Samir Chopra October 19, 2009

Of fielding and statistics

Fielders do not figure on cricket scoresheets except for when they take catches

It's not too fantastic to imagine a scoresheet, that along with the batsmen and bowlers' names, also records the fielders' names as well © AFP

A few weeks ago, I wrote a little piece suggesting cricket take a leaf out of baseball's book and maintain statistics for fielders. The practical difficulty with this suggestion is that cricket scoresheets do not contain this kind of information: fielders do not figure on scoresheets except for when they take catches. The runouts and boundary saves they make, the catches they drop, the misfields the inflict on their team are all missed.

But for a few years now, a scoresheet has been present which could potentially address this difficulty. I am referring to the Cricinfo ball-by-ball commentary, which currently records brilliant fielding, catches, drops, some misfields (if they are particularly egregious), and sometimes information on the fielder.

Consider the following excerpts from the Ashes:

9.4 Harmison to Ponting, 1 run, oh dear! Huge run-out chance missed there by Ian Bell! Ponting prods to cover and takes off for a single, but it's misjudged and Ponting had given up on making it as Bell's throw bounced over the stumps. Ponting was about three metres out there. Enormous chance missed.


26.5 Broad to Ponting, no run, prodded out to third man - no, brilliant stop from Anderson at gully! He's a lithe and brilliant fielder for a fast bowler.

In this commentary/scoresheet, besides the usual recording of dot balls, runs, batsman and bowler, we have information on the fielders, on what they did or did not do. Thus the Cricinfo ball-by-ball is in fact, an annotated scorer's sheet, which could be used to generate the kind of fielders' statistics I had in mind in the piece linked above.

Of course, the annotations in the Cricinfo commentary are voluntary; they are placed there by the commentators on duty at that time and the level of detail can vary. The commentary still does not record fielders' names when there is no error as in:

25.2 Harmison to Watson, no run, shorter delivery, slapped to point

Here, we do not know who was at point, and thus we have no way of finding out, for instance, whether a particular fielder commits more errors at point and is better placed somewhere else. Adding this information would certainly add to the burdens of the (possibly already overworked) commentator/scorer. But it's not too fantastic to imagine a scoresheet (suitably tweaked to make the commentator's task easier), that along with the batsmen and bowlers' names, also records the fielders' names as well.

How would fielders' statistics be extracted from such a scoresheet? That task would be made considerably easier if the commentary facilitated the use of keywords that would allow for automated processing of the commentary transcript (another requirement would be a form-like entry for fielders for each delivery). Hopefully, such a tweak to the commentary software would not be too involved.

Fielder's statistics for too long have been ignored in cricket. Instead, we are left with a host of entirely subjective statements like "he is worth 30 runs in the field" or "his fielding has declined over the years" and so on. Quantification and recording of fielder's statistics would not only allow for comparison and record-keeping, it would also permit a ranking and recognition system for fielders that is long overdue. Annotated commentaries like the Cricinfo version point the way forward in this regard.

Samir Chopra lives in Brooklyn and teaches Philosophy at the City University of New York. He tweets here

Comments have now been closed for this article

  • testli5504537 on January 27, 2012, 9:22 GMT

    Smart thinking - a clever way of looikng at it.

  • testli5504537 on October 30, 2009, 12:36 GMT

    To all those who THINK Jonty Rhodes was the best (And he was a very good fielder) I have just one thing to say : S O B E R S

  • testli5504537 on October 26, 2009, 17:41 GMT

    Sounds well-intentioned but I am somehow not enamored by this. What ails Cricket is its complexity..

    Imagine teaching an alien what football is - 11 people on either side trying to hit the ball without using legs on a goal.

    Try explaining cricket..It's way too complex a sport that makes it difficult to spread. Why add all these unnecessary complications. We all know Jonty Rhodes was the best..No need to quantify it...

  • testli5504537 on October 25, 2009, 9:05 GMT

    I think you may have a problem with run outs for example. In his recent success in the ODIs against England Ricky Ponting accepted the accolades but ruefully pointed out that in each case he had missed the stumps with an earlier throw. The point is that he and Ian Bell are in that position entirely because of their world class record as fielders known to their team and coaches. They are bound to miss some but they also threaten the batsmen with their lightning pick-ups. The runs they prevent and cut off just by their speed of reaction aren't measured. I remember Fletcher saying he didn't mind a failure at a difficult catch or a near miss throw at the stumps, what he objected to was the fielder who decided not to expose himself by not trying.

  • testli5504537 on October 22, 2009, 8:01 GMT

    Samir, I just happened to watch the semi-final yesterday between NSW and the Vics and saw that they were displaying stats such as 'misfields', 'runs saved', catches dropped and catches for some of the players. Wonder whether it was for the whole tournament or just the match! Anyhowz, looks like Modi is reading ur blog! :)

  • testli5504537 on October 21, 2009, 23:07 GMT

    i understand there are major problems with trying to do stats for fielding.but do we need any? i remember the great fielders and im sure other people remember them aswell [border rhodes m.waugh logie etc} plus the modern greats.but we also remember the bad ones.really i was always told fielding is something we can all do well maybe some modern players could reminded of this too.

  • testli5504537 on October 21, 2009, 3:15 GMT

    Runs saved on the boundary or elsewhere, mathematically reduce to the difference between the expected outcome and the observed outcome. The treatment of the problem on hand via a mathematical expression brings with it a set of challenges. It does not consider other variables like,the stage of the game, tiredness of the batsman, batsman's perception of a fielder, availability of a back up position (ball beating slip could be fielded by someone at thirdman) and so on. It is therefore fool hardy to try and capture the number of runs saved through the course of the game for all instances on which a ball was fielded by someone.On the other hand a contribution of a fielder could be better judged by the delta between the time when a run out/ catching opportunity was missed to the eventual score of the batsman. Catching rates, run outs per hundred opportunities could be other statistics that you could measure for the fielder.

  • testli5504537 on October 21, 2009, 0:52 GMT

    Total agreement. I even lament the lack of a good compilation of "the best fielders". I have dvd's of batsmen and bowlers,but give me one showing the best fieldsmen stopping runs,and/or catching. Once they've departed their feats are lost! It is such a huge component of the game, and the analysis is generally light to nil apart from the keeper's record. All players should have their worth measured in the field and commentating would be helped with quantitative stats in front of them, giving an insight into the whole team not just the jontys or pontings

  • testli5504537 on October 20, 2009, 21:23 GMT

    Recording runs saved by fielders sounds grand in theory, but how will the number of runs saved by fielders be determined and applied equally across all fielders in all matches? What if in one match, Cricinfo commentator 'A' determines a stop in gully as saving four, but in another match Cricinfo commentator 'B' determines a similiar stop in gully as saving three? Stats for misfields, clean fields, overthrows, catches taken and dropped and run outs made and missed would be ok, but it would appear recording runs saved would be left too much to the interpretation of the commentators.

  • testli5504537 on October 20, 2009, 14:43 GMT

    The problem is that most fielding events are qualitative. Was that a bad miss on the boundary or a good try? What's more is that nowhere in the game are runs not scored recorded, so taking note of that seems rather problematic to me. I mean you could take a tally of dropped catches but how would you record David Warner's effort for NSW on the boundary against Somerset? He caught the ball but was going out of bounds and thus threw it back in.

    What is a crime is that we don't keep track of what would be the qualitative statistics. Namely: Run Outs Direct, Run Outs Affected.

    We do record byes but it's almost impossible to get a keeper's record in this regard without going scorecard by scorecard.

    Also, I've sometimes thought that there is some justification in recording overthrows against a fielder, instead of, as at present, against the bowler. The batsman still scores them - they put pressure on the field - but the fielder making the throw concedes them.

  • No featured comments at the moment.