The sun is beating down on a parched Old Trafford in the hottest English summer on recent record. Derbyshire are 81 for 2, one ball short of halfway through their chase of 158. Wayne Madsen and Calum MacLeod have put on 51 for the third wicket, and look untroubled against Lancashire's spin-heavy attack.
MacLeod has 30 off 28; he's yet to free the shackles, but he is a slow starter in T20, his value drawn from batting deep into an innings. After hitting the first ball he faced from Keaton Jennings - an inconspicuous medium pacer in the short form - for four, MacLeod looks in control.
Jennings drags the final ball of the over down the leg side, and MacLeod, trying to flick him into the legside, lifts his back leg. Dane Vilas, Lancashire's experienced, bearded South African wicketkeeper, reacts sharply. Unsighted by MacLeod, he jumps over to the legside, takes the ball cleanly, and whips off the bails. MacLeod is way short of his ground, and the wicket brings about a Derbyshire collapse: they manage only 64 for 4 in the second half of their innings, and fall 12 runs short of a gettable total.
Vilas' stumping changed the game. It was a difficult chance, which required not just quick hands but perfect footwork, and he pulled it off. And yet, on the way out of Old Trafford, the conversation is not about the moment that turned the run chase. Inevitably, Liam Livingstone's quickfire 44 and young tyro Toby Lester's four wickets are the details that are remembered; Vilas' involvement in the match is noted as much for his scratchy 11 with the bat as his magic with the gloves.
Last winter, Yorkshire's Jonny Tattersall had a problem. He had a mountain of 2nd XI runs to his name, but had not been given a run of games for the first team. Leaving Headingley was not an attractive option, but there was no point staying to play second-team cricket. His coach, Andrew Gale, had an idea. Tattersall had always been a good fielder, and came from a family of wicket-keepers. The club knew they would be without Jonny Bairstow for most of the 2018 season, and Andy Hodd, though highly respected across the county circuit, was not a heavy runscorer and was not getting any younger.
Gale recalled: "We had a chat with his at Christmas and said 'why not just give keeping a go until March and see where it takes you? If it's not working out and it's not for you, then we've not lost anything.'"
Tattersall, a diminutive man with an impudent, toothy grin, described it as a "cheeky" career move, and it has paid off for him: he is now Yorkshire's first choice with the gloves when Jonny Bairstow is on England duty, and kept wicket in all of Yorkshire's T20 games this season. The perception is that he held his own behind the stumps and his decent form with the bat means the move has paid dividends.
For those with half-full glasses, his story shows the benefit of hard work and innovation; for their half-empty counterparts, it is a damning statement on the standard of wicket-keeping at county level that Tattersall can pick up the 'art' at Christmas and be a county's first-choice only months later.
Richard Barker, the analyst who was instrumental in Northamptonshire's T20 revival between 2013 and 2016, thinks that change is coming. Since the advent of T20 some 15 years ago, wicket-keepers have almost invariably been viewed as stoppers, or, to borrow a phrase from ESPNcricinfo's Jarrod Kimber, point fielders with gloves on. They are picked as batsmen, who can do a job behind the stumps. A true specialist wicket-keeper who can bat in the top six - Ben Foakes of Surrey, or Worcestershire's Ben Cox - is a welcome bonus.
Barker predicts a shift. "As bowlers are now all expected to bat; data such as how many balls on average a number 7 or 8 is likely to face could lead to a revival of the wicketkeeper as a specialist," he told The Guardian last year. "Some baseball teams have a personal catch for their star pitcher. If he's unlikely to face more than half a dozen balls and there is enough firepower in the line-up, there is a case for picking the best wicketkeeper and having him stand up to all but the quickest bowlers to prevent the batsman from leaving his crease."
Put simply, he thinks the increased role of data in T20 cricket will bring about the revival of the specialist keeper.
It is a strategy that has already been used to great effect, with Hampshire providing the best example. In the 2010 and 2012 finals - both of which they won - their wicket-keeper Michael Bates, then a quiet young pro, was due to bat at No 10. He played 15 times across the two tournaments, and didn't bat once. Bates was faultless keeping to the spinners, and also stood up to the medium pace of Dominic Cork, Dimi Mascarenhas, and Chris Wood. Batsmen were unable to use the crease as they would have liked, and across the two finals, Bates let through one bye in 40 overs. Bates was instrumental to their success.
But at the end of the 2012 season, Hampshire signed Adam Wheater from Essex. The Hampshire hierarchy recognised that Wheater was nowhere near Bates' level with the gloves on, but he was an attacking option at the top of the order. Wheater might have missed a couple more chances behind the stumps, but he scored more runs, and allowed the team to pick another bowling option down the order. Two years later, Bates was unceremoniously released, and now, aged 27, he is a specialist wicket-keeping coach.
"The advantage we had as a team back then was the attack we had very much suited me being up to the stumps a lot of the time," he says. "We had Dimi Mascarenhas leading the attack, and we had a lot of spin.
"As much as a spinner is creating chances, if you've got someone behind the sticks who's not capable of taking those chances, then it negates the point"
"As much as a spinner is creating chances, if you've got someone behind the sticks who's not capable of taking those chances, then it negates the point."
Of course, Bates is right: high-quality spinners and pace-off bowlers are more likely to create chances for their wicket-keepers standing up to the stumps than an attack full of quick bowlers in the shortest format. That much is evident from this season's Blast: the keepers with the most dismissals - Vilas (12ct, 6st) and Durham's Stuart Poynter (13ct, 3st) - both stand up more than usual thanks to spin-heavy strategies.
But using dismissals statistics immediately raises issues. While Vilas and Poynter may have taken the most catches and effected the most stumpings, there is no publicly available data as to how many chances they have failed to take.
Similarly, both have benefitted from being part of a side that creates plenty of opportunities thanks to a confident and able bowling attack. Northamptonshire's keepers, Ricardo Vasconcelos and Ben Duckett, took only 7 catches behind the stumps between them this season compared to Vilas' 12, but that was influenced by their bowlers' impotence. Indeed, with such small sample sizes, it is difficult to draw valid conclusions from the available dataset.
That is not to say no data exists. Opta manually collates data on catches and stumpings in the Blast, but their three levels of difficulty - easy, medium and hard - are subjective and tend to be submitted on face value alone. This provides an issue: there is an element of subjectivity in analysis of wicket-keepers that doesn't exist in other parts of the game, as what one analyst considers a regulation chance could be very difficult to another, and vice versa.
Similarly, most models manage to punish keepers who attempt difficult skills. If Poynter stands up to Chris Rushworth, who typically bowls 75-80mph in the Blast, and misses a legside stumping chance, few models account for the fact he has been bold enough to stand up to a seamer, with all the benefits that entails in terms of restricting the batsman, and instead the miss registers against Poynter as worse than if he had been standing back.
Various freelance analysts and bloggers have devised their own metrics to compare glovework, using byes/leg byes conceded and runs scored by players after drops and missed stumpings as metrics. But again, these rely on unscientific data and cannot accurately reflect the value of a wicketkeeper - any adjustments made will again be subjective.
This leaves room for inaccuracies; and while adjusted models like England's 'weighted averages' in Test matches are being used, those models have had far more time and effort devoted to them than keeping ones. There are just so many variables and so few constants: it is hardly surprising that Cricket Australia's 2016 attempt to implement a 'fielding average' stat never took off.
And outcomes do not always reflect the course of events. At the start of this season's Blast, Poynter stumped Leicestershire's Neil Dexter off Imran Tahir's bowling, despite initially fumbling and then dropping the ball. If Dexter had been more alert, Poynter's error would have been logged as a missed stumping; instead, despite his ordinary glovework, it went down as successful.
Not all chances are created equal. "If Kevin Pietersen walks past one from an offspinner, and a part-time keeper, there to strengthen the batting, misses the chance, and Pietersen goes on to make 75 off 30, he wins the game," says Darren Berry, a former Victoria gloveman who has coached Rajasthan Royals and Adelaide Strikers. "If the keeper takes it, it changes the entire game. Maybe the ball doesn't get through to the keeper much, but the times it does, it is often very important."
But such moments are rare. There might be a few instances when a keeper has a difficult chance to dismiss their opposition's best batsman over the course of the season, but a single bowler can have 10 or 15 in a single innings. As teams try to maximise their resources, they think there is more sense investing in a better bowler and a better bowling coach than in weakening their batting by picking a specialist keeper and hiring a coach for him. Around the world, it is noticeable how few franchises have a specialist wicket-keeping coach on their staff.
"As of yet, there isn't actually any way to quantify the quality of a keeper," says Bates. "It's such a difficult thing to quantify. If someone was able to come up with some sort of strategy or data collection that was able to do that, I think that would certainly help the argument for having a really good keeper behind the sticks, especially in that instance when you have a pace-off seam attack or a couple of quality spinners.
"But until you've got data to back that up, you're going on coaches' opinions alone really."
"You've got a lot of coaches making decisions on keepers who don't necessarily know that much about keeping"
Of course, data is by not entirely absent when teams consider which keeper to pick. The problem as far as the specialist is concerned is that the only data thought to be sufficiently reliable is their batting. While it is difficult to compare Keeper A and Keeper B's glovework in terms of value to the team, knowing that Keeper A averages 25 with a strike rate of 150 and Keeper B averages 15 at a strike rate of 110 means the easy choice for most coaches is pretty clear.
Bates laments: "You've got a lot of coaches making decisions on keepers who don't necessarily know that much about keeping - they're looking at it with a fairly untrained eye. If you've got a keeper who is whacking it around and getting runs - high average, high strike rate etc. - that's easily quantifiable, whereas how they can affect a game with the gloves alone is based purely on opinion."
The inference is clear. Until data that is not only comprehensive but also trustworthy and readily available is developed, Barker's suggestion that the specialist keeper will make a comeback has an element of wishful thinking.
Even when that data is properly developed, there is not necessarily much reason for the specialist to be hopeful. One IPL analyst suggests that the very best keepers are worth approximately three runs more to a team over the course of the average T20 innings compared to the very worst. While that might mean someone like Bates is slightly more valuable than a keeper with similar batting ability but worse with the gloves, it does not bring about much hope that the rebirth of the specialist keeper lies around the corner.
Perhaps Barker is right. More detail and clarity in wicketkeeping data could transform the way T20 sides are structured. The specialist gloveman could make a return on particular surfaces, and the 'stopper' might be jettisoned. It is not implausible, and the speed at which T20 has evolved in the past five years shows that nothing should be written off.
But to reach the point of change requires several things. The data must become better. Teams must be receptive to it, and it must prove that picking a specialist keeper is a beneficial strategy. When you bear in mind that players like Bates are now few and far between with a generation of young keepers encouraged to focus on their batting or risk never making it to the top, the return of the specialist looks like little more than a pipe dream.