What has gone wrong for Pakistan in T20Is over the last two years?

There's a line in the finale of the American version of The Office that feels relevant not only to the world in general - especially the last year - but also to cricket fans. "I wish there was a way to know you're in the good old days, before you've actually left them," the character Andy Bernard says.

It was true of Pakistan's mid-'90s ODI side, true of Misbah-ul-Haq's Test side, and most obviously true of Sarfaraz Ahmed's T20 side.

In the three years following the 2016 World T20, Pakistan played 36 T20Is, winning 30, and yet were not taken seriously. Their rise, and fall, coincided with the longest period in the game without a T20 World Cup since the event began. Their achievements were usually appended with caveats: their wins came in the UAE (even though only 11 of Pakistan's 40 T20Is during Sarfaraz's captaincy were played in that country); they cared too much about bilateral series, and T20s were all about domestic leagues; they were playing second-string sides (though ten members of the England team they beat in 2016 had played the World T20 final just four months prior, and ten of the 13 Australians in the side that was clean-swept by Pakistan had played in the IPL, and the New Zealand team Pakistan beat away had ten players with 20-plus T20Is).

And this wasn't just ex-players or fans sniping away - that would be par for the course. The PCB itself didn't take these achievements seriously, sacking Sarfaraz after one bad series, despite three years of continuous success. And even that lost series was defined by an overhaul that wasn't needed. Predictably, when that overhaul failed, Misbah, the new head coach, further undermined the achievements of his predecessors - to him, three losses in their last 13 T20Is somehow represented a losing streak.

Eighteen months on, Pakistan's T20 fortunes elicit a nostalgia stronger than Andy Bernard's.

Where did it go wrong?

Conservatism up top
In the summer months of last year, as the world moved from one lockdown to another, a conversation brewed on cricket analytics Twitter: was it time to do away with anchors?

Theoretically it makes sense. T20 is still played in a conservative fashion; it is seen by too many as an offshoot of 50-overs cricket, not as a separate format. The argument was encapsulated in this article by Karthik Krishnaswamy, the headline of which was the equivalent of a red rag to a bull.

It is an argument most progressive T20 thinkers would side with, albeit in hushed tones. Seeing drastic changes in other sports makes one question accepted notions within your own: basketball had its three-point revolution; baseball has had its barrage of three-true-outcomes hitters, who take fielders out of the game by aiming only for home runs, walks or strikeouts; all football codes have had their own analytics revolutions. The key difference is that those sports provide, a degree of homogeneity that helps in terms of analysis in a way cricket, reliant so much on its pitches, lacks.

On pitches where a score of 180-plus is par, an anchor could be a liability. Once the par score goes below that, the anchor becomes a necessity. We've seen the value of the anchor in T20 leagues over the last year played on a limited number of pitches, from the CPL to the IPL.

It's not always the pitch, however. As with the Pakistan side, the case for an anchor is also made by the lack of faith a team has in its batting depth. (From here on, the period from post-2016 World T20 to January 2019 is referred to as phase one and from January 2019 to March 2021, phase two.)

You can gauge the state of a country's domestic cricket system - or at least its blind spots - by the foreign players selected in its franchise tournament. The eight Big Bash franchises this year employed seven different Asian spinners. Twelve of the 24 foreign players selected for The Hundred are predominantly bowlers, an indication that England's white-ball bowling depth doesn't quite match up to their white-ball batting depth.

In the PSL draft every year, foreign top-order hitters are a prime commodity. Even in their successful first phase, Pakistan were backbenchers in the powerplay. It could be argued they often played on slow surfaces, but considering both Bangladesh and Sri Lanka feature in the top three above, it seems a flimsy excuse.

This has continued through phase two.

In both periods Pakistan were one of the two bottom teams on scoring rate in the powerplay. It worked because in phase one Pakistan decided to conserve wickets up top and succeeded: their powerplay average in that phase was the third best (behind Afghanistan and England). Since then they have been much poorer at protecting wickets.

Top-order conservatism in T20s is a high-risk strategy. At its best, it reduces the game for your batting unit from a 20-over match to a 14-over one, ensuring that even if you don't win many games in the batting powerplay, you don't lose many either. But if the top order fails, the entire deck comes crashing down. Considering Pakistan's middle-order frailties, it's a strategy that makes sense for them.

But even as Pakistan's top order has risen in the last few months, there has been a call for a more progressive style of play, crystallised in the Babar Azam-Mohammad Rizwan vs Fakhar Zaman-Sharjeel Khan debate. Considering the top three usually play over 50% of all deliveries in a T20 game, that is where Pakistan's batting needs to catch up. So the debate goes that one of Rizwan or Azam should be paired alongside a "hitter" like Sharjeel or Zaman, or that both Fakhar and Sharjeel ought to open.

Except, the numbers show there isn't much to this debate.

The concept of a top-order hitter alongside Azam or Rizwan is appealing, but the reality is that neither Zaman nor Sharjeel is that - both are essentially glorified anchors. Each of the four strikes in the mid-120s in the powerplay. Sharjeel has the highest ceiling, with a post-powerplay run rate of almost 9.50 but a powerplay average of 17 indicates that the probability of him getting to that scoring rate is way lower.

The quickest of the four in the powerplay is Rizwan, who goes at 126, a run rate of 7.56 per over. For contrast, eight of the 12 Full Member sides have scored at over eight an over in the powerplay in phase two.

Pakistan could look at the table above and conclude that there's no need to change the Azam-Rizwan combo and that they could do with Zaman and Sharjeel lower down. Of course, there's a difference between batting in the eighth over when you've been there from the start, and arriving at the fall of a wicket with the bowling team's tail up. Each of these four players has a much higher post-powerplay strike rate when batting in the top order than when batting in the middle order. Yet it isn't such an outlandish idea to have them in the middle, such is the state of Pakistan's middle-order batting.

Middle-order malaise
In raw numbers Pakistan's middle-overs phases have been far better than their powerplays. They have been an average or above-average team for most of this era.

However, these phases have been dependent on that top order batting long and scoring quicker the longer they bat, and in 2020, on Mohammad Hafeez. The latest iteration of Hafeez is the prime reason Pakistan have actually kept up with the world in those middle overs.

Pakistan began this five-year period with the likes of Sarfaraz and Shoaib Malik in their middle order. Unsatisfied with accumulators, they looked for a power-hitter and have since experimented with Asif Ali, Khushdil Shah, Haider Ali, and even tried to force Iftikhar Ahmed into that role. None have stepped up. Since Iftikhar made a couple of scores in Australia in 2019, Pakistan have played 21 T20Is, tried over a dozen players at five and six, and only once has a player made more than 25. Forget averages or strike rates, that's a level of failure that is hard to comprehend. For context, even the allrounder, Faheem Ashraf, has scored three 30s in this period, batting at seven and eight.

This has led ubiquitous former Pakistan players still playing domestic cricket to peddle a false argument that selectors are unfairly obsessed with picking "PSL hitters" over "domestic performers". In doing so they ignore that Shah averages 45 at a run a ball in his six-season-long List A career, and that Asif Ali has averaged 52 at a strike rate of over 110 in 50-overs cricket in the last four domestic seasons, and that Haider Ali averages over 45 in both first-class and 50-overs domestic cricket in his two seasons. And that Iftikhar, a man who has never really set the PSL alight, is one of three players with at least 50 innings (along with AB de Villiers and Virat Kohli) to average over 50 with a strike rate of over 90 in his List A career.

It isn't unknown for cricket teams to follow a selection formula like deep states secretly running countries employ: they put in place a new regime and at the first sign of trouble start wondering whether the old one was really that bad. Sure, the old guys aren't what they dreamed of, but right now there's not much else that can be done. And so:

This is simply comparing Malik and Hafeez to the overall mean - forget about comparing their numbers to those of the like of Glenn Maxwell, Keiron Pollard, Rishabh Pant or David Miller. Pakistan went back to Hafeez very soon after discarding him, and it has paid off. What are the odds they'll go back to Malik too?

In a spin
Much of Pakistan's success over the past five years has been built around their multi-pronged spin attack. They have been aided by helpful conditions, but both Imad Wasim and Shadab Khan stepped up in Pakistan's rise. And the subsequent decline of both has been one of the causes for the team's overall fall.

For all their legacy of fast bowling, the past decade has been defined for Pakistan by their being at the forefront of the powerplay spin revolution: no team in history has opened with a spinner more often than Pakistan; no spinner has shared the new ball across formats more than Hafeez; and in T20Is no one has opened more often than Imad Wasim.

Wasim reached the top of the T20 rankings almost entirely on the back of his new-ball expertise, but the Wasim of phase one and two are different bowlers.

This change in numbers, however impressive the economy rate still is, isn't to be regarded in isolation. After a decade during which spinners' fortunes were high in the powerplay, there has been a movement back towards pace bowling in the last couple of years. Teams, both national and franchise, are learning that the value of top-order wickets is higher than they assumed and so a run-saving spinner isn't always the positive he once was.

Since the white Kookaburra tends to swing for only the first two or three overs, giving spinners those early overs ends up being counterproductive. So the powerplay spinner has become a shock tactic, perhaps best exemplified by the fact that even in the 2021 half-season of the IPL, with spin-friendly wickets, less than 14% of the new-ball overs were bowled by spinners (compared to, say, the 2018 IPL, when that number was 26%). Instead, as the Chennai Super Kings showed with Sam Curran and Deepak Chahar opening the bowling, new-ball swing bowlers are back in vogue.

In that light, Pakistan dropping Wasim for Mohammad Nawaz made sense. Nawaz may never be the powerplay option that Wasim is, but he is a better bowler in the other phases. And with the rise of Shaheen Shah Afridi and the return to form of Hasan Ali, Pakistan have pace resources for the powerplay.

Mohammad Wasim, Pakistan's chief selector, specifically noted how Wasim needs to improve his middle-overs bowling - a clear indication that he was thinking of a move to pace. And yet, his words have not been backed by the actions of the team management. In the absence of Imad Wasim, Nawaz has played eight T20Is and shared the new ball in six.

A similar trend has emerged with legspin, and for Pakistan it has been exacerbated by Shadab's lack of form and fitness since the start of 2019. His overall decline mirrors that of many of his contemporaries.

Since the 2016 World T20, eight of the top 20 wicket-takers in T20Is are wristspinners, only one of whom has had a better phase two than phase one. Only Yuzvendra Chahal and Kuldeep Yadav have had as sharp a decline as Shadab; one is now out of the Indian team and the other has question marks around him for the first time in years. Shadab can't be far behind.

But in the absence of many real middle-overs enforcers, and the fact that the World T20 is going to be in Asia, Pakistan are still going to play a wristspinner. The only question is whom.

If you consider only the tournaments that each of them have all played, it's clear that Shadab remains among the best options for Pakistan. Based on the data below, Zahid Mahmood is the likeliest to replace him: from being the best T20 bowler to being the first legspinner in a decade to take more than 50 first-class wickets in a season, he has been the outstanding leggie in the Pakistan domestic game over the last two seasons.

Of course, these aren't exactly like-for-like comparisons - the biggest advantages Wasim and Shadab have is that they can be significant contributors with the bat.

And there is the wild card. Pakistan spinners have always depended heavily on their captains, from Abdul Qadir being so beholden to Imran Khan that he ended up naming his eldest son after him, to Misbah making Yasir Shah's career after having captained him domestically for half a decade. It is with this in mind that Usman Qadir's case is to be assessed. He goes back a long way with Azam and his quality as a wicket-taker under Azam in T20s is clear from his numbers too.

Any one of Shadab, Mahmood or Qadir should be fine, as long as Pakistan realise that T20 legspin isn't what it was even four years ago - teams have employed more lefties as hitters in the middle overs to counter it (which is also why Pakistan should consider Zaman and/or Sharjeel in the middle order more seriously than they actually will).

So this is where Pakistan now stand. Over 18 months they have gone from wondering whether their all-conquering side could translate their success at a T20 World Cup, to wondering what it will take to get back to that level. It's clear that a misuse of their own resources, a drop in form of several star performers, and perhaps most significantly, an inability to keep pace with an ever-evolving format have all contributed to the decline.

Even their pace bowling, the one safety net Pakistan have had in T20Is, has question marks over it. Despite the emergence of Afridi and the return of Hasan Ali, Pakistan's pace attack has gone from being one that took wickets more often than the global average (while going at over a run per over better than that in phase one), to one which is pretty much average by most metrics.

Some of that is down to a change in personnel, but there's also the simple fact that in the Sarfaraz era, every bowler had been assigned just the right role: it rarely felt as if Pakistan were bowling a fast bowler in the wrong phase. Now, take the case of Haris Rauf. He is one of the few fast bowlers to go at a lower economy at the death than in the powerplay (but is twice as likely to take a wicket at the death), while being an above average middle-overs bowler. Yet 30% of the deliveries he has bowled in T20Is have been in the powerplay.

That sums it up, really. A change in personnel, a lack of resources, yes, but more than anything, a failure to understand those resources and maximise them. And until Pakistan return to being the team that was bigger than the sum of its parts, you can see their T20 fortunes going the way their 50-over fortunes have gone in the past two decades.