First-over specialists

Tim Wigmore @timwig recently wrote a piece on players specialised in the "art of the first over"... the surprisingly common phenomenon of otherwise part-time bowlers being called on to open the innings for the bowling team

The article featured some of my own analysis. Although it is fair to say that both 'featured' and 'analysis' are probably strong words to use. My numbers contributed to just a tiny aspect of the overall argument and that contribution boiled down to just a simple stat comparing the economy between full-time and part-time bowlers in the first over

Can't resist an excuse to use a picture of Chris Gayle, even in a post about bowling

Can't resist an excuse to use a picture of Chris Gayle, even in a post about bowling

This is not me grumbling. The article does a much better job than I ever could of explaining the subtleties behind the strategy... a strategy that I should admit I had never even noticed until Tim pointed it out over some beers on a Friday afternoon, highlighting a match that I had actually attended to demonstrate his point. There is a lot of subtlety to consider: match-ups dictate how risky it is to expose a part-timer during the fielding restrictions, whilst the value of the flexibility afforded by such a cheap first over is devilishly tricky to quantify

One particular area where the numbers cannot provide insight is the behind-the-scenes thinking that goes into teams using such a tactic. It may seem obvious, but reading that "it's about 'this is probably the best match-up for this situation and at the same time we're pinching an over'" (CPL Director of Cricket) makes it a lot easier to appraise the tactic itself

However, in this post I want to delve a bit deeper into the numbers themselves and apply an analytical lens to establish what these numbers tell us

In the numbers that follow, I define a part-time bowler to be any bowler who averages 2 overs or fewer per match across the duration of an entire tournament. This means that I am limited to matches which belong to an easily identified tournament. The tournaments considered here are all the World Cups, all IPL seasons, all Big Bash, Pakistan Super League 2016, and the T20 Blast 2016. Everything else should be fairly self-explanatory

The first thing to note is that it is definitely a 'thing' to bowl part-timers in the first over. This may not even seem worth stating but it is always good to see the numbers backing up the common sense. The tactic of bowling part-timers is used more in T20 Blast and Big Bash matches (5.5%) than in the IPL (3.8%). But it is used most in World Cups between International teams (6.8%)

Chart showing the usage of part-time bowlers during a match

The next question is the initial cost of bowler a less skilled bowler in the first over of the Powerplay. And predictably, a part-time bowler is less effective than a full-time bowler, both in terms of taking wickets and conceding runs. That sacrifice for using a lesser bowler is roughly 1.5 runs in the over

Except from the Cricinfo article on "The art of the first over"

Except from the Cricinfo article on "The art of the first over"

Is that a lot? To give some context, there are only two bowlers* who I have as rated 1.5 runs per over better than average during the 2017 IPL season (minimum 120 balls). To phrase it slightly differently, there were only two bowlers in the IPL who represent a greater upgrade over an average bowler than simply not playing a part-timer in the first over

And this doesn't account for the reduced wicket taking threat. On average, a non-specialist bowler takes 0.07 wickets fewer in that first over. If we take a fairly conservative assumption that a wicket is worth about 5 runs then this equates to an additional 0.35 runs forfeited by bowling the part-time

* Chris Morris and Andrew Tye combined for 25 wickets in 15 matches at an economy of 7.29 and 2 MoM awards during the regular season. These might not seem like world-conquering numbers but they look even better when you consider that 55% of the bowling occurred either during the Powerplays or at the death. Note that I am not saying that they are necessarily the best two bowlers in the IPL, only that they performed best over a relatively short period

The Benefit

of course, this doesn't tell the whole story, there is clearly a benefit to sacrificing value in the very first over. The argument is that most teams will need to bowl a part-timer at some point anyway in order to fill the full 20 overs. Most teams wait for a favourable match-up to present itself during the course of the match but using the first over seems like a valid alternative

Accepting the assumption that we will need to a part-timer at some point, the question becomes whether or not it is better to use them immediately or wait until the middle overs of the match. The charts below show the difference between using part-timers and specialist bowlers for every over, in terms of both runs (top) and wickets (bottom)

Charts showing the difference between part-timers and specialist bowlers by over in both the first and second innings. Differences in both runs and wickets are shown

The are a few interesting stories in these charts, although we need to be careful of the small sample sizes. I particularly enjoy seeing the part-timers taking wickets at the death in the first innings. Batsmen feel the opportunity to swing for the fences and end up getting caught - to such an extent that they actually score fewer runs in the last over than they would if they had faced a regular bowler

04 final chart.PNG

But most of the part-time action occurs in the first over (green) and the middle overs of an innings (blue). There does not seem to be a big difference between these two options

This suggests that teams are already doing a decent job identifying the correct opportunities to employ a part-time bowler. All else being equal, I suspect that the middle overs are still the best time to use a part-time bowler. And this aligns with the fact that the middle overs are, indeed, when they are used most often (about 8% of overs)

On the other hand, bowling a non-specialist in the first over is usually a bad idea but sometimes (about 5% of the time) the match-ups are favourable enough that the team can deploy a part-timer without sacrificing any more than they would at another point in the innings

The one area where part-timers do seem under-utilised is in the seventh over of the second innings. It seems that batsmen use the seventh to reset themselves regardless of who is bowling. A non-specialist only concedes 0.5 runs more than a specialist would. As it is, Over 7 sees teams use part-times just 6.5% of the time - less than in Overs 8-14. The numbers suggest that 6.5% is too low

In the future, I'd love to be able to break this down further based on the exact match-ups between batsman and bowler. As it stands, my numbers can't divine much beyond how often teams deem it fit to deploy this specific tactic. For now, I am happy to rely on an actual journalist to capture the nuance I can't access:

"There is no right answer about whether a team should bowl one of their weaker bowlers at the start of an innings. Like all T20 strategy, the decision is determined by a mixture of meticulous forward planning and instant in-game judgement. But any young T20 batsman who can bowl useful medium pace or spin would be well-advised to learn from Stevens: mastering the art of bowling the first over would boost their worth in the Blast and leagues around the world"