Picking Hameed

The season is over  - ended in fine style by Middlesex at Lords. The next event on the English cricket fan's horizon is the Bangladesh tour. One of the really exciting things about that tour will be watching how Haseeb Hameed fares. He has been announced as part of the squad and will almost certainly open alongside Cook.

The case for Hameed's selection grew steadily over the summer. Evidence of his fine batting accumulated run by run and the expectation of fans was pumped up superlative by superlative in newspapers and commentary boxes around the country. England have selected a gifted teenager to open the batting. What could be more exciting?

But here's my question... how did they actually pick Hameed? Because it seems very odd to me that they overlooked openers who scored more and averaged more. I hope it turns out well for Hameed and for England but I'm not convinced by the decision and I think that looking at the process of how it happened tells us a lot of interesting things about English cricket.

The hype

Hameed's selection was announced in early September and there swiftly followed a hot blast of cricket press attention. If you live in cricket world then you won't have been able to ignore it (although it probably hasn't made much of a dent if you don't). Sitting back and reviewing some of the things said in the press it's hard not to see them as hype.

First of all there was the background-to-greatness stuff like the Telegraph telling us that Hameed was a "once in a decade" batsman who grew up practising in his living room with his Dad throwing balls from the sofa - as if his end of career biography was already being written.

Then there was the Guardian who opened their report of the selection by quoting Mike Newell, England selector and Notts supremo, saying that Hameed was born to open in Test matches.

Others compared him to Mike Atherton (another Lancastrian opener) and almost everyone has pointed out how rare it is for England to pick a teenager - allowing more comparisons, this time to Brian Close. Probably my favourite bit of hyperbole though came from ESPN Cricinfo.

On their influential website Paul Edwards penned a piece which was titled "A Chateau Lafite cricketer in a wine-box world". The article wasted no time in getting very emotional "Watching Haseeb Hameed bat is to be reminded of one's innocence and Blake's 'echoing green'"... wrote Edwards and he went on to explain why "...because for all the technique, the coaching, the selection of shot, there is in his play a palimpsest of his childhood."

Leaving aside the fact that it's a prime candidate for Pseuds Corner, Edwards' piece is a classic example of the hype that has gathered around Hameed. A Chateau Lafite cricketer? Hameed hasn't played a single international game and his career consists of only a season and a half of county cricket. Yet he's described as if he was already a legend of the game. It all seems to amount to praise for acts that are yet to be accomplished.

If he had averaged 90 for Lancashire this season then perhaps it would make sense. But he didn't. He averaged 49.92. It's good but it's not the best out there, not for openers and not by a long shot and that includes England qualified players.

The numbers

Hameed for his part has come across as modest and very clearly focussed on his own game and can't be blamed for the over-heated opinions of others. And yet the hype has come thick and fast. So why? Why is there such a weight of expectation around him?

Because if you're looking for an opening batsman to replace Hales, then why not the ones who are actually averaging the most this year? Why not Keaton Jennings from Durham or Nick Gubbins from Middlesex? What have they done wrong?

They both averaged more and scored more runs in the aggregate. Jennings averaged 64.5 (1548 runs in total) and Gubbins averaged 61.26 (1409 runs in total). They're both a long way ahead of Hameed who averaged 49.92 across a total of 1198 Championship runs.

And when I say a long way ahead I mean it. Jennings' average is 29% higher - Gubbins' 26% higher. These don't seem like statistics you can just wave away.

Red ink inflation?

Maybe the reason for the Hameed selection lies in the fact that averages aren't everything and that over the course of a season averages can be inflated and misconstrued.  So let's have a look at that....

One of the things that can really inflate an average is the not outs column - red ink inflation. It happens all the time. A batsman gets lucky with a series of not outs and finds his average pleasantly padded out. So what about not outs for Hameed, Jennings and Gubbins - who comes out on top?

Actually the red ink inflation argument doesn't help Hameed at all. Jennings and Gubbins haven't had any more not outs than he has. In fact Gubbins has had two fewer. And a more sophisticated adjustment for not outs doesn't change the picture either...

If you wanted to adjust for red ink inflation, here's what you might do. Count all the innings as statistical "outs" even if the batsman was in fact not out. Then all you'd need to do is adjust for the unfairness of a batsman with a high average getting left stranded not out on a low score. And you could do this fairly simply by rounding up his below average not out scores to be the same as his average.

This way 200 not out becomes 200 and out. And 3 not out for a player averaging 42 becomes 42 and out. It's a bit like the map projection in geography that distorts shape in order to be true to size - this method distorts the runs total in order to be true to the "actual" average.

So what happens when you do that? Well Hameed's average drops by 5.55, Jennings by 6.63 and Gubbins by 2.55. That still leaves Jennings (adjusted average 57.87) and Gubbins (adjusted average 58.71) well ahead of Hameed (adjusted average 44.37).

Home ground advantage?

So what about home ground advantage? Because this is the other thing that could affect an average - one batsmen playing half his matches on a flat home track while another has to play home matches on bowler friendly wickets.

This season the average team score for completed innings in Hameed's home matches was 314. While the same number for Gubbins' home matches was higher at 345, in Jennings' home matches it was actually slightly lower at 312. Trying to take into account home ground advantage favours Jennings as much as Hameed.

The narrative argument

So why Hameed? What is it? Because it clearly isn't the statistical evidence. 

Well if you're being cynical you'd say it's a narrative thing. he's just one of a group of talented opening batsmen but what marks him out is a narrative that's been constructed around him. Not one that he's constructed himself but a construction nonetheless.

The building blocks are simple enough. He's very young. He's come up from the Lancashire leagues and scored runs all along the line. And of course he's a British Asian player at a time when not enough of them are breaking through and when a lot of people suspect that more could be done and that high profile role models would help enormously.

That narrative is only helped by the fact that when you see Hameed interviewed or look at his Twitter feed he comes across as humble, grateful and mature for his age.

Passing the eyeball test

So maybe it's narrative. But if you don't buy that there's another fairly obvious reason for his selection. Simply put, the people who have seen him play in the flesh have "seen something". They've judged his technique and temperament to be better than that of Jennings or Gubbins or any other county opening batsman qualified to play for England.

And that's interesting because it's probably what the selectors have done here - made eyeball judgements and listened to the eyeball judgements of others. All of which has helped them to understand that his technique and temperament are better.... except there's a problem.

If Hameed's technique and temperament are better than Jennings' or Gubbins' then why doesn't he average more than they do? If he's a better player why doesn't he play better?

Because the problem with judging his technique and his temperament is that that's all subjective. I mean what is a good technique? Presumably the best technique is the one that scores runs and doesn't get out. And what is a good temperament? Well again presumably it's one that helps you to score runs and not get out. This year Hameed has been very good at those things but he hasn't been quite as good at it as Jennings or Gubbins and that's clearly what the stats tell you.

Picking on a hunch

So I'm struggling to understand the Hameed selection. That may seem a bit harsh given that he's had such an excellent season and shown a lot of talent. But it isn't really about being anti-Hameed or pro-Jennings or anything else. It's about being pro understanding selection. It's about being pro picking players based on performance and anti picking on hunches.

Because I think picking on hunches takes you into dangerous territory. Unconscious bias and ingrained assumptions come into play and you miss good players even though they're performing well right in front of your eyes.

I'm assuming here that the selectors don't have access to a giant secret database of stats (and maybe they do and their analysis is a lot more sophisticated than what I'm able to do at home with a simple spreadsheet). But if we assume that they don't have a secret statistical weapon then they have picked Hameed on entirely subjective criteria - they have picked him based on a hunch.

Is that unfair given how good his numbers are? Well, if the numbers tell you that Jennings and Gubbins are better opening batsmen and you still pick Hameed then it's a hunch. And at a time when so many other sports are advancing their understanding of the game through statistics it makes me scratch my head.

Not trusting the data

Frankly it looks as if the selectors don't trust county cricket to be a good barometer of success. And that might not come as a surprise to a lot of cricket fans. After all the summer started with the chairman of the ECB suggesting that one of the key competitions was "mediocre" and ended with a £1.5 million promise to each county as long as they sanctioned a competition that wouldn't include them (city-based cricket).

Of course the ECB and the selectors will feel that their attitude is born out by the history of player selection. They would point to players picked for England based on county stats who underperformed and players picked on a hunch who went on to do great things.

They would point to Hick who was a giant for his county and a minnow for his country, and they would point to Trescothick who was picked on Duncan Fletcher's hunch and went on to score plenty.

So I can understand why selectors are so willing to pick on subjective factors rather than hard stats. But frankly it's a mindset that doesn't really convince me. I think it leads to players being picked on style rather than substance. And I think that over time it corrodes the link between the county game and Test cricket.

Sometimes it even seems to get to the point where good players are overlooked because of good county careers. Sometimes it's as if there's something cheap about scoring runs or taking wickets time after time, day after day in the country's premier competition. As if this somehow makes you less fit to play for England in the long term.

For a young player wanting to play Test cricket it's not impossible to imagine that the optimal strategy to getting picked would be to get into England youth and Lions teams nice and early and spend as little time as possible actually playing for their county. All of which is a huge shame. 

Nevertheless it seems to be happening and I think we're starting to see the results with some of the players getting picked for the Test team. I think James Vince was a good example of this over the summer. He has a long way to go as a player and may come back to have a successful Test career but frankly the fact that he was picked to play for England following a season in which he averaged 28.58 while players like James Hildreth who averaged 53.46 (only Bairstow, Bell and Voges averaged more) was ignored is telling.

Rounding up

So there we are - I hope the hunch of picking Hameed turns out to be justified and that he scores gallons of runs in Bangladesh and India this winter. Who knows he may even justify the Chateau Lafite comparison. In which case I'll drink a bottle to celebrate.

Nevertheless in the absurd parallel universe where I select the England team Hameed would never have been on the plane. I'd have taken Jennings with Gubbins as a reserve. More important than my fantasy selection though is the question of how these decisions are actually made. The next time an England selector picks someone because of technique and temperament instead of numbers I want to know why.