India’s fielding was poor in the first Test, but just how bad was it?
It frustrates me that there is no effective fielding measure in cricket. We count the number of catches, but this doesn’t tell us how difficult a catch was or how many were dropped.
A possible solution is to assign a runs value to each fielding event. If a catch was taken or missed, how difficult was it? How likely would a typical fielder have been to catch it? And, therefore, how many runs were saved or cost? Applying these questions to every potential catch, stumping and run out and every piece of ground fielding results in a fielding score.
I’ll go into more detail below, but the headline results for the first Test are:
| Innings 1 | Innings 2 | |
|---|---|---|
| Australia | +29 runs | +22 runs |
| India | -40 runs | +15 runs |
So, all up, Australia finished the match on plus 51 runs, India on minus 25 runs, a difference of 76. This confirms the prevailing opinion that India’s fielders had a shocker.
However, while the overall result comes as no surprise, some of the individual players’ figures contradict the popular perception of how they performed.
Take the wicketkeepers. On the face of it Tim Paine had a better game than Wriddhiman Saha, taking seven catches to Saha’s zero. But this is misleading, as six of Paine’s catches were straightforward and the seventh – his catch to dismiss Hanuma Vihari – was of only moderate difficulty. A competent wicketkeeper would be expected to take them all. Paine put down a difficult chance off Pat Cummins and all told finished with a fielding score of plus one.
Saha missed two difficult chances. He also assisted in a straightforward run out of Mitchell Starc in the first innings and cleverly ran out Matthew Wade in the second innings, and the credits he received for these pushed his overall result to plus 15.
So while Paine deserved player of the match because of his splendid 73 not out, Saha actually out-kept him. Yet in the absence of a fielding measure, it is being reported that Paine kept well and Saha was a bit wobbly.
Tim Paine (Photo by Mark Kolbe/Getty Images)
So how does my method actually work? I’m glad you asked.
There are two components to consider when assigning a runs value to every chance taken or missed: how many runs a wicket is worth and how difficult a chance was.
In the last decade of Test cricket just over 30 runs have been scored for each wicket and so 30 is the number I used. To assess the difficulty of an opportunity I asked, “What percentage of the time would a standard, decent fielder have taken the chance?”. This percentage was then applied to the 30 runs wicket value and a credit or debit arrived at. See below for a worked example.
This is a subjective process but precedents exist in baseball – a much more statistically sophisticated game – where human judgment determines whether a player gets given a hit or a fielder gets an error. And subjectivity could be removed with technology – ball tracking data could be compiled to enable a more definitive assessment of how difficult a chance was.
Ground fielding was more straightforward. If a fielder saved runs when I judged a typical fielder would not have, the fielder received a credit for the amount saved. And if a fielder failed to save runs when a typical fielder would have, the fielder was debited. So a superb chase and dive that turns four into three results in a one-run credit but if a wicketkeeper misses a simple take that leads to four byes, this incurs a four-run debit. Again, it’s subjective, but again, technology could be used to fix this.
As an example of how the method works, take the first-Test performance of Mayank Agarwal, whose name appears three times in the data.
Agarwal failed to haul in a four in Australia’s first innings. I judged a typical Test fielder should have kept it to three, hence Agarwal was debited a run.
Agarwal dropped a straightforward chance off Paine, a chance I deemed would be taken about 80 per cent of the time. This mistake is therefore deemed to cost 80 per cent of the 30 runs assigned to a wicket, so 24 runs are debited from Agarwal’s fielding score. Finally, he took a straightforward chance to dismiss Marnus Labuschagne in the second innings, a catch I assessed would be dropped 20 per cent of the time, so he saved his side 20 per cent of 30 runs, which is six runs.
All up, Agarwal’s contributions were -1 run, -24 runs and +6 runs for a total of -19 runs, a poor fielding game.
If this method were applied consistently match after match (eventually with the assistance of technology), it would be valuable, and not just in Test cricket – at the IPL auction a superior score might mean a difference of hundreds of thousands of dollars to a player’s sale price.
The specifics of the methodology can be debated, but eventually a fielding score should be able to sit alongside batting and bowling averages as a fundamental benchmark of performance. The sport would be better for it.
The Late News
Roar Rookie
Interesting idea. Might need a little tinkering.
Kicker
Roar Rookie
Greatest fielder for me is Colin Bland. Those who have seen him will attest to his brilliance.
matth
Roar Guru
Good stuff, now if you can just watch and catalogue the fielding for the other 2395 tests, that would be great. See you in a decade! :laughing: :laughing: Seriously though, this appears to be an effective but also easy to understand measure. Well done.
Nick
Roar Guru
And yet, a standard far lower than what MLB players are held to.
Chris Johnson
Guest
I like the concept. The reporting around Saha is likely because he missed two chances which the bloke he is keeping out would likely have taken, and Pant is significantly younger, a significantly better batsman and more athletic (at this stage of their careers). Baffling to me why they went for Saha, even if he is a better keeper to the spinners. Hope they stick with him though!
JamesH
Roar Guru
I like your idea but... "So while Paine deserved player of the match because of his splendid 73 not out, Saha actually out-kept him." I don't think you can conclude that based solely on your current scoring method. Players can only field/catch what comes their way. Paine missed a tough chance where he only got the tip of his glove to the ball, but otherwise did virtually everything that could have been expected of him. Saha was presented with a few more difficult opportunities, some of which he missed (including a similar one to Paine's that he didn't even touch, but should have) and some of which he took. That doesn't mean he out-kept Paine, it just means he was challenged more than Paine was. It's not Paine's fault that he didn't get the chance to run someone out by flicking the ball back between his legs. Your method seems reasonable for determining whether or not a player had a good or bad game in the field (although I'm not quite sure how Saha got +15 and Paine only got +1) but it's not particularly useful for a comparison between the performances of different fielders. There needs to be some weighting of the frequency of opportunities actually presented to the fielder before you can make such a comparison, and I don't think that's always going to work because some fielders just have very quiet, simple games, for reasons beyond their control.
Andrew
Guest
One thing you could do to make your 30 value for a batsman as an average is to skew the value front ended. Eg. Batsmen no. 1 gets 55 value, no. 2 gets 50, 3-45 and 11 gets 5. This would give a more accurate value as a number 11 you wouldn't expect to do too much damage with the bat whereas a number 1 could cost 55+ but in terms of average 55 is not a stretch.
Gray-Hand
Roar Rookie
Fair point.
Paul
Roar Guru
I see where you're coming from in your example using Shaw & Gill, but a few things spring to mind. The product the author has included here, needs a lot more fielding variables added in, eg if a throw comes into the keeper and a guy makes the yards to back up and stop any additional score, he should be credited with runs saved. There are no doubt dozens more. The other issue is where guys field. First slip for example, might not touch the ball at all in two completed innings, or might only field the ball to prevent it from going to the boundary after an edge. Other positions, eg point, cover, fine leg, might touch the ball dozens of times in an innings. There needs to be a way of making this type of scenario more equitable, so Shaw who is a gun outfielder who tocuhes the ball often, isn't unfairly advantaged over Gill, a gun slipper who only touches the ball to throw it back to the bowler.
Brian
Guest
this only looks at catching from mid 2000s to mid 2010s http://www.sportstats.com.au/articles/droppedcatches2016.pdf Interesting Graeme Smith comes out top having dropped only 14% of what came his way. Worst was Umar Gul who took just 11 of 25 potential catches. There is one caveat, bowlers are more likely to drop catches due to the difficulty of chances when created of their own bowling. Australia's best was Michael Clarke who went at 21% and Ponting 22%
dungerBob
Roar Rookie
I can see your point but what if you get a guy who has a habit of dropping the big guns but snaffling the tail. I'm not sure that's ever been the case but I think it makes sense to use both metrics to find out.
Brian
Guest
It could be useful say Shaw first class fielding average is +70 and Gill's is -90 it could explain why one is selected when no other reasonable explanation can be found Would be even more useful for keepers who we currently eveluate on their batting average and the vibe of how tidy they are behind the stumps.
Brian
Guest
That's wrong because it measures the team not the fielder. If Aus is hosting Afghanistan and Paine drops a sitter but then only loses say 3 runs before Cummins finds another edge that doesn't make him a better keeper then the Afghan keeper who drops Steve Smith but then watches Warner make 300 because his bowlers aren't good enough to create any more chances. The generic 30 runs is a much better way of working out the fielder's worth
JOHN ALLAN
Guest
You have probably given the BBL “Masterminds” food for thought for next season. Just when the viewers are getting used to Rule Change No. 1421.
Dexter The Hamster
Roar Rookie
Thanks Paul, was a nice read. I guess the issue is that cricket is fairly unique and fielding is not easily quantifiable (as you are obviously aware). Some have already commented that dropping Kohli is not the same as dropping Bumrah. The fielder who dives but still gives away a run trying to save a boundary on the ropes, is not the same as the fielder who is having a snooze at mid-on and lets a ball bounce off the knee and lets in a single, when the batter was on 0 off 10 balls and feeling pressure. But in both cases they are -1. I cannot see a way that any system can measure the game situation in any meaningful way. But as a measure of pure "who is the best fielder", I can see plenty of merit in it, and would happily embrace it.
Ruairidh MacDonald
Roar Rookie
Unbelievable! I’ve been working on an article and system for measuring fielding as well! Yours is nice, using the average value of a wicket is something I didn’t think about. I’ll keep going with my article and publish it eventually, just want to make sure you know it’s not a copycat!
Paul Dennett
Roar Rookie
True but this is the same for batting average and bowling average. Paine's 70-odd runs in the first Test were worth more than many centuries but they'll only count as 70 to his average.
Mango Jack
Roar Guru
Interesting to ponder if fielding stats were able to be calculated retrospectively, who would be considered the greatest fielder of all time. Probably more so than batting or bowling, the athleticism of the modern cricketer has improved fielding incredibly. Likewise, the introduction of limited overs cricket. So it's almost certain that the greatest fielder would be a player from the last 30 years or so. Jonty Rhodes would be up there, as well as Ponting. The worst? That's easy. Inzamam Al-Haq, take a bow!
Gray-Hand
Roar Rookie
Not sure that you can assign a value if 30 runs to a wicket. Runs don’t have a set value. 50 runs on a difficult pitch might be worth 100 on a flat road.
Mango Jack
Roar Guru
Or that stupid pressure meter, or whatever it's called.