Watching over the stars
The night-watchman is Test cricket's security guard, protecting the stars in the lineup during unhelpful conditions, and saving their turn at bat for a better time. But this is also an experiment with some obvious risks - considering that the night-watchman must bat in conditions that are deemed difficult for the star he's standing in for. Y Anantha Narayanan looks at the history of this experiment.
Gillespie's recent exploits and the uninformed views of a few writers consigning Boucher to the ranks of night-watchmen have made us go back and re-look at the analysis of Night-watchmen.
This analysis is based on the earlier analysis we made on the Test batting positions. In Test Batting Position Analysis, we looked at several aspects of batting positions. Just as a refresher the BPI(Batting Position index) is an average of the positions the batsman has batted in, keeping the two opening positions as 1. We set aside the particular case of the night-watchmen, which I take up here, with special emphasis on Gillespie, who re-wrote a few record books.
I have always wondered at the idea; it seems odd that a better batsmen could ever be protected by one who is obviously his inferior. But certainly there was a time when night-watchmen regularly padded up an hour before close, and would walk in at the fall of a wicket. In some cases, more than one would appear; I've seen at least one game where after both 'night-watchmen' had failed to make it through the day, the regularly scheduled player was forced to put in an appearance anyway, this time with his team two extra wickets down!
Anyhow, the Australians, led by Steve Waugh, changed things. A top order batsman is expected to be at the crease at 11 AM or 17.44 PM, and do just as well at each time. Overall this seems to be the more rational attitude, although most other teams still follow the old idea. The Aussies too, haven't always stuck with Waugh's thinking; since taking over from him, Ponting and Gilchrist have often steered their own course. Especially Ponting has reverted back to the more traditional method of sending in a night-watchman.
Our interest here is analytical. First, to ask: who is a night-watchman, and thereafter to understand the merit of sending one in to deal with various situations. Let's start with identifying night-watchmen correctly; i.e. before we make any analytical comment about this role and those who have played it, we must first be able to tell whether a player is really standing-in for a better batsman, or 'pinch hitting' up the order for other reasons. This is a very difficult determination to make, because these are all sorts of reasons why a player might bat out of his position. Also, there is very little data available on things like the time of the day when a batsman came to the wicket.
A simple starting definition may be that a night-watchman is one who bats earlier than his intended position. Then we are bound to ask "what about Gilchrist? If he opens the batting for Australia when the team sets off looking for quick runs, should we treat that as equivalent to a night-watchman?". Gilchrist's Batting Position Index (discussed in detail in part I of this series) in Tests is 6.68, indicating that he is a batsman who has batted at No.7 most of his career. Or consider Wasim Akram, with three Test centuries under his belt and a BPA of 8.1, batting at number 3 or 4. In order to do a correct job of selecting true night-watchmen for our analysis, it is necessary to define a number of parameters other than batting position alone. These must account for the quality of batsman as well as the specific instance.
It may be easier to say who is not a night-watchman. For instance, we might say that any batsman whose career batting average is higher than 25.00 cannot be classified as a night-watchman; his wicket would be of considerable value even in a difficult situation where one might want to protect the better players. (JFM Morrison and Parker, a rare breed of specialist batsmen with a career average below 25.00 get taken care of because of conditions explained later). similarly, we might say that any batsman who has scored more than one century in Test cricket cannot be considered as a night watchman even if his average is less than 25.00. A late-order batsman can luck his way into one century, as Agarkar did, but not two. This filter would avoid considering the following batsmen as night watchmen. This is the list of batsmen who have scored 2+ centuries and have a batting average of between 20 and 25.
================================================== India Ramchand G.S 33 1180 24.58 2 New Zealand Parker J.M 36 1498 24.56 3 Australia Gregory S.E 58 2282 24.54 4 Australia Benaud R 63 2201 24.46 3 New Zealand Pollard V 32 1266 24.35 2 South Africa Sinclair J.H 25 1069 23.24 3 England Illingworth R 61 1836 23.24 2 West Indies Ganga D 34 1421 24.08 2 Pakistan Wasim Akram 104 2898 22.64 3 Australia Lindwall R.R 61 1502 21.15 2 England Evans T.G 91 2439 20.50 2 New Zealand Taylor B.R 30 898 20.41 2 New Zealand Watson W 38 939 20.41 2 ==================================================
It's also true that hardly any captain would protect one batsman by sending in another of similarly high caliber. I.e., if a player is normally wont to bat at positions 1-6 (or even 7), then the likelihood of him being chosen as the night-watchman is somewhat small. This would amount to filtering out those whose BPI is less than 7.00 cannot be classified as a night watchman. Without this restriction, Morrison and Morgan will create problems as also Benaud and Illingworth. The reverse situation is also valid, that a night-watchmen, however selected, must bat at positions 1-6; there's not much point labeling a #10 batsman batting at #7 a night-watchman.
Finally, and most important of all, an innings will be considered as a night-watchmen innings if the difference in batting position between a player's normal spot and the one he is sent into is greater than or equal to 3.0. This is admittedly qualitative, and therefore needs some fine tuning, especially because hardly anyone bats always in the same spot lower down the order, unlike for the higher order players. Most true bowlers have played at 8, 9, 10 and 11 sometime or the other, and this tends to reduce their BPIs a little. Jason Gillespie, for example, has had a few bats higher than #9, and this has brought his BPI down just a shade under 9, to 8.98. Yet, in the recent Chennai test he went in at No.6 and played what is arguably the greatest night-watchman's innings in history. The slight revision proposed to account for such errors in filtering is that for any batsman whose BPI is between 7.75 and 8, the BPI is adjusted to 8, and this is repeated for positions 9, 10 and 11.
With these restrictions we have a list of 613 night-watchmen innings in the 1799 tests played so far. If you want to browse through the whole list, the data are presented here in a few different ways. It is of interest that a few innings which quaified as night-watchmen innings in last year's analysis have now gone out. Perfect example is Nikky Boje whose Test batting average is now up to 26.00, into the realms of a Test batsman and gets caught in the 25.00 filter.
Who has been the best night-watchman in history. Easy to guess. Gillespie, in 9 innings has scored a total of 327 runs at an average (no doubt aided by the unbeaten 201) of 40.87. More relevantly, he has faced a total of 1040 balls in these 9 innings, an average of 116 balls per innings. He wins the title hands down.
What has been the best night-watchman innings played. Take your pick. Gillespie presents you with a short list of two, his match-saving effort at Chennai and the mammoth 201 against Bangladesh. As far as I am concerned, I would take the Chennai effort any day. It was a water-shed innings and changed the course of one of the most important series of recent times.
Now for the contentious task of determining whether the night-watchman experiment has been a success or not. Assuming that no one would be stupid enough to send a night-watchman an hour before close of play, we are looking at a possible maximum of around 8-10 overs to be played during the evening. We must also assume that the night-watchman should play out a few balls the next day. A ballpark figure is that if a night-watchman bats for 30 balls, he has more than done his job since he has been in the middle for around 45 minutes. 57 of the 135 batsmen whose night-watchmen innings has "balls played" information, have been successful on this basis. This works to a decent figure of more than 40%.
Of course for many of the night-watchmen innings "balls faced" information is not available. In such cases a figure of 10 runs can be taken as the minimum for a night-watchman to be considered a success. It must be remembered that most of the night-watchmen innings are slow stone-walling types. Of course some one might score 23 in 29 balls (as Kamran Akmal did in 2003) and get out before draw of stumps or score 7 runs in 50 balls. These things even out. 214 of the 478 batsmen whose night-watchmen innings have only runs scored information have been successful on this basis.
Out of the total population of 613, it can be deduced that 270 have succeeded in their task, making the success rate of the night-watchmen exercise around 40%. Is an experiment that is likely to succeed 40% of the time worth it? Perhaps. Certainly, the success stories are very significant, as is the case with Gillespie, Hoggard, Tudor and Larwood. One great factor in these night-watchmen decisions is that the expectations of success are not high to begin with. If they succeed, one counts that as a bonus, and if not, no great loss has occurred. On this basis, 40% seems quite good.
I should probably change my opinion and conclude that an experiment, using less accomplished batsmen, with a success rate of around 40% is a reasonably successful one.
Y Anantha Narayanan