Compare the Pair: SA vs Aus Batsmen

Published in

Sporting Chance Magazine

11 min readMar 9, 2018

The South African test series is already proving to be a cracker, with the ageing stars of South Africa coming up against a young and fiery Australian side determined to continue their good record on South African soil (having not lost a series their since South Africa’s reintroduction to the Test arena).

The two bowling attacks are equally the most lethal in world cricket and so it may come down to which team’s batting line-up is better able to cope with the conditions.

Let’s therefore break-down and compare both team’s batting stocks.

Rather than simply comparing line-ups, batting rankings or batting averages, I took a closer look at the main batsmen who will take part in the upcoming series. Each player has been compared in the following ways: ability to start an innings, how good they are when their eye is in, and how good they are at converting starts to big scores. There are a number of metrics that were used to devise each rating but to save boring you I won’t go into too much detail except for particular points for each batsmen to give an idea as to why their rating is how it is.

To form these metrics, I analysed the 85 most prolific run-scorers in test cricket history and came up with statistics that contributed to their ability to start an innings, ability once set and ability to make a big score. The ratings were then averaged and each player given a z-score (comparison of their rating to the average — a score of 0 means the player is average, -1 means they are 1 standard deviation worse than average and 1 means they are 1 standard deviation better than average). The players below are basically compared to the average of the highest run-scorers of all time so if a player does better than average then that is a good achievement.

Let’s have a look at how each player stacks up…

The Australians

David Warner

Warner is an exceptional starter — being much less vulnerable early in his innings than most batsmen. This is a great attribute for an opener to have, as he really does see off the new ball. It is interesting that despite his non-traditional, aggressive style, he remains effective at starting his innings. Maybe he is helped by the typically attacking fields that are set early on in an innings, which gives him a chance to start his innings with boundaries.

4.58% of his innings are ducks, which is much lower than average (6.5%)
He is only a 11.45% chance at going out before getting to 5, compared to the global average of 15.63%
25% of his scores are below 11 (his Q1), which may seem like a lot, but the global average Q1 is 9, hence although only 2 runs more, it is a 22% improvement on the average
He only improves his average by 5.5% when his eye is in (from 48.78 to 51.46), which occurs almost immediately after getting off the mark.
He is also not as good at cashing in on starts with just 2.3% of his scores above the global outlier range — compared to the average mark of 3.94%.

Usman Khawaja

Khawaja mirrors the global average very closely. He is slightly quicker at getting set than average, taking approximately 5 runs to get to his eye-in average of 55 (a 20% improvement). Khawaja’s similarities to the global average makes him both ordinary and yet unusual, because most players are better or worse in some categories. Nothing really stands out about Khawja’s pattern of play apart from the fact that it is so close to the global average.

Steve Smith

Smith’s radar confirms what we already know, he is one of the all time great batsman. Smith’s brilliance is his combination of talents. He has an exceptional starting ability, an incredible eye-in ability, and a great capacity to make big scores. This rare combination means that he is not only far harder to dismiss early in his innings but also once he is settled. Of all the batsmen analysed, only 6 were better starters (Bradman comes in just behind Smith), and only Bradman was better when set and at going big (Bradman’s z-scores for those two categories are 4.35 and 5.27 compared to Smith’s 2nd best of 1.63 and 1.61). Smith reaches peak performance once he gets to about 9. Some of his more impressive stats include:

3.6% of his innings are ducks, 9th overall
Q1 of 13.5, equal 6th overall
8.11% chance of making less than 5, the best overall
25% of Smith’s scores are above 97 (his Q3), which places him behind only Bradman (whose Q3 is 169)

Shaun Marsh

Marsh almost breaks the chart, his starting ability is almost 4 standard deviations worse than the average player. To put this in context, the worst starter in the group of 85 retired players analysed was Marvin Attapattu, who was just 2.3 standard deviations away from the average. The radar actually confirms everything that is said about Marsh — coaches say he is one of the most fluent and best bats on his day, which is true — he actually performs better than average in all other categories, its just that he is woefully vulnerable at the start of his innings. Here are some numbers that paint the picture of his inability to start an innings:

16.33% of his innings are ducks or 1 in 6 — this is the worst recorded
He is a 34.69% chance to make less than 5 — more than a third of his innings. This is 10% worse than the next worst recorded (Kapil Dev)
53.5% chance to make less than 20 — again the worst recorded
Q1 is 1.5 — remarkably, 25% of his innings are below this mark

So whilst he has an overall average of 40.87, more than half of his scores are below 20. He improves his average to 55.3 after getting to around 4, which is a 33% improvement — so if Marsh can survive his first 4 runs then he becomes as good a player as Khawaja or Warner.

Mitch Marsh

Not yet established as a test quality player

Mitch Marsh has been picked for the majority of his 24 test matches on potential rather than output, and hence his statistics do not make for pretty viewing. As can be seen in the radar, he is worse than average in every category by a significant margin. We can cut him a bit of slack because he is an all-rounder being compared to batsmen, and he is early in his career still, but he will want to improve in all areas if he is to lock down the number 6 position for Australia. Interesting to note that he is still a better starter than his brother, but doesn’t really improve from this base. Due to his poor ability when settled, it is hard to ascertain how long it takes to get his eye in, but probably within the first 4 runs of his innings.

10.26% of innings are ducks
Q1: 3.5
Q3: 33.5 almost 10 runs worse than any of the 85 batsmen analysed, this shows that only 25% of his innings are above 33.5 compared to over 52% of Smith’s being above this value.

The South Africans

Dean Elgar

Somewhat surprisingly for an opener, Elgar is not a great starter. He improves his rating as his score increases — only slightly worse than average once settled and at going on to get a big score. It takes him around 12–15 runs to get to his peak improvement of around 30%

Q1: 7.5
46.26% chance to make less than 20, 7th worst overall
Avg at zero: 42.07, avg at Q1: 46.48, avg at median: 53.15, avg at Q3: 62.43. This shows how he is one of the few batsmen that actually does continually improve over the course of his innings — the longer he bats, the harder he is to get out.

Hashim Amla

Slightly more vulnerable early in his innings

Amla is a quality batsmen who has shown he can make really massive scores — but his weakness appears to be early on in an innings, where his rating is below average. It takes him about 17 runs to improve to his average eye-in level and so the Australians will want to start very well to him and really concentrate on dismissing him early, otherwise he is a very capable batsman.

40.9% chance to make less than 20
2nd best eye-in ability of batsmen in this series
2nd best at going big of batsmen in this series

AB De Villiers

The radar reflects De Villiers’ natural talent — he is one seemingly able to walk in and thrive instantly no matter the batting conditions. Where he doesn’t quite compare to the greatness of Smith is his ability to then capitalise on his starts. He is still above average in these categories, but his starting ability is almost a standard deviation better than the average. Because he is such a good starter, he appears to be virtually settled immediately, but he does improve his output marginally after he gets to 4 runs.

3.83% of his scores are ducks — 10th best overall
34.23% chance of making less than 20, the best in this series
Avg at Q1: 53.12, Avg at median: 50.56, Avg at Q3: 55.86. His average remains pretty steady over time.

Faf Du Plessis

Du Plessis follows a similar pattern to De Villiers, but on a smaller ability base. He is pretty much an average starter and average even once his eye is in, but he drops away significantly when his ability to convert starts into big scores is looked at. His highest score is just 137, and whilst he is probably best remembered by Australians for his match-saving 376 ball innings on debut, he is rarely the kind of player who can bat for a long time and refuses to be dismissed once his eye is in. Again, similarly to De Villiers, he is at peak ability very early in his innings — meaning he is pretty much as likely to get out when he is on one compared to when he is on 50.

7.69% of his innings are ducks — higher than average
37.67% chance to make less than 20 — much lower than average (once off the mark he is a good starter)
Average is improved by only 4% — he doesn’t become much more difficult to dismiss once his eye is in

Quinton De Kock

Probably limited by batting with the tail

De Kock is the pivotal wicket-keeper batsman who has saved many an innings with his aggressive strokeplay. He is a decent starter, and dangerous once settled but rarely goes on to post really big scores. A lot of this is probably due to circumstance, coming in at six or seven means that he is commonly batting with the tail and hence plays more aggressively, or the team’s innings finishes early and so he is unable to showcase his ability to make really big scores. This aggressive nature probably also partly explains the slightly poorer starting ability — with De Kock not waiting until he is settled to play his shots, and hence giving chances early in his innings. However, De Kock gets settled very quickly, and so this poor start really doesn’t last long and is only really reflected by an increased chance of making a duck.

8.51% of his innings are ducks — much worse than average (6.5%)
Highest Score: 129*
Three 100s in 47 innings — doesn’t often go on to get big scores
Q3: 64, this is just as good as many other batsmen with far higher averages — so shows that De Kock’s average of 39.26 is artificially lowered by a lack of high, outlier scores that most other batsmen achieve

Temba Bavuma

Hasn’t established himself at Test Level

Bavuma is probably in a similar situation to Mitch Marsh in that the selectors want him to achieve because of his potential, but he hasn’t quite been able to deliver yet. He is actually not bad at starting his innings, but hasn’t shown an ability to make the most of starts or go on and get big scores as of yet. Like other South Africans in this series, Bavuma doesn’t improve much at all over the course of his innings and is just as likely to be dismissed early in his innings as he is once his eye is in.

9.52% of his innings are ducks — very high percentage
Median: 21, a very low score, only the Marsh brothers are lower in this series (equal with Elgar)
Q3: 50.5, again very low — only better than Mitch Marsh
One hundred in 42 innings (102*)

Summary

The South African and Australian batting orders are almost opposite to each other in terms of style. Australia’s best starters bat in the top 4, whilst South Africa’s top order of Elgar and Amla are relatively poor starters that improve as their innings progresses. Similarly, Australia’s lower order of the Marsh brothers are very poor starters but (in Shaun’s case at least) improve once their eye is in. In contrast, De Villiers, De Kock and Du Plessis are all good starters but don’t generally go on and get big scores in the way that their top order does.

This leads to an intriguing battle where if Australia’s top order is dismissed cheaply, they appear vulnerable to a spectacular collapse. However, the Australians’ better starting top order may also be an advantage as they are better equipped to see off the new ball in a series where fast bowlers will be deadly. Conversely, Australia will feel they have a chance to take a few quick early wickets in South Africa’s innings with their poorer starters at the top of the order.

South Africa are perhaps better equipped to deal with a top order collapse with Du Plessis, De Villiers and De Kock all capable starters and able to stem the bleeding. A key to the series will be Australia’s top order combined with Shaun Marsh — if the top order succeeds then it doesn’t matter as much if half of Marsh’s innings are below 20 and the other half good. If the top order fails then Australia will need Marsh to improve his starting ability to avoid another disastrous collapse that the Australians have been prone to over the last few years.