US Open Semi-Finals Recap

backhands down the line—control defeats power—rushing forehands

Djokovic def. Shelton 6/3 6/2 7/6

Before this match I tweeted my thoughts on this first semifinal:

“Hot and humid day session with the home crowd. Best possible conditions for Shelton if he’s to have a chance. Will need the first set to keep the crowd involved. Djokovic is too good of a front runner. I think Novak will try to keep Ben moving: he’s [Shelton] still a little heavy and ND is so good at changing directions and keeping players from feeling set on their shots. Shelton needs to find his forehand early and go hard with it. No point screwing around.”

This was another masterclass from the Serb, who dismantled Shelton’s power game with his ability to blunt the big serve and then turn the screws from the back. One of the talking points of the tournament has been Shelton’s serve power (twice he sent down 149mph thunderbolts this fortnight. Today he thundered down a 144mph second serve). Truly huge serving has always drawn an excited gasp from the crowd the same way long drivers do for golf, but as they say in that maddening game: “Drive for show, putt for dough”.

The same notion holds for tennis, although it isn’t exactly clear what tennis’ putting equivalent is. Perhaps it can simply be one’s ability to redirect groundstrokes at 70mph. This isn’t particularly sexy or exciting to casual observers, in fact it sounds downright benign, but much like compound interest, it pays off handsomely over time, and it didn’t take long for the cracks to appear in Shelton’s game. He was broken in the sixth game, committing four unforced errors that were all the result of pressing too hard and too early against elite defense. In contrast, Djokovic was using his innocuous rally pace ball, the below example off the backhand wing in that silent killer fashion: down the line, catching opponents leaning on the crosscourt anticipation. It’s rarely a winner, but nearly always earns a short ball or forced error. No one else extracts so many errors from rally-ball pace placed in aggressive areas. No one.

Often Novak hits this with enough precision so that it effectively becomes an inside-out backhand; more extreme than just straight down the line, the ball actually is fading away from the centre of the court. I touched on this in the Alcaraz/Djokovic Roland Garros semifinal, the Djokovic/Tsitsipas Australian Open final, and the Djokovic/Paul Australian Open semifinal.

So while the power was squarely on Shelton’s racquet, the ability to dictate the terms of the match largely rested on Djokovic’s all-absorbing strings:

Midway through the first set Djokovic had given nothing away, and for all of Shelton’s vicious rips, he only had a -6 ratio to show for it. Brutal.

There is still a lot of upside for Shelton. First of all (and I don’t have any map data on this yet) it seemed to me that Shelton’s serve accuracy isn’t great yet. He doesn’t quite stretch the court as well as he could—especially as a lefty. When he did, he looked great:

Marrying 149mph missiles with this kind of accurate and kicking wide serve is what will make Shelton a nightmare for returners. Look how wide Djokovic gets pushed to deal with the width and spin from a 104mph serve. Tonight he didn’t quite find the lines as closely as he needed to.

Marrying that huge power with sharp and accurate angled serves would make him that much harder to break.

Secondly, he lacks great intangibles—slices, block returns, volleys, drop shots—those gritty cat-and-mouse elements when you’re flying in the forecourt by dead reckoning and require slick feet and slicker hands. But Shelton is a great athlete, I like his technique, and we can expect him to add these elements and refine his physique as he gets accustomed to life on tour and grows with experience. It is his first year on tour, after all.

In the end, while Djokovic let a routine third set become a little more complicated, Shelton was just too raw to present as a genuine threat to the 23-time slam champion, and his ability to be consistent while moving was tested beyond his means. As predicted, Djokovic made his younger foe do most of the legwork, and came out with a more favourable winner:unforcer error ratio:

Djokovic: “The Grand Slams are the ones that motivate me the most to play my best tennis, perform my best tennis. I knew prior to the quarter-finals that I would play an American player and that is never easy. To control the nerves and be composed in the moments that matter. Today things were going really smoothly for me and then he broke back and it was anyones game at the end of the third set. This is the kind of atmosphere we all like to play in, so I am really, really pleased with this win today."

---

Medvedev def. Alcaraz 7/6 6/1 3/6 6/3

The opening games of this match featured themes we have covered before: Alcaraz’s forehand susceptibility and tendency to bleed errors out wide; and his ability to serve and volley on Medvedev’s deep return position. Tied at 3 games a piece in the first set, all six of Alcaraz’s unforced errors had been committed from the forehand wing, and he’d let slip several chances on Medvedev’s serve by mistiming a handful of second serve returns. At 4-4 in the first set, Alcaraz had ventured forward 15 times already and won all 15 points, with the vast majority of these volleys played short, even when faced with unfavourable positions, in a bid to get the Russian taking the longest route possible to the ball.

Daniil isn’t even on the screen when returning this one. Playing it short also limits Medvedev’s pass options; he can’t be as aggressive with his swing and has to hit up steeply on the ball.

Although their previous two encounters, at Indian Wells and Wimbledon, had been blowouts in favour of the Spaniard, Medvedev has overall been playing well in 2023, in part attributed to a string change that allowed him to get more spin on his forehand. In my US Open preview I wrote:

“I think Medvedev’s losses recently haven’t been bad; he’s run into form players in de Minaur and Zverev and he’s played some big points poorly. Overall he’s still a menace at the US Open…”

Tonight, he showcased what a hardcourt menace he is when he gels his brick-walled groundstrokes so relentlessly. He reminded most of us—me included—that you should never count out a US Open champion with a great record against world number one players.

Despite lacking firepower—a weakness that probably cost him that famous Australian Open final against Nadal—the Russian has exceptional racquet face control, courtesy of minimal wrist action and very little lag. His backhand down the line—a feature of Djokovic’s game also, with both men having takebacks that start from inside the line of the ball—was on fire this evening, catching Alcaraz off guard over and over again as the Spaniard looked for ad-court forehands:

Robbie Koenig: “That backhand down the line has been kryptonite for Alcaraz this evening.”

Medvedev gets a great takeback from well inside the line of the ball and uses what I call an “arm-led” backhand, where the hips are passengers that get pulled through passively by the arms.

Medvedev: “I said I needed to play eleven out of ten, and I played twelve out of ten.”

Ultimately, I felt Alcaraz was a little too aggressive and impatient for the first few sets here. His Wimbledon victory was so great because he played a contained brand of tennis for much of the encounter. When he finds the balance between patience and tornado offense he is a different player. Yet, at just 20 years old it’s natural that such harmony is yet to be achieved consistently, especially given how exceptionally broad his repertoire already is. Still, this match should serve as a reminder that you can expect to be punished if you lack discipline this deep in slams. What turned the match around in the third was in part a willingness to take some heat off and build the points. The average rally length jumped from around 3.5 shots to 4.5 shots during that momentum swing.

Koenig: “He went into Medvedev beastmode. Tactical masterclass today, the way he used his backhand down the line and forehand crosscourt to breakdown Alcaraz’s forehand.”

“He [Alcaraz] couldn’t bring his A-game and he wasn’t allowed to bring his A-game.”

Federer 2006—perhaps the greatest forehand of all time—and Alcaraz at Wimbledon 2023. You can see how much higher Alcaraz’s elbow is, and consequently, how much more “inverted” his racquet head is, with the tip of the racquet facing his opponent. Federer has a lower elbow and a racquet head with less to “undo” into contact. It must be noted that the swingweight and weight of your frame also dictate what is possible/preferable. Alcaraz is reportedly using quite a light stick, so such inversion is not really an issue (Dan Evans is another example), in fact, it may be what he naturally needs to do in order to create enough power with a lighter frame. Still, I think this highly inverted takeback with a light racquet that employs a very fast swing is liable to mistime the ball, especially with Alcaraz’s ultra-aggressive mindset.

The rap sheet shows just how tidy Daniil was, and how erratic Alcaraz was:

In a repeat of the 2021 US Open Final, Medvedev will take on Djokovic.

Medvedev: "The challenge is to play a guy who won 23 Grand Slams and I have only one. When I beat him here, I managed to play better than myself and I need to do it again. There is no other way."

Always cagey, what has often separated these two are two key areas:

1. First serve percentage

2. Forecourt ability: slice/drop shots/volleys etc.

Because both are natural counterpunchers, the rallies tend to extend when the free points on serve dry up. As both defend so well with their stock groundstroke games, their ability to execute short slices, drop shots, and sneak volleys has often played a crucial role in unsettling the other. Djokovic has come out on top recently in such matches, winning four of the last five. However, with Medvedev’s newfound confidence in his forehand and improved ability to dictate with that wing, we might see the Russian surprise us once again.

Let’s hope both men bring their best.