Is there a good time to lose wickets in T20s?

There seems to be some ambiguity about the worth of a wicket in T20s. Especially if they come at an expensive cost. A spell of 2/24 would always seem good but the opinion seems to be divided when it comes to figures resembling 4/40. 

Some people think that they come at a huge cost while others think that a wicket is a wicket and it always adds value. A game as complex as cricket does not allow strict notions to remain unquestioned for long. One day you might think that strike rate is of paramount importance and the other day this elusive game throws an LSG vs RR match at you where the most prudent approach was to keep wickets in hand. 

There is always some nuance that you will end up missing while making a statement about how the game should be played. It’s almost like a British economist saying ‘Whatever you can say about India, the opposite is also true’, only replace India with cricket. 

This debate about the importance of wickets in T20s is one among many that need a bit of nuance to be fully comprehensible. In this piece, we have tried to see the consequences of a wicket falling in T20s. Most of what we have done revolves around the scoring rate. This is because T20 batting is an optimisation problem where you have limited resources (120 balls) to maximize the output (runs). Theoretically, it can be said that run rate is the most important metric in T20s.

What happens when a wicket falls? 

We have compared the run rate 10 balls before a wicket falls and 10 balls after it. The matches considered are T20s post 2018 across major leagues. 

What we see in the graphs below is fairly obvious. The fall of a wicket leads to a dip in run rate in both innings. The drop is across phases except for the first two overs. 

That exception is because if you lose a wicket or two early in the innings, you’re already at rock bottom and the only way forward is up. What is interesting is that the drop is consistent even in death overs where one would assume that the batters have the highest intent to hit. In any case, one can conclude that the fall of wickets inevitably leads to a drop in run rate in most of the matches even in the phase where there is hardly any question about the intent of the batters. 

Is there any condition where the fall of a wicket actually leads to an increase in the run rate? Yes, there is. 

The condition is when the scoring rate in the last ten balls has been slower than the scoring rate of the innings at that point.

We can see that in the middle of the powerplay and especially in death overs, a fall of a wicket does not lead to a fall in scoring rate when the batting team had slowed down in the last ten balls. This might seem very obvious but please note that overall, there is a drop in run rate after a wicket falls at the death. This goes on to explain that when the scoring rate has picked up recently, a wicket is detrimental to the batting team but it is actually fine for the batting team to lose a wicket when the scoring rate has gone down. 

In the recent CSK match in which Rahane went ballistic, the first two wickets fell as soon as the scoring rate in the last ten balls became less than the scoring rate of the innings and it was quite evident in the way their batters were showing intent. While in the same match, KKR kept losing wickets when their scoring rate in the last ten balls was above the scoring rate of the innings. Four of them falling in the second half of the innings made it worse for them. CSK ended up winning the match.

In another extreme where 154 was defended by LSG against RR, they lost their first two wickets when the scoring rate in the last ten balls was lower than that scoring rate of the innings. Keeping true to the trend, their scoring rate picked up after that. Though these wickets fell in the middle overs. RR also lost their first two wickets in the middle overs. There was one difference - RR lost their first two wickets when the scoring rate had picked up in the last 10 balls and was above the innings scoring rate. According to the trend, they lost their wicket at the wrong time. 

Most of the conversation around the success of an innings (please note that we said innings, not match) can be pinned down to the timing of the wickets falling. One can of course say that take ten wickets and get done with it but that would only happen in a bowling paradise and as far as T20s are concerned, a bowling paradise is as real as Mordor. There are good moments and bad moments at which a wicket should fall for both sides. A change in scoring rate in the last ten balls is one of the ways to give some flesh to this conjecture. 

As cautioned above, one should always be wary of generalizing in cricket. In the same match, there were two instances where RR lost the wickets at appropriate times (i.e. when the scoring rate in the last 10 balls had gone below the scoring rate of the innings) - one in the 14th over and one in the 16th. But their scoring rate did not see any lift, which is against the trend we saw. I am sure there would be more such matches where such exceptions exist. 

After all this, the question of how players should approach this comes up. Should Kohli get out deliberately if the scoring rate in recent balls has gone below the innings scoring rate? Well, the simple answer to this simple question is that he should play the conditions. Only that the conditions also include the recent drop in scoring rate, regardless of the innings.