There have been a few extraordinary run chases in this IPL. Andre Russell chased down 53 in the last three overs on March 24 and again on April 5. Kieron Pollard chased down 133 in the last ten overs on April 10. Rishabh Pant made a requirement of 111 in 60 balls look easy on April 22.

The brief nature of each T20 innings makes batting first and second to be different propositions, far more so than in Tests or ODIs. Scoring rates tend to be well north of a run a ball as a rule. While batting first, this means that setting a "good" or "competitive" total can mean pursuing anything between 180 and 220, depending on the conditions - a difference of two runs per over. By contrast, in ODIs, the common contemporary dilemma is whether to target 280 or 330 - a difference of one run per over.

In a T20 chase, the temptation seems to be to try and stay in the game for as long as possible and hold back for a final assault. This is a legacy of the fact that most players are trained to play longer-form cricket in their formative years. Teams typically tend to throttle back from the seventh to the 11th over, especially if they have had a strong Powerplay. The conventional wisdom appears to be that games are won or lost in the last ten overs of the chase, and that it is important to stay in the game, especially over the last five overs.

MS Dhoni is perhaps the high priest of this approach. Teams seem to organise their batting line-ups to suit this method as well. The top four positions in most, if not all, IPL line-ups consist of specialist batsmen who represent international teams in the more traditional cricketing formats.

However, some teams, like Mumbai Indians, appear to take a different approach and never throttle back if they can help it. There's some evidence to suggest that this latter approach is better because T20 chases seem to be decided in the first ten overs of the chase, and not in the last ten.

The table below gives an overview of chases in the IPL from 2008 through April 29, 2019. At the outset, this data suggests that losing three or more wickets during the first ten overs of a chase is both infrequent and disastrous. Chasing teams appear to preserve wickets because when they fail to do so, the probability of them losing increases. For instance, of the 32 occasions when a team chasing nine runs per over has lost three or more wickets in the first ten overs, only two times have they won.

This appears to support the idea that an old-fashioned limited strategy from the 1980s is alive and well, at least in spirit, in contemporary T20. The conventional wisdom in ODI cricket in the 1980s was that while a chase could fall apart in the first 25 overs, it couldn't be won in that period. As a consequence, teams sought to keep wickets in hand, allowing the required rate to rise up to seven or eight runs per over during the last 10-15 overs.

The advent of the pinch-hitter changed this somewhat, but the pinch-hitter never enjoyed sustained success in ODI cricket. Rather, the strategy that became far more successful involved promoting the best batsman (or at least, the most talented stroke-maker) in the side to the top of the order, allowing them to make the most of the fielding restrictions to score hundreds. To a large extent, this has been adopted in T20 as well, with Ajinkya Rahane, Kane Williamson, Virat Kohli, Steve Smith, Shreyas Iyer and other specialist batsman batting in the top three spots with the aim of anchoring the innings as strokeful batsmen rather than out-and-out hitters.

In successful chases of targets where the required rate is nine runs per over (targets from 170 to 189 in 20 overs), successful teams attain the required rate by the sixth over on average and maintain it steadily until the end.

Unsuccessful chases where the chasing team has lost up to three wickets by the end of the tenth over have an innings trajectory where, despite keeping wickets in hand, the chasing team is never able to catch up with the rate. In these games, the asking rate has risen to 10.9 runs per over by the 11th over on average, nearly two runs per over worse than in successful run chases. This suggests that success in chases does not come from allowing the asking rate to keep going up to protects wickets. Rather, it comes from attacking from the start and keeping the asking rate in check.

This pattern for success and failure holds true for chases where the asking rate is eight runs per over or ten runs per over (controlling, as usual, for the loss of wickets). An analogy can be drawn with track athletics. Middle-distance runners strive to develop a late "kick" - a burst of acceleration during the last lap of the race that allows them to pass their competitors. Much of the tactics in a middle-distance race come down to managing the race to get to the final lap with enough strength for that push. By contrast, 100m sprinters reach their peak speed by the 60-metre mark, after which the greatest sprinters distinguish themselves by decelerating less than their opponents. The T20 chase is like a sprint, while a 50 over run chase was conceived in the 1980s like a middle-distance race.

This idea has implications for the way in which chases are set up. Should teams reserve their best power hitters for the last few overs of the T20 chase? Or would it be better if they promoted at least one or two of their best power hitters into the top four? Are specialist batsmen like Williamson or Kohli wasted in T20? Would they be better off trying to play, as much as possible, like power hitters?

The table above gives the success rate of IPL run chases of various sizes in which a team has at least seven wickets in hand after the tenth over. The data suggests that having to chase somewhere between 80 and 100 in the last ten overs is achieved 63% of the time, while chases of 110 in the last ten overs fail five times out of six.

For targets between 150 and 210, teams that score 40% of the target in the first ten overs are more likely to lose than to win, even in cases where they have at least seven wickets in hand. On the other hand, teams that score at least 50% of the target in the first ten overs win more often than they lose in the IPL.

At this point, the following different situation might be considered: a team that has scored roughly half the required runs but lost four or more wickets. There are only five games (1, 2, 3, 4, 5) in the IPL in which a team chasing over 150 has scored within ten runs of half the target at the halfway mark in the innings, losing four or more wickets; the chasing team has won two and lost three of these games. In contrast, there have been 318 instances of teams being more than ten runs behind the halfway total after ten overs but with at least seven wickets in hand; the chasing team has won 143 and lost 175 of these games. What is striking here is not the even split of the results, but the rarity of the first situation.

All this suggests that avoiding risks to keep wickets in hand in the first ten overs of a T20 run chase is not the best idea, but it remains a popular idea. More significantly, it shows that as a rule, T20 chases are decided in the first ten overs of the chase, not the last ten. Teams are less likely to go all out in the first ten, losing wickets and therefore the game, than they are to try to preserve wickets for a late dash at the cost of an escalating asking rate. The late dash rarely works, even though it is frequently set up.

Thus far, T20 has revealed an interesting practical limit to run-scoring given the rules of cricket. Teams seem to be unable to sustain scoring rates of 11 or more runs per over a period of ten overs with any reasonable certainty. It will be interesting to see if hitting techniques advance to make two runs per ball commonplace. If they do, it might make the approach of preserving wickets at the cost of allowing the asking rate to escalate more viable. But today, such an approach is doomed to fail because it gives the side defending the target a significant, often decisive, advantage in the first half of the run chase.