Why aren't there more true allrounders like Shakib in ODIs?

"I think he's got a bit of a point to prove and he probably thinks that as well. He seems to have been a little bit forgotten." That was Bangladesh coach Steve Rhodes on Shakib Al Hasan before the start of the 2019 World Cup. On commentary during one of Bangladesh's early games, Mark Nicholas remarked that most people would be surprised to see Shakib at the top of the rankings.

It seems that just being the world's best allrounder for a generation is not enough to grab attention.

To a Bangladesh fan, he has been the alpha and omega of their cricket, a 50-foot giant. And for most hardcore fans, it is hard to forget about Shakib. He is in every T20 league and has been around in international cricket for more for a decade. But for more casual fans, he's almost invisible. Bangladesh cricket is still somewhere on the margins for them. There has never been a global star from Bangladesh, even if Tamim Iqbal, Mustafizur Rahman and Shakib have been incredible at times.

For most of Shakib's career, Bangladesh have been a non-entity, as either an easy team to beat or a lower-tier one doing just okay. Often the media from major nations have downplayed his achievements because he is a spinner on helpful tracks. And while Shakib's bowling skill, left-arm fingerspin, is a useful part of cricket, it's no one's idea of sex on wheels. For all his talent, he doesn't possess any mystery balls or unique ones, just a good spinning ball and a slippery straight one. And even as he has made big money in the IPL and other T20 leagues, most of his batting is done in the middle overs - cameos for a few balls.

He has been a remarkable player for the longest time. Since 1999, no one has been involved in more balls per match than him (86.3). The only other player we know who averaged over 80 deliveries a game was Jacques Kallis. (Wicketkeeping top-order batsmen would also be involved in these many balls, but we don't have stats for how often they take the ball.)

Shakib is an above-average batsman and bowler in ODI cricket. Since his ODI debut, he is second on the wicket-takers' list, and 13th on the run scorers' one. For over a decade now, his presence has allowed Bangladesh to include another batsman or bowler in any line-up, and there is no reason for them ever to pick a bits-and-pieces player as their fifth bowler or to risk a dud No. 7.

He may not be the best player in the world, but by being this efficient in batting and bowling, he might be the most important.

Shakib contributes significantly in all three formats. He has been a must-pick for fantasy IPL teams and has consistently been at the top of the ICC's Test rankings. He is the allrounder every team wants to create. But for all the advancements in sports science, data and training, and the free-market influence of T20, there still aren't that many allrounders in the world.

Everyone wants an allrounder, but not many actually exist. Just a handful of players have ever been good enough to be picked as a batsman or bowler in their team for an extended period. Only South Africa's Aubrey Faulkner played over 20 Tests and averaged over 40 with the bat and under 30 with the ball (from 1906 to 1924).

The fifth bowler can be the most significant risk in a team. While it is nice to think your three or four front-liners can win the game for you, ODI pitches are set up to nullify bowlers, and the white ball doesn't hold up. Batsmen have learned to score at nearly a run a ball while keeping their wickets intact, which means the fifth bowler is almost always in play.

If you take out your two best bowlers and look at the bowlers from positions three to six, you can identify the bowling attacks with good fifth and sixth options. If you calculate the cumulative averages of bowlers from Nos. 3 to 6 for each team playing in this World Cup (since the end of the 2015 World Cup), Afghanistan (25), India (32.8) and South Africa (32.8) are the only three teams under 33. Australia (40), Sri Lanka (42) and West Indies (43) are the three over 40. The economy rates more or less correspond.

A poor fifth bowler doesn't just affect their own figures. If you have a weaker third or fourth bowler, they don't get the chance to bowl the easier overs since they have to carry the load for the fifth. You also get less flexibility, because if you know you have to nurse one player through their spell, you can't make as many tactical changes as you would like to. Before the ICC changed the rules for the middle overs, with only four fielders allowed outside the 30-yard circle, and England's batting hacked them, you could hide your fifth bowler. The window to hide now is between overs ten and 30, and that's only if you aren't playing big-hitting teams like England or West Indies.

Then there is the fact that your fifth bowler might not be a ten-overs-a-game player, which means you will need a sixth bowler, either as insurance, or to combine with another player to fill that fifth slot. In a perfect world, your sixth bowler is a tactical variation. For example, England have Moeen Ali as their fifth, Ben Stokes as sixth, and Joe Root as a back-up if they need more spin. But most teams don't have that flexibility. Their sixth bowler is a part-timer who is thrown the ball when a match-up or conditions are in their favour, or when everything has gone wrong.

Usually when picking an allrounder, you're willing to take a slight drop-off in batting for their all-round skills. In ODI cricket, you can risk a bowling allrounder at seven. The No. 7s of all teams at the 2019 World Cup have faced only 20 balls per innings on average since the 2015 tournament.

But when you have a short batting order, there is a psychological effect: your top order will play within itself, either subconsciously or often deliberately. And that doesn't include the horrible times when your top order is ripped out and your No. 7 has to put an innings together to save the game.

When you bat deep, it allows your batsmen to play unleashed, as England have shown while becoming the first team to score at more than a run a ball in ODI cricket for a sustained period. That's not downplaying their actual batsmen - they have been fantastic, but they have batted in games where Adil Rashid, who has ten first-class hundreds, is at No. 11. When you play a team that bats deep, your bowlers have to bowl to get wickets, because otherwise the batting team can attack as they please in the last 20 overs. And when bowlers attack, they often give batsmen even more balls to score off.

The dream for any team is to bat until nine, with seven bowling options, but most allrounders come with caveats. Most allrounders are not all-round.

The 2D description of an allrounder is someone who averages more with the bat than with the ball. By that description, Don Bradman is an allrounder. It also ignores the wicketkeeping allrounders. In ODI cricket, Kallis has the largest positive differential between batting and bowling averages among players who have played a minimum of 50 games and taken at least 50 wickets, but Darren Lehmann is second. Such a metric also doesn't take into account strike- or economy rates.

The most obvious kind of allrounder is the true allrounder - someone who can legitimately bat and bowl well enough to be picked in an XI, and this is a massively subjective claim. The only way to really test it would be if the player got an injury that stopped them from doing one of their skills and was still picked. At this World Cup, the two true allrounders are Shakib and Afghanistan's Mohammad Nabi. They would be in their teams - perhaps not consistently - as either a batsman or a bowler. No one else from the other teams can make that claim.

Mohammad Hafeez comes close, but would he be picked as either batsman or bowler? It's the package that makes him worthwhile, so while Pakistan fans might seem annoyed at Hafeez's existence, the fact that he is involved in over 61 balls a game tells us how valuable he can be - even if perhaps he's not positively affecting all these balls. Hafeez is the perfect example of a player who may not be picked for either skill automatically but who, as a composite, is more than a bits-and-pieces player.

There are plenty of all-round package cricketers like Hafeez - Ben Stokes and Hardik Pandya are the obvious ones, while Moeen, Jason Holder and Andre Russell all have strong claims.

Then there are the players skilled in one discipline, but only okay, or slightly below average, in another - like Mitchell Santner, Imad Wasim, Andile Phehlukwayo, Glenn Maxwell, Jimmy Neesham and Bhuvneshwar Kumar.

Some batsmen bowl a tiny bit - Mosaddek Hossain, Kedar Jadhav, JP Duminy, Mahmudullah and Aaron Finch. And there are plenty of bowlers who can hit out to score runs - Pat Cummins, Rashid Khan, Isuru Udana, Nathan Coulter-Nile, Jofra Archer, Tom Curran and Ashley Nurse.

What we seem to have had less of in recent times is the real bits-and-pieces players - those not up to the mark, or even close, in either skill. The great South African fielder Derek Crookes averaged 14 with the bat and 40 with the ball. The idea at that time was that his overall package was worth enough. Perhaps the most bits-and-pieces player in this World Cup is Jeevan Mendis, who, based on recent form, has probably been picked more as a front-line bowler than for his batting. There's also Fabian Allen of West Indies, who, as it stands, would likely not be selected to bat top seven or bowl ten overs, but who can be a top-seven batsman in the future. He is probably here as a back-up spinner and a super dog fielder.

The bits-and-pieces allrounders of ODI cricket's history seem to be disappearing. Look at the difference between England's allrounders today and the ones they had in the '90s. Players like Ronnie Irani, Gavin Hamilton and David Capel didn't possess one skill that made them international cricketers. It was the package combined that did. Whereas, Adil Rashid, Chris Woakes and Moeen are all international-quality in at least one skill.

Match-ups and faster scoring rates are factors now. The more variety you have in your attack, the better. For example, you want a batsman who can spin the ball the other way to your main spinner. That means the number of allrounders available should be at an all-time high.

At times it feels like today there are more allrounders than in the past, but maybe that's because England have so many. If you look at their line-up against West Indies in Southampton, you see that eight of their players have some level of all-round skill. Only Mark Wood, Eoin Morgan and Jason Roy don't. If you look at the squads for this World Cup and include wicketkeepers as allrounders (only Afghanistan's replacement keeper, Ikram Alikhil, is not one), there are around 60 to 65 players who, you could say, have all-round skills. That ranges from Shakib's all-round mastery to Finch's slow, naked full tosses. That's over a third of the players on show, but how many of them are successful?

T20 has also changed cricket's landscape when it comes to finding or manufacturing all-round talent. Colin Ingram was a young legspinner when he first broke into professional cricket. Then he became a keeper, before ending up as a specialist batsman who bowled legspin. Like Joe Denly, he went back to bowling leggies regularly because it was another skill that helped in T20.

There is a commercial reason to use these secondary skills in T20. Earlier, players would often come into the game with all-round skills, then focus on their primary job, so the other one fell away. Now even having the ability to bowl an over a game in T20, or hit sixes every eight balls in the death, is worth money.

Without T20 cricket, Sunil Narine would be a sloggy tailender. There is a financial reason for him to bat better, and he has spent years honing that ability. T20 takes a skill from an accident in the nets or a game and makes it a financial imperative.

Michael Bevan only bowled in 27% of his ODIs, though his springy, unpredictable fast wristspin was a useful skill. Unless injury intervenes, today you would expect players like Bevan to be used more, partially because of tactics but also because they take those skills more seriously.

Back then, Bevan, who would have been picked for Australia even if he couldn't bowl, didn't need those skills since he was already on the team sheet as a batsman. But today a player like that, say, Root, needs those skills, not just for the odd over to help his country but because of the free market. You only need to see Root around the England nets, trying carrom balls, doosras and knuckleballs to know that that's true. Here is one of the world's best batsmen practising his part-time spin looking for that something extra. There are others in this tournament pitching in with their secondary skills, like Coulter-Nile and Rashid Khan with their hitting.

Of course, it's also clear that cricketers are becoming specialist ever more narrowly - death bowlers, top-order cameo batsmen, No. 7 non-bowling hitters. White-ball cricket seems to crave the best of the best now. Part-timers and bits-and-pieces won't do, but because of this quirk of the game where you need a fifth bowler, there are still compromises that have to be made if you can't find a true allrounder. If the maximum amount of overs per bowler in limited overs was increased from 20% to around 25%, the situation would be very interesting for allrounders. Until then, teams will continue to hope a genius, or a few, will arrive on their doorstep.

When you try to look at the numbers, there is no easy way to quantify what an allrounder is. Many players who may be allrounders at the domestic level don't play as allrounders at the top level. And Denly was never seen as an allrounder until just before being picked for England.

Shiva Jayaraman, ESPNcricinfo's senior stats analyst, and I came up with a metric. We looked at players who average 18 balls a game when bowling and bat 18 deliveries without being dismissed (wicketkeepers included here). Our system is not perfect; there are a few dogged tailenders who slip in. Plus, not everyone will agree that Lehmann's sidearm sliders make him an allrounder. But these players are involved in two disciplines for a decent amount of time per game.

In ODI history, 1033 players have played a minimum of 20 matches, and of those, 313 are batting and bowling allrounders, going by the 18-ball metric; 79 keep and bat decently. That means 38% of players have all-round skill in ODI cricket.

The question is: are there more allrounders today than before? By our metric, of the 194 players who have played over 20 ODIs between the 2015 and 2019 World Cups, 38% again have all-round abilities.

White-ball cricket has never craved allrounders more or rewarded them with more cash, but even then, the numbers have not changed. Most teams dream of a unicorn like Shakib and just make do with what they have.