In this post, I will generalize the idea for the period from January 1, 1990 onwards. I have done the following:
1. I looked at all the cricket grounds that have hosted Test Matches since January 1, 1990, and organized these according to the visitors' batting average at each ground.
2. I created two separate lists. In the first list, I included all the grounds where the visitors batting average was 32.5 runs per wicket or below. In the second, I included all the grounds where visitors batting average was above 32.5 runs per wicket.
3. I then compiled two lists of batsmen. The first list included all batsmen who had scored at least 1000 runs in Away Tests played at venues where the visitors' batting average was 32.5 runs per wicket or less. In the second, I added these players record in Away Tests at grounds where visiting batsmen had scored more than 32.5 runs per wicket.
The table is available below.
Before arriving at the 32.5 figure, I tried a number of other thresholds. In Test Cricket since January 1 1990, I found that a visiting team loses a Test Match 52% of the time when it scores32.5 runs per wicket (or 650 runs for 20 wickets) or less, while it loses only 26% of the time when it scores more than 32.5 runs per wicket.
The 32.5 runs per wicket measure should not be seen merely as a measure of the flatness of the wicket. It is, rather, a measure of the general difficulty of the conditions of play - which involve a combination of the strength of the opposing team, the strength of the opposing bowling, the quality of the wicket etc.
Cumulatively, I found the 65 batsmen who have scored at least 1000 runs in Away Tests in conditions where visiting teams have scored 32.5 runs per wicket or less, on average, scored 40 runs per completed innings (this is the conventional batting average) and survived 82 balls per dismissal. The same batsmen, in Tests where visiting teams scored more than 32.5 runs per wicket, scored 52 runs per completed innings and survived 101 balls per dismissal.
The table below is sorted by Batting Average in the first case (32.5 runs/wicket or less). Note that Andy Flower, who tops the list has made 1726 runs at 66.38 at grounds where all visiting batsmen, from all visiting teams, in all positions in the batting order have made on average 32.5 runs/wicket or less. So he is being measured not in relation to his Zimbabwean team mates, but in relation to all visiting batsmen who batted at those grounds from January 1, 1990 onward.
Only 5 batsmen average 50 in both cases during this period - Flower, Samarweera, Cook, Tendulkar and Saeed Anwar. Damien Martyn is the next best player on the list, while Chris Gayle is surprising prolific.
The Survival Rate (SuR) figure reveals at least as much as the batting average statistic. Sehwag, Gayle, Gilchrist and Jayasuriya appear as anomalies. They have poor survival rates in both cases (under 70 in all cases). Nearly all grounds in Australia feature in the first list, while nearly all grounds in Bangladesh feature in the second list.
This method gives rise to the possibility of a number of interesting analyses. One of the things I hope to do in the future is to develop an adjusted batting average. If you notice in the table below, some players play a far higher percentage of their Away Tests in conditions where visiting batsmen have struggled than others. For example, Michael Atherton played 68 out of his 78 Away innings at difficult venues (about 87%). while Tendulkar and Dravid have played 71% of their Away innings (108 out of 152 in each case) in difficult conditions. For this, the notion of "difficult" conditions - for example, is 32.5 really a better threshold than say 35 for Away batting? - as well as a few other distinctions need to be considered.
More on this later.
Full table after the jump
Records of batsmen since January 1 1990 at Away Test venues where teams have scored 32.5 runs per wicket or less
surprised to see Gayle above the likes of Ponting and Kallis.
ReplyDelete