Ben Stokes elected to enforce the follow-on in the Second Test between England and New Zealand, with assistant coach Paul Collingwood later arguing that it was the ‘aggressive option’ - and therefore the one best fitting England’s Bazballers.
New Zealand were bowled out for just 209 in their first innings, and even that was down to a captain’s knock by Tim Southee, who threw the willow about for a brutal 73 straight out of the Stokes-McCullum playbook. Leading by 226, Stokes asked New Zealand to bat again and saw his bowlers toil on an improving pitch as openers Conway and Latham put on a partnership of 149 for the first wicket, which was the foundation of New Zealand’s overnight 202-3 (now only 24 behind).
At the time, I felt that the better - and, actually, more aggressive - option for England was to bat again, particularly given the talent and speed of scoring in the team. There are few demons in this pitch currently, compared to days 1 and 2, and I think the hardest time to bat in the future will be the second half of day 5, as the pitch begins to wear. I know this England side will back themselves to chase any score, on any pitch, but there’s no need to make things harder for yourselves.
I’ll come onto what I think this decision means for the rest of the match later, but this decision made me curious - what are the relative success rates of a team asking the opposition to follow on?
What is the follow-on?
The follow-on is dealt with by Law 14.1, which reads:
14.1.1 In a two-innings match of 5 days or more, the side which bats first and leads by at least 200 runs shall have the option of requiring the other side to follow their innings.
So far, so simple: if you bat second, and fail to get within 200 runs of the oppostion, they can ask you to follow on. This holds as true if you’ve been bowled out for 40 in reply to 250, as it does if you make 400 where your opposition scored 601.
So a side that even has the option of making the other follow on is necessarily at a significant advantage in the game. If you follow on:
You’ve probably been bowled out cheaply (certainly relative to your opposition), suggesting they’re the better side in a given game; and
You still need to make at least 200 just to make the opposition bat again.
So what should a captain do?
We would, therefore, expect sides with the option to enforce the follow-on, to win a significant proportion of these games. And they do. There have been (excluding this Test, because its result is not yet known), 406 matches where one side has had the option to enforce the follow-on. This is about 16.3% of all Tests ever played.1
Where the follow-on has been available, it’s been enforced 72.66% of the time. So Stokes’ decision was not exactly mould-breaking in the context of all Test cricket. What’s interesting is that sides who are in a position to enforce the follow-on, regardless of whether they do, actually win an average of 80.79% of the games they play.
Sides enforcing the follow-on win 78.64% of the time - an overwhelming majority, for sure, but less than the average for a side in their position. So enforcing the follow-on actually makes you less likely to win the game. Sides who elect to bat again, by contrast, win 86.49% of the time - nearly eight percentage points more.
So far from being the ‘aggressive’ option, enforcing the follow-on actually makes you less likely to win. What it does do, conversely, is make you more likely to draw. Teams enforcing the follow-on draw more than 20% of the time - that’s one match in five. Where the follow-on is not enforced, only 11.71% of matches are drawn. So enforcing the follow-on is actually the conservative option, doubling your chances of drawing the match.
At this point, you might be thinking, sure - making the opposition follow on increases the likelihood you’ll draw. But surely it also makes you less likely to lose?
Well, yes - but barely. There have been three occasions where the side enforcing the follow-on has lost (amusingly, if you’re English, on each occasion it was Australia on the receiving end).2 Where the follow-on has not been enforced, the team who held the lead has lost on two occasions (albeit from fewer matches). But express this as a percentage, and that loss chance goes up only from 1.02% to 1.8% - still vanishingly unlikely.
The history of the follow-on, therefore, tells a clear story - enforcing it is the conservative option, all but doubling the probability of a draw while only reducing your odds of defeat by a tiny margin. If you want to win - or to take the aggressive option - you should bat again.
But what about this match?
Now, I’m quite happy to admit there are times when you should obviously enforce the follow-on.
Let’s say you’re a non-Asian country playing away in India. You win the toss and bat first, and you bat well, making 600 over two days on a pitch that was flat to begin with, but breaks up. You bowl India out for, say, 350 (putting them 250 behind) in another day and a bit, so it’s lunch on Day 4 and the pitch is really starting to turn. If you bat again, you might get rolled for 70, giving them 320 to win - unlikely, to be sure, but not impossible. Whereas, if you enforce the follow-on, even if they bat well and score that 320, you have a tiny target to chase on a poor pitch, which you’d back yourselves to get because it’s easier to manage your resources (wickets in hand) when you have a target to chase then if you’re batting third, trying to score enough runs while keeping time in the game to bowl India out.
Or, you might be playing in England and you made 450 in a day and a half during your first innings; then the rain comes and your opposition scramble to 230 over the next couple of days with rain interruptions and a day completely washed out. So you get to tea on Day 4, or even the start of Day 5 and your best chance of victory is to put the opposition in and hope to bowl them out again, rather than taking half a day to set them a nominal target and see them bat out for a draw.
So the follow-on is a situational call, but unless it’s blindingly obvious you should enforce it, like the second case above, I’m minded to say you should bat again. Actually, even in the India example, I’d bat again - but I can see the logic in enforcing the follow-on there.
In the case of this match, I think England should have batted again. The forecast for the rest of the match is mostly fine (there are scattered showers on Day 5 in Wellington) and there was plenty of time left in the game when England bowled New Zealand out yesterday. Why not bat again, make a quick 200, and set the Kiwis 450+ to get, with two days left in the match? Nobody’s ever chased that many in a Test match, and they’d have had to bat two days to draw.
Instead, New Zealand go to Day 4 effectively -24 for 3. They’re in a tough spot, sure, but Kane Williamson is one of the best in the world, and if he makes a hundred from here England could be made to chase 200-250. Unlikely, but certainly possible. Again, England - especially this England - will back themselves to chase that score, and probably would 95 times out of a hundred: but they might not.
Enforcing the follow-on, in this case, actually opened an avenue to victory for New Zealand that simply didn’t exist before. The Bazballers took the risky option, but in this case the risk is not matched by reward.
If you’re still not convinced, let’s take the reasonable worst-case scenario to England batting again. Let’s say they were bowled out for 70 within a session. Their lead would have been 290-plus, a score that New Zealand have beaten only once this series, in their first innings of 309 at Mount Manganui. England would still have been favourites.
Let’s Bazball, smartly
Whichever way you slice it, England will almost certainly emerge from the Basin Reserve clutching the trophy, a 2-0 series win, and an eleventh win from twelve games. They’ve played some super cricket in this series against a good (but not great) Kiwi side, and it’ll be a first win in the country since 2008.
I love the brand of cricket they play and am more than happy to risk losing in order to win. They just need to make sure their decisions actually do make them more likely to do that. In general, enforcing the follow-on makes you less likely to do both - and in this case, it opened a path for New Zealand to get back into the game. I want to see England play funky, attacking cricket, but I want them to be smart with it, too. In this case, Mr. Collingwood, they didn’t take the aggressive option.
Data source: http://www.howstat.com/cricket/statistics/Matches/MatchFollowOn.asp?FollowOn=T&Result=Won
Those matches, if you’re interested:
Australia v. England, Sydney, 1894 (England won by 10 runs)
England v. Australia, Headingley, 1981 (England won by 18 runs)
India v. Australia, Eden Gardens, 2001 (India won by 171 runs)