English-speaking countries, is not readily available. Trying to compile data, I wound up with a total of ten seasons covering leagues in Spain, Germany, and the United States since 2012[2].

The distinction between “reserve team” and “non-reserve team” in North America can be slightly arbitrary: I did my best, erring towards considering teams independent. For example, Chivas USA and New York City FC did and will not appear on my lists; nor do USL Pro or USL PDL affiliates which are more like parents/feeders than full farm clubs. In the great scheme of thing potentially controversial cases are heavily outnumbered by clearcut Bayern Munich II/Chicago Fire Premier types.

Season League Level Avg. Attend/G Reserve Teams Non-Res Attend/G Reserve Attend/G Diff # Diff %
2012-13 Liga Adelante Spain 2 6724 2 6998 3990 3008 75.39%
2013-14 Liga Adelante Spain 2 7879 2 8328 3395 4932 145.26%
2012-13 3. Liga Germany 3 6162 2 6616 2077 4539 218.52%
2013-14 3. Liga Germany 3 6071 2 6556 1707 4849 284.15%
2012-13 Regionalliga Germany 4 1022 27 1288 390 898 230.62%
2013-14 Regionalliga Germany 4 1139 25 1380 524 856 163.36%
2014 USL Pro USA 3 3114 1 3308 597 2711 454.03%
2012 USL PDL USA 4 488 5 455 1026 -571 -55.63%
2013 USL PDL USA 4 588 7 578 686 -109 -15.81%
2014 USL PDL USA 4 590 9 606 482 124 25.83%
Averages 2563 2981 794 2188 275.59%

It’s not even close. At the same level, independent clubs are massively more popular than reserve teams, even considering cheaper (or free) tickets for reserve football, and this sample including the reserve sides for some of the world’s biggest clubs.

Look at Spain. The two reserve teams in the Liga Adelante in 2012-13 and 2013-14 are as huge as you can get: Barcelona B and Real Madrid Castilla. This is first-rate soccer. The current Real Madrid Castilla team includes three full internationals and Barcelona B has four. Both also have a handful of players who we’ll see on the senior Spanish side someday. And the attendance? Barça B had a middling year in 2012-13 but, on average, both these world-class development sides drew crowds that would shame an NASL team. (Most La Liga reserve sides, including Real Madrid Castilla this season, play in the Segunda División B, a level down, where attendance numbers are not reliably available.)

The two reserve teams in the German 3. Liga, Borussia Dortmund II and VfB Stuttgart II, boast big senior sides. But attendance-wise they finish behind almost everybody. In 2012-13 Stuttgart and Dortmund were second-last and last, respectively, in attendance. In 2013-14 Borussia Dortmund II improved to fifth from bottom, but still well behind 14th-place SV Wehen Wiesbaden (who they?!) while VfB Stuttgart II brought up the rear.

The largest group of reserve teams for which I had attendance data was in the German Regionalliga, made up of five regions and over 90 teams. In 2012-13 only three reserve teams (FC Bayern München II, 1. FC Köln II, and TSV 1860 München II) finished above the median in Regionalliga attendance. 15 of the 25 worst-supported Regionalliga teams, and all of the last seven, were reserve teams. Not bad when only 27 reserve teams played in the division.

It’s the same story in 2013-14. Three Regionalliga reserve teams (TSV 1860 München II, FC Bayern München II, and Hertha BSC II) again finished above the median attendance. 14 of the 25 worst-supported teams, and again all of the last seven, were reserve teams. Some of these sides drew truly atrocious crowds. 2012-13 SC Freiburg II got 164 fans a night, which would have embarrassed USL PDL.

Over in the United States, one reserve team operated in USL Pro last year: the Los Angeles Galaxy II. They did not draw flies, despite offering season tickets free with the MLS package and independent seats starting at US$72[3].

North American fans will be inspired, however, by USL PDL. In 2012 and 2013 the PDL affiliate teams actually drew better than the independent ones, and in 2014 they were darn close. This bucks the trend in Spain and Germany, and might mean that North America’s different culture and greater familiarity with minor-league teams will bring more success.

But I will respond with three words: the Portland Timbers. When it comes to reserve team popularity Portland is an exception; Portland is always an exception.

In 2012, the Portland Timbers U-23s were the third-best supported team in USL PDL. In 2013 they were third again, and in 2014 they were actually second. Portland’s U-23s regularly beat USL Pro teams in the attendance race. This is a credit to Portland fans, but it also weighs unusually heavily in our table; it takes only a few well-attended games to drag up the average number when such a small proportion of the league is reserve teams.

To demonstrate Portland’s distorting effect, let’s remove the Portland Timbers U-23s and the best-supported independent team all three years, the Des Moines Menace, from the USL PDL list and see what happens.

USL PDL Attendances 2012-14 (without Des Moines and Portland)
Season League Level Avg. Attend/G Reserve Teams Non-Res Attend/G Reserve Attend/G Diff # Diff %
2012 USL PDL USA 4 393 4 400 243 157 64.36%
2013 USL PDL USA 4 505 6 526 262 264 100.67%
2014 USL PDL USA 4 503 8 546 167 380 227.99%

Take away those maniacs in Portland and USL PDL lines up a lot more with Europe. Well-supported Cascadia rivals Seattle Sounders had a USL PDL team in 2013 and 2014 and have had below-average attendance. The Vancouver Whitecaps had a PDL team (and quite a successful one) for almost a decade, and their attendance is regularly in the basement.

Note as well that USL PDL attendances are not entirely reliable. Many teams, especially badly supported ones, do not report their attendance for all games. Orlando City U-23, who draw two- or single-digit crowds, reported only one game in 2013 and none at all in 2014. The Chicago Fire Premier/U-23 miss a couple games every year. Games not reported are not included in these tables, but would lower all average numbers and disproportionately hurt affiliated teams.

Obviously nothing in this post is related to player development: the most important job of a reserve team. But those looking to reserve teams to grow soccer in Canada and the United States should look elsewhere. Fans can get behind their own club even at the lowest levels but reserve teams? They just don’t care.

(notes and comments…)

Son of Forward Perspective!

By Benjamin Massey · September 23rd, 2014 · 3 comments

Bob Frid/Canadian Soccer Association

Bob Frid/Canadian Soccer Association

Last year, when I wrote about the Whitecaps a bit more often, people were prematurely jazzed about the Whitecaps’ forward corps. Seeing signs of human happiness around me, I chugged a bottle of Haterade then sounded the horn of perspective. Sure, Camilo was the real deal, I said, but Kenny Miller probably isn’t. (The very next day, Martin Rennie signed Miller to a contract extension. Max lulz were had.) I also had a word or two in defense of Darren Mattocks[1]. My statistical predictions were largely overtaken by events, with Camilo fucking off to Mexico and Miller heading back to the United Kingdom.

Now, I haven’t watched 400 minutes of Whitecaps soccer this year so I’m not the most qualified son-of-a-bitch to repeat that exercise. But clearly nobody else is going to do it. When not excreting clickbait bullshit about women daring to be soccer fans while having breasts, our mainstream soccer writers have banged the “omg teh fukking witecaps forwards!!!!” drum like Keith Moon on speed. I even braved the Vancouver Southsiders Facebook page and forum to check the mood and came back with eye herpes. People, it is fair to say, are unhappy, crying things like “why didn’t we keep Camilo by offering him a bigger-money contract, like the one he walked out on?” and “why didn’t we sign Leo Messi?” and “how can I blame Vancouver conceding field goals every other week on Darren Mattocks?” and “I’m seven years old and can’t remember before mid-August, why have the Whitecaps never scored?”

So fuck it. How terrible are the three young forwards, Erik Hurtado, Kekuta Manneh, and Darren Mattocks, who are causing this drama factory to run overtime? I have no idea, I don’t watch the games, but here’s a table anyway. All statistics from MLSSoccer.com.

2013 2014
GP MIN G A PKG SD SoG SoG% S% SD/90 SoG/90 GP MIN G A PKG SD SoG SoG% S% SD/90 SoG/90
Hurtado, Erik 15 489 0 2 0 13 3 23.08% 0.00% 2.392 0.552 25 1493 5 2 0 42 17 40.47% 29.41% 2.532 1.024
Manneh, Kekuta 20 764 6 2 0 38 15 39.47% 40.00% 4.476 1.767 24 1009 3 1 0 47 18 38.30% 16.67% 4.192 1.606
Mattocks, Darren 20 911 3 1 0 45 14 31.11% 21.43% 4.446 1.383 26 1664 6 3 1 51 23 45.10% 26.09% 2.758 1.244

We have three players who saw fewer than 1,000 minutes last season. Short sample sizes suck. But at a glance two guys are trending as expected.

Erik Hurtado is still sort of crummy. Nothing like a hot streak to make a coach think you might be the real deal, but luckily he’s cooled off in time to not get a big raise. 1.0 SoG/90, a forward who’s doing that might be tolerable as a depth guy, but why bother when there are so many underappreciated NASL and USL Pro players around with something to prove and a modicum of technical ability? Long Tan just made the USL Pro All-League Second Team[2], but there’s probably a decent forward around there somewhere.

Only Darren Mattocks has genuinely regressed, losing 1.7 shots directed per 90 minutes. He’s dropped an entire Gershon Koffie in offense! Admittedly, this is with Canadian Soccer Jesus, one of MLS’s leading playmakers last season, having dropped back to d-mid to keep Matty Laba from getting lonely. But this is also with Pedro Morales existing and somehow not straining a hamstring for weeks. I suspect that, looking at Mattocks’s 2012 figures and bearing in mind his nasty case of shoot-from-anywhereitis at points last year, the SD/90 we’re seeing now is in line with Mattocks’s actual skill.

Even Mattocks’s season has been acceptable. Not great, not what you want from a first-rank striker, but 0.325 goals per 90 minutes runs 43rd in MLS among players with at least five goals. That sounds worse than it is: there are a lot of Max Urrutis/Luis Silvas/Michels who are banging in penalties, shooting over 60%, and having the one good goal-scoring season they’re going to live off for the rest of their hilariously overpaid careers. There’s no reason to think Mattocks’s percentages aren’t sustainable, and this level of performance is well worth holding onto. Not that they suggest any danger of a 20-goal-carrying-the-team-on-his-back season.

I’ll tell you one thing I am pretty darned confident of. Kekuta Manneh is a serious soccer player. He’s still the youngest on this list. He turns 20 in December. He’s not progressing by the numbers, but the team’s exploded so much around him and his playing time has been in such fits and starts I’m not sure what we could expect. Anyway, his current level of performance is fine, he just needs the shots to start going in and that’s only a matter of time.

If it were my team, I’d throw Manneh at the top for the rest of the season as long as his body could take it, have Mattocks as my second guy, and send Erik Hurtado to a nice house in the country where he could play every day and eat all the kibble he wants.

(notes and comments…)

Fall 2013 FC Edmonton Shooting Statistics

By Benjamin Massey · September 15th, 2014 · No comments

After the 2013 spring season, the North American Soccer League switched boxscore providers. The former company listed each player’s shots directed and shots on target over the course of a game. But starting in the fall of 2013 Soccerway took over, and their publicly-available resources list team shots but not player shots. This is a pretty considerable blow for those of us who like trying to look at player statistics.

So, very slowly, I rewatched all of FC Edmonton’s 2013 fall season games to count the individual shots directed and shots on target for each player in each match. I finished up a couple weeks ago, and the boxscores are available for public use at http://www.maple-leaf-forever.com/boxscores/nasl/2013/. I hope to do the same thing for the 2014 NASL season, eventually.

If you have any corrections contact me by e-mail, Twitter, or however you like. The rest of this article is just me musing on collecting these figures.

This was the first time I’ve done anything like this and it was a pretty interesting experience. Official team shooting statistics were still available; I was able to compare my results to the league’s numbers. There were some pretty immense differences. It brought home to me how, even in fairly clear-cut statistics like this, the subjective factor is very real.

Take, for example, a blocked shot. A forward cranks a shot and it’s deflected out straight off his boot by a defender. A forward shoots and it’s deflected out well away from him but not yet near the goal. A forward lets fly and it’s hacked off the line. Which of those is a shot directed? Which of those is a shot on target? Calling some of them shots directed requires mind-reading, knowing that he was trying to shoot rather than cross. And a blocked shot is almost always a low-percentage chance; if we’re using shots directed as a proxy for offense, which we pretty much are, then “rewarding” a guy for low-percentage shooting is irritating (though when Neil Hlavaty rifles a shot twenty feet over from thirty yards every other game we don’t have much choice). On the other hand, a ball hacked off the line is a five-alarm scoring opportunity and certainly needs to be counted somewhere.

I’m just a duffer figuring it out based on what I think “should” be. The NASL, naturally, is more professional. The league employs a couple groups of statistics-keepers, found by the teams, for each market. But there’s a factor in my favour: I’m looking at recorded matches, so have access to instant and unlimited video replay. It’s not quite so easy for somebody sitting up at Clarke Stadium. So I might well be more accurate in certain cases; there were a few times when I could correct the commentators, for example, on who took a given shot.

My instinct, and an offhand comparison of my numbers to what I’ve seen in other leagues, suggest that I may have undercounted. (But 2013 FC Edmonton was a defensively decent side that lacked creativity and was no offensive juggernaut.) If I’m going to err in any direction I’d rather err low, so I can live with that.

Dwayne De Rosario and Canada’s Attack, By the Numbers

By Benjamin Massey · October 16th, 2013 · 2 comments

Canadian Soccer Association

Canadian Soccer Association

Yesterday, while we were watching the latest Canadian national team debacle (a 3-0 loss to Australia), Duane Rollins called me out on Twitter. I was complaining about Dwayne De Rosario’s lack of contribution, as is my wont, and Rollins took exception. Accusing me of “anti-Toronto bias”[1] (which I cannot honestly deny) and “tall poppy syndrome”[2] (here I must protest), Rollins challenged me to “back up your assertion that he has under delivered with evidence.”[3]

It’s a good idea. In the past I tried to compile spreadsheets for the Canadian men’s national team, such as I do with other teams I follow, listing players’ minutes, shooting, etc. I ran into three problems: firstly, the vast number of players involved with the national team in a given span made the spreadsheets a bitch to maintain; secondly, the availability and reliability of statistical information was often poor; thirdly, the relatively limited number of minutes each player got made such exercises of dubious value anyway.

That said, in 2012 and 2013 the Canadian men played an unusually-full schedule of 20 matches, of which nine were competitive. As Rollins pointed out, MLSSoccer.com has begun providing decent statistical roundups of these matches since about the 2011 Gold Cup. Were I to take a limited sample of players who we would expect to be Canadian offensive threats, and compared them to each other and De Rosario, we might be able to get somewhere.

My argument was not that Dwayne De Rosario is the worst player on the Canadian roster or anything ridiculous like that. I hate the bastard, I carp about him on Twitter out of proportion to his importance, but I’m not a complete lunatic. I think I summed up my belief fairly when I tweeted “of course [De Rosario]’s not ‘the problem’. I pick on him on Twitter because I don’t like him. My sober analysis is not all about him.[4] He is, however, certainly part of ‘the problem’, which is that Canada needs more good players and he usually hasn’t been one.[5]

But what do the numbers say? I’d been asked that question, and as spreadsheet maestro of Canadian soccer I was obliged to answer.

I compiled match information for almost every game played by the Canadian men’s national team between June 7, 2011 (the beginning of the 2011 Gold Cup) and October 15, 2013. My source for player shooting information was MLSSoccer.com. There were five matches for which I did not have player shooting statistics: March 25, 2013 v. Belarus, August 15, 2012 v. Trinidad and Tobago, October 7, 2011 @ St. Lucia (World Cup qualifier), and September 6, 2011 @ Puerto Rico (World Cup qualifier). These matches are excluded from the shooting rate statistics you see below. For example, if a player played 900 minutes, and 90 of them were in a match I don’t have shooting statistics for, his shots/90 minutes would be calculated as (shots/810) rather than (shots/900). The minutes a player has played for which shooting statistics are available are listed under “StatsMin” in the following table.

If anyone can provide me with player shooting information to clear up the missing matches, please do let me know.

To make this task practical given my time and resources I chose thirteen players who, over the span of matches, were or will be significantly associated with the Canadian MNT attack. As you can imagine there’s been a lot of roster turnover and only seven of my thirteen played so many as ten matches in this span. The sample size is therefore very poor and I hope nobody will rely on this as the gospel, but it is a starting point.

One further note regarding penalties. To my knowledge, no Canadian player missed a penalty in the timeframe we are studying (the last penalty misses I know of were against Ukraine in 2010). Therefore when I calculate non-penalty shooting rates, I simply subtract the number of penalties made from the number of shots. That said, only two players made penalties in this timeframe: De Rosario and Will Johnson, and Johnson made his in a match without shooting data. So his shooting rates are unaffected by his penalty. Make sense?

GP Strt Min StatsMin G A PKG G/90 SD/90 SoG/90 NonPKG/90 NonPKSD/90 NonPKSoG/90
De Rosario, Dwayne 18 18 1363 1153 5 0 4 0.330 3.044 1.171 0.066 2.732 0.859
Edwini-Bonsu, Randy 3 3 211 142 0 0 0 0.000 2.535 1.268 0.000 2.535 1.268
Friend, Rob 2 0 35 35 0 0 0 0.000 0.000 0.000 0.000 0.000 0.000
Gerba, Ali 2 1 89 89 0 0 0 0.000 3.034 2.022 0.000 3.034 2.022
Haber, Marcus 9 6 529 418 1 0 0 0.170 2.799 1.938 0.170 2.799 1.938
Hume, Iain 10 5 427 247 4 0 0 0.843 4.008 1.093 0.843 4.008 1.093
Jackson, Simeon 21 13 1195 990 4 0 0 0.301 3.273 0.909 0.301 3.273 0.909
Johnson, Will 18 18 1352 1064 3 3 1 0.200 1.269 0.338 0.133 1.269 0.338
McKenna, Kevin 13 13 1170 1080 1 0 0 0.077 1.000 0.167 0.077 1.000 0.167
Occean, Olivier 9 8 595 505 4 0 0 0.605 4.277 1.248 0.605 4.277 1.248
Osorio, Jonathan 6 3 314 249 0 0 0 0.000 0.723 0.000 0.000 0.723 0.000
Ricketts, Tosaint 25 13 1405 1179 4 1 0 0.256 2.061 0.916 0.256 2.061 0.916
Simpson, Josh 10 9 772 654 2 1 0 0.233 1.789 0.826 0.233 1.789 0.826
Teibert, Russell 6 4 412 294 0 0 0 0.000 0.612 0.306 0.000 0.612 0.306

De Rosario comes off better in context than I thought. Still not good, for a player whose only value on the pitch is in attack, but better than expected.

As we may have guessed, De Rosario’s goalscoring rate from non-penalties is inferior to almost our entire sample (his one non-penalty goal was off an Atiba Hutchinson trick free kick[6], though that was an excellent attacking game from DeRo besides). He hasn’t scored a goal from truly open play since 2008, which is a long damned time for an attacking player. But his shooting rates, while poor, are not catastrophic. Among players with more than a few hundred minutes De Rosario’s 0.859 non-penalty shots on target/90 ranks behind Marcus Haber, Olivier Occean, Iain Hume, Tosaint Ricketts, and Simeon Jackson.

That’s five players who, just by the numbers, deserve to start ahead of De Rosario. So is DeRo hurt by not playing against minnows? The Canadians had three immense shooting games in their first round of World Cup qualifying where everybody got a crack: home to St. Lucia, at St. Lucia, and home to St. Kitts and Nevis. De Rosario appeared in all three of those matches, playing 235 of a possible 270 minutes. The one game that he was substituted out of, St. Lucia 0 – 7 Canada, was one of the games with no shooting statistics, and therefore is not counted in anyone’s shooting rates. DeRo had all the opportunities his peers had and more than some: of the players who lead De Rosario in shots on goal/90, only Simeon Jackson played more minutes in those three games. Marcus Haber got only eight minutes, and Haber’s games, while limited by a lousy sample size, stand out for the calibre of opposition he’s got shots against.

De Rosario is the oldest player on this list: Kevin McKenna is two years younger but is already out of the national setup for 2012. He is relatively one-dimensional, although not as badly as some of his opposition. His chief offensive skill in the past five years for his country has been hitting penalties, and while that is a genuine talent and one a team misses when it’s absent, it’s also not worth an automatic place in the starting eleven.

In his favour he plays a variety of attacking positions, which makes comparing his shooting rates to out-and-out strikers like Haber or Occean unfair (but then where are the assists, Dwayne?) His non-PK shots directed rate is somewhat better, leaving Ricketts in the shade. Benito Floro trusted De Rosario enough to put the armband on him for the first half of yesterday’s match, and as Duane Rollins insisted, that must count for something. There’s something to be said with having players of experience around the team, even if they’re no longer primary contributors, to connect newer players with Canada’s past, and if any such player is going to get the benefit of the doubt it should be the team’s all-time leading scorer. By no means are De Rosario’s numbers so poor that he must be dropped.

This is all comparing De Rosario to his peers within Canada, and you’ve probably noticed Canada has not got a great attacking national team. Let me put it this way. Major League Soccer players who are near De Rosario’s 0.859 non-PK shots on goal/90 this season include Darel Russell (0.877), Corey Hertzog (0.841), and Kenny Miller (0.825). Of these players only Miller is a starter for his offense, and Miller is both a) riding an unsustainably high shooting percentage which makes him look better than he is, b) a player who provides much more to the Whitecaps attack than just goals, unlike DeRo with Canada. Hertzog played a similar attacking role, Russell is a bloody defender. It’s a place for middling offensive forwards, and players with numbers like that seldom get big minutes in MLS.

Going entirely by the numbers, Dwayne De Rosario is not part of Canada’s best or probably even its second-best attacking lineup.

(notes and comments…)

Forward Perspective

By Benjamin Massey · July 23rd, 2013 · No comments

Benjamin Massey/Maple Leaf Forever!

If readers were to take one lesson away from this website’s many rambling posts, I hope that it would be “perspective is key”. A few games is not a long enough time to draw any conclusions. Even a full season is, in statistical terms, not long at all. An MLS season is only 34 games; seasons in other leagues with 82 or even 162 games are still impacted by luck and other non-reproducible non-skill factors which make players and teams appear different from what they are. Why would MLS be exempt with its shorter, unbalanced schedule? Sometimes a player is on a season-long hot streak, sometimes he has improved but not as much as it looks like he’s improved, and sometimes he’s unlucky. We’ve seen it time and time again, with one-hit wonders getting undeserved money and teams giving up on slumping talents who come to life in their new homes. The wider a perspective we take, the shorter the odds we make such a mistake.

This year, the Vancouver Whitecaps have gotten great performances from two of their key forwards. Camilo is turning in a career year, fighting for the MLS scoring lead while delivering a fistful of chances every game. Whereas last year he was allegedly one-dimensional, always trying to cut back and shoot or hammer it from a free kick, this season he’s been mad, bad, and dangerous to know. Meanwhile, Kenny Miller isn’t in the same race for the scoring title but he’s also trundled in six goals, some absolutely spectacular, while in 2012 much of the Vancouver fanbase (including me) wouldn’t have minded seeing him released. A Daily Mail report saying Miller may sign an extension with the Whitecaps[1] has been greeted with almost unanimous support.

On the other hand, after a promising rookie season Darren Mattocks has fallen into the tank. He scored a gem of an insurance goal against the Seattle Sounders but, otherwise, there hasn’t been much to speak of from him beyond a series of misses and getting passed by everybody on the depth chart. A knee injury has ruined his chances so far to exploit that Seattle appearance but, prior to that, he has struggled just to get playing time. Mattocks’s preseason target of a 20-goal season[2] has been quoted for all the wrong reasons. Plenty of fans, and not a few pundits, want Mattocks traded. The dominant analysis these days seems to be that he’s a one-dimensional speed merchant, one who will never reach the lofty projections so many made in 2012.

So three Whitecaps forwards whose fortunes have turned around 180 degrees from 2012. Camilo and Miller on the rise, Mattocks in the basement. How about some statistical perspective, a look at a somewhat larger picture? Below is a comparison of the major offensive statistics for these three players in 2012 and to date in 2013. As always, regular season only. This table reflects only first assists (for more on my fight with second assists, see this June article)[3].

2012 2013
GP MIN G A PKG SD SoG SoG% S% SD/90 SoG/90 GP MIN G A PKG SD SoG SoG% S% SD/90 SoG/90
Camilo 28 1719 5 7 0 64 24 37.50% 20.83% 3.351 1.257 19 1319 13 3 3 71 31 43.66% 41.94% 4.845 2.115
Mattocks, Darren 21 1300 7 0 0 39 16 41.03% 43.75% 2.700 1.108 13 623 3 1 0 34 11 32.35% 27.27% 4.912 1.589
Miller, Kenny 13 716 2 1 0 11 4 36.36% 50.00% 1.383 0.503 12 1021 6 0 0 22 9 40.91% 66.67% 1.939 0.793

So if you’ve been watching the games this year and thought “I think Camilo and Miller have definitely improved this season,” you’re right. See, statistics aren’t so complicated.

Let’s look at Camilo first. His shooting rates have gone through the roof. If he kept it up through the end of the season, Camilo’s 2.115 SoG/90 would be the best shooting rate among players with over 1,000 minutes since Yura Movsisyan’s 2.417 in 2008[4]. Extreme values are always hard to maintain and this is a considerable improvement on Camilo’s previous career best, 1.341 SoG/90 in 2011[5]. With an extremely high 43.66% of Camilo’s shots attempted being on target, it’s fair to say luck is on his side. That said, Camilo’s historically a fairly abundant shooter and there are reasons why he’d generate more offense this season. Another year’s maturity (he’s 25 and entering his prime), more MLS experience, and most importantly a far improved midfield which, between Nigel Reo-Coker and Russell Teibert, is capable of generating the ball movement and accurate passing through the middle third so important to a forward like Camilo who can’t really do it all himself.

Camilo’s high shooting percentage is inflated by going 3 for 3 on penalties. Remove them and Camilo is shooting 35.71%: still too high, not outrageous. A more reasonable shooting percentage still has Camilo on eight or nine goals this year in a relatively-modest 1,319 minutes. That would be among the league leaders (including high-percentage guys like Mike Magee and Jack McInerney) but with hundreds fewer minutes than his rivals. He is, in short, having a first-class season, one which can be trimmed but not cut down by any of the statistical tools at my disposal. It may still be a “career year”, he may still be unable to repeat it, but what a year it is. It is between Camilo and Marco di Vaio for the title “best forward in the league in 2013”.

So when we talk about perspective with Camilo, we’re not demolishing his 2013 season but building up his 2012. Playing frequently off the bench and seldom getting a good run of starts, Camilo scored only five goals in 1,719 minutes, not good enough for the team-leading striker he looked like he could be in 2011 and has been in 2013. However, his shooting percentage of 20.83% provides a clue to what went wrong. If Camilo had his 2013 non-penalty shooting percentage of 35.71% that year he would have scored eight or nine goals: enough to be considered a respectable team scoring leader, as well as doubtless earning him more minutes which may have pushed him into double digits. Add in those seven assists and we would have been very pleased with Hypothetical 2012 Camilo. Instead, his shooting percentage was in a funk so people could post, quite seriously, on message boards that they didn’t think Camilo was an MLS-calibre striker.

Perspective, people. This may be the best year of Camilo’s career, but it’s his third good year in MLS and he’s proven his quality. We were fooled in 2012 because his shooting percentage was in the dumps. I hope that, next time Camilo hits a slump, we shall not be fooled again.

On the other hand, we have Kenny Miller. His shooting rate has improved, but from “very bad for a striker” in 2012 to just “bad for a striker” in 2013. His respectable goal haul is on account of a remarkably high shooting percentage. Miller has a 61.54% shooting percentage in his MLS career, and that MLS career amounts to 1,737 minutes. Just eight minutes more than what Camilo played in 2012, when his shooting percentage was so unsustainably low.

Some readers will be tempted to argue that Miller is a unique player, one who can sustain a shooting percentage which would mark him as the best finisher in the recent history of Major League Soccer. Please don’t. I remember hearing Seattle Sounders fans, after last season, saying Eddie Johnson would maintain his 56.00% shooting percentage because he’s a natural finisher who scored a disproportionate number of goals with his head and didn’t really have any “lucky” shots trundling across the line. This year Johnson is shooting 26.67%[6]. I saw it coming, not because I’m an expert on the Seattle Sounders but because 56.00% is too much to be sustained. The same, I’m afraid, applies to Kenny Miller. We may love the guy, we may think he has a lot to offer besides his shooting (and he does), but if he wants to keep scoring goals he has to generate more shots. If Miller keeps up such a percentage over the long term, it will be unprecedented. There is no grey area here: either Miller is due for a decline, or he is such a great finisher that he is 50% better than the next-best, that otherwise-inalienable statistical laws do not apply to him.

I think he’s due to decline. Which, again, does not subtract from the other things Miller does well. His ability to hold the ball up, his high work rate in midfield and defense, and the quality of his positioning opening up more room for players like Camilo, all of these are real contributions which will not show up in his SoG/90. I think Miller gets some credit for Camilo’s career year. He also scores some beautiful goals, which do not help the Whitecaps win games any more than ugly ones but are worth a hell of a lot to us fans. Of the three forwards I am assessing today Miller is certainly the best all-rounder. But he is not shooting at the rate of a quality MLS goalscorer.

Having gotten the obligatory Maple Leaf Forever! pessimism out of the way, let’s move on to Darren Mattocks. Poor Mattocks. He’s a victim of the NCAA soccer development model, one which let him coast on his speed and his remarkable foot-eye coordination rather than develop the technical skills and soccer smarts which could have made him into a star. Yet for all that he sure piled in some goals in 2012. This year, though, he has shown next to nothing. What’s gone wrong?

You can see immediately, as I have pointed out in this pages before, that Mattocks was never going to keep up his 2012 shooting percentage. 43.75%, with no penalties or free kicks helping him out? Forget it. And sure enough he hasn’t (indeed, if you hear one complaint about Mattocks this year, it’ll be all the gilt-edged chances he’s missed). The laws of the statistical universe have avenged themselves upon young Mattocks, at a time when Camilo is tearing the league apart and guaranteeing Mattocks a place on the bench.

So here’s the good news. Firstly, although it’s a staggeringly small sample, Mattocks has improved his shooting rate. 1.589 SoG/90, if Mattocks could keep it up (he’s only played 623 minutes this year so who knows?), would be very respectable. Had Mattocks gotten the same minutes as Kenny Miller this year he’d have five goals. Not twenty, for sure, not among the league leaders, but reasonable enough for a playoff team’s second-best striker in his second professional season. If that improvement in his shooting rate is legitimate (a massive “if”), Mattocks would be a worthy MLS starter.

Second, while last year Mattocks’s shooting percentage was too high to conceivably be sustained, this year it is unreasonably low. 27.27% is bad enough (it was much worse before that nice Seattle goal), but it’s down with speed merchants like late-period Omar Cummings and Sanna Nyassi and might, reasonably, be excused. However, high-level MLS scorers tend to have shooting percentages above 31%[7] and, with a powerful shot and the aforementioned foot-eye co-ordination I believe Mattocks belongs around the average rather than with the extreme low end. He’s also getting only 32.35% of his shots attempted on target this year, another low figure. In short, luck hasn’t been on Mattocks’s side this year, and it was way off before that excellent goal against Seattle.

If he gets playing time, Mattocks should rebound. Not to the level of his luck-driven 2012 season but certainly to respectability. Minutes look hard to come by with the way Camilo is playing but injuries, slumps, and squad rotation happen; patience is the key.

My advice to Martin Rennie, and my notes to the fans who’ve gotten this far, is as follows. First, if you want to extend Kenny Miller at designated player money, think twice. Long-term you will not get an elite number of goals out of him unless he’s got another 0.7 SoG/90 hidden in his other boots. Probably he won’t even generate enough goals to be considered a good second-line striker. So ask yourself how much you think Miller’s many other abilities are worth. If what he does besides score is worth a designated player slot to you then Godspeed, but definitely do not give him major money for the sake of his scoring.

Second, be slow to trade Darren Mattocks unless a great offer comes in. Right now we would be trading Mattocks at the bottom of his value. He’s almost nowhere to be found on the goalscoring charts and his reputation is probably lower than it has ever been. At some point he’ll get into a rich vein of form and the luck which has so far eluded him will come back; when he’s banging a few goals in and people are impressed rather than disappointed, that’s the time to make a move. You may even want to keep him depending on Camilo’s fate. The numbers suggest that Mattocks is due to rebound, given time, and while the best day to trade him may have been the last day of 2012 he’s still better than he looks at the moment. Selling low and buying high is bad strategy.

Third, this season is such an improvement on Camilo’s past two that I won’t believe he can keep up this level of production until he proves it. But his 2012 wasn’t nearly as bad as some fans thought and, of course, we remember his quality in 2011. If anybody on the Whitecaps is an established MLS asset it’s Camilo Sanvezzo. 2011 Camilo was probably worth designated player money, 2013 Camilo definitely is, and only his shooting percentage was keeping 2012 Camilo from the league elite. He’s only 25 years old. If the only way to keep Camilo in Vancouver is to make him a designated player then I’m fine with that.

(notes and comments…)

PDO for Teams in the MLS West Through July 22

By Benjamin Massey · July 22nd, 2013 · 3 comments

Presented with slight comment: the 2013 season PDOs for all teams in the MLS Western Conference through July 22, 2013. The first installment was on May 6 and the second was on June 20.

Team GF GA SoGF SoGA S% Sv% PDO
Chivas USA 18 34 66 110 27.27% 69.09% 96.36
Colorado 26 24 87 90 29.89% 73.33% 103.22
Dallas 27 27 98 109 27.55% 75.23% 102.78
Los Angeles 32 25 106 72 30.19% 65.28% 95.47
Portland 30 18 97 86 30.93% 79.07% 110.00
Salt Lake 33 20 115 94 28.70% 78.72% 107.42
San Jose 21 32 84 105 25.00% 69.52% 94.52
Seattle 22 21 68 77 32.35% 72.73% 105.08
Vancouver 33 28 93 88 35.48% 68.18% 103.67

For those not familiar with PDO, it is a back-of-the-envelope calculation which helps indicate how lucky a team is. Shooting percentages and save percentages in soccer tend to regress to the mean: an extremely low or extremely high value is very, very unlikely to be sustained. By definition, the sum of a league average shooting percentage and a league average save percentage will be 100. Therefore if we add a team’s shooting percentage to its save percentage and see how far it is from 100, we have a fairly good idea how far a team’s percentages are below/above the mean. This number is called a team’s PDO number. The number was invented for hockey but has been applied in soccer as well by advanced statistics mavens in several countries, including this site.

Therefore, teams with high PDOs should expect to fall back to Earth; teams with low PDOs should expect them to rise (though I suspect that teams with extremely high or extremely poor shooting differentials may be more prone to extreme PDOs). The highest PDO in Major League Soccer last year was the Seattle Sounders with 109.12; the lowest was Chivas USA with 87.78.

Since last we spoke Los Angeles (very slightly), Portland, and Salt Lake have moved further away from our magic 100.00 mark, while all other teams have come closer. Take a close look at Portland in particular. If you believe that Donovan Ricketts, a career 72.77% goalkeeper who turned 36 a couple weeks ago[1], is going to keep up his 79.22% this year I would be very much interested in hearing your reasoning. Perhaps you’d like to consult the Seattle Sounders fans convinced that Michael Gspurning’s 79.73% last year was the real deal and that they’re a team which is always going to have a high PDO for whatever reason because magic. (Portland is a good team and even if their PDO regresses they’ll be fine, but Seattle is a lousy team and watching their PDO come back to earth makes me touch myself.)

We should probably give Salt Lake a little slack for high PDO, though only a little. They’re on a +1.000 SoG/90 differential so far this year and spend a fair bit of time on the wrong end of score effects, which will cause a bump in their PDO. Salt Lake is in an argument with Los Angeles for “scariest damned team in the Western Conference” (the Galaxy’s SoG/90 differential? +1.619! I’m expecting them to start whipping teams like they did Vancouver this past weekend any day now. Carlo Cudicini is really not working out for them at all).

Since this is a Vancouver-based blog, let’s give the Whitecaps special attention. Vancouver went from a 95.24 PDO on June 20 to a 103.67 PDO today. Not coincidentally, they’ve been winning a lot of games in that time. Now, in the big picture 103.67 isn’t too far out of true: the Whitecaps’ current position is nearly fair. Most of their PDO increase has come through raising their save percentage from “low” to “about average”, which is fine. But watch that shooting percentage: 33.33% is the highest shooting percentage in the Western Conference. Maybe not today, maybe not tomorrow, but eventually fewer Whitecaps shots are going to go in and we’re going to see a rehash of annoying articles asking why none of Vancouver’s forwards can finish like we had late last year and early this one.

(notes and comments…)