Viewed as a standalone event, you could call it a poor shot. A full ball outside off stump, leaving Wiaan Mulder late, and a loose-looking drive edged to the keeper. You could pick holes in the execution of the shot: front foot not far enough forward or far enough across, front shoulder open, all this opening up a dangerous distance between the line of the head and the line of the ball.

Most wickets in Test cricket aren't standalone events, though, and this one definitely wasn't. These were the three other balls Mulder had faced in that over: a full tempter, left alone; an in-ducker from a good length, jabbed at from the crease and inside-edged through square leg for a single; a short, rising, away-swerving ball in the corridor, left alone.

Similar line, different lengths, movement in both directions. A classic Mohammed Shami over, inducing all the uncertainty and scrambled footwork of the wicket ball.

The full, drive-me nature of the wicket ball itself, though? Not necessarily what you'd term classic Shami.

No, the widely-held image of Shami is of a bowler of virtuoso skill and a hypnotic fourth-stump line who discomfits the best batters in the world without always getting the reward for it, thanks to a natural length that is perfect for subcontinental pitches but a touch short for surfaces with a bit more bounce in them. A beat-the-bat length rather than a find-the-edge length. During India's recent tour of England, the TV commentary team endlessly mulled over the relationship between his length and his notorious lack of luck on his previous visit to the country in 2018.

And there was some truth to this, at least during India's 2018-19 cycle of away tours. Enough truth for it to become a one-line description of Shami the bowler: the guy who beats the bat but doesn't find the edge.

But as all lavishly gifted cricketers tend to, Shami has evolved, and that one-liner - hugely reductive even when it contained its grain of truth - is now out of date. It was out of date even in Adelaide last year, where his stump-to-stump relentlessness, aided by a strong leg-side field, played a massive role in Steven Smith only scoring 1 off 29 balls in the first innings.

But Shami didn't dismiss Smith, or anyone else, and all of India's other bowlers took wickets in that innings. Then Shami was the last victim in 36 all out - retired hurt with a broken arm - and everything that came before was a distant memory.

Shami was India's most successful fast bowler in the World Test Championship final against New Zealand, and took 11 wickets at 27.54 in the Test series that followed against England. Superb numbers, but his bowling wasn't a headline act in either of India's two wins: his utterly unexpected final-day half-century overshadowed his three wickets at Lord's, and he was rested at The Oval.

Perhaps it's inaccurate to say Shami's evolution has gone unnoticed. But it's slipped under the radar somewhat, without a defining performance to get everyone talking about it. Until Tuesday.

Even before India bowled on Tuesday, you probably expected Shami to trouble South Africa. India had collapsed in the morning from 272 for 3 to 327 all out, and the primary cause for this had been the quickening of this Centurion pitch, which had made its uneven bounce more dangerous. And uneven bounce makes Shami's natural lines and lengths more dangerous, as he's shown plenty of times in the past - in India, of course, but also while picking up five-wicket hauls at Johannesburg and Perth in 2018.

As a standalone event, the most memorable of Shami's five wickets on Tuesday was straight out of his uneven-bounce playbook. At the start of his fourth over, he had got a back-of-a-length ball to rear unexpectedly at Aiden Markram, and the opener had done well to yank his gloves and body out of the way. The fifth ball of the over pitched on a not-dissimilar length - a few inches fuller, perhaps, but not a whole lot fuller - and skidded through low while straightening off the deck. Markram opened up completely, with both feet off the ground, as the ball, a miracle of geometry, slid past outside edge and back thigh and clanked into the top of off.

This was the sort of delivery Shami was expected to produce, over and over, on this Centurion pitch. But it wasn't quite as straightforward as that. South Africa had slipped to 32 for 4 against the new ball, and India's collapse had come against the second new ball. As the ball grew older and softer, it misbehaved less and less.

The control numbers told the story. South Africa's batters managed a control percentage of 77.78 in overs 1-15. Thereafter, it rose to 83.90 - this despite the lower order facing a lot of balls.

Conditions, like cricketers, evolve. On this day, Shami showed the range of skills to keep getting wickets even as uncertain bounce became less of an issue. There was some seam movement, but not a lot of it compared to, say, the English pitches India have lately played on, so Shami even deviated from the perfectly vertical seam he's known for, occasionally employing one canted slightly in the direction of the slips in an effort to find outswing.

Perhaps it's inaccurate to say Shami's evolution has gone unnoticed. But it's slipped under the radar somewhat, without a defining performance to get everyone talking about it. Until Tuesday

Seam movement brought Shami his first two wickets - inwards to castle Keegan Petersen off the inside edge, outwards to bowl Markram. Replays of his third and fourth wickets suggest it may have been late swing rather than seam that consumed Mulder and Temba Bavuma, both batters drawn to the front-foot and into their doom by full lengths in the corridor. By this time, the Kookaburra's seam had flattened significantly.

Then came wicket number five, and wicket number 200 in Test cricket. Around the wicket and wide of the crease to the left-handed Kagiso Rabada. Inward angle, followed by the ball holding its line. A flat-footed poke, and Rishabh Pant falling to his left to grab the edge low down. A trademark Shami wicket to bring up a cherished landmark.

At stumps, the host broadcaster put up a graphic that showed Shami had pitched 61% of his deliveries in the "short" area. And it was accurate. Shami had used the shorter lengths extensively, both to test the pitch for possible vagaries of bounce and to push batters back to try and wrong-foot them against his fuller follow-ups. A legitimate strategy, and one that worked magnificently.

Therein, then, lies the danger of averaging out a bowler's lengths, and using that average length as an indicator of larger truths about that bowler. As much as there was a grain of truth to the criticism of Shami's lengths three years ago, that criticism also flattened out a lot of nuance.

You don't get 200 Test wickets by bowling the wrong lengths. You certainly don't get them at an average of 27.10 and a strike rate of 49.4. Among the 80 bowlers who've taken at least 200 Test wickets, only seven have better strike rates. Kagiso Rabada, Dale Steyn, Waqar Younis, Malcolm Marshall, Allan Donald, Mitchell Starc, Fred Trueman.

All those names, and Mohammed Shami. A champion bowler.

Karthik Krishnaswamy is a senior sub-editor at ESPNcricinfo