Tuesday 2 July 2013

The Difficulty of the Double Save.

Just after the turning of the new millennium, Stoke found themselves striving to escape from the third tier of English football, preferably by way of promotion. Therefore, visits from the likes of Liverpool required pre-season largess or a favourable cup draw and both occasions transpired in the 2000-01 season. Of the two games, the November meeting at the Britannia stadium in the League cup proved to be the most memorable. Unfortunately, for all the wrong reasons from the Stoke City viewpoint.

An 8-0 defeat was scant reward for a Stoke team that had struck the visitor's post after barely a minute and such was Liverpool's displeasure at a narrow pre-season defeat, that the Merseysiders were still pouring forward with the game well won. There is no mercy rule in football, so an inept Stoke offside trap ensured more goals and Stoke's misery was further compounded when a confused City scoreboard operator awarded the Premiership team an extra ninth goal for effort.

It is perhaps a measure of Liverpool's fall from the top table that, along with Arsenal from the elite, that they have found Stoke a much tougher nut to crack in the Potters' Premiership career. Anfield wins have been relatively rare and outside of the League cup, Liverpool have yet to redress any of Stoke's three home wins with an away victory of their own. Stoke Liverpool matches have followed the now discarded Pulis blueprint to the extreme. Little Stoke possession, almost always outshot, defence to the fore, but seldom defeated.

City's 1-0 win at the Britannia in 2011-12 was typical fare. Stoke led when referee Clattenburg decided that Jamie Carragher rather than gravity had sent John Walters to ground in the penalty area and then the home side survived a shot differential of nearly 20-2 to claim three points.

One particularly noteworthy shot sequence occurred after 61 minutes. Stoke were caught exposed by an Enrique long ball, leaving Henderson a free run at Begovic in the home goal. Three shots, two saves and a Shawcross block later, the ball broke to Adam. Two shots, one block and another save later and Stoke were spared the consequences of their uncharacteristic attacking ambition.

The Reds had used up a quarter of their shots in barely ten seconds of continuous play and to the two sets of watching fans the sequence of play had been either hugely frustrating or nerve wracking, roughly in equal measure.

The shot sequence invites many questions surrounding the use of goal attempts when analysing games and matches, but the most obvious revolves around the increased difficulty faced by keepers as they attempt to keep out secondary efforts. Certainly the ovation that Begovic and Shawcross received following their efforts demonstrated that the Stoke fans recognized a magnificent feat of defensive football.

Intuitively, a keeper should be best prepared to face an initial shot, but less so if another attempt rapidly follows on. A keeper would appear to have limited control over where he palms away a shot (or he would never provide an opposing forward with a follow up opportunity), so he is likely to be away from his ideal saving position and often recovering from having to go to ground when attempting to counter a rapidfire , second attempt.

To test the theory I used the familiar shot location model to compared the actual outcome of goal attempts, whilst including a term to signify if the shot was an initial attempt or followed on almost immediately from a save by the keeper.

Once again the added factor does appear to significantly alter the likelihood of success. In my dataset comprising 1,000's of normal shots and 100's of rebounded efforts, if the keeper has been involved in the previous save, the striker has around a 30% greater chance of scoring from the rebound compared to an identical "normal" attempt. This figure may not hold in larger samples, but an advantage seems likely to persist.

Secondly, it would appear that the wholehearted involvement of the goalkeeper is required to elevate the conversion probability of the second chance. Rebounded efforts that result from a defender blocking the initial chance don't show a statistical difference from normal, first shots. A blocking defender appears to allow the keeper much more time to set and prepare himself for any follow up shot.

So this, potentially adds another layer of complexity to a location based shot model. It also adds to the ability to more accurately evaluate a keeper. For example a keeper who benefits from a defence that is adept at clearing or blocking the secondary opportunities that he provides for the opposing strikers may not be similarly lucky with his defensive colleagues should he change clubs. We may also be able to better appreciate a striker's skill (if it exists) at anticipating rebounds and bagging himself numerous gilt edged chances along the way. Anticipation, while others merely look on may set some strikers apart from teammates.

Therefore, it would appear that a double save is significantly more than the sum of the individual parts, a fact already appreciated by fans worldwide. For the best example of such a skill, showcased on the then biggest domestic stage of the FA Cup final, check out the legendary Jim Montgomery moment against Trevor Cherry and Peter Lorimer from 1973.

No comments:

Post a Comment