It was an innings designed to start a Twitter riot. Perhaps never in cricket history has one man been so lambasted for doing exactly what he was meant to do.

Ajinkya Rahane was recalled for the 2016 World T20 semi-final against West Indies in Mumbai. He was selected to do a very specific job: prevent India's innings from getting destabilised by early wickets, harry between the wickets, and set the game up for the explosive batsmen beneath him. He did just as he was asked. He scampered twos frenetically, both for himself and for Virat Kohli. He sought to give Kohli the strike whenever possible. By the time Rahane was dismissed, India were 128 for 2 and assured of reaching a total that looked imposing.

Only, what Rahane did according to orders may well have lost his team the match. He took 35 balls for his 40 runs. On one of the finest T20 batting pitches in the world, he batted until the 16th over and mustered only two boundaries - one off an outside edge through third man. India reached 192 from their 20 overs but the ease with which they got there - and that it cost them only two wickets - betrayed how resplendent the conditions were for batting.

Eighteen balls into their response, West Indies had lost two wickets too. But T20 batting is far less about protecting wickets than about using resources efficiently, as West Indies proved with a clinical chase that took them home with two balls to spare. There were many heroes for West Indies that day - Lendl Simmons, Andre Russell, Samuel Badree - and also, perhaps, Rahane. He had scored at 6.85 an over whereas West Indies scored at 9.96. CricViz's player impact found that Rahane's innings was worth minus 6.5 runs on India's total - making it more damaging than when some players are dismissed first ball.

The notion of defensive decision-making - in medicine, the stock market and beyond - is that humans don't make decisions that are optimal. Instead, they make decisions to "cover their ass", as Gerd Gigerenzer argues in Risk Savvy. These don't maximise the chances of success - but they do minimise the chances that failure will lead to a backlash against them. "Choosing a second-best option is not stupidity or bad intention," Gigerenzer writes. "Defensive decisions are imposed by the psychology of the system." As John Maynard Keynes observed: following conventional wisdom keeps you from getting fired.

What this tendency for self-preservation means in limited-overs cricket is clear. Analysts - in both T20 and 50-over cricket - believe that wickets are overvalued, and teams are routinely too cautious at the start of their innings. Consider the opprobrium that greets a team when they are bowled out for 100, say, in a T20. Clearly such an innings involves poor execution. Yet arguably worse is getting, say 160 for 5 off 20 overs - executing the team's plans well - and still being defeated easily, which reveals flawed strategy.

"Every ball, you have to re-evaluate the situation for the next ball," Phil Simmons, who was West Indies' coach when they won the last WT20, has said. "In Test matches you assess by sessions; in one-day cricket you assess by overs; in T20 cricket now you assess by balls. Every ball is an event."

In T20 the margins by which games are decided tend to be small. With fewer wickets lost per innings, batsmen must reorder their mindsets - to place less emphasis on playing themselves in, and simply less value on their own wickets. The challenge can be particularly great when a side bats first. With imperfect information about what they need to do - what score is likely to prove enough, for instance - batsmen tend to edge on the side of caution. This helps explain why teams who bat first in T20s have only won 47% of matches since the start of 2016.

© ESPNcricinfo Ltd

"I've seen instances where it works for some teams," says Brendon McCullum. "A la Sunrisers Hyderabad - where they have a strong bowling line-up and therefore don't need 'overs' from a batting point of view. They just need a score and their bowling will give them every chance. In most cases, though, I would say when guys play like this it is more likely a selfish move than a structured plan."

The sheer insecurity of the global T20 circuit often makes for a tension between the interests of player and team, and therefore leads to match-losing innings. Besides the T20 Blast, which runs alongside the rest of the English county game, no other leagues last even two months. Players flit in and out in between international commitments - or, for those below the elite, in and out when there are vacancies because of someone else's international commitments. Contracts of more than two years are highly unusual. Even in the IPL, the three-year contracts agreed by players in a "big auction", such as this year, give their franchises the option to release a player after one or two seasons.

All of this creates an incentives problem. Imagine a top-order player batting at the end of a season in one league, hoping to get a contract soon in another. The needs of the team may dictate that he attack with alacrity from early on. But he may reason that should he do so there is more chance of him being dismissed cheaply, rendering him less attractive to franchises. He may play himself in for longer before accelerating - increasing the chances of recording a high personal score, even if this might not increase the chances of his side recording a high total.

"If you look at the strike rate for the first five balls of a T20 innings you will see a trend of self-preservation," says Mohammad Khan, the general manager of Jamaica Tallawahs when they won the 2016 CPL. "You will also see this same mentality in first-innings and second-innings splits - and even among opening batsmen. Because of the financial nature of domestic T20 and the sheer incompetence that exists in the selection process, players have a default setting that leads to self before team."

Scouting in T20 leagues is less sophisticated than in other sports - like baseball, say - making it harder to contextualise performances. For instance, imagine that a team in the Bangladesh Premier League is considering signing a player playing in the T20 Blast in England. Only a small proportion of T20 Blast games is televised. Should a batsman make a high score, but after starting too slowly at a time when the conditions were easiest to bat, and harm his team's chances in the process, the Bangladeshi franchise would have no way of knowing; all they are likely to be able to see is the batsman's final score and strike rate.

"There is no doubt there have been occasions where batsmen have been caught up in their cause against the team's," explains one leading T20 coach, speaking anonymously. "The pressure of holding their position in the side can also influence a player's approach, by being conservative… Players in franchise cricket around the world are very aware of their performance, it's their currency. In some cases this can have a negative influence."

In the 2016 World T20 semi-final, Ajinkya Rahane followed the brief given to him of anchoring the innings, but his 40-ball 35 eventually cost India the game © IDI/Getty Images

Paddy Upton, who has coached T20 franchises throughout the world, believes that match-losing innings fall into three broad groups. The first is what he characterises as "selfish players looking to score runs for themselves" - putting their own narrow interests above those of the team.

The second group of match-losing innings are not the result of selfishness but rather of inexperience. "The overly defensive, conservative approach of overvaluing their wicket can come from a player still trying to make the mental adjustment from the longer formats."

Finally, there is what Upton terms a situational awareness error: simply misreading the situation in a game, which can come from either a player, their team or - like Rahane versus West Indies - both. (This type of error can lead to innings that weigh a team down; it can also translate into the opposite problem: not valuing a wicket highly enough. In Upton's words: "They attack when they actually could have gone deeper, or tried to go deeper when they should have gone earlier.")

When batsmen go too slowly at the start of their innings, this can actually be an attempt to do the "noble thing" - trying to set a game up for themselves to win it for their team: to do the job yourself, rather than leave it to someone else. But it is also infused with jeopardy: if they are dismissed at an inopportune time it can suddenly leave their team-mates too much to do and not enough time to do it. A slow innings can be played for the noblest of reasons and damage the team in the same way as one played out of unadulterated selfishness.

In the 2017 IPL final, Steven Smith seemed to have paced the chase to perfection, adjusting his game to the demands of a slow pitch, and ensuring that Rising Pune Supergiant had wickets in hand in case they were needed for the final push. This narrative was only a couple of boundaries from being inked into history. Smith - with support from Rahane, who scored a little quicker than him - set up what would ordinarily have been a cruise for Pune: 33 from four overs with eight wickets in hand. Yet at this stage Smith had only 33 from 38 balls.

After reaching his half-century, Smith was dismissed in the frenetic final over, which Pune began needing 11 with seven wickets in hand. Losing with seven wickets in hand going into the last over - especially when the victory target was "below par" in Smith's estimation - required not just a cock-up in the final balls of the game. It required, instead, a deeper failing in a team's strategy, and a fundamental miscalculation of the relative worth of runs and wickets in T20.

© ESPNcricinfo Ltd

Even Chris Gayle, the Bradman of T20, has played innings deeply damaging to his team: 3 off 16 balls in the 2012 World T20 final, for instance. The Gayle calculation is essentially that caution at the start of his innings, of a sort that would be ill-advised for lesser players, makes sense because he is capable of brutal acceleration. Such an approach is inherently hazardous, for it means that if Gayle is dismissed before becoming set, he is more of a drag on his side than batsmen who make few runs while being more aggressive from the start of their innings.

On 23 occasions over his career Gayle has faced at least 15 balls in an innings while scoring at a rate at least 25% lower than the overall strike rate in the match. His side has lost 13 of these games.

Similarly, MS Dhoni has played at least 15 balls in an innings with a strike rate at least 25% under the match average 22 times, 13 of them in defeat. Other players perform even worse by this metric - perhaps the closest we have to a statistic for the tendency to play match-losing innings in T20. Eoin Morgan has played 20 such innings; 17 of these have come in defeat. Sixteen of Andre Fletcher's 19 innings of this type have come in defeat. Sourav Ganguly played 14 such innings - almost one in five of his total T20 innings - and 11 came in losses, showing his struggles adjusting to the faster pace of T20 [see table above].

As T20 evolves, match-losing innings should become less common. The increase in average scores indicates that batsmen are becoming more adept at hitting from earlier in their innings and learning to place less of a premium on their own wickets.

Savvier metrics can also help. England use a system that evaluates every innings in T20 and ODI cricket in terms of each ball's impact on the team's expected total, giving each player a match impact score. For instance, in the first innings of a T20 the average number of runs scored off the final ball of the innings is 1.7. A player who fails to score off it would thus have a match impact of minus 1.7 for that ball.

"So the idea of a match-losing innings is sort of built into that," says Nathan Leamon, the England analyst and one of the creators of the CricViz system. "We've largely moved away from traditional averages and strike rates in white-ball cricket, although we look at those things because they give you qualitative indications about how someone plays. But the value of the individual performances we look at is how it affects the overall team total. So we dive into what you mean in terms of match-losing innings."

By CricViz's analysis, since the start of 2017 there have been 22 innings rated as costing their team over 15 runs [top ten presented in first table]. Two innings by Dhoni, at minus 27.6 and minus 21.7 - both in wins - are rated the most detrimental to the side. One drawback of this system is that it penalises batsmen who score at below the average match rate during comfortable chases: for instance in July, Virat Kohli scored 20 not out at under a run a ball after coming in at 130 for 2 in the 13th over in pursuit of 160; India still won with ten balls to spare. The system is also arguably harsh on batsmen who give a more explosive team-mate the strike.

Chris Gayle often starts slow before nitro-boosting the innings, but it's a risky tactic and one that proves costly if he is dismissed early instead © Ashley Allen - CPL T20 / Getty

T20 franchises analyse the best time for their most impactful batsmen to begin their innings, and determine who should bat next based on the overs remaining. The idea is to guarantee that their most dangerous players at the death enter at the optimal time. Yet one roadblock remains to ensuring that the appropriate batsmen are batting at a particular time: wickets need to fall at the right time. If they do not, T20 hitters can be wasted.

Consider Andre Russell in the CPL this year. His season began with one of the greatest T20 innings of all time: entering at 41 for 5 in the seventh over against Trinbago Knight Riders, he scythed 121 not out off 49 balls to haul Jamaica Tallawahs over their target of 224.

And yet when Jamaica next played Trinbago, nine days later, Russell only entered with ten balls left in the first innings because of a fourth-wicket partnership of ten overs. Jamaica were defeated with one of the most belligerent T20 batsmen barely given a chance to bat at all. They wasted their best single resource - Russell's batting. They did so again in their very next game. Chasing 157 to win against Barbados, Jamaica ended up on 154 for 3, somehow contriving to lose by two runs despite having seven wickets in hand. This time Russell was completely unused; Ross Taylor made 26 not out off 28 balls - an innings worth minus 16.1 on CricViz's match index.

On both occasions, Jamaica could have used Russell earlier had they retired out a batsman in the middle, which is permitted by the laws of the game. The great obstacle to retiring out batsmen, perhaps, is one of the tenets of match-losing innings: defensive decision-making.

A coach who dared to retire out a player during a fraught chase and still lost the game would expose himself to far more criticism than one who didn't retire out a struggling batsman in a defeat. Failing conventionally is much better for job security than failing unconventionally. And retiring out a batsman would, if nothing else, be an easy way to guarantee another Twitter riot.

Tim Wigmore is a freelance journalist and author of Second XI: Cricket in its Outposts

© ESPN Sports Media Ltd.

 

RELATED ARTICLES