Triple 20: Examining The Impact of the NBA’s Big Three Concept

Ever since Kevin Garnett, Ray Allen, and Paul Pierce joined forces ten years ago to win the 2008 title for the Boston Celtics, the maxim in the NBA has become that you need a Big Three (or, in Boston’s case, The Big Three Featuring Rajon Rondo) in order to have a chance of winning a title.

While it took awhile for this to truly get off the ground, as the Lakers and Mavericks, who won the next three titles, were very clearly driven by a single superstar/decent cast paradigm with Kobe Bryant and Dirk Nowitzki, the idea truly flowered in 2012, when LeBron James, Dwyane Wade, and Chris Bosh got the ball rolling in Miami.

Later, whether it was Kawhi Leonard taking the mantle from Tim Duncan as the head of a new triumvirate in San Antonio while Duncan, Manu Ginobili, and Tony Parker won one last title together in 2014, the Stephen Curry, Klay Thompson, and Draymond Green machine in 2015, LeBron again with Kevin Love and Kyrie Irving in 2016, or Golden State trying out a Big Four with the addition of Kevin Durant last year, the rule has been abundantly clear:

No team wins with one or even two star players anymore.

Which is great, but how do you quantify it? How does a Big Three roll from game to game rather than across an entire season?

Last week, we showed that bad teams who don’t have constellations of stars on their roster can improve their chances of winning with a seven-man-or-more “Total Team Effort”, getting guys into double figures. The average team that gets one of those has a winning percentage in the “good luck winning even 30 games” range, but they win nearly half the time when they spread the ball.

This is the Newton’s Third Law version of that. Namely…

If three (or more) of your players each score 20 points or more, does it improve your chances of winning, and if so, by how much?

We’ll call this a “Triple 20” for purposes of giving me something to call it in the Breakfast Special.

Let’s crack open the case of numbers delivered to us by Basketball Reference and have at it. This is through games of November 11 and we’ll come back to this as the season goes on. I’m working on putting a catalog of these games (and Total Team Efforts, and anything else my fevered little mind can dream up for counting stats) together to make this easy, but in the meantime, it’s just poring over box scores until Basketball Reference makes them searchable by “x players with y stats” criteria.

The complete list, you can find here; it’s too dry to include in its entirety here.

There have been 35 total games so far this season in which three players on one team have scored at least 20 points. In two of those cases (Oct. 17, when Houston beat Golden State, and Nov. 7, when Cleveland beat Milwaukee), both teams have crossed the threshold.

Assigning a win and a loss (since indeed that was the outcome for the two teams) to those “ties” in the raw data, that yields a 27-10 record. Over an 82-game season, that’s 60-22, give or take a win. Clearly this is a powerful edge, and teams that can do it consistently are going to be teams that win ballgames.

Well…hold on for a second.

After all, Golden State’s done it more times this year than any other team; they appear on the raw list a whopping seven times. But in those seven games, they’re 5-2, going 5-1 in their other six games this year. What gives?

Well, for one thing, half a game doesn’t tell us much, but what it does tell us is that a team with a .769 winning percentage wins 71.4 percent of the time when they get three guys over 20.

But, if you take away the season opener (when Nick Young was the third 20-point scorer) and restrict it to just the Curry/Durant/Thompson axis of three-pointers that terrorizes the league in nearly half of the Warriors’ games, that percentage rises to .833. That’s the difference between 63-19 and 68-14. Five wins over a season when the margins between “No. 1 overall seed” and “greatest of all time” are thinner than they look.

But what about Boston? They’re 11-2 and only appear on the list once, and with two rookies, Jaylen Brown and Jayson Tatum, joining Kyrie Irving. With Gordon Hayward injured, the chance of an Irving/Hayward/Al Horford triple-20 drop to literally zero, yet there they are…but even then, they won that game, so it’s .833 (10-2) vs. 1.000 (1-0.) Small sample, but part of the overall total.

And Oklahoma City? Russell Westbrook, Carmelo Anthony, and Paul George were supposed to set the world on fire, but instead, those guys have two Triple 20s, and they actually lost a game in which all three of those guys plus Steven Adams all scored 20 plus; Minnesota beat them 119-116.

Speaking of Minnesota, how many games have Jimmy Butler, Karl-Anthony Towns, and Andrew Wiggins combined to produce a Triple 20? Zero. Nil. Nada. Their lone Triple 20 was Towns, Wiggins, and Jeff Teague.

Let’s continue to break down the patterns here. How many of those 35 games featured a true “Big Three” in the sense that fans understand it’s a team built around three stars sharing the ball and the glory?

Well, there’s six for Golden State (the Curry/Thompson/Durant games.) Two for Oklahoma City, including the Adams Game. Two for Portland (Damian Lillard, C.J. McCollum, and Jusuf Nurkic.) Two for Washington (John Wall, Bradley Beal, and Otto Porter.) Arguably one for Orlando if you believe that Evan Fournier, Nikola Vucevic, and Aaron Gordon qualify as a Big Three on a team that is, after all, 8-5 and the current fourth-best team in the East.

Mind you, in those 13 games, the team with the Big Three went 11-2. This really gets to the heart of the value of a true Big Three. That’s a 69-13 pace, which just so happens to be Golden State’s average regular-season record in their three Finals years.

Which also strips the data down to another key point. Most of the rest of those Triple 20s are teams with two stars and one guy who happened to step up that night.

Taking the data apart and examining cases where exactly two of the three guys with 20 were genuine “any casual fan recognizes them” stars and keys to their team, now you’ve got a more interesting data set, because it includes teams like Houston, where James Harden and Eric Gordon are the scorers and they’re 4-0 in Triple 20 games. Cleveland got one of these out of LeBron James and Kevin Love; Indiana got two out of Victor Oladipo and Darren Collison. Charlotte’s got a 2-1 record out of Kemba Walker and Dwight Howard.

Condensing the data to include only teams with two stars (so that you don’t get a three-star team that failed by its own standards, like the Swaggy P Game from the Warriors or even Orlando’s game where the third man after Fournier and Vucevic was Jonathon Simmons off the bench), and defining “two stars” as guys who tend to show up together with a different third man like Harden/Gordon or even Phoenix’s Devin Booker and T.J. Warren combination, well, it goes a little something like this.

Best reckoning on this is 11-2 where the star pairing is clear either from statistical repetition (Indiana, Phoenix) or popular consensus (Charlotte, Houston, New Orleans, Memphis). Houston is 4-0, Charlotte is 2-1, the Pacers are 2-0, the Suns are 1-1, and the Pelicans and Grizzlies 1-0 in these circumstances.

So the point there is that if you’ve got two primary scorers in an offense who can reliably go for 20 and either a third star who’s less of a scorer (guys like DeAndre Jordan, Rudy Gobert, Al Horford, and the ur-example Chris Bosh) or a bench guy who can step up and be Big Three For A Day, you’re going to win a lot of ballgames and contend for a title.

Which leaves the wild card; teams that are 5-6 under all other circumstances. That’s a Golden State star failure in a loss and one for Orlando in a win accounted for, along with a Minnesota win that involved Towns and Wiggins but not Butler.

The wacky teams in the remaining scenarios? Teams like the Lakers, Celtics, Heat, Sixers, Cavs (twice, and both times involving LeBron but with four unique players in the other slots), Bucks, and Spurs. Total combined record in their games: 3-5.

Go down the list, and except for the utterly weird Lakers Triple 20 that involved Kyle Kuzma, Jordan Clarkson, and Kentavious Caldwell-Pope, what do you see?

That’s right. Guys like LeBron. Kyrie. LaMarcus Aldridge and Giannis Antetokounmpo and Joel Embiid. Guys who are the ultimate focal point because of the way their offense works, guys who score points in bunches with fluid contributions from their teammates. (or, in LMA and Kyrie’s case, because guys like Kawhi Leonard and Gordon Hayward are hurt, otherwise they’d be in the second tier.)

So do you need a Big Three of scoring threats from any point on the basketball court to win ballgames?

Well, it’s a nice-to-have, and Warriors fans certainly love watching it.

But if you have two competent scorers and a third star who does other things like grab a ton of rebounds or dish a bunch of assists (you’ll notice that Toronto doesn’t appear on the list because Kyle Lowry and DeMar DeRozan handle the scoring), that’s just as good as having three guys who go for 20 a game.

Otherwise, you’re just a team having one weird night in the box score and more often than not losing the contest on the court.