The Role of Age in White Ball Batting Performance
Does aging impact ODI batting perf?, Have T20s changed this?, and How hitters and rotators might age differently
A few weeks ago, I wrote about the role of age in red ball batting where we saw how some naive methods of looking at aging have inherent bias, dispelled the myth of the typical batter peaking beyond 30yrs of age and delved into some insights on how the level of competition determines how red ball batters peak and age. I also chatted with Jarrod Kimber about it on his podcast.
Here, I use the same methodology and build on the findings with data from the shorter formats. I intentionally lean more on ODI data and only leverage modeling work that I personally produced a couple of years ago in T20 cricket to avoid conflicts with my work at Zelus Analytics.
Comparing Test & ODI Aging
We saw that the typical red ball batter peaks in their early 20s, while really good batters relative to the level of competition continue to grow and peak in their late 20s to early 30s. They also decline at a slower rate.
Continuing to look at the batting average as a measure of performance, we can compare the aging curves for ODI and Test cricket.
Similar to aging in Test cricket, we see that the average ODI batter, who is only a year younger than the average Test batter, is also past their peak. However, their decline from peak is much less steep. The peak for a typical batter in ODI cricket is much flatter i.e. it lasts a lot longer than the peak for a Test batter. Another interpretation of this is that the level of competition is lower in ODI cricket; hence the later peak and slower decline, which matches the purist intuition.
Let’s test this notion further. With Tests, we saw that the top Test batters have had a prolonged and later peak. If ODI’s have a lower level of competition, the top players should not have a much flatter aging curve with a later peak than the typical ODI player. This is exactly what we see.
Although the separation between the average player in the respective format and the top player at the average player age is similar in both formats, age impacts the typical Test player much more drastically than it does the typical ODI player. The reverse is true for players in the top 100. The top Test players age quite gracefully. The top ODI players’ aging trend is similar to that of the typical ODI player. At least in part, this is because there is more separation between the best and the average in Tests.
Other Skills for the Shorter Format
So far, we have been looking at the batting average as a measure of performance. However, in the shorter format, the speed at which batters score is also relevant, arguably more so than the number of runs scored.
So let’s see if the aging curve for the strike rate as a measure of performance trends differently.
Clearly, SR seems to decline less steeply than the batting avg. It is important to keep in mind that the two measures are on different scales and changes in each have a different impact. The scales matter less because we are looking at values relative to the expectation and they happen to be in similar ranges for both metrics. But the impact of a 10 point difference in SR is different from the impact of a 10 point difference in the batting avg. is different.
The scoring rate has two key components — boundary hitting and strike rotation. Since one is a complement of the other let’s look at Non-Boundary SR and compare it with Overall SR.
When it comes to strike rotation, aging causes a very gradual decline in the SR. From the average age of 28 to 35 years of age, the overall SR declines by ~5 runs. ~2 of these are from running between wickets. So ~3 should come from boundaries.
Beyond 35, we see that the non-boundary SR declines more steeply, at a similar rate to the overall SR. This indicates some tradeoff with boundary trends. Batters must be rotating less but not scoring fewer boundaries. We already know that they score fewer runs (lower batting avg.), so they could be playing more failed cameos or holding up an end (anchoring?) more.
However, these point to some batter roles that weren’t always part of the one-day game. Let’s see if our scoring rate aging trends changed with T20s picking up steam. I chose to look at pre-2012 and post-2012 segments here.
Quite unsurprisingly, batter aging has adapted to the evolving trends in the game and the changing roles that batters have been put into. We see much flatter SR aging trends and only slightly steeper, if that, non-boundary SR aging trends post 2012. Again, this indicates a change in boundary trends, and indeed, we see this very clearly in the aging curve for Boundary % over expectation, the proportion of runs scored via boundaries above average.
We see that in the modern-era, batters have been able to improve their boundary hitting even as they age while their ability to rotate strike has continued to decline with age as it has in previous years. This could very well be an artifact of the focus on hitting in training. It is interesting to see that younger batters see a decline in boundary % before it starts picking up in their late 20s. If we look at the batting average above expectation, we see that these younger batters also tend to go through growth in overall run accumulation skills.
So, to summarize some of what we have seen, ODI batters in the modern era go through the following phases:
- Early 20s: Improve run accumulation by learning to play longer innings at the cost of boundary hitting. They are essentially learning how to pace an ODI inning after having started off taking a risk too many when they are younger.
- Late 20s: Get a somewhat stable set of performances at peak average, and a reasonably high SR.
- Beyond 30: Strike rotation suffers and they try to make up with hitting more boundaries, but as a result of the higher risk undertaken, their average suffers. The boundary% goes up, but the number of runs scored drops as well and so the boundary count isn’t going up. The number of balls faced isn’t dropping quite as quickly as the runs scored and so SR doesn’t go up.
The impact of aging that we see after 30 is accelerated after 35, when we see that risks undertaken come off even less in terms of dismissals, but more so in terms of dot balls, resulting in the steeply declining strike rate that we saw earlier. Here is the dot ball % aging curve that reinforces this inference:
When “Other Skills” Become Core Skills in the T20 format
Given that we are seeing a difference in ODI aging trends for skill areas that may be attributable to the rise of the T20 format, let’s look at how these curves compare to T20 aging curves for some of this core set of measures.
For T20 related expectations, which I used to compute the metric (e.g. SR) above expectation (plus-minus) measure, I used the models I had developed here.
Let’s start with the SR above expectation. The average age of a batter is almost identical for ODI batters post 2012 and T20 batters. So we are looking at the measure relative to the average batter at 29 for both curves respectively.
We see that batters past the average age decline rapidly in T20 cricket, whereas in ODIs they are able to decline much more gradually, especially until 35. This speaks to the demands of the game. What is an essential skill in ODI is a core skill in T20s and so the competition is much more fierce.
If we look at strike rotation, which is arguably a less critical skill for T20s, we see that the converse is true — the T20 aging curve is almost flat until the batter hits about 33 years of age, whereas ODI batters age quickly after 30.
This reinforces what we saw with my piece on red ball batter aging; that age influences performances when the relative level of competition is more cutthroat. We saw the top Test batters peaking later and aging more gradually than everyone else and Test batters playing in domestic first-class Cricket faring similarly. Here, we see a similar effect translate across skill types based on the competition level demanded by the format for that particular skill.
Looking at Pace vs Spin
With Test Cricket, we saw that aging curves were different against pace and spin. There was a distinct improvement phase against spin that was missing against pace, but the decline against pace was a lot more gradual than it was against spin when looking at Tests since 1999. Here, inferences relevant to the modern game are most interesting to us. So, we will stay focused on ODIs since 2012 and all T20s for the relevant measures.
Aging against pace and spin have been nearly identical as far as ODI batting averages are concerned. But when it comes to scoring rate, the aging curve against spin is almost flat.
This is interesting given the discussion we have been having around the core or primary skills and “other” secondary skills in the context of aging. The 50 over format is often considered to be a bridging format between Tests where the average is the primary batting metric and T20s where the SR is arguably more critical. We have seen that T20 batters age more steeply than modern ODI batters in terms of SR. Let’s look at T20 aging against pace and spin in light of the above.
While batters still seem to age slightly more steeply against pace, the difference is very subtle and probably within noise. Moreover, there is no indication of aging stabilization as there is with ODIs.
Note, that considering primary skills for white ball, these curves are quite different from the aging curve differences that we saw in Tests.
Seeing the above curves for Tests, I speculated that survival is more reflex dependent against pace whereas batting against spin requires more active decision making and footwork which results in a steeper decline after a period of growth. For ODI averages, this clearly doesn’t seem to hold. One explanation might be the extent to which the pitch and conditions come into play in Tests vs ODIs. Another is, as we have seen before, the level of competition.
Going back to scoring rates, as we have seen before, in addition to how fast they score, how batters score is interesting to look at and even more so when it comes to pace vs spin. Let’s look at this for T20s where we know, intuitively as well as based on what we have seen so far here, that it is a core skill and where batters are required to push the limits more than they are in ODIs.
We have already seen that the rate of overall SR reduction against both pace and spin relative to the average/expectation, is similar. Here, we see that as batters age, they start doing better on strike rotation against spin and have fewer dot balls as a result. This could be a result of batters being put in anchor roles as they age (anchors tend to play more deliveries in the middle overs which tend to be bowled by spinners) or a result of them attempting to make up for a diminishing ability to hit boundaries against spin. Against pace, we see a rapid decline in strike rotation that starts before the average age of 29. The question is whether this decline in strike rotation is the primary reason for a decline in SR or whether an additional decline in boundary hitting contributes to it.
Looking at this from both the run contribution of boundaries (boundary%) perspective as well as the boundary rate (boundaries per 100 balls) perspective, we see that there is a steady but more gradual decline against pace than there is against spin. This shows that hitting gets much harder against spin than it does against pace as batters age. However, the additional decline in hitting ability against spin can be somewhat compensated by improved strike rotation.
To recap, we saw evidence that reinforced the inference that age related performance trends are relative to the level of competition. When a skill is harder to execute well and when there is competition from peers executing that skill or a challenging opposition, age plays a big part and the age related decline starts very early. This is the case for core skills rooted in the demands of the format and skills where there is less separation between the best and worst players. We saw that boundary hitting trends in ODIs have been influenced by the T20 format and we categorized aging related performance decline into three phases. Finally, we looked at aging with respect to scoring rates under both the strike rotation and boundary hitting paradigms and then split these by pace and spin to see that it gets much harder to hit boundaries against spin, but it is possible to improve strike rotation.
If you enjoyed this piece, check out more of my work at Boundary Line and follow along here & on twitter @amol_desai
