All Rounders

I’ve often wondered about all-rounders in test cricket – how much they’re worth, and where exactly is the tipping point between them being worth more than one specialist and less than one specialist.

England’s current test team is deemed to have a lot of all-rounders, though I think we should get the past the notion that wicketkeeper-batsmen are all-rounders (in particular keeper-batsmen, like Buttler, who aren’t even keeping). It is a given now that any test keeper worth his place should average 30+, hopefully more like 40 – it’s not really added value.

For a couple of decades, England was obsessed with finding the new Ian Botham. In that period, there were a lot of all-rounders given a go, none of whom were up to the job – the likes of Derek Pringle, David Capel, Adam Hollioake, Dermot Reeve, Chris Lewis … Lewis, unlike the other four, was close to being a test bowler level. Dominic Cork and Darren Gough, who were called all-rounders early on in their career, became specialist bowlers who could bat a bit. Crucially, their bowling was good enough on its own for a place in the side.

That is the fundamental truth of nearly all all-rounders – they have to be good enough to earn a place as a bowler in the side. With two (ish) significant exceptions, all the great all-rounders were more bowlers than batters.

Why is that? I’d put it like this  - a bowler (particularly a seamer) who is not good enough is, essentially, wasting overs that a better bowler could be bowling. Of course, this is not always true, but, very often, 4 or 5 bowlers can do all the work in the innings without getting excessively tired – only sometimes do test innings require a bowler who can “take up an end” while the actual wicket-takers recuperate.

Whereas this is a little less true of batsmen, certainly not in the same way. A guy who averages 30 is, simply, only worth 5 runs less than a guy who averages 35 – there is less, as such, being wasted by having an all-rounder who’s not quite as good as a specialist batter than an all-rounder who not quite as good as a specialist bowler. eg if Stokes has a really tough time with the bat for a year but carries on bowling well, they won’t look to replace him, but if his bowling goes to pot, and his batting just carries on as is, at a point, that seriously affects the dynamic of the side …)

I hope that makes sense.

I think, early on in their careers, all-rounders realise that it’s the bowling that will keep them in the side, that is the crucial, day-in day-out, this-cannot-fail, side to their play. For most of the great all-rounders, it is not fundamental to their side’s success that their batting comes off, unlike their bowling. So I reckon they concentrate more on their bowling.

If you look at Stokes, his bowling has come on in tests, in the way that maybe his batting hasn’t. He needs to be better than England’s other options as 3^rd/4^th seamer.

There are a lot of cricketers who are close to being all-rounders but choose not to be because they recognise that attempting to be will take away from their strongest suit – on the batting side, the likes of Joe Root, Pietersen, Atherton, Viv Richards, Allan Border, on the bowling side, the likes of Stuart Broad, Malcolm Marshall and Gough. The true all-rounders are, I suppose, the ones that can hang on in there with both strings.

Having said that, in modern cricket, it has really dawned on people that “let the batsmen score the runs” is self-defeating, and that lower-order runs are utterly, vitally important to team success – good test teams can really only afford one utter rabbit – 8, 9 and 10 need to be players who can both hang around and occasionally change the course of the game. Broad, early on, and Swann, were magnificent at that. Now, I think, England are hoping for the same from Curran and Rashid. Those late-order runs have been key to the series with India.

But, nevertheless, the test will come when they really need to earn their corn as bowlers. Are Curran and, in particular, Rashid good enough bowlers to play test cricket? This is the question a lot of people are asking. We’ll find out, most likely, in the winter. Rashid has been used sparingly, but fairly effectively, this series. In Asia, he (or whoever the spinner is) will need to bowl more, will need to fulfil a different role. Thus far, his statistics might suggest that he’s not a good enough spinner for test cricket.

But, here an important statistical note is required. It is a little different for spinners than seamers. A seamer averaging 40 is not worth their place in the side, a spinner may be, not just because of paucity of other options, but also because, on slow wickets on hot days, spinners are needed to bowl and bowl. If you look at Rashid, his baptism of fire in test cricket required him to bowl several long spells against top-class Pakistani cricketers in the UAE – ending up with a 1-150 in a way that a young seamer would simply not be required to bowl such a long, thankless spell.

Looking at Moeen Ali (an interesting type of all-rounder) proves this. It feels like Mo has bowled a lot of match-winning spells for England, that he’s shown he’s a difference-making test cricketer time and again. Yet his bowling average is still only 39.

Even Shane Warne, considered the greatest bowler of the era, only averages 25 (compared to McGrath’s 21). Some highly acclaimed spinners like Abdul Qadir, Mushtaq Ahmed, Danish Kaneria, Monty Panesar all average over 30 in tests. It’s those long but necessary spells that distort it.

So, I’d say, going back to Mo, that, despite a batting average of 32 and bowling 39, he is a true all-rounder. He’s a difference-maker with both skills in the way that, say, David Capel was not.

But … then we look at his disastrous tour of Australian in 2017-18 and we can see that he singularly failed as the side’s solitary all-rounder.

I think this is where it gets interesting as to how the make-up of a test side has changed in the last decade, and why, as England stand now, I think the more all-rounders the better.

These days, cricketers are taught to take more risks, be difference-makers. They come off or they don’t. This is rather different from eg Flintoff (prime), Cairns, Vettori, who were consistent cricketers. With Flintoff in his golden period, you knew what you were getting (lots of 50s, lots of tight 3-fors), you could bank on it and build your side around it – you didn’t need any other all-rounders. Having one proper all-rounder when you’ve got quality, consistent batsmen and quality bowlers is just fine.

Nowadays, England have several players who can potentially influence a game with bat and ball, but will not do so every game – hence, if you play 4 or 5 of Stokes, Ali, Woakes, Curran, Rashid, Broad, they can cover for each other’s failure, but there’s a good chance that two or three of them will come off with bat and/or ball (as opposed to eg Derek Pringle, who could do both reasonably well, but neither to the extent of dramatically altering a test match).

Generally, the rule of thumb for an all-rounder is that their batting average is higher than their bowling – I’d add that you can’t really be a true all-rounder if you’re batting average is under 30 or your bowling average over 40.

Thus, Stokes is doing all right, Ali’s there on a technicality, the best all-rounders in world cricket are bowlers who make more runs than you’d think, like Jadeja and Ashwin, and, the purest all-rounder currently, who has no option but to always try to make an impact with bat and ball, Shakib al-Hasan of Bangladesh, who averages 40 with the bat and 32 with the ball.

Of the great bowling all-rounders, Hadlee was the nearest to being a bowler who batted a bit (batting average 27), likewise Wasim Akram, Shaun Pollock’s stats are surprisingly spectacular (32 and 23), Tony Greig’s excellent (40 and 32), Botham & Kapil Dev outstanding, and Imran Khan (37 and 22) pretty much the best.

Then there are the outliers, Sobers and Kallis -  the only two who managed to be world-class batters and still do a job with the ball. I never watched Sobers (whose bowling average on its own was a good but not amazing 34) but, I assume that, like Kallis, if he’d concentrated on bowling, he’d have taken 400ish wickets at under 30. So good were they with the bat that everything they did with the ball was a bonus, and yet they still had the statistics of frontline bowlers.

That is rare.

If you’re trying to design a batting order, I’d say, without compromise, batsmen up to 6 need to average 35+, 7 probably 30+ (depending on whether they’re keeper), and bowlers at 8, 9, 10 and 11 should average under 35.

Within that, there is space for a little flexibility – a spinner who can bat pretty well (average 25+) and averages 37+ but takes key wickets. A number 6 like Stokes who averages low 30s with bat but can really change the game and also can bowl.

At the moment, Rashid’s figures (20 with bat, 40 with ball) are not quite good enough, they’re a bit Pringle-esque, though also not far from Emburey, and if he can get that to 24 and 36, I’d say that would be worth him having in the team above a slightly more consistent specialist spinner.

I’d say all these all-rounders are flooding test cricket at the moment because of ODIs and also because of the realisation of how important late-order runs are.
To answer my question, what’s the optimum number of all-rounders (assuming they average 30 with bat and sub-40 with ball). I think, though of course it varies depending on their quality and on whether you actually have strike bowlers, that there’s no point in having more than will give you 6 proper useable options with the ball, but you simply can’t have enough people that can bat competently.

Nearly all world-class all-rounders either start up as, or end up as, primarily bowlers - (even ones who start off thinking they're batters, like Stokes and Flintoff) - so I think the term all-rounder doesn't quite mean what it used to. Any truly world-class specialist bowler is not going to be left out of a test side, but a large proportion of bowlers now are not that far from averaging 20ish with the bat. They're almost all-rounders by default.

P.S. I had a further thought, based on personal experience, and I wonder if it holds generally true. Early on in schoolboy cricket, you get a lot of the best players doing everything, opening the bowling and batting. As it progresses though, I think specialism is encouraged more - it's seen as almost indecent for one or two people to do everything for a team.

I was pushed, as I reached 15/16, towards bowling being my stronger suit, which I totally went along with, and certainly always believed it was. I became a bowler who could bat a bit, or at best, a bowling all-rounder.

And it was only when my bowling went to pot in my early 20s that I put a bit more into batting, made quite a few good scores, and realised that I'd been pretty good at it all along. (equally, it may well have been down to my own diffidence. I loved the security of being a bowler - you knew you'd get, nearly every game, at least some stretch of a chance to influence things, whereas batting high up was a) nervewracking and b) could be over in a minute).

I only mention this because I wonder if there's a culture within youth cricket coaching where not everyone is encouraged to be as good as they can be at everything. I'm sure doing only one thing very well suits a lot of players, but I wonder if, for many years, a lot of all-rounders fell by the wayside.

Now, because it's such a necessity for T20s and ODIs, it may be that young cricketers are really being allowed to focus on all their skills right through their development.

And ... here's a balanced team of "all-rounders" from since I've been watching cricket (not all batting 6)

Jayasuriya
Gayle (Sehwag and Gooch also acceptable bowling openers)
Kallis
Tendulkar
Shakib-Al Hasan
Botham (though I prefer Flintoff)
Dhoni (WK, also a bowler!)
Imran Khan
R Ashwin
Richard Hadlee
Wasim Akram