Finishing and Hitting the Target in the MLS.

The first attempt I made at looking beyond the commonly available football stats of the day, namely goals, used shot and save data from the MLS. The quality of play may not have quite matched that seen in the Premiership of the day, but the amount of data available far outstripped that that was commonly available to the UK newsgroups that preceded blogging.

Attempting to tease the luck from the talent in the shot saving percentages seen in the likes of Kevin Hartman, Tony Meola, Joe Cannon and Tim Howard was a lot easier than trying to sensibly argue who England's current stopper should be. So a belated h/t to Big Soccer, where around half a dozen stat enthusiasts hung out in the dim distant past.

A recent tweet from the influential Steve Fenn, a must follow at @SoccerStatHunt, reminded me of the excellent work that is being done by the guys at http://americansocceranalysis.wordpress.com/ notably, Harrison Crow (@Harrison_Crow). They are collecting and also sharing shot data in the current MLS. So a major h/t to them, the first attribute is fairly common, but the second is extremely rare and most welcome!

The availability of data is the major bottleneck is blog based analysis. Methodologies are fairly standard, but weight and credence to any conclusions only comes with increased sample size. It is fairly easy to develop a novel methodology, but the limited data can still make you look dumb.

Back in the day, shot attempts and outcome was the limit of the data, but the the volume of the data, stretching over seasons and, in the case of keepers, their longevity, still made analysis possible, if with a slightly wider error bar attached. Increased shot volume, it was hoped would even out issues of shot and chance quality, that did not exist to such as degree in either the controlled pitcher/batter contest in baseball or the more restricted playing area of hockey.

 I don't have an MLS photo. Instead here's Clint Dempsey celebrating Sounders' Interest (and a Goal against Stoke).

Nowadays, the still flawed gold standard from blogging shot analysis is data with x,y co ordinates, but often devoid of even the tiniest hint of defensive pressure, except in the most dedicated of collectors. Which is why the MLS data dump at American Soccer Analysis is so welcome. It improves greatly on shot data of the past by partly bridging the gap to professionally collected and protected data with the subdivision of shots into zones. Usually, slicing and dicing sample size leads to noise and over fitting, but ASA's venture may sacrifice sample size, but greatly increase uniformity of events within those smaller samples.

Applying one of my shooting analysis methods to ASA's improved data was therefore both sensible and a nostalgic treat. Broadly, this method assumes that shot outcome is common to each MLS team and centered around the league average. Any apparent deviation in shot accuracy percentage or conversion (and there is bound to be some) is going to be down to random variation and a talent gap in performing these tasks between sides. Quality of opportunity is hopefully controlled by ASA's use of shooting zones. So if we see a wider range of outcomes in the attempts each side made, compared to a random draw using league averages, we can possibly conclude that random variation isn't the only factor at work in deciding the shooting pecking order.

The sectors used along with the data are all available at ASA's site, so I urge everyone to seek it out there, but for partial clarity the sector descriptions are sector's 1,2,4 and 5 are central to the goal and more distant with increasing number and sector 3 is wide within the area and sector 6 is wide to the flanks.

I have taken shooting data from the site for every game played by every side in 2013 and compared the spread in accuracy (in terms of shots that require a save), conversion rates (goals scored) and the undesirable ability to see shots blocked that was recorded by each side against the type of spread expected from those shot numbers if team talent was universally the same in each sector and variation of outcome was purely luck driven.

Do Sector Outcomes Suggest Factors Other Than Random Variation are at Play in the MLS?

Sector taken from American Soccer Analysis Site.	Does Accuracy Deviate from Random?	Does Conversion Rate Deviate from Random?	Does Avoiding Blocked Shots Deviate from Random?
1	Yes	Yes	Barely.
2	Strongly	V Strongly	Random.
3	Yes	Yes	Random
4	Yes	Random	Yes
5	Random	Random	Random
6	Yes	Random	Random

The results are tabulated above. Using shot data from 2013, there does appear to be some evidence that team conversion rates may show a talent differential when strikers are closest to goal. As attempts move further from goal (in the case of zones 4 and 5) and much wider out to the flanks (in case 6), that differential appears to disappear and outcomes become consistent with the average overall conversion rate for the  MLS. In short, skill may exist inside the box, but outside you're hoping to get lucky....in the MLS at least.

A talent for greater (or lesser) shooting accuracy as measured by an attempt requiring a save appears to survive to greater distances and angles or it may show a tactical approach whereby a side is required to "make the keeper work" in expectation of a follow up rebound....Or everything may be the result of insufficient detail contained in the current, admirable data.

I know very little about the specifics of the current MLS, other than Dallas produce technical adept players and Seattle has the coolest kit, but others may make sense of Philly being the best opportunity corrected finishers in sector 1( closet to the goal) and Portland the most efficient in sector 2.

Random variation is ever present in the data, but recourse to this concept as a catch all when a side over or under performs against the league norm, may be less (or more) than fair to player and coaches alike, especially in the absence of any evidence that the talent gap at the very top level has disappeared completely.

To reiterate here's the link to American Soccer Analysis.

Predicting Penalties.

Nothing illustrates the seemingly random nature of penalty awards than the experience of Stoke over the last month when they have been awarded and have conceded precisely zero spot kicks, but could have possibly accumulated nearly half their recent average seasonal total in each category over that timescale.

The first non decision came early in the match against Fulham as Ireland was felled yards from goal by a clumsy Hangeland lunge from behind. More a penalty than not in my unbiased eyes, not so in the judgement of Roger East. An opinion backed up by the innumerate, self styled refereeing expert, Graham Poll, who cited Ireland's "theatrical" fall as justification for East turning down appeal. A clear case of artistic impression nullifying a 80% goal scoring chance.

Jonathan Walters was then allowed the benefit of the advantage rule to ride an Amorebieta foul outside the box, only to be fouled a second time inside the area. Having blown for the second foul, East then convened an NFL style brains trust meeting of officialdom before returning to the spot of the first offence.

Almost inevitably East then denied Fulham a clear spot kick, when Stoke's Marc Wilson took the easier option of kicking the opposing forward rather than the ball. Minutes later East blew for halftime, mentally plotting his route to London Road, Peterborough, the reward for his eccentric application of the game's laws.

Following the now obligatory complaint to the refereeing assessors, Stoke were rewarded for their unfortunate Craven Cottage experience with the appointment of World Cup final referee, Howard Webb for their following home match with WBA.

The Baggies have long been Stoke's punching bag and the Potters have amassed win after win over their near rivals. But on this occasion, WBA more than held their own, especially in one second half period when Youssouf Mulumbu powered into the box in a move reminiscent of Ireland's dart against Fulham. As with Hangeland, Mulumbu's pursuer, the fleet footed Charlie Adam chose to end the danger by raking his studs down the calf of the opponent.

Mulumbu's fall was much less theatrical than Ireland's, although the contact was comparable. Adam flung his arms wide in an expression that footballers take to signify innocence, but is interpreted by everyone else as a cast iron admission of guilt. In short, the incident passed every criteria for a penalty to be given, but in the October spirit of non application of the Laws, Webb awarded Stoke a goal kick instead.

Webb cannily apologized to WBA after the goalless match, citing being un-sighted as mitigation, hanging out to dry his nearside assistant, but avoiding a potential visit to the wilds of Huish Park, rather than the plusher surroundings of Stamford Bridge.

Four, cherry picked incidents, all of which would result in a penalty kick more often than not, but on these particular, actual occasions, none were awarded.

Penalties are extremely valuable events in a low scoring sport, such as football. Their very award has a longterm expectation of nearly eight tenths of a goal, it is then up to either the taker to drive the favourable probabilities to 1.0 or the keeper to redeem a flawed teammate. They are also fairly rare events, a team rarely manages to receive or concede double figure spot kicks and occasionally the penalty spot remains unused by a side over an entire 38 game season.

Incidents of award or non award are rarely as clear cut as the compromised decision making that has followed Stoke around during October, but the often used phrase of "that would have been a free kick anywhere else on the pitch" is not entirely without merit. An official, I would imagine is naturally reluctant to exert as much influence on a match result as a penalty award does unless he is confident of his decision. Making a decision that merely restarts play in a non dangerous area is far removed from making one that usually, directly and visibly, contributes to the final result.

Penalties, therefore have all the ingredients of being extremely difficult to predict. They are rare, contentious and important decisions. The forward correlation from one season to the next is very poor. Tottenham, complete with the game's biggest "flop" can go 38 EPL matches in 2012/13 without being awarded a penalty, but then make penalty goals a primary method of acquiring points in the following season, where six points have come directly from successfully converted kicks so far in 2013/14.

The causes for Spurs' barren spell in 2012/13 may (or may not) have been partly particular and unique to that season. The media perception of Gareth Bale as a diver is as simplistic and as lacking in nuance as is the current branding of him as a flop by Spain's finest. But the widely reported perception may have influenced refereeing decisions when he was apparently fouled inside the box during his last season at the Lane.

Similarly, in the more usual case of Stoke, if they continue to draw or commit reckless and clumsy challenges inside the box, their penalty ledger will start to see entries both on the credit and debit side of the book, although it may take longer than 38 matches to be seen in their raw penalty record.

In short, a single season is insufficient for luck, bias, incompetence, competence, excellence or play acting on an Olivier-like scale to even out and consequently the number of penalties awarded in year N is a very unreliable indicator of likely penalties next term.

Stoke and Spurs do their best to combine for a near penalty kick.

However, lack of correlation between outcomes across the seasons does not mean that there is no strong correlation, merely that the timescale and the rarity of the events prevents that correlation from showing itself. Broadly speaking, the better attacking teams get more penalties and they attain those levels mainly, but not exclusively by possessing the ball inside the box. All four of the cited incidents of non penalties in this post saw the man in possession getting kicked off the ball by a defending player.

Therefore, whenever an attacking player has the ball inside the box, there is potential for a spot kick should the defender be unkindly and the referee kindly disposed. These events are much more numerous than penalty kicks. The best sides touch the ball over a thousand times inside the box over a season compared to around 600 for the Premiership's perennial struggling teams. Any penalties awarded are likely to be a random draw from the individual interpretation of potential incidents that may occur where a side has possession in the box and a defender is trying to deprive them of the ball.

Creating opportunities to induce a lunge from an opponent in the box is therefore a much more common event than an actual penalty kick award. How often the attacking team cashes out that lottery ticket depends on the ebb and flow of a multitude of different factors over a single season ranging from the timing and venue of the event, to an official's desire for more diverse and interesting travel opportunities. But season on season the same type of sides "buy " more tickets by their repeatable level of possession of the ball inside the box. Even if their yearly penalty kick return doesn't always fully reflect their "outlay" in the short term.

In short, penalties don't correlate well over seasons, but the underlying cause of why they are awarded (using touches inside the box as a decent proxy) probably do. So if we want a guide as to how many penalties a side will get to take next season, looking at this season's final tally is not the best place to start. The relationship between touches inside the box and penalty kicks awarded provides an avenue to produce the expected number of spot kicks a side might have won, devoid of particular, situational and largely unrepeatable, random factors, such as minor aberrations from either player or official (also known as the reasons we avidly watch sport).

Data is now an even rarer resource than was previously the case with the enforced closure of many re-sellers, so this study is partly compromised by lack of numbers. But the 1000+ touches inside the box by Arsenal and Manchester United in 2011/12 implied expected penalty counts of around six each rather than the recorded respective totals of 3 and 11. In 2012/13 their respective totals were 6 and 7. Overall a penalty prediction based on expected figures from touches in the box gave better N+1 estimates than actual figures in around 75% of cases, although, quite naturally neither actual or expected penalty totals for Spurs in 2011/12 came close to predicting their duck egg in 2012/13.

Overall though, penalties represent another case where in looking to predict rare events, the more numerous cause or causes of the required outcome and the relationship between the two is almost always the best point of reference, rather than the actual event frequency itself.



  

England Qualify For Brazil 2014. How Hodgson Defied The Odds.

Waldemar Fornalik and Roy Hodgson led their respective teams out at Wembley on Tuesday night at the conclusion of the 2014 World Cup qualifying campaign. Of the two, Fornalik, looked relaxed and at ease and was almost certain to be sacked regardless of the result against England. Roy Hodgson knew that only a win against Poland would prevent a unwelcome visit to the November playoffs before the trip to Brazil could be confirmed.

The odds had been stacked in favour of Hodgson reaching game ten in a relatively comfortable position at the initial drawing of the 2014 qualifying groups. Despite this kind draw, the anguish etched on his face was evident, as forward after forward snatched shots wide of a gaping goal in the opening half. Jan Tomaszewski sat in the stands and the specter of 1973 hung over a nervy opening half hour.

Back in the more relaxing surroundings of Rio in July of 2011, an out of place France in pot two had been avoided during the initial draw for the nine qualifying groups and Group H had pitched England in with four lowly ranked teams from pots three to six, with only Montenegro, ranked second to France in pot two, appearing to represent a less than ideal group rival.

The random element in drawing qualifying groups, often years before the final match is played has been excellently documented by Simon Gleave here and the mechanics of the procedure illustrate how arbitrary group makeup can become. Poland, England's final group opponents were one ranking place away from being in pot 5 rather than pot 4. In addition as Simon points out, Ukraine may have started the draw process as FIFA ranked 45 and a place in the bottom two of pot 3, but they ended the campaign, over two years later ranked 20th in FIFA and a seeded playoff side.

So the fluid nature of national sides can see the protracted timescale of a qualifying campaign transform the strength and difficulty of each group.

The pained, first half expressions on Hodgson's face were possibly down to an appreciation of the chances of a favourable outcome materializing on the night. A 73% chance of an England victory was a huge comfort blanket against a national expectation that tends to think in more stark outcomes, where defeat (or a draw) is virtually inconceivable, but that still left ample room for a two game shootout in November if randomness drew 74 or above.

Even greater levels of  World Cup "certainty" were overturned at Wembley in 1973. Coincidentally, and largely irrelevantly, also against Poland, as the product of numerous, low quality chances, repelled by a massed defence and marshaled by a future Polish politician, could only equal the goal count from one, clearer cut Jan Domarski shot on the rapid counter.

Certainty of outcome, as Hodgson may have reflected, only arrives after 93 minutes of actual play, no matter how much ability and national entitlement is bought to the table. Somewhere in the mass of possible, probable and improbably outcomes on Tuesday, lurked a quickfire Polish break that ended with a shot that squirmed under a largely untroubled English goalkeeper....but on this occasion, under the full glare of live, HD TV, the ball rolled uncomfortably wide of the target.

Defying the odds is usually expressed as being exclusively the preserve of the underdog, but every result defies some outcome, even if it is "only" the 27% chance that Tuesday evening's result would send Hodgson's England into an unwanted two legged playoff. In topping Group H, England turned the combined uncertainty and myriad of possible combinations of results from thirty qualifying matches played across 13 months into the certainty of confirmed qualification.

It is difficult to make an intuitive guess at the likely true size of the task faced by each team at the outset of qualification. The average FIFA ranking of England's opponents on the day of the draw was 84 and this had been pushed out to 89 by last week, as Montenegro fell and Ukraine rose. The task faced by England ranged from the near certainty of victory at home to San Marino, up to barely deserved favouritism on the road in Ukraine.

The final margin of group victory, a single point, hardly denoted a simple, alarm free passage.

Simulating the Group H qualifying matches, using the relative abilities of each side at the time of each individual match, allows us to record the range and frequency of both finishing position and points totals that may have been achieved by each team. With it comes a better understanding of the odds Hodgson's England defied to claim their spot in Brazil.

The 22 points actually gained by England was their second most likely points haul (~10% of the time), behind 23 (14%). 75% of the time, the 22 points were sufficient to send England straight to Brazil, either as outright winners or winners on goal difference. 20% of the time they fell into the playoffs when amassing 22 points. Once every 60 occurrences, 22 points was insufficient to place them any higher than 3rd place in the group and their World Cup campaign ended, as in 1973 with a final match against Poland. It is only when 25 points are breached that the need for a playoff begins to entirely disappear. A perfect 30 points certainly guarantees qualification, but England only met that target once in every 50 or so of the simulated leagues.

The lowest ebb for England was a fifth place finish in the six team group, occurring once every 200 or so iterations with a points total of around 13. It is a sobering thought for all aspiring national team coaches, that you can be demonstrably the best side in the group, favourite in all 10 of your matches and still, on very rare occasions, finish below Montenegro, Ukraine, Moldova and Poland.

No amount of iterations could force San Marino above England. 21 minutes of barely believable supremacy against the founders of the game in Bologna seems to be as good as it will get for FIFA's perennial minnows and England's national footballing pride was maintained, at least in the virtual world of number crunching and spreadsheets.

Possible Finishing Positions For England From Simulations of Group H.

Position in Group H.	Frequency per 100.
Top.	61
2nd.	22
3rd.	12
4th.	4.5
5th.	0.5
6th.	0

The bottom line, for a nation obsessed with World Cup qualification is the 61% figure, denoting the likelihood that England might turn a favourable draw into a top of the table finish. A further 22% of the time England fell into the playoffs and as one of the likely seeded sides, they would expect to progress to the finals by this route well in excess of 50% of the time. So as a conservative estimate, Hodgson has just completed a campaign where his chances of successfully negotiating a route to the 2014 finals could be put at around 75% prior to a ball being kicked.

It may not have felt like it on Tuesday night, as Hodgson waited for Rooney to temporarily discarded his padded headband before heading England into the lead, but his side went into the final match with an improved chance of qualifying compared to their chances at the outset. Their 73% chance of winning the match, combined with a 27% chance of entering the playoffs if victory wasn't achieved, equated to a near 90% chance of appearing in Brazil via the route of either automatic or playoff progression.

The possibility of a hoped for outcome had increased, but the anxiety had been compressed into one single match, rather than spread over a 13 month campaign.

Poland, as widely expected sacked Fornalik and his coaching staff following their 4th place finish and a respectable 13 point haul. A 3rd or 4th place finish was almost equally their most likely outcome in the group. Fornalik's side had a near 70% chance of occupying one of those positions in the final group table, but they did fall 4 points below their most likely tally of 17 league points.

Where Hodgson had inexorably, if at times unconvincingly, dragged his team over the line, Fornalik failed to build on his side's meagre 10% chance of qualifying and with it went their World Cup hopes, his job, those of his collegues and judging by his demeanor on Tuesday night, the ever present pressure of turning possibles into probables and finally into certainties.