Episode 15 of The Pitch Side Experts Podcast, featuring CricViz analyst Freddie Wilde.
The Podcast
In the latest episode of The Pitch Side Experts Podcast, presented by former West Indies fast bowler, turned commentator Ian Bishop, former Australia all rounder, turned coach Tom Moody and CricViz Analyst Freddie Wilde, the trio discuss why and how pace bowling is dominating Test modern Test cricket. You can listen to the full episode of the podcast below on Spotify. The Pitch Side Experts is available on all the primary podcast platforms, including iTunes and Google Podcasts.
The Data
CricViz data is used throughout the episode. Some of the key statistics and bits of analysis are outlined below.
Firstly, analysis of pace bowling averages is used as a stimulus for the discussion. A three year rolling average of pace and spin averages in Test cricket shows that the current three year pace average of 26.09 runs per wicket is the lowest in a three year block since the 1950s. Furthermore, the gap of around 13 runs between pace and spin averages is one of the biggest in the post-war era.
Bowling averages are the product of two things how fast runs are scored – economy rate, and how quickly wickets are taken – bowling strike rate. The reason pace bowling averages are so low in the modern era is because strike rates are at a historic low. Wickets are being taken faster now than ever before.
One theory as to why batting averages are so low is that modern batsmen are too attacking. However, CricViz shot-type data suggests that in the last three years—when batting averages against pace have been at their lowest—batsmen have actually been attacking less often than in the ten years prior to that when attacking shot percentages were as high as 26%, while now they are around 22%.
Shot-type analysis suggests that batsmen’s defences are being breached more often now than at any point since such data has been recorded. Defensive shots per dismissal is essentially a measure of defensive solidity and this shows that defensive shots per dismissal in the last three years have been around 50 whereas in the ten years prior to that they reached as high as 75.
The paucity of defences is such that in 2020 a batsman is just as likely to get out defending as they are attacking. The line graph below shows the percentage of wickets that have fallen to pace according to whether the batsman played an attacking or a defensive shot and this year those shot-types each contribute to 38% of dismissals. The remaining 24% is made up of what we call ‘rotating’ shots, neither full-blooded attacking or defensive shots, and the rare occasion when a batsman is dismissed not playing a shot at all. If you include ‘no shot’ as a defensive shot then 2020 is the first year on record when batsmen are more likely to get out defending than attacking.
The fact that defences are being breached more regularly does not necessarily imply poorer batting. It could be the consequence of tougher conditions and, or better bowling. Using CricViz’s eXpected runs and wickets model, which evaluates the difficulty of batting by analysing ball-tracking data and assessing how likely any given delivery is likely to take a wicket or concede runs by comparing the characteristics of those deliveries to similar deliveries in our database and looking at the runs and wickets produced by them. This model shows that in the last few years eXpected pace bowling strike rates are lower than they have ever been. eXpected strike rates enable us to analyse the quality of batting by comparing the eXpected figure with the actual strike rate. For example, in 2018 the eXpected strike rate was 55.5 but the actual pace strike rate was 51.8 – meaning that the batsmen underperformed because they got out 3.7 balls more often than we would expect. The chart below shows the eXpected and actual pace bowling strike rates for every year since 2006. This shows that pace bowling is better than ever before but in 2018 and 2019 batsmen were still underperforming slightly. So far in 2020 they are beginning to fight back.
Closer analysis suggest that the low eXpected pace bowling strike rates appear to be a consequence of tougher conditions and better bowling. Using ball-tracking data we can show that in the last few years pace bowlers are finding less swing than in the decade previously but are finding significantly more seam movement and seam movement, because it happens later, is more dangerous than swing. The table below shows ball-tracking data from around the world across the last 15 years.
Pace bowlers are getting better too. In Test cricket balls on a good length (6.25 metres to 8 metres from the batsman’s stumps) and a good line (a 40cm channel outside the line of off stump) are the most effective areas to bowl to most batsmen and in most conditions. The ability to hit these areas is not the only thing that defines a good Test bowler but it is a big one and analysis of ball-tracking data suggests this is something that bowlers are getting better at doing. The percentage of balls on a good line and length in the last few years has been higher than ever before.
Freddie Wilde is a CricViz analyst, @fwildecricket.
