Stats Analysis

The ideal T20 team today looks like Pakistan's 2007 and 2009 World Cup sides

They were ahead of the curve over a decade ago, but haven't quite followed that blueprint since

Hassan Cheema
Hassan Cheema
Shahid Afridi is mobbed by his team-mates after he guided Pakistan to an eight-wicket win, Pakistan v Sri Lanka, ICC World Twenty20 final, Lord's, June 21, 2009

Shahid Afridi was not Pakistan leading run-scorer or wicket-taker in their victorious 2009 T20 World Cup campaign, but he was the pillar around which the rest of the team rallied  •  Richard Heathcote/Getty Images

As Pakistan have stumbled their way through the back end of the Asia Cup to two heartbreaking losses in the World Cup, every aspect of the team has been debated over. Yet this World Cup has been defined by a question that is ever more familiar in non-sports discourse in Pakistan: why don't we have what others do? What does a Pakistani T20 side that's up with the zeitgeist even look like?
If we were to create the ideal, data-driven T20 side, it would have: two to three top-order hitters, two to three middle-overs specialists who are good spin-hitters and bat deep, followed by allrounders who create the depth that allows those above them to play with freedom. For pace, you'd want a powerplay specialist fast bowler, a death-overs specialist, and another fast bowler who can do both. Among these three, you'd want express pace and a left-armer. For spin, you'd want bowlers who turn the ball either way and can bowl across phases, plus additional bowling options to create positive match-ups. Six or more bowling options and batting that lasts till eight.
In other words, the ideal T20 team today would look almost identical to Pakistan's 2007 (runners-up) and 2009 (winners) T20 World Cup sides.
The late 2000s are a dark period in Pakistan's cricket history. They went four years without winning any Test series. They lost ten of their 15 bilateral ODI series, with four of their five wins coming against Bangladesh, Zimbabwe and West Indies. They dealt with the death of a beloved coach during a World Cup, lost hosting rights, and had their players banned for, variously, using performance-enhancing and recreational drugs, spot-fixing, scuffing up the pitch, and conspiring against their captain. The 2007 World Cup was a forgettable experience, and while they made it to the knockouts in the 2009 Champions Trophy, the semi-final loss led to fixing accusations. Yet in the middle of all this, Pakistan stumbled upon the perfect way to play T20 cricket.
The top order: hitters over anchors, please
There are a handful of players from those Pakistan teams who would have had different careers if they had been ten years younger, but no one more so than Imran Nazir, who was the lynchpin of the 2007 T20 World Cup side. He finished with a career strike rate just shy of 150, a figure that would have made him a franchise globetrotter today. A lot of those runs were made in the lower-quality Indian Cricket League and on the Pakistan domestic circuit, but even at the highest level, Nazir's method was successful. Until 2010, for example, only Yuvraj Singh and Andrew Symonds scored more T20I runs at a higher strike rate than Nazir.
Opening alongside him was Mohammad Hafeez, who had scored over 700 T20 runs at a 30-plus average and a strike rate of 160 ahead of the 2007 World Cup. The Nazir-Hafeez partnership was, statistically, as attacking as any team can hope for, even if it came together through trial and error than through any grand strategic plan.
Pakistan began the 2007 World Cup with Salman Butt as opener, but dropped him ahead of the semis. In 2009, they started with Butt and Ahmed Shehzad as openers, but ended it with Kamran Akmal and Shahzaib Hasan at the top, going from two anchors to two hitters in the middle of the tournament, showing a willingness to change their flawed plans when needed. Even though Shahzaib failed to make his mark at the international level, Pakistan had figured out how to construct their team: they preferred failures from the batter who finished his T20 career with a strike rate of 138 (Shahzaib) to one who finished with 113 (Butt).
But their inherent conservatism prompted them to switch back to anchors every time a major tournament came around. This trend was best evidenced in Nazir missing the 2009 and 2010 T20 World Cups while Butt, with a strike rate of 83 in the 2007 and 2009 tournaments, started as first-choice opener.
As so often with Pakistan, it was less a question of personnel than intent, and no one personified this more than Hafeez. From being a top-order hitter before 2007, he became something entirely different the following decade. He captained Pakistan in two T20 World Cups and his skills improved, but as his poor strike rate shows, intent matters. And he wasn't the only Pakistani top-order hitter who failed on that count.
Pakistan and Hafeez had the right answers on how to bat up top, even though they refused to learn from their failures or successes. But for two glorious events, they got it right, however brief and accidental it may have been.
Batting against spin: get the match-ups right
From 2000 to 2016, the overall average for batters at Nos. 3-5 in ODIs was 34.3 and the strike rate 76.4. This period coincides with the one-day career of Younis Khan (average of 31.2 and strike rate under 76), arguably Pakistan's greatest batter in Tests, but a below-average one in ODIs.
Then there was Misbah-ul-Haq, whose limited-overs batting generated the sort of debates that Babar Azam and Mohammad Rizwan's partnership does today. Since the start of 2000, 59 batters have scored over 5000 ODI runs, but only four have done so at a lower strike rate than Misbah.
Of those 59 batters, Shoaib Malik stands 51st on average and 34th on strike rate.
In an era when T20 was still seen as a shortened ODI rather than a distinct format, Younis, Misbah and Malik were the backbone on which Pakistan built their T20 success, preferred even over better one-day players. None of the three would ever make the best ODI XIs of their era, but Pakistan had understood T20 cricket before the rest of the world did. And that's not just hindsight speaking; after the 2007 final, Rashid Latif wrote about why Pakistan had been so successful in that tournament, lessons that remain relevant 15 years later.
What this trio instinctively grasped was that the format required them to target their positive match-ups. None of them scored at over seven per over against pacers in those two tournaments, but they made up for it with their expertise against spin. Across the 2007 and 2009 World Cups, they scored over 400 runs against spin at an average of 43 and a strike rate just shy of 140.
But 2009 was the last T20 World Cup that Younis played in; Misbah was dropped before the 2012 edition; and Malik cratered the way Hafeez and Akmal did, striking at under 90 and averaging under 16 against spin over the three T20 World Cups between 2012 and 2016.
As the T20 World Cup went from being a tournament that Younis compared to the WWE to being a marquee event of the international calendar, the added pressure meant a reduction in the intent that had brought Pakistan success. The world caught up to Pakistan, except Pakistan had now regressed. They quickly went from being one of the best batting teams against spin to one of the worst.
A decade on, Pakistan are still struggling to find batters who can attack against spin. The ones they have are considered too old, too unfit, or not recognised as batters at all (like Shadab Khan and Mohammad Nawaz).
The worth of the low-value wicket
Much of the aversion that ex-players have towards data-driven T20 has to do with the language it employs. Those scoffing at a low-value wicket would have previously lauded the benefits of pinch-hitters. Both are essentially the same thing, the newer term a more accurate, if corporatised, version of the older.
Here too Pakistan were ahead of the game. Shahid Afridi was neither Pakistan's top run-scorer, nor the highest wicket-taker at the 2009 T20 World Cup, but the tournament was defined by him. Younis' decision to promote him halfway through the tournament was what led to them winning the title.
Afridi's is an interesting case, the following tables highlighting how miscast he was.
Pakistan had someone who was the best middle-overs hitter in their history, while not even being the best death-overs hitter in his own team. Across his T20I career, excluding death overs, Afridi's strike rate against pace was 141, and against spin 157. His numbers in ODIs (where ball-by-ball data is available) follow the same pattern. These stats scream of a batter who should be first in towards the end of the powerplay or immediately after it. Sure, Pakistan had those batters who could attack spin, but none of them could hit like Afridi. Few in history have been able to.
In the semi-final and final of the 2009 T20 World Cup, Afridi scored 49 off 39 balls against pace (SR 126) and 56 off 35 against spin (SR 160). At the time those innings were seen as uncharacteristically mature, unlike a real Afridi innings, but looking back, that should have been his permanent version. They remain the only fifties he scored across 56 World Cup innings.
If Afridi had been born in 2000 rather than 1980, his career arc would have looked entirely different. Across franchise cricket, he would have been routinely utilised at three or four. The 2009 World Cup would not have been the exception, but the rule. He ended up batting at those positions in only 16 of his 91 innings, but thankfully for Pakistan, three of those were in 2009.
Eventually Younis' instinct coincided with what the data would have pointed to. And as with so many things, Pakistan stumbled on the most efficient way to play.
Fortunately, Pakistan would learn from this and never miscast an allrounder by playing him too far down the order ever again. Nope, never, especially not Shadab, who didn't bat at four for Pakistan until his 74th T20I, despite a stellar record for Islamabad United* there.
Start with Mohammad Asif, finish with Umar Gul
In an ideal world, a pace unit is built of multiple Jasprit Bumrahs or Shaheen Afridis - bowlers who are exceptional across phases of an innings, and otherworldly in at least one. But most bowlers aren't that complete a package. Considering those resources, teams aim to maximise every bowler's 24 balls in the phase their skillset is best suited for (even if the norm is to have pacers who can bowl two up front and two at the death).
Thirteen pacers bowled 20 or more overs in the first T20 World Cup. Two of them stand out for how they were used.
No fast bowler bowled a higher percentage of his overs before the halfway stage than Mohammad Asif; none bowled more in the second half than Umar Gul. This too was not a strategy that Pakistan came into the tournament with, but one they struck on halfway through. It made sense to have Asif, the preeminent new-ball bowler in the world, to get through his quota before the tenth over; but six of the first seven overs Gul bowled in that tournament were in the powerplay. After that he wouldn't bowl a single over in that phase for the rest of the tournament, instead coming only towards the back end of the innings.
Across the first two World Cups, Gul bowled 14.1 overs at the death and conceded a scarcely believable 5.85 per over. The game changed a lot in the next decade and no one has those sorts of death numbers anymore, but even in his era, Gul was one of one. His greatest contemporary, Lasith Malinga, went at 6.85 per over at the death in those first two World Cups. Among bowlers who bowled more than six death overs in those two World Cups there was only one other who went at under 7.30.
With Asif and Gul as leaders of the two halves, Pakistan could build the rest of the unit around them - spinners in the middle and Sohail Tanvir to plug in the remaining slots and provide the left-arm angle. In 2009, Pakistan no longer had Asif (banned again), but Abdul Razzaq deputised for him exceptionally well (five wickets in 12.3 overs at less than a run a ball), and Mohammad Amir was a sexy upgrade on Tanvir.
The irony, looking back at it in 2022, is that the one thing those pace units lacked was extreme speed. It's not that they didn't have such bowlers then, but Mohammad Sami was considered too wayward, and Shoaib Akhtar was at the tail end of his peak. Also, Akhtar was sent home from the 2007 World Cup for hitting Asif with a bat, and withdrawn from the 2009 squad because, the PCB claimed, he had genital warts.
The supporting spin act
One of the more interesting aspects of looking back at the first T20 World Cups was how dominant elite spinners were then. Five of the top seven wicket-taking spinners in those tournaments went at under a run a ball, with Afridi barely above it.
Neither Afridi nor Saeed Ajmal (12 wickets at 5.82 across 2007 and 2009) was easy to line up and hit with the spin, which made them ideal support acts for Gul and the other fast bowlers.
Ajmal went at six runs an over at the death in those first two World Cups (he bowled only four overs in that period). And as back-up, Pakistan had part-timers in Hafeez, Malik and Fawad Alam, who combined to bowl 35 overs in those first two tournaments - 15 balls per match - while going at under 8.50 runs per over.
Pakistan had as complete a T20 attack as any team could hope for. They didn't have the data but they had experience and intuition. A lifetime later there are still lessons to be learnt from that.
*The author is the strategy manager for Islamabad United at the PSL

Hassan Cheema is a sports journalist, commentator, and strategy manager of the Islamabad United PSL franchise. @mediagag