The Question of Intent

Email: Sportsanalyticsadvantage@gmail.com

If this article has given you insight into the data that Sports Analytics Advantage can offer cricket teams around the world in formulating team strategies, selection, draft or auction plans, or any other work, please feel free to get in touch at sportsanalyticsadvantage@gmail.com.

22nd October, 2020.

The subject of intent has been discussed on numerous occasions in the current IPL, with there being more than several examples of players and teams seemingly lacking intent with the bat.  Notable recent team examples include Chennai Super Kings totalling just 125-5 from their 20 overs against Rajasthan Royals, and Kolkata Knight Riders limping to 84-8 against Royal Challengers Bangalore.

Following the CSK loss to RR, captain MS Dhoni said that 'this season we were not really there', while KKR coach Brendon McCullum mentioned that 'we spoke at length before the game about trying to be positive and show some strong intent', that 'unfortunately tonight, if anything we probably lacked intent', and they were a little bit timid in their approach.  The fact that an IPL team lacks intent is seemingly rather odd from the outset.  Teams get their pick of the best players from around the world, and the overwhelming majority of T20 watchers would say that the IPL is the highest standard domestic league.  With this in mind, the purpose of this article is to discuss the different types of intent, the impact of those and why IPL batsmen lack intent.

First of all, I want to look at the different types of intent, with there being two main types - boundary-hitting, and non-boundary runs.  These are pretty obvious with boundaries being earned by attacking shots, and non-boundary runs coming down to dot ball avoidance.  However, there is often a trade-off between the two types of intent - most boundary-hitters aren't above-average rotators of the strike, and the same goes for elite rotators of the strike - they aren't usually above-average boundary-hitters.  This usually comes down to a mentality and strength issue - some players willing to risk 'dotting up' in order to attack with boundaries (e.g. they play and miss more) whereas the opposite type of players are willing to earn regular runs via running between the wickets, while not being either willing or able to find or clear the boundary rope on an above-average basis.  

Some players also have a dynamic where they look to start slow and subsequently accelerate when well-set, while others are happy to tee-off from close to the start of their innings.  Both strategies have different risks attached.  The latter strategy - attacking early in an innings - can lead to loss of quick wickets but has the obvious positive trade-off with potentially quick high scoring.  The former strategy (starting slowly before accelerating) is probably perceived as being more risk averse and sensible by many in the game, but actually could be even more riskier - if a player plays slightly above a run a ball or worse for a considerable spell of balls (say 12 balls or more) then they risk wasting a considerable proportion of their teams batting resources if they are then dismissed.

This difference in the dynamic of various batsmen is evidenced by the chart below, which looks at boundary percentage and non-boundary strike rate for the 38 current players facing 100+ balls in this year's IPL:-


Here we can see that the likes of Rohit Sharma and Shane Watson have solid boundary percentages just shy of 20% - around 2-3% better than the average tournament figures - but with low non-boundary strike-rates.  Essentially they score a large percentage of their runs in boundaries, while Virat Kohli and David Warner are taking the opposite approach with a low boundary percentage but high non-boundary strike rates.  Kieron Pollard and AB De Villiers are close to the ideal top-right hand corner with strong boundary percentages and non-boundary strike rates, indicating that they have maximised their scoring options in this year's IPL.

As well as these noted players there were also some further players with mediocre or unimpressive figures for one or, worse, both metrics.  These players have lacked scoring intent so far in the competition, either via boundaries or non-boundaries, and this will inevitably contribute to a low overall strike rate.  With these players it might also be prudent to wonder whether they are either batted in their best role by their team or indeed, whether they simply lack the ability to succeed at this level - more on that later.

It is also worth noting that these players lacking above-average levels of intent are unlikely to positively impact teams even if they have additional skill-sets such as being an all-rounder.  All-rounders tend to bat in the 5-8 slots (often 6-7) and these roles typically demand a high level of intent given that batsmen in these spots are often batting in the death overs.

So, which type of intent is better?  Boundary-hitting or dot-ball avoidance?

The chart below illustrates the relationship between boundary percentage and team runs for teams batting first (with the capability of facing 120 balls) in the IPL in the 2017-2019 seasons:-

Here we can see that there is an extremely close relationship between innings boundary percentage and innings runs.  
Looking at the chart, we can also see that if a team scores in excess of 20% of boundaries, then they will almost certainly score 170+, and the overwhelming majority of team scores of 150 or below feature boundary percentages of 15% or below. Very few teams with below-average boundary percentages scored 170, and no team broke 170 with below 15% boundaries in their innings. No team broke 180 with below 16.67% boundaries (an average of one boundary every six balls). Only two of the 38 teams scoring 190+ scored fewer than 20% boundaries in their innings, and the lowest boundary percentage for a 200+ total was 18.33% (and an incredibly low 20.83% dot ball percentage) giving even more evidence - if you somehow needed it - that boundary percentages are an absurdly huge driver of high scoring rates.

How do dot balls (balls faced by batsmen where they do not score any runs) influence team scores? The chart below illustrates innings runs versus dot ball percentage for the same sample of first innings:-

While there is a broad relationship between dot ball percentage and innings runs, there are many more outliers than that between boundary percentage and innings runs. For example, 10 of the 38 teams scoring 190+ still had 40+ dot balls - an unimpressive number. On the flip side, even some teams scoring around the 150 mark had fewer than 40 dot balls, but in these instances they were hamstrung by very low boundary percentages. 

A look at potential peak scores based on dot ball figures is also useful. It’s very difficult to score even around the 160 mark with poor boundary percentages, although Sunrisers Hyderabad twice featured with some outlier results here. They scored 160 with 10.83% boundaries and 25.83% dot balls, and also managed a 159 score with 12.50% boundaries and 35.83% dot balls.

Interestingly, though, it is possible for teams to score 170+ even with 50 dot balls. In isolated matches in the last three seasons of the Indian Premier League, Mumbai Indians scored 173 with 20.00% boundaries and 51 dots, while also, quite incredibly, posted 213 with 25.83% boundaries and 49 dots. Kolkata Knight Riders scored 175 with 20.83% boundaries and 55 dots, while Kings XI managed to score 170 with 19.17% boundaries and 56 dots.


So, from these numbers, we can establish that it is quite possible to score an above-average score even with a high dot ball figure or percentage, but it is virtually impossible to score an above-average score without a higher than average boundary percentage. Therefore, we can conclude that avoiding dot balls is desirable for teams in their pursuit of big totals, but a high boundary percentage is necessary to achieve this.

Giving too much weight towards dot balls and rotation is a dangerous philosophy - an approach seemingly advocated by the aforementioned Kohli and Warner this year - yet I often hear coaches and commentators alike discussing these areas as if they are of paramount importance. It is difficult to know why this is the case, but reasons for this could span several diverse areas, such as ‘conventional wisdom’ formed without utilising any data, crossover from other formats such as Test or ODI cricket, or simply that it is potentially easier to coach a player to rotate better than to coach boundary-hitting (which usually is hugely driven by physical strength and player mentality) in the short time-frame that franchise league coaches have players to work with in a given tournament.

A further way of looking at this is looking at individual batter data.  While looking at boundary percentage versus non-boundary strike-rate is useful to ascertain a profile for a batsman in how they score their runs, I want to go further now and look at the implications of the two metrics on a player’s strike rate. The chart below illustrates the boundary percentage versus strike rate for IPL batsmen facing a minimum of 300 balls in the tournament between 2017 and 2019, with the intersecting lines showing the mean tournament figures:-

The first obvious conclusion that many readers may draw is that there is a fairly strong relationship between a player’s boundary percentage and their strike rate, with the distribution of players generally sloping from the bottom-left (low for both metrics) to the top-right (high for both metrics). It is certainly reasonable to state that a player’s boundary percentage and how they score their boundaries is a considerable driver towards their overall strike rate. 

This is also apparent when looking at the players with the highest strike-rates, and in particular, Sunil Narine and Andre Russell. The two West Indies all-rounders recorded, by some distance, the two highest boundary percentage figures in the Indian Premier League during this time period, and they also recorded the two highest strike rates, also by a reasonable distance. Interestingly, though, while Narine recorded a higher boundary percentage than Russell, it was Russell who recorded the higher overall strike rate. The primary reason for this was that Russell hit almost double Narine’s already stellar 10.40% six-hitting percentage, and had a six percentage of over 8% greater than any other batsman facing 300+ balls in the Indian Premier League between years 2017 to 2019 - so when Russell hits a boundary, it’s much more likely to be a six, while when Narine hits a boundary, it’s much more likely to be a four.

Essentially, a player can be as good a rotator as they can possibly be, but without a strong boundary percentage and/or six-hitting percentage, they won’t be able to record high strike rates over a long period of time. This is evidenced again in the above chart, with only two players - Kane Williamson and Sanju Samson - being able to record below-average boundary-hitting percentages (and only just) but being able to record above-average strike rates. Every other player who faced 300+ balls with a boundary percentage below the tournament mean was unable to generate a strike rate in excess of the tournament mean.

As you can see from the two graphs, there are in excess of 20 batsmen in the Indian Premier League with below-average boundary-hitting percentages, and there are various reasons (justified or not) for their recruitment and selection. However, paying large sums for these players is not recommended, unless they offer additional skill-sets (captain or genuine frontline bowler, for example).  Certainly, teams need to be very mindful of recruiting numerous below-average boundary-hitters, unless they have an elite bowling attack which is capable of regularly defending 150-160 type totals. 

Following the CSK defeat discussed above, captain MS Dhoni stated 'Result is always the byproduct of the process' - he's spot on. So why do IPL teams - and indeed T20 franchise teams around the world - have players who seemingly lack intent, or have the 'wrong type of intent'?

The responsibility of this mostly has to lie at the hands of the decision-makers (coach, general manager, owner) at the teams, who are making the decisions to recruit players.  Some of these players recruited are ill-equipped for the task ahead, and it's unfair to blame the players for struggling when they are essentially set up to fail because they shouldn't have been recruited in the first place (or shouldn't have been recruited for such high salaries making them marquee players).  Why these players are recruited is an interesting discussion and something I have already mentioned in great detail in a previous article for The Cricketer magazine.

No-one is forcing these teams to sign these players - there is a free market for players - and there are various recruitment metrics which would be accurate at displaying player intent both at the beginning and throughout a batters innings.  There are various players around the world who have high levels of intent but who are not currently recruited at T20 franchises and it is 100% the decision-makers choice at franchises to sign slow-starting, low intent players, or players with poor boundary-hitters as opposed to players with more intent.  As cricket begins on its slow path towards recruitment efficiency, this situation will change in the future, but at the current point in time, numerous players are recruited who it is difficult to justify from an intent perspective.



If this article has given you insight into the data that Sports Analytics Advantage can offer cricket teams around the world in formulating team strategies, selection, draft or auction plans, or any other work, please feel free to get in touch at sportsanalyticsadvantage@gmail.com.

Comments