Disclaimer: Cricket, like other sports, always brings out an uninhibited stream of opinion/criticism/advice from its fans. This cerebration is sometimes backed by facts, other times by statistics and almost always laced with a healthy dose of emotion. So it is with this piece below. I am not doling out blinding insights based on rigorous facts – merely playing with numbers to make some interesting points up for debate.
This post was inspired by two thoughts that seemed to enter my head simultaneously while I was in the shower. (Aside: the shower and the pot have got to be the two most inspirational places for me). Here are the thoughts:
- The first was if you were the parent of a young teenager aspiring to be a fast bowler in India, how concerned should you be?
- The second one, related to the first was just how much of an impact is made by bowlers in winning you Test matches in the modern age.
My early hypothesis on the two thoughts without having looked at any data (and pretty much formulated while I was still in the shower) were as follows:
- If your son is a bowler and wants to be a Test cricketer in India, you must nudge him towards improving his batting or get him a spin coach. If he is adamant on bowling pace, you must consider emigrating to Australia, SA or England!
- Despite the modern game becoming increasingly batsman friendly, Test cricket is still about having a quality bowling attack that can take 20 wickets on any surface without giving away too many runs.
Having overspent my time under the shower, I had to get out and felt compelled to delve deeper into the two questions. Thus, armed with data from Cricinfo’s Statsguru, my modest analytical skills and some free time, I played around to see the impact that Test bowlers have had on their team’s performance across conditions. I created some boundary conditions to keep things simple:
- I have looked at Tests played after the last ODI World Cup in 2011. So the period under consideration is 1 April 2011 to 25 February 2013
- I analysed the performance of the top 5 countries in the ICC Test Rankings – South Africa, England, Australia, India and Pakistan. I could have looked at the last 4 years since the test rankings are like a rolling 3 to 4 yr average but I was more interested in isolating and benchmarking India’s sketchy Test performance post World Cup 2011. Also, I could’ve looked at two more nations but I am lazy and the point is well made with five countries.
- I divided bowlers into seamers and spin bowlers. This poses some interesting moral questions in the grey zones; for example, Jonathan Trott and Michael Hussey end up being classified as seamers and Kevin Pietersen as a spinner!
- To further disaggregate bowling performances, I looked at Home and Away Test matches for each country. There were 6 Tests played by Pakistan in UAE (neutral venue) in this period (3 v Eng and 3 v SL). Now, UAE is a veritable home ground for Pak with the country getting significant crowd support in these venues. I have classified the 6 tests as Home tests for Pakistan and as Away tests for England.
An interesting picture emerges from the analysis:
Firstly, 4 out of the top 5 Test nations have extremely good bowling averages – all under or around 30 runs per wicket. India (with overall bowling average of 35) has the worst average among the top 5.
Aside: The reason I say that a bowling average for a team of below 30 is crazy good is because of the following. For this period, the minimum number of wickets picked up by any team was 286 by Pakistan. All other teams picked between 320-350 wickets in this period. To understand how difficult it is to pick up so many wickets at a sub-30 bowling average, sample this: Anil Kumble (619 wkts) had a career bowling avg of 29.65, Brett Lee (310 wkts) that of 30.80 and Makhaya Ntini (390 wkts) of 28.82! These are individuals who were champion bowlers of their sides. Now for a team to replicate that average having 5-6 bowlers of which there will be 1-2 crap bowlers like Samit Patel or Mohammad Sami is quite significant.
This is quite closely correlated to the outcomes or win % for the teams. South Africa has the best overall bowling average (26.07) and has the highest win % of 67%. India with an average of 35.19 quite unsurprisingly has a win ratio of 33%. (See figure 1)
Of course, when you de-average these numbers and see how the bowling averages change between Home and Away Test matches, the figures are all the more revealing. Most teams have, for obvious reasons like home conditions, pitches suited to home bowlers, crowd support, a better bowling average in home matches that rises as they travel abroad. Some interesting observations:
- England manages to reverse the trend and has a slightly better average away (29.15) than at home (31.49)
- Australia and Pakistan’s averages increase ever so slightly as they tour abroad – by 4 and 5 runs per wicket respectively
- India and South Africa’s bowling average increases by over 10 runs when they travel abroad. However, SA have an exceptionally low home test bowling average of 21 so a 10-run increase in Away Tests still takes them to a very respectable 31. India, on the other hand goes from a Home test average of 30 to away average of 41! Again this is quite closely correlated to outcomes. (See figure 2)
Interesting and Scary Take-away: How worried should India be for its upcoming SA tour at the end of 2013? If these averages were to hold true, a crude estimate (overlooking the finer nuances of de-averaging innings wise scores, Sachin’s potential retirement, yada, yada) would suggest that SA batting first would score an average of 410 runs against India (applying India’s away bowling average) and would then bowl India out for around 210 (SA’s Home bowling average) which is just enough to enforce a follow-on and also (again applying SA’s home bowling average) enough to bowl out India again to inflict a humiliating defeat. Or defeats, considering there will be at least 3 tests. Now I do wish that none of this materializes but it is still reason to be afraid. Very afraid.
Having established that there is a correlation between winning Tests and bowling well, I’ve tried to go one level deeper into what is responsible for the efficacy (or lack of it) of some of these bowling attacks. For each country, I looked at the share of wickets picked up by spin and seam bowlers and how that changes in home and away conditions. Also, how do spin and seam bowling averages vary between home and away conditions for each country? Refer to figure 3 and figure 4 below.
The charts are quite revealing to understand the dominance of South African bowling.
- Seamers take ~87% of all wickets for SA whether playing in home or away conditions. Also the bowling average of seamers while bowling home or away varies between 19 and 29 runs per wicket. So while SA spinners have a ridiculously high average of 56 in away conditions, it doesn’t matter since the seamers are going to take most of the wickets anyways.
- Contrast this with India – while spinners understandably take 78% of the wickets in Indian conditions but only 26% of the wickets in away Tests. This is, arguably, due to the lack of world – class bowlers (e.g. Anil Kumble) who can take wickets on any track. Also while seamers take 74% of the wickets in away tests, they do so at a relatively high average of 36!
- The other sub-continental team Pakistan does a much better job of sticking to its strength. Spinners take ~65% of Pak’s wickets in all conditions and do so at an average of 22-25!
- England presents a very interesting example of nurturing a balanced bowling attack. In seaming conditions at home, fast bowlers get 82% of the wickets at an average of 29. When England travels, the spinners take over and get 60% of the wickets at 27! It’s probably why I am willing to bet that they would likely be the better side in both home and away editions of the Ashes this year.
The key takeaway from the above analysis is that in order to win Tests, you need one of the following:
- Seam or spin bowling attack that works in all conditions – like Steyn/Philander/Morkel for SA or Ajmal/Rehman/Hafeez for Pakistan
- Or, which is tougher, you have a balance of top-notch spin and seam bowlers sharing the load irrespective of conditions – like Anderson/Swann/Broad/Finn/Panesar for England.
Another (and the last) analysis is to look at the concentration or reliance of teams on star bowlers. In business, you can neither be too reliant on one supplier nor should you be spread too thin and have a long tail of suppliers. Likewise, in cricket, you can’t have one bowler taking most of the wickets or be in a situation where the top 4 bowlers don’t account for bulk of the wickets. The ideal, in my opinion and you can draw the line in the sand differently, is if you have your top bowler taking ~25-30% of the wickets and the top 4 taking at least ~75-80% of all wickets. Now look at the concentration analysis below (figure 5) and you get further evidence why SA, Pakistan and even England do so well bowling wise. SA, England and Pakistan have star bowlers who take 25%+ of the wickets for their side and also have a bowling quartet that can take 75-80% of the overall wickets. This provides a great combination of having individual top-class bowlers as well as sufficient bowling depth.
Hence, in summary:
- Strong bowling performance has a good correlation with winning Test performance across conditions. Note that I am not advocating a causality between a Test win and good bowling, just a strong correlation
- India’s poor Test performance abroad can partly be explained by the lack of world-class bowling options that can deliver the goods across different conditions
- Strong bowling sides typically have a seam or spin strength that delivers wickets even in alien conditions. England seems to have the most balanced attack capable of doing well in seaming or spinning tracks
- Strong bowling sides have 1-2 star bowlers in the side that take ~25-30% of their wickets as well as the depth in their top 4 to take a majority of the wickets
- India sucks on points 3 and 4 and is good reason for concern as it prepares for a challenging away tour of South Africa
I realize that apart from the analysis, the real question is how do you develop a bowling attack and nurture talent to win you Tests. It’s something I have views on but would rather save it for another post. Also, looking at bowling is just looking at part of the bigger picture – there are other factors that decide the fate of a Test. However, bowling is, unsurprisingly, a critical area and one that will make a big difference in who wins the Ashes this year and how well (or badly) India cope in the South African tour.
Other thoughts / criticisms / analysis insights / suggestions are highly welcome.