AB de Villiers takes a catch at the boundary to dismiss Lahiru Thirimanne

Does posting top fielders like AB de Villiers on the boundary towards the end of an innings result in more catches?

© Getty Images
7

Comment

Five stats cricket could use

Our game loves its statistics but is it time it borrowed some from other sports? A few suggestions

Peter Della Penna |

1) Tracking fielding errors
A standby commentary cliché when showing a replay of a catch is, "He doesn't drop many." Well, just how many does he drop? One every five matches? One in every 20 chances? Does he have a streak of chances taken? How many matches since his last drop?

In 2010, Casey Kotchman of the Seattle Mariners set a record, going two years and 2379 consecutive fielding chances without an error. It's doubtful that anyone in cricket has the foggiest clue as to who has the longest equivalent streak of clean fielding.

Charting misfields, drops, errors, and the percentage of total chances converted - both on a team and player basis - would give us a more conclusive idea of the best fielders. Tracking catches per chance would tell us if captains are putting their best catchers in areas where they are most likely to make a big impact.

AB de Villiers and Kieron Pollard might field at point or in the covers early in a game but towards the end they are typically sent out to patrol long-on, long-off or deep midwicket. Does this result in more chances taken? How many runs do they save in the deep? And how many chances go down at fine leg or third man because captains are trying to hide weaker fielders in those regions?

2) A modified Earned Run Average
Baseball shrewdly employs the earned run average, or ERA, to show the effect a fielding error has on a pitcher's performance. Any runs scored courtesy of an error are not charged to the pitcher.

The ERA could be modified for cricket. First, a batsman who has been dropped - or has survived a run-out or stumping - would have his runs tabbed after the miss. Over a period of time, these "unearned runs" would put his average in perspective. The stat would also tell us which batsmen offer more chances and which ones make the most of their chances.

More importantly, an ERA can give us an idea of the bowlers who are most punished for poor fielding. Think of a situation where a batsman is dropped off a pace bowler in the first over of a Test in the subcontinent. The bowler sends down four more overs before being taken off. The batsman goes on to smash a spin bowler, who bowls for most of the day, to all parts of the ground.

So even though the pace bowler was denied the wicket, the spinner is more adversely affected statistically: he is being taken for runs by a batsman who should have been out even before he started his spell. An ERA stat (modified for cricket) can track the runs a batsman scores after he has been dropped - both off the bowler who was denied and off other bowlers. This would give us a more rounded view of the day's play and ensure bowlers are not judged solely on the basis of runs conceded and wickets taken.

The ERA is also vital when a team is nine wickets down. For instance, earlier this year Ian Bell dropped Ishant Sharma at The Oval with Ishant on 1 and India 95 for 9. Ishant finished 7 not out, and one might be inclined to say that Bell's miss cost England only six runs. However, MS Dhoni at the other end helped India claw up to 148. Which means that the drop actually cost England 53 runs, all of which should be marked as "unearned". If tabulated consistently, this stat would tell us which teams are profiting and hurting most because of such errors.

Negative impact: if you rated him on net contribution, it could be said that Matt Prior was responsible for England's loss at Lord's to India

Negative impact: if you rated him on net contribution, it could be said that Matt Prior was responsible for England's loss at Lord's to India © Getty Images

3) A plus/minus or net contribution
Plus/minus - a measure of a player's defensive liabilities against offensive output - is an established stat in ice hockey and has gained traction in basketball over the last few years. In cricket, a plus/minus could be an essential evaluation tool for wicketkeepers, particularly in an era where pure glovemen have been cast aside in favour of specialist batsmen who can keep.

In 2013, blogger Hassan Cheema did a superb analysis of Kamran Akmal, the former Pakistan wicketkeeper. Akmal's poor glovework had been a running joke but no one had analysed its true impact. Cheema tabulated Akmal's runs scored versus runs conceded and found he had contributed 4.60 runs per Test between 2006 and 2011. (He calculated this on a match basis rather than an innings basis, because there were Tests where Akmal fielded once and batted twice, or vice versa.)

Earlier this year ESPNcricinfo reader Tony Hastings brought up the proposed stat in relation to Matt Prior's struggles. In the first two Tests against India, Prior conceded 159 runs (46 byes and 113 extra runs off four dropped chances) and scored 40. So his plus/minus or net contribution per Test was 40 minus 159, divided by 2, which is -59.50. Considering India beat England by 95 runs at Lord's, that's a significant impact.

Prior's replacement, Jos Buttler, conceded 45 byes and 15 runs - off a lone missed chance - and scored 200 runs in the next three Tests. So Buttler contributed 46.67 runs per Test. Weighing the two against each other, Buttler's selection resulted in a 106-run positive swing per Test in England's favour. One could argue that if Buttler had played at Lord's, England might have won the series 4-0.

For India, Dhoni scored 349 in five Tests and had the most 50-plus scores of any Indian player. He was the second-highest scorer in the team. However, ball-by-ball commentary and video analysis show that he conceded 210 runs - 56 byes and 154 runs in four missed opportunities (two catches, one stumping and a run-out) - and his net contribution was 27.80 runs per Test, about 20 fewer than Buttler. (This reflects poorly on Dhoni's wicketkeeping, particularly because England batted in only seven out of a possible ten innings in the series.)

4) Distance covered
An increasingly visible statistic in soccer, distance covered is particularly useful for midfielders whose contributions can't be measured by goals and assists alone. It's also useful for assessing a player's hustle in winning back balls in defence, as well as in kickstarting attacks and providing service to strikers.

Other sports are latching on to this. ESPN The Magazine recently asked the men behind Hawk-Eye to develop a distance-based metric for tennis. Some of the results, based on four recent Grand Slams, were instructive - Novak Djokovic covered more distance on the fast courts at the US Open in 2013 than on the clay at Roland Garros this year; Roger Federer returned 70% of serves from inside the baseline (compared to 42% for Djokovic and 5% for Rafael Nadal).

Measuring the distance covered by cricketers would add tremendous value in ODIs and T20s, where players steal singles and twos and where fielders are moved to key catching and run-saving areas at various stages. In some T20s, where more than half the side doesn't bat, this stat can quantify the value of every player on the field. It would be particularly useful when the two teams are inseparable in terms of batting and bowling and when fielding and fitness may tilt the scales.

Virat Kohli may have had a relatively modest strike rate in the World T20, but his high scoring percentage more accurately reflected his value

Virat Kohli may have had a relatively modest strike rate in the World T20, but his high scoring percentage more accurately reflected his value © Getty Images

5) Scoring-shot percentage
Basketball and American football typically apply an efficiency measure, whether it is shooting percentage in basketball or completion percentage for a quarterback. A similar metric can determine a batsman's efficiency.

While a bowler's strike rate tells us how often he takes a wicket, a batsman's strike rate is a curiosity, particularly in T20 matches, where an entire team's innings can last only a little over 100 deliveries. Batting strike rate is also a crude measure because boundaries obscure a batsman's failure to rotate strike.

In the World T20 in March, Virat Kohli was the leading scorer, with 319 runs, but his strike rate, 129.14, put him 27th among players who faced 40 or more deliveries. The most relevant stat, though, was his scoring percentage of 72.5% - the joint second among those who batted in the top five. When he wasn't striking boundaries, Kohli still kept the score ticking and refused to allow bowlers to build pressure with dot balls.

The best example of this was his unbeaten 72 off 44 balls in the semi-final against South Africa. He struck five fours and two sixes, and his strike rate was the third-best in his own team that day, behind Rohit Sharma and Suresh Raina. But Kohli scored off 41 of his 44 deliveries - a stunning 93% scoring percentage. By comparison, Rohit scored off 54% and Raina 60%.

Contrast Kohli with Yuvraj Singh, who scored off 54.9% of the deliveries he faced through the tournament. In the last group match against Australia, Yuvraj struck a 43-ball 60 with five fours and four sixes - an innings that many characterised as a return to form, even though his scoring percentage was a moderate 58%. In the final, when he was vilified for his 21-ball 11, Yuvraj's scoring percentage dipped to 52.38%. In the same match Kumar Sangakkara - another No. 4 who had had a wretched tournament - scored off 27 of the 35 deliveries he faced (77%) to produce a match-winning 52.

Although this stat is more useful for analysing limited-overs performances, there is scope in Test cricket too. In the first England-India Test this year at Trent Bridge, M Vijay scored 146 off 361 balls with 25 fours and a six, and Joe Root scored 154 off 295 balls with 15 fours. But Vijay scored just 27 singles and had an overall scoring-shot percentage of 16%. Root, with 57 singles, had a scoring-shot percentage of 30.5%. Despite hitting far fewer boundaries, Root found more ways to score.

Fast-forward to the fifth Test at The Oval, where boundaries dried up for Vijay. He scored 18 off 64 and 2 off 16. His scoring-shot percentage was about the same, at 17.50%, but it was hard to relieve the pressure because he wasn't rotating the strike. Contrast that with Root's unbeaten 149 off 165 balls in the same Test, where his scoring-shot percentage was an even more robust 41% thanks to nine twos and six threes. In fact, Root's 35 singles were just ten fewer than what India's entire line-up managed in both innings of the hopelessly one-sided Test.

Peter Della Penna is ESPNcricinfo's USA correspondent. @PeterDellaPenna

 

RELATED ARTICLES

 

LOGIN TO POST YOUR COMMENTS

  • POSTED BY Peter Della Penna on | October 30, 2014, 4:36 GMT

    @Kartik Saboo On - Your observations are valid. As the article points out, the stat is better designed for limited overs cricket because of the Test tail-ender scenario you've pointed to, which occurs much less often than limited overs cricket. Even so, that is more of a rare occurrence and over the course of a career the true stat value would balance anomalies like that out. As for your point on the variations according to batting order, this has also been taken into consideration and it is correct to say that scoring shot percentage for T20 cricket in particular is more accurately reflected when analyzing players from batting position to batting position. For example, most openers at the 2014 World T20 had a SS% in the 50-55% range. Meanwhile, players coming in at 6 & 7 tended to show up in the 65-75% range. Within each batting position though, the statistical comparisons are extremely enlightening as to who the more efficient openers and efficient finishers are.

  • POSTED BY Kartik Saboo on | October 29, 2014, 3:55 GMT

    Loved the first 3 ideas!

    However, I have an issue with the usage of Scoring-shot percentage (your final suggestion).

    Whenever a proper batsman is batting with a tail-ender, he/she will prefer owning the strike as-much-as-possible (except for the last ball of the over where he/she would take a single to own the strike next over). Thus, his/her Scoring-shot percentage would decrease whenever he/she is batting with the tail, despite being more responsible in his/her role as a front-line batsman.

    Another drawback of Scoring-shot percentage is when in Test matches an opener tries to see off the new ball and thus, will leave/defend many deliveries. Now if the comparision of Scoring-shot percentage is considered among various openers of different teams, then it would be fair. But when the comparision would be among various batsmen from all batting positions, it would be unfair since an opener will tend to have a lesser Scoring-shot percentage compared to say a No.5 or a No.6.

  • POSTED BY contrast_swing on | October 25, 2014, 9:49 GMT

    Clearly you do understand the nature of statistics and especially cricket statistics. Data and more data is not an answer.

  • POSTED BY BillyBlue on | October 24, 2014, 5:18 GMT

    LOVED the article!!! Being in the USA & an ardent cricket fan, I have wondered about these exact statistics often for cricket, but you sir, have definitely penned them very beautifully! Kudos!!

  • POSTED BY dkrstar on | October 9, 2014, 13:46 GMT

    very interesting thought, some of them could be used.. would be fun then to see who actually is a 'value' player for his team.

  • POSTED BY espncricinfomobile on | October 7, 2014, 22:42 GMT

    ezcellent article / impressive observations

  • POSTED BY espncricinfomobile on | October 7, 2014, 4:35 GMT

    Brilliant and entrepreneurial.