The Week in Sports II

 A pulsating first test and the flaws in DRS.

As a massive cricket fan this weeks viewing has been dominated by the first Test of the Ashes. It wasn’t always cricket of the highest quality but, in terms of drama, entertainment and tension, it was one of the great test matches of all time, and in the outstanding debut performance of Ashton Agar, it had a story line that would have been thrown out of a Hollywood studio for being less believable than Sharknado. It had everything you could want in a cricket match; from records being broken, to devastating bowling and belligerent batting, from dodgy umpiring and outraged controversy to the arrival of a star in the making. 

Ashton Agar's performance was incredible. To perform that well from number 11 was outrageous, to do it on debut was inconceivable. The audacity and nervelessness displayed were hallmarks of youth. His shot selection and judgement of length an indicator of serious talent. Never before has an English crowd wanted an Aussie to score a century but there were groans of disappointment when he was caught on the boundary for 98. His was the highest score by a No.11 in the history of Test cricket and it was part of the highest ever tenth wicket partnership in Test cricket, ably supported by Phil Hughes. Agar also had the skill to take a couple of wickets including the England captain, Alistair Cook. A great start for a great talent.

The world's best?

His performance was only topped by the brilliance of England's best player, Jimmy Anderson. 10 wickets in the match on a pitch designed to suit the spin of Graeme Swann does not quite tell the full story. The 13 over spell to win England the game on the fifth morning was nothing short of heroic, his abilities highlighted by the ease with which Australia scored off England's other bowlers. Special mention must also go to the high quality performances of Peter Siddle and Ian Bell, both players who have previously been much maligned in the media. Siddle is commonly dismissed as an 'honest trier,' completely overlooking his accuracy and skill. After his 8 wickets in the match, maybe a little more analysis will be applied to his skill and a little less on his effort. By contrast Bell is often criticised for only scoring stylish but easy runs against poor sides. His gritty and determined century in the second innings put England in a position to win the game.  It was without doubt a brilliant match, full of the unexpected, and, when I was watching on TV, it was brought to me by Sky.[1]

It is an accepted wisdom that TV has influenced all sport in thousands of ways since its arrival but in none is that influence so marked as in cricket. Since they took over the summer Test matches from Channel 4 in 2005, Sky has had a TV monopoly on English cricket. As in many sports, fixtures are now timed for maximum TV viewing rather than the enjoyment of fans actually attending the match. Cricket tours are shorter, (why play warm up matches for the benefit of quality Test cricket when you can play another lucrative and pointless T20 match?) and often played at ridiculous times of the year to fulfill the contractually obliged quotas demanded by Sky, (why else would you choose to play Test matches in the rain of April and May?)

Nerdy types, otherwise known as physicists, have long postulated that in quantum mechanics the nature of something changes when it is observed.[2] Cricket has certainly changed since the advent of Sky’s coverage.  Technology such as Hawkeye and Hotspot that were originally introduced by Channel 4 and Sky for viewer enjoyment have now been incorporated by cricket in DRS (decision review system).  Specifically designed to reduce controversy by getting more decisions correct, it would be reasonable to argue that in this Test the DRS has caused the controversy.

The technical issues all started with the dismissal of Joe Root in England’s second innings. Joe Root was strangled down the leg side with ball seeming to lightly brush his bat. After chatting with Alistair Cook, Root decided not to review the decision and walked off. This brought Jonathon Trott to the crease and he was promptly given out LBW first ball to Mitchell Starc. Trott decided to use the DRS and review the decision. All the front-on camera angles appeared to be inconclusive as to whether he had hit the ball. The crucial angle would the side-on hotspot camera. Unfortunately there was no footage from that camera as it was being rewound at the time to show that they couldn’t find any evidence of Joe Root hitting the ball and that if he had reviewed, the decision would have been overturned and he would of stayed in. As the Hotspot camera and machine work in a similar way to an old VHS, it cannot both record live footage and rewind at the same time.[3] Without Hotspot to save him Trott was confirmed out.

This incident shows more than a few flaws within the DRS system. The first problem was when Joe Root walked off. It is widely acknowledged by former players of the game that ‘you know when you have hit it.’ If Joe Root had not hit it he would certainly have reviewed the decision immediately. That suggests that he knew that he had hit the ball thus explaining why he walked. So why didn’t the Hotspot camera show the nick? This should put serious doubt on Hotspot’s ability to detect faint edges, thus rendering it almost useless as a part of the DRS.

A secondary problem is that the laws of cricket give the benefit of the doubt to the batsman. Without the Hotspot camera in his review, Trott was denied that benefit. Instead the 3^rd Umpire confirmed up the original decision of the umpire and gave him out. This brought two separate laws into conflict with each other, an obvious problem. Trott was rightly furious that he had not had all of the DRS at his disposal and controversy reigned.

In England’s second innings Ian Bell and Stuart Broad put on a major partnership that would give England a commanding lead. Australia had wasted their reviews on two dubious LBW shouts earlier in the innings and were therefore out of reviews. This turned out to be crucial when Broad blatantly edged Agar to Michael Clarke at slip. How the umpire missed it is beyond my comprehension, but miss it he did. Whether Broad should have walked is another argument but it brought into focus two problems; the reliability of the Hotspot Camera and  whether two reviews is enough or too many. Once again Hotspot had failed to spot an edge behind and if Root’s had been a faint nick, Broad’s was a fat edge. Though the decision couldn’t be reviewed, a riot would have ensued if he had been given out and it had been overturned on DRS due to the evidence of Hotspot.

Originally the system was brought in do away with howlers such as this. Some, including the BBC’s cricket correspondent Jonathan Agnew, argue that by giving teams two reviews, it encouraged teams to gamble with them rather than saving them for overturning obviously bad calls. Others argue that Australia should have had more reviews to use and this would have prevented the controversy from occurring. Both theories have their merit so whilst there is no consensus this flawed system will remain unchanged.

Rev counter is... IN THE RED!

Sky’s newest gizmo, the Rev Counter, which debuted in this match appears to be totally pointless except as a way to elicit thoughtless commentary  from the FECRC (Former England Captain’s Retirement Committee) when it poops up on the screen.[4] Such delights included constant exhortations of “Its in the red!” and Andrew Strauss ridiculously questioning whether Graeme Swann would forget about bowling normally and try to just spin the ball as much as possible in competition with himself. In all, I am pretty sure this adds nothing to the game and Sky may have agreed with me as, by the end of the match, the rev counter graphic was half the size it had started at.

Without any doubt there are definite problems with the DRS system as it stands. Though flawed its defenders correctly argue that more decisions are now being given correctly. Its critics counter by pointing out its deficiencies. In the end it will all come down to viewer entertainment and it is hard to argue that Hotspot and Hawkeye, despite their flaws, do not enhance the viewing experience. The very act of observing cricket through the medium of Sky has changed the game. In itself this is not a worry as sports evolve all the time. The worrying thing is that it’s now changed the game to such an extent that it’s affecting the outcome of the game.

---

[1] Other, no doubt better commentated, channels are available in other countries… The BBC does great radio coverage on TMS.

[2] The guy who made this theory, Heisenberg, would have been a great fan of the DRS system I feel. He would also probably have loved the complexity of the Duckworth Lewis method as well.

[3] Nerdy readers will note that Heisenberg’s principles of quantum mechanics argues that something can be in 2 places at once due to the uncertainty principle. In this case the DRS was also in two places at once.

[4] I didn’t mean pops. I meant poops.

'Moneyball' in Cricket?

On the day that the Ashes starts I thought it was about time to do a post on Cricket. I will be writing my weekly round up of the week's sports tomorrow and will be including some thoughts more directly related to the Ashes there. 

Billy Beane. Doesn't look like Brad Pitt.

The rise of advanced statistics in cricket has been a long time in coming. Similar in many ways to cricket, baseball has been at the forefront of statistical analysis in sport since the mid 1990s. Made famous by the film ‘Moneyball’ starring Brad Pitt, baseball had its statistical revolution, known as sabermetrics, under the guidance of Sandy Alderson and Billy Beane, the general managers of the Oakland Athletics. Sabermetric principles focused on analyzing players using objective evidence measured from in-game activity. By recording every action that each player took in each game, Alderson and Beane were able to create either entirely new statistics to rate players or use the existing statistics alongside new qualifiers to give a more accurate picture.[1] In turn this allowed them to build up a far more detailed picture of which players were the most effective to a team’s success then their competitors. The principles of research and analysis allowed Beane and the Athletics to remain relevant for long periods despite having one of the lowest payrolls in baseball.[2] Given the similarities between the two games, why hasn’t cricket followed the same trend until recently and where might we begin to see similar analysis begin to creep into cricket?

For many years Cricket had been dominated by the most simplistic statistical thinking; Averages and strike rate (and additionally for bowlers – economy rate) were the only statistics taken into any sort of consideration. The problem with these statistics is that they take no account of game situation, pitch conditions, overhead conditions, quality of opposition and type of opposition (i.e., left arm spin, right arm seam etc). For example, whilst Kevin Pietersen averages a quality 49 in test cricket, his average against left arm spin is a modest 38. Stuart Broad has an overall bowling average of 31.93, which improves to 27.51 in England but balloons to 43 in Asia (including two tests against enthusiastic minnows Bangladesh). Would it not be sensible to consider picking other another batsman who doesn’t exhibit this weakness when facing quality left arm spin bowling.[3] Broad’s has struggled in the sub continent for a while now and a large enough sample size in evidence to suggest that an alternative should be found by England. Even these simple manipulations of the statistics seem to be beyond England’s selectors who seem to prefer a rigid team selection to a more squad-based system where players are picked according to their various strengths.

Alan Wells is dismissed by Curtley Ambrose.
Is it better to be picked once and dropped than
never to be picked at all?

Until recently even these basics statistics were often ignored if an international selector managed to watch a player put in a stellar performance.[4] Small sample sizes were ignored and any player on a decent run of early season form was considered for the England team with no thought given to temperament, conditions and quality of opposition. Despite players like Vaughan and Trescothick being successful additions to the England squad, a policy of picking players purely on the basis of recent good individual performances, rather than long-term weight of achievement, largely led to the shambles that was English Cricket circa 1990-2000. One cap wonders like Mike Smith, Gavin Hamilton, Simon Brown, Joey Benjamin, Neil Williams and Alan Wells were all unlucky to only be given one chance, (or maybe they were just lucky to get a chance at all), though none endured the humiliation that Ian Blackwell suffered in being dropped after his only test match for the truly inept… Liam Plunkett.[5] All were unfortunate victims of the horrific lack of consistency endemic in England’s selection; a policy that closely resembled teaching kids to swim by just chucking them into a pool and allowing those not naturally gifted to drown.

Only the arrival of Nasser Hussain, Duncan Fletcher and the new standard of professionalism that they brought with them saved English cricket from the inconsistent selection that had plagued it for so long. Alongside the consistency required to build a strong team, Fletcher, Hussain and, later, Vaughan used technology to help England improve and analyze their game. The rise in professionalism and analysis has coincided with a rise in the England team’s fortunes. England has sports analysts, Nathan Leamon for tests and Gemma Broad for ODIs, whose sole role is analyze data and come up with plans to combat opposition players.

Jimmy Anderson and Steve Harmison recently spoke on the Tuffers and Vaughan radio show and highlighted the advanced use of specific plans to individual batsmen that they faced. When specifically questioned by Mark Chapman as to how he would get out Ramnaresh Sarwan, for example, Anderson responded that Sarwan is LBW candidate early on and that he would try to “run on back into him.” To AB De Villiers the plan would be to “make him play with a straight bat” as he scores heavily with cross bat shots. 

His answers show that plans to get out different batsmen are created in two ways, firstly through analyzing any technical deficiency, (in Sarwan’s case above that he gets his front foot too far across early on and ends up playing round his pad) and secondly through Hawkeye pitch data and a batsman’s average when facing balls pitched in certain areas. De Villiers scores heavily when facing anything short pitched and so the plan is always to keep it up to the bat to force him to play straight. It must be noted that De Villiers is such a quality batsman that Anderson’s method of bowling to him is more of a way of restricting his scoring rather than targeting a weakness.

An example of the Hawkeye pitch map. This one appears to to show
Zaheer Khan's left arm seam bowling in a particular match to right
handed batsmen from both over and around the wicket. Whilst
conceding fewer runs going over the wicket, Khan has been more
successful going around the wicket as he has taken two wickets
shown by the 2 white dots.

The use of Hawkeye and similar ball tracking technologies is at the heart of the new advance in analytics in cricket. By being able to track the flight and pitch of every ball bowled in world cricket, statisticians are able to record the strike rate and average of each batsman in different pitching areas and finishing points.[6] For example analysis of the finishing points of a certain bowler could help Eoin Morgan’s shot selection outside off stump, the pitch map could allow Mitchell Johnson to locate the pitch (just a tad) more regularly and Jimmy Anderson can even end up seeing where his deliveries pitch.

Analysis and research into the game is continuing apace and its only going to accelerate in the future, so where is it headed? I feel that a squad system is more likely to become commonplace. Players will be used in a rotation policy slightly reminiscent of football. This will allow for squad depth in bowling and batting departments to cover for injuries and allow management to pick teams in a slightly more horses-for-courses way. These ideas are commonly used in county cricket with young players picked in short formats of the game to gain experience and experienced players picked in the more important county championship matches.[7] Bowlers will be on limited over counts similar to pitch counts in baseball in order to manage workloads. Other statistical elements will slowly work themselves into the game as captains and coach’s search for even more sophisticated ways of gaining an advantage.

Traditionalists fearful of the total dominance of cricket by analytics should not worry too much though. Cricket is a far more cerebral game than baseball and often a captain’s feel for the game will capture a wicket far more quickly than stubbornly sticking to a statistical plan that may not be working. Making use of the statistics in sensible ways is far harder than their creation, so whilst cricket may be appropriating statistical ideas from baseball the complexity of cricket will make it harder to totally analyze the value of each decision made.  Cricket is so complex that it will never be as comprehensively analyzed as baseball but their is definitely some work that can be done. Digital decision-making may be useful but sometimes a little analogue thinking can get you a wicket much more cheaply. Or, if you are an Australian, you can just punch someone in a bar and hope that will put him off his game.

---

[1] OPS –on-base plus slugging - is an example of using two old statistics combined together to form a newer, more accurate measure of a batters value to the team.

[2] For example in 2006, the Athletics finished with the 5^th best record in Major League Baseball despite having the 24^th lowest payroll of the 30 teams.

[3] For the record, in this case, I think they should consider it, then forget about it. KP is too good to get dropped. But they should definitely be considering things.

[4] Two famous examples where this strategy paid dividends were Marcus Trescothick and Michael Vaughan. Trescothick was famously picked for England after scoring 167 in a low-scoring match at Taunton in front of Duncan Fletcher, soon to be England coach, despite averaging in the low 30s in his career until this point. Trescothick went on to average 43.79 in tests and 37.37 in ODIs as a destructive opening batsman for England. Similarly Michael Vaughan also performed much better for England than Yorkshire with his test average standing at nearly 5 runs better than his first class average (41.44 – 36.95).

[5] Unbelievably JJ Ferris took 13/91 in the match and never played again!

[6] Only 3 of the international teams have statisticians at the moment – India, England and Australia.

[7] England are also following this route with their own T20 side. The large majority of the side is very young with only KP a regular in the test side.