Sunday, January 13, 2013

How Do ODI Batsmen Perform In Specific Batting Positions?

The tables in this post show the Performance of a player in a particular batting position, in comparison to the performance of the top 8 batsmen in the line up. The first four column are self explanatory. The way to read this table is that Sachin Tendulkar, when opening the batting, scored 13 runs more for his dismissal in the average ODI game, when compared with batsmen in the top 8 batting positions on both sides in that game, at a strike rate which is 13% better than that of the top 8 batsmen.

By comparison, Saeed Anwar score 9.4 runs more than the top 8 batsmen in games that he played at a strike rate which was 9% better.

A 9% superior strike rate means that the player would score 9 runs more per 100 ball faced, than the batsmen he's compared to.

One interesting thing I found with this data is that Number 3 batsmen tend to score slower than openers and Number 4 players. Even Viv Richards, who scored 32% faster than other batsmen in games that he played at number 3, scored 45% faster than other batsmen in games that he played at number 4.

Inzamam-Ul-Haq scored more than 2000 runs at Number 3, 4 and 5, scored 3% slower than other batsman when he batted number 3, 1.5% slower at 4, and 6% faster at 5.

The trade offs for players who took more chances (Gilchrist, Sehwag and Jayasurya for example) are clearly visible. A low Performance score reveals that these players tend to make runs in conditions where others also make runs.

What the figures also show is that the great finishers (Dhoni, Bevan), are not great finishers because they score quickly, but because they know when not to. Bevan, for example, scored 13% slower than other batsmen when he batted at number 4, but scored 4% faster than others when he played at number 6. His batting average was about the same in each case. His Performance rating was superior at Number 4 than at Number 6.

The Performance (PERF) of a player is not the same as an average. It takes into account when a player scores runs, and when he doesn't. Runs made in a low scoring game are worth more, not outs in a low scoring game are worth even more, as the following formula ensures:

Performance = (Player's Runs) - (Player's Dismissal)*(Runs/Wicket for given ODI)
Runs/Wicket for a given ODI is given by Total runs scored in ODI / Total wickets lost in ODI.

A high performance score suggests that a batsman tended to control his team's innings very often. Only 4 batsmen scored more than 2000 runs in a single batting position, and did so scoring 10% faster and performing 10 points better than their team - Tendulkar (opening), Lara (opening and at 3), Viv Richards (at 3 and 4), Shane Watson (opening), Andrew Symonds (at 5), Graeme Hick (at 3) and Zaheer Abbas (at 3).

There are 4 obvious names there, and three not so obvious ones. Of these, I wonder whether Shane Watson can sustain his form.

The tables below are sorted by aggregate.


  1. Hello,

    Well done on a nice blog! I enjoyed browsing the site and reading some of the articles.

    The reason for my email is to see whether it might be possible to display our link details on your homepage?

    Please take a look at our high quality and high traffic website at

    Please let me know what you think.



  2. With the exception of the openers, I wonder, but don't know, how significant it is to analyse a batsman's performance by his position in the batting order. The batting order number tells us how many batsman have batted before him and how many will come in after he has. But it's no more than a proxy for the most important variable - the state of the game. That, I concede, is a less easy metric to extract from scorecards and databases, but a combination of 1st or 2nd innings, overs completed and strike rate at the point the batsman begins his innings, could be used to create some common scenarios with which to assess each batsman's performance. I do have a thing about batting order: .

    Other than that, your usual impeccable analysis!