It's a good thing Mitchell Johnson will be playing in South Africa over the next few weeks. The last 16 months haven't been a memorable time for him: in 11 Tests during this period, Johnson has taken 32 wickets at an average of 42.03. For any frontline bowler, that's a generous average; for a leading strike bowler of a team, those numbers could be grounds for dropping. Fortunately for him and for Australia, their next campaign is in South Africa, which, over the last few years, has been one of the most favoured countries for fast bowling.
Since the beginning of 2006, fast bowlers have averaged 29.12 in South Africa, and more than 32 in every other country. Part of the reason is probably the strength of the home team's pace attack: fast bowling has traditionally been South Africa's strongest suit, and the fact that they play at home every season obviously means their performances will influence the overall numbers more than those of any other side. Johnson himself had a series to remember in 2009, taking 15 wickets in three Tests at 25, and roughing up batsmen with his short stuff. He has already made a mark this time too in South Africa, taking 3 for 20 in the rain-shortened first ODI in Centurion.
What the table below shows is that some of the countries that used to offer favourable conditions for pace bowling don't quite assist fast bowlers to the same extent anymore. The most prominent example is Australia, where the fast-bowling average since 2006 is 34. Even that is an improvement on the corresponding numbers between 2000 and 2005: the fast-bowling average in Australia during that period was 36.34. That was the second-highest, next only to India.
What's also clear from the table is that in South Africa the ratio of wickets taken by pace to those taken by spin is the highest, which means spin makes the lowest percentage contribution in terms of wickets. That, again, is partly due to the lack of attacking spin-bowling options for South Africa - Paul Harris normally plays the defensive role - but it remains to be seen if the advent of Imran Tahir changes that ratio to some extent. The average for spinners in India and Australia is even higher - it's more than 40 in both countries - but that's explained by India's ability to play spin really well, and by Australia's lack of quality spinners since Shane Warne.
|Host country||Pace-wkts||Average||Strike rate||Spin-Wkts||Average||Strike rate||Wkts ratio-Pace:Spin|
Most fast bowlers who've played in South Africa over these six years - both home bowlers and overseas ones - have enjoyed the conditions. Several of them have done better in South Africa than in other countries during this period. The list below compares the averages of bowlers in South Africa with their overall bowling averages since the beginning of 2006.
On the tour of South Africa in 2006, for instance, Stuart Clark took 20 wickets at less than 16, though his overall bowling average is excellent too. A better example is India's Sreesanth: in two series in South Africa, in 2006-07 and 2010-11, he took 27 wickets at less than 29; in all other countries since 2006, he has averaged 41.66, with only 60 wickets in 21 Tests. Similarly, Brett Lee's bowling average in South Africa is almost 10 lower than his overall average during this period.
The one top bowler who doesn't conform to this trend is James Anderson. On the tour to South Africa in 2009-10, Anderson took 16 wickets in four Tests, but at an average of 34.25; his overall average since 2006 is 29.58.
|Bowler||In SA-Tests||Wickets||Average||Strike rate||overall ave since 2006|
With fast bowlers doing so well in South Africa, it isn't a huge surprise that opening partnerships don't tend to thrive there. In all Tests since 2006, the average opening partnership in South Africa is 30.23, which is almost five runs fewer than the next-lowest, 35.04 in New Zealand. In fact, the partnership average in Australia is about 42% higher than that in South Africa, and the rate of century partnerships is almost twice as high: there've been 11 in 105 innings in Australia, compared with six in 108 in South Africa.
In the last few games, though, the stats have been looking better for opening pairs: they average 38.16 in the last four years - there are three other countries (excluding Pakistan, which has hosted only five games during this period) that have lower averages. In fact, on India's last tour of South Africa, there were century stands for the opening wicket in successive innings of the Centurion Test, while on Australia's last tour there, Simon Katich and Phil Hughes added a massive 184 in Durban. For the most part, though, South Africa remains a country that offers more to the fast bowler than most other countries. If that holds true this season as well, then Johnson surely won't be complaining.
|Host country||Innings||Runs||P'ship average||100/ 50 stands|
|West Indies||89||3393||38.55||7/ 19|
|Sri Lanka||95||3582||38.51||8/ 22|
|New Zealand||86||2839||35.04||5/ 14|
|South Africa||108||3235||30.23||6/ 18|
S Rajesh is stats editor of ESPNcricinfo. Follow him on Twitter