Aaron Finch was certainly facing a dilemma going into the final over of last nights match against the Sixers.
Four wicket’s down, the Sixers needed 8 runs to tie and 9 to win off the final over. Brad Haddin was on strike and the set batsman, having made a measured 23 off 19. At the non-strikers end, Jordan Silk had made 6 off 7.
Finch had the option of bowling Pattinson, Tremain or one of his two part time spinners in Short and Cooper.
Who would you have bowled in this situation?
Before answering, let’s take a look at some stats.
Across the IPL, BBL and CPL, spin has historically only been used to bowl the last over of the second innings 13% of the time…and for good reason. The economy rate for spin in the last over of the second innings is 11.0 runs per over, compared to 9.7 runs per over for non-spin bowling.
So at first glance, “non-spin” looks like the right option. The bad news however is that the runs per over for both spin (11.0) and non-spin (9.7) in the 20th over are both on average higher than the 9 runs the Sixers needed to win the match. So, the Sixers were definitely sitting in the box seat. And how pretty was the box seat they were sitting in?
Historically across all CPL, IPL and BBL matches, the batting team has won 76.7% (roughly 3 out of 4) of all matches where they have gone into the final over requiring 8 or 9 runs to win. The chart below shows the win rates for the batting team across all ranges of runs required in the last over.
For any given range of runs required in the last over, the historical win rate has always been higher for the batting team against spin bowling compared to non-spin bowling. Another strong case to opt for non-spin in the last over!
So what we have shown so far is the actual win rate for teams batting second, given a range of runs required off the last over split by spin and non-spin bowling.
What about the expected win rate? What is the expected win rate?
Suppose the batting team requires 13 runs to win off the last over, and the person chosen to bowl the last over historically goes for an average of 14 runs per over in the death overs. Then even though teams batting second only win around 40% of the time they are chasing 13 runs, in this example, because of the bowlers poor economy rate, the batting team should be expected to win. The chart below shows that the historical win rate for batting teams “expected” to win, given the bowlers poor economy rate, is indeed higher than the normal win rate without taking into account the bowlers economy rate. For example, chasing 10-15 runs to win in the final over, the batting team will win on average 48% of the time. This increases to 75% of the time if the bowler has a higher average economy rate in the death overs than the runs required.
So where does that leave us?
Right. The Sydney Sixers required 9 runs to win off the last over. Of the two non-spin bowling options available to Finch, Pattinson historically goes for 11.0 runs per over in the death overs, while Tremain goes for 9.3 runs over in the death overs. Both had an economy rate higher than the runs required, so the chance of the Sixers winning with either of those two bowling would have been at least 83%. At face value Tremain might appear to be the better option however he was going at 12.7 runs per over from the 3 overs he bowled in the match to date.
So what about the two spin bowling options available to Finch: Cooper and Short?
Firstly, neither had bowled in the death overs in a T20 match before. Also, given the earlier stats shown on spin bowling in the final over, neither really looked an attractive offer even though the two overs they bowled between them earlier in the match only went for a combined miserly 7 runs!
To answer this question, it is important to look at the strengths and weaknesses of the batsmen at the crease, Haddin and Silk.
Against “non-spin”, both Haddin and Silk score at an average of 8.6 runs per over. Therefore facing either Pattinson or Tremain, who both go at more than 9 an over at the death, it should be expected that Haddin and Silk would be able to get the Sixers over the line with the 9 runs required.
Against spin the story is slightly different. The man on strike, Brad Haddin, scores at 9.3 runs per over against spin, while Silk only scores at 6.2 runs per over. Haddin would be expected to get the Sixers over the line if he remained on strike, while the Renegades would have hoped to concede a single to Haddin and focus on keeping Silk at the strikers end. The question would be whether Cooper or Short had the nerve and skill to deliver in the final over.
So back to the original question. Who would you have bowled in this situation?
It would be a brave captain to have turned to one of his part time spinners to deliver the final over. However, given that stats, given the bowling form to date in the match, and of course given the wonderful gift of hindsight, it might be been a better calculated risk to have used Short or Cooper over James Pattinson.