Stephen Curry does winning things

The more we learn about what works, the more he matters

BY HENRY ABBOTT

This week on TrueHoop we dive into the debate about the greatest basketball players of all time. In some ways, it’s a pointless exercise. At an All-Star game in the early 2000s I asked NBA All-Stars like Shaquille O’Neal and legends like John Thompson to pick their five. By the time I got to Grant Hill, he just laughed, refused to play along, and said “that’s just barbershop talk.” 

Grant is wise. These lists are not very important. Not inside basketball. Maybe ten people deserve to be in the top five. Maybe a hundred do. That’s the trick. Every single way you do it, it’s an outrage to leave out the blood, sweat, tears, and brilliance of some of the best players ever to play the game we love. 

These lists matter a little more outside basketball, though. What are the names people who don’t follow basketball need to know? (I don’t know much about Broadway, for instance. But my daughter makes sure I know a few names like Philippa Soo and Ben Platt.) 

My five, unveiled on BRING IT IN on Monday:

  1. LeBron James

  2. Michael Jordan

  3. Stephen Curry

  4. Wilt Chamberlain

  5. Tim Duncan

Then, naturally:


My safest decision: Michael Jordan. A no-brainer, and not just because I grew up watching the majesty of his six championships. Nowadays, there are plenty of good assessments about which season was Jordan’s finest. 1990-1991 is a popular pick. At 27, Jordan led the Bulls to the first title of their first three-peat. They won 61 regular season games, rolled to a title through Magic Johnson and the Lakers with only two losses all playoffs long. Jordan was, by every measure, the best offensive player in the world, and by the crappy defensive stats of the day you could argue he was the most effective defender too. In the Finals, the Bulls were better at a rate of 125-102 when Jordan was on the floor. 

That year, our eyes were feasting on quite possibly the greatest basketball season in global human history. It’s the kind of stuff people point to in documentaries like “The Last Dance” and all kinds of books, podcasts, and magazine stories. This is when the player made the myth. 

But it’s a myth that people knew all that in real time. The idea it was the best ever mostly came later. I just toured the newspapers.com archive from 1991, when those wins were fresh, and couldn’t find anyone saying that Jordan was the best ever. Instead, there’s stuff like Pittsburgh coaching legend Hank Kuzma telling the Post-Gazette “If you’re talking today you’d have to say Michael Jordan. All time, I’d have to say Wilt Chamberlain.” When Jordan retired the first time in 1994, most people didn’t say MJ was better than Wilt, Magic, or Kareem. 

Maybe no one has ever walked off the court hearing that was the best performance ever. It’s easier to place a performance in history when it’s history. Which is wrong. Sometimes it is the best performance ever.

What a bummer of curmudgeons. At the very time Jordan was outplaying Russell, Chamberlain, and Kareem, sportswriters—in their wisdom—watched those games, talked to Jordan in the locker room, and wrote stories about how he might improve, what he was missing, or what was undone. (In some cases, those same writers are saying, now, that Jordan was the best ever.) 

This was eye-opening to me as I lived it, reading those stories all through Jordan’s career, and writing about players like LeBron and Steph. Let me not make the same mistake. I don’t need decades to pass before I celebrate the historical weight of a thing. Sports, like everything, evolve and improve. Now we have people shooting from the logo with the game on the line, seven-footers crossing people over and shooting game-winning 3s. World records fall just about every decade in just about every sport. It’s not only possible that today’s players are the best ever—it’s likely. I’m open to it. 


At the beginning of this season some whip-smart machine learning engineers—Harrison Chase and Tony Liu—did a rigorous examination of sophisticated measures of NBA performance. One interesting takeaway that aligns with the work of just about everyone who has dug into it deeply: Players peak way younger than we thought. It’s hard to find examples of NBA players who improve after the age of 25. 

The most famous and successful players might be 30 or older. But in almost every case, they stopped opponents better, scored more efficiently, ran higher, jumped faster, had better plus/minus stats when they were 25. LeBron James and Jimmy Butler might dominate our TV screens and the ball with the Finals on the line, but both were probably better years ago, before we appreciated them as much, and let them dominate the ball, the Finals, and our psyches.

These are also biases in coaching and commentary from coaches. I think often of this Grantland insight from Zach Lowe in 2013, back when Jonas Valanciunas was a rookie in Toronto:

Valanciunas, like most rookies, misses rotations, overhelps, and commits other sins of positioning on defense. Coaches hate that stuff, and they’ve often nailed Valanciunas to the bench in crunch time in favor of Aaron Gray — a fundamentally sound player who lacks NBA athleticism.

The numbers in large part disagree with that tactic, at least as it relates to Valanciunas’s defense. The Raptors’ defense has been better with Valanciunas on the floor.

Re-capping: Valanciunas is better than Gray (so much more effective that later Lowe quotes a team official saying “who cares?” if Valanciunas makes mistakes) but is benched because Gray is … better? This is deep, psychologically. One issue might be that the thing Valanciunas is good at does not require or involve coaching. It’s antithetical to it. Gray follows the program. His success is an endorsement of the coaching staff. How many coaches would say, “he doesn’t follow our instructions, but we just keep playing him because he’s great?”

If you put Damian Lillard in a time machine, and dropped him into an NBA game from the 1980s he could lead a dynasty … if the coaches could resist benching him for his first logo jumper and doing everything “wrong.” He knows stuff they do not. They might not love him for it.

Over my career, which started before basketball analytics, there have been a few common themes: Coaches overrate the value of timeouts, playcalls, and pep talks. Old-school players overrate hand checking and flagrant fouls. Everyone overrates whoever they loved when they were ten years old.

Which is all human and natural. But it results in a world that roughs up game-changing pioneers. Nobody is well positioned, ego-wise, to say that guy is fucking incredible in ways I barely understand

But the numbers don’t have egos or loyalties—and every once in a while, that’s exactly what they say. For a while, people said Jordan was a showboat or a ball hog, but the numbers on the scoreboard started showing the Bulls were the best. 

Allow me direct your attention to Real Plus-Minus (RPM), a stat that begins with the final score, then removes the parts of the game when a player is out of the game. Then the numbers are tweaked a bit more to account for the other nine players on the floor (did you do well against LeBron?), and again to factor in points, rebounds, and your other box score statistics. Although no one says this stat is perfect, it’s excellent. The inputs include everything likely to matter, and it makes a good guess weighting the different components. Daryl Morey listed this stat first in assessing the meaningful things to come out of the MIT Sloan Sports Analytics conference. 

Because it factors in all these things that go into basketball, it’s the kind of stat that favors an all-rounder like LeBron. And indeed, LeBron does very well in Real Plus-Minus—he’s almost always top five. 

Plus/minus factors in defense, size, and toughness. So surely, this would be the stat to expose the weaknesses of putting small, shooting specialist Stephen Curry high on your all-time list, right? 


Stephen Curry has the best Real Plus-Minus scores of all time. The stat only goes back to 2013, but since then, a period of peak-LeBron, Steph’s numbers are incredible. Over six years, Steph was first in the whole league four times, often by a lot. The other two years, he was second out of the whole league, his worst rank on record. In 2014-2015 he had the best-ever score of +10.92, then in 2015-2016, his score was even better, at +11.37. 

This is the numbers saying: That guy is fucking incredible in ways we barely understand

Steph might be scrawny and he might have grown up with privilege and he might not have a lot of the things we once assumed would win games. But boy does he win some fucking games. 

LeBron’s best ever season was +8.63. 

Was Steph riding Kevin Durant’s coattails? Durant before, during, and after playing with Steph will help answer that. But I’ll say this: Durant has been excellent, at times as high as third in the league, but has never had anything close to Steph’s effect on lineups. His last year in Oklahoma City his RPM was his highest on record, at 6.49. That year Steph was 11.37. When they played together for three years, both of their numbers slipped—Steph’s to numbers between six and eight, Durant’s to numbers between two and five. 

That doesn’t tell me everything I need to know, but it does account for every second he was on the court, how his team did, and who else was on the court trying (and failing) to stop him. It does credit rebounds and punish turnovers. It tells me that there was a race to win the game, and this dude won it. If we dig way deeper, zero in, do video analysis, I already know we will not find long stretches where he hurt his team. If we did, the numbers wouldn’t be like this. He is OK on defense, which is not true of every scorer, and he melts the offensive record book into a gooey puddle. 


Stephen Curry is 32 years old. There’s close to zero chance his best basketball is yet to come. I would suggest that he will either partner with a superstar and win more titles, or not and not. (This puts him very close to the LeBron schedule; we saw LeBron without a superstar teammate last season.) In other words, Steph’s like MJ in the mid-1990s. His most brilliant work is behind him, and either is or is not enough to earn him a place in the pantheon. We don’t have to wait to know his key accomplishments: 

  • Set the all-time record by winning 73 out of 82 games.

  • Won three titles in four years during peak LeBron. 

  • Led a team that outscored playoff opponents by 245 points over 602 minutes when he was on the floor in 2017.

  • Changed the meaning of the words “point guard” forever.

  • Made it not just OK—but required—to shoot 3s off the dribble.

  • Set the record for the best Real Plus-Minus ever.

  • Made a generation of young players love shooting.

There will be a lot of “but he doesn’t …” responses. And maybe they’re brilliant; I’d like to hear them.

But I’d just add—if he’s so bad, how come he’s so good?

Fun exercise: As Jarod pointed out, my top five are one player from every position. Just imagine if, somehow, the lineup of Stephen Curry, Michael Jordan, LeBron James, Tim Duncan, and Wilt Chamberlain could suit up together and play a game. First, I don’t care if you have five Teen Wolves, this team will win. Picture it. The final score, the highlights, the email newsletters the next morning. Be honest. Even in your imagination Steph’s probably the high scorer, right? Out of this lineup? Are you sure he doesn’t belong?

The memo of Steph’s career: Stay curious! We are still learning about this game. It might be that what works isn’t exactly that we thought it was.


The GOAT debate is going on all week on BRING IT IN.