Stephen Curry versus Kyrie Irving. Who is better? Who is more valuable to his team?

This is the eternal debate around every water cooler in America, and while it’s mostly just fun, it also has underappreciated implications within the betting and DFS markets. Take the Curry/Kyrie example: The former is likely the better player in a vacuum, but if he misses a game the Warriors still have Kevin DurantKlay Thompson, and Draymond Green to fill in the gaps. If Kyrie misses time, inferior players like Marcus Smart and Terry Rozier are forced into more action.

Kyrie missed his first game of the year last week against the Philadelphia 76ers. It’s unclear what the betting market thinks he’s worth to the spread: It opened at Celtics -4 and closed at Celtics -3. The total moved in a big way, closing at 203 after opening at 211. The line was posted at 10 am ET, and Kyrie was ruled out at 5 pm ET. It was a bit surprising he actually sat, but it’s possible the line was shaded toward the 76ers at open given his questionable status. Sports Insights reported that he became a Tier 2 player after Gordon Hayward‘s injury, which would put his worth at about four points to the spread.

That said, creating a rigid pre-season tier system doesn’t always account for team dynamics or current injuries. Is it possible that Kyrie is actually worth more than four points specifically to the Celtics because 1) he is one of the best shot creators in the league and 2) the Celtics are sorely lacking in that skill in the rest of their roster? It was only a single game, but the Celtics’ numbers in that game versus Philly suggests maybe Kyrie is in fact undervalued:

New Metrics to Measure Offensive Creation

The problem is how to measure this dynamic. You could look at on/off court numbers, but those are skewed by role. Sure, a backup PG could play really well and generate efficient offense while he’s on the floor versus other backups, but that doesn’t mean he’ll be the same player if inserted in the starting lineup. You could look at assist rates, but that likely overrates PGs who generate assists simply by passing to a teammate coming off a screen and underrates the high-level creators who devastate defenses with penetration, gravity, and elite vision.

Thankfully, our friends over at Nylon Calculus — a hat tip specifically to Ben Taylor and Kevin Ferrigan — have created two metrics that help measure actual shot creation within an offense. These metrics, called “Box Creation” and “Offensive Load,” more accurately paint a picture of how valuable a player is to his offense based on shot creation (follow the links for a more robust description and explanation of how the statistics are calculated). Below is a chart that shows the NBA in terms of both metrics versus VORP. You can break the graph into quadrants: The top right is the elite superstars; they are valuable and create offense. Bottom right is elite role-players; they have a high VORP but don’t create shots. Top left is shot creators who aren’t particularly efficient or valuable, and the bottom left is players who have average or mediocre value and don’t create offense. As you might expect, the majority of the NBA is in that quadrant.

Let’s now look at team-by-team basis.

Notable Team Situations

Take the Boston Celtics again. Kyrie leads the team with a 10.3 Box Creation mark and 48.2 Offensive Load mark. That ranks 10th in the league, although that again undersells his value to the team. The Celtic with the second-highest Box Creation mark is Al Horford (6.3), who obviously isn’t going to fill in for Kyrie when it comes to perimeter offensive creation. The guys to do that are Marcus Smart (4.1) and Terry Rozier (2.9), who rate poorly in this aspect of the game. When Kyrie is out, his absence is felt in a big way, as without Hayward he’s the only efficient shot creator on the team. Also, this might mean we should slow down on the Jayson-Tatum-is-a-superstar-wing narrative. Both his college data and his creation metrics in his rookie season suggests he’s a potentially elite 3-and-D role player (maybe a better Otto Porter who can more effectively attack close-outs) rather than a No. 1 option.

Houston, meanwhile, has both James Harden (18.1 Box Creation) and Chris Paul (13.3). They both rank in the top-five, which means that the Rockets have plenty of shot and offensive creation even if one them is out of the lineup, as has been the case often this year. Does that mean that they won’t miss a beat when one is out? No, but it does mean they’re perhaps in a better spot to weather the storm than Boston without Kyrie. Even their third-best creator — Eric Gordon with a 4.5 Box Creation mark — can help. He has a higher creation mark than any guard on the Celtics.

This final chart shows the Box Creation and Offensive Load marks for each team’s top-three players:

Russell Westbrook was bumped down from a Tier 1 player (worth five points to the spread) to a Tier 2 player this season, mostly because the addition of stars Paul George and Carmelo Anthony. However, it seems especially Melo is no longer the shot and offensive creator he was in his prime: Westbrook has a massive 15.0 Box Creation mark, and George’s 5.1 mark, while second on the team, is lower than many other team’s secondary creators. Brooklyn’s Caris LeVert has a higher mark of 5.8, for example. That isn’t to say PG is a poor player; it just suggests that perhaps Westbrook is still Tier 1-level important to the Thunder’s offense. In terms of creation, he’s still the sun of their offense.

Each team dynamic is quite interesting. For example, the Raptors are a lesser version of the Rockets’ creators: Both Kyle Lowry and DeMar DeRozan have Box Creation marks of at least 8.0. As long as one is on the floor, they’ll likely be able to create decent offense. The Clippers have had a million injuries this year and for a while were missing five of their top-seven players, and yet they were still able to stay afloat thanks to the offensive creation of Lou Williams. The Spurs are perhaps the most interesting team of all: Pau Gasol and Manu Ginobili lead the team in Box Creation. Maybe San Antonio is just an anomaly — they’ve been frustrating bettors for years with how little any one player has seemingly meant to a single-game spread — and they don’t need the ball in any one player’s hands. Perhaps a well-run offense can deal with that crutch.

Conclusion

We’re in the golden age of NBA analytics, but it’s still early. As we gain more knowledge of how certain players affect offenses, we’ll get better at predicting the sport. The NBA goes in cycles: We really valued how much a player like Kobe Bryant or Michael Jordan could “get a bucket,” but now we’ve moved to valuing uber-efficient basketball that emphasizes ball movement and corner 3s. It seems that perhaps we’ve moved too far and have now devalued primary offensive creators. Those guys are still the most valuable players in the league, and it’s possible — especially with situations like the Celtics, where they are lacking secondary creators — that those guys are somewhat undervalued in the betting and DFS markets currently.

And finally, here’s a quick cheat sheet of offenses that will be most affected when their primary creators miss time. These are situations in which the team’s top player either has an outsized creating role or the team lacks a great secondary offensive creator. Some are obvious, and some might be surprising, like with the Damian LillardC.J. McCollum dynamic. That, in particular, deserves more study and suggests that perhaps McCollum is better suited for a secondary playmaking role and is better at getting his own shot rather than creating for an offense. Until next time, your cheat sheet:

  • Atlanta Hawks (Dennis Schroder)
  • Boston Celtics (Kyrie Irving)
  • Charlotte Hornets (Kemba Walker)
  • Cleveland Cavaliers (LeBron James)
  • Detroit Pistons (Reggie Jackson)
  • Memphis Grizzlies (Tyreke Evans)
  • OKC Thunder (Russell Westbrook)
  • Phoenix Suns (Devin Booker)
  • Portland Trail Blazers (Damian Lillard)
  • Washington Wizards (John Wall)

Photo via Brad Penner-USA TODAY Sports