Introducing The F5

A refreshing take on the NBA

Welcome to The F5, a weekly newsletter about the NBA told through charts and data.

(If you’re wondering why you’re receiving this email it’s probably because you signed up for my Substack a couple of months ago when I first posted a tutorial about how to scrape Reddit comment data.)

---

Today, I want to talk about some early MVP trends.

But before I do that, I need to get some housekeeping out the way.

About this newsletter

The meat and potatoes of this newsletter is free. That means you will be able to read it or delete it without ever paying a single cent.

However, I am offering some perks to those who sign up for a paid subscription of $5 a month (or $50 a year). You can find all the details here, but this is the gist:

• At the end of every month I’m going to give away 50 percent of whatever I made from premium Substack subscriptions that month to a randomly selected premium subscriber.

  • That means if you are my only premium subscriber (hey Mom!), then you’re guaranteed to win $2.50.

  • If you refer a friend and they sign up for a premium subscription you will receive two entries into that month’s lottery.

  • The more people who subscribe, the bigger the winnings — so tell a friend!

• Premium subscribers will receive the R code I used to make the graphics in this newsletter.

  • I cannot promise the code will always be clean or easy to read, but I can promise to respond to any questions you have.

I am not expecting to make much money from all this. Honestly, I would consider this newsletter a success if I made enough to cover the cost of NBA League Pass. I’m just glad you’re here.

---

Back to the MVP

One of things I’ve noticed about the MVP race over the last few years is that the early season frontrunner has usually maintained their lead over the field from start to finish. Giannis Antetokounmpo and James Harden — the winners of the last three MVPs — coasted to mostly uncontested victories after hot starts to their seasons. So even though only a handful of games have been played it’s worth taking stock of the field of MVP candidates.

There are two places you can look on any given day to get a good sense of who the real MVP contenders are — sportsbooks and Basketball-Reference’s MVP Tracker. Each provides a different perspective and looking at them together can paint a reasonable landscape of the MVP race.

Because there’s a million different sportsbooks out there all with their own odds it’s helpful to look at the average of a few different ones. Fortunately, the Action Network does exactly that by offering a “consensus” look across multiple sportsbooks. Below is a chart showing the ten players with the best odds to win MVP as of this writing.

Each point on the chart represents the odds for the corresponding player to win the MVP from one of the sportsbooks tracked by the Action Network. The darkest point for each player represents the consensus across all sportsbooks tracked.

The consensus has Luka Doncic as the odds on favorite at +400. That means a bet of $100 would win you $400 plus the original $100 bet if Doncic were to win. Doncic as the favorite makes sense after finishing fourth in voting last year and leading the league’s most efficient offense. The case for him to win this year is straightforward. All he needs to do is:

• Maintain or build on his production from last season [❌]

• Pilot another top offense [❌]

• Hope that none of the other usual MVP suspects have a standout year [❌]

So far, Doncic hasn’t checked any of those boxes. He’s performing worse in almost all major statistical categories compared to last year and his Mavericks are a league average 14th in offensive efficiency, according to Cleaning The Glass.

I would expect Doncic and the Mavericks to turn things around by no later than when Doncic’s running mate, Kristaps Porzingis, returns from injury, but I believe the biggest obstacle between Doncic and the MVP is the third box which is entirely outside of his control. Several other players, including previous MVP winners, have come roaring out of the gate this season.

Still, Doncic’s odds to win MVP are nearly the same as they were at the start of the season. Based on what we know now, I am not sure they should be. One of the problems with looking just at sportsbooks is that they can be slow to react to what is happening in real-time. For something that responds more quickly to new information we need to look elsewhere.

---

Basketball-Reference.com hosts the only public, data-driven MVP tracker that updates daily. There, players are sorted by their probability to win the award based on a statistical model that takes into account a handful of variables. Although I don’t know the exact details of the model, I have a good sense of what it considers based on a blog post from 2009:

I looked at 54 years of voting results to determine what factors the voters deemed most important when selecting the MVP. In the end, four factors proved to be much more important than anything else: team wins, points per game, rebounds per game, and assists per game.

In other words, a few box score stats and team record can predict with a decent degree of accuracy the likely contenders for the MVP award.

It is important to note that Basketball-Reference’s tracker is not attempting to predict who will be the winner at the end of the season, but rather it’s predicting “if the season ended right now, these are the players that would likely be in contention for the award.” However, at some point in the season those two should converge.

Here’s who Basketball-Reference’s model believes are the top contenders for the award based on the first few weeks of the season:

Joel Embiid, who is not in the top ten for MVP odds (he’s +2,500), is the favorite according to the Basketball-Reference model after averaging more than 23 points a game to start the season and anchoring the league’s top-rated defense. What’s going on here?

Players on teams that have gotten off to good starts — like Joel Embiid and the Philadelphia 76ers — have their MVP chances inflated in the model due their team’s early season win percentage. Meanwhile, players on teams that have gotten off to lukewarm starts — like Luka Doncic and the Dallas Mavericks — are nowhere to be found.

So to correct for this issue, I recreated the MVP Tracker using a logistic regression based on MVP voting behavior since 1981, which was the first year the media voted on the award. Just like with Basketball-Reference’s model, I controlled for points, rebounds, and assists per game. The main difference between my model and Basketball-Reference’s is that I used a team’s projected win percentage (specifically, FiveThirtyEight’s projections) as opposed to their current win percentage. After all, the MVP is decided by the standings at the end of the season, not in January.

Here’s what I found:

With James Harden, Nikola Jokic, Giannis Antetokounmpo and LeBron James as the headliners this list passes the smell test since it is effectively a list of the best players from the best teams, which is usually a prerequisite for winning the award. Meanwhile, Joel Embiid drops from the MVP favorite in Basketball-Reference’s model to the player with the fifth best chances of winning in my model, in large part because his team is expected to win closer to 65 percent of their games instead of their current rate of 85 percent.

Because this model does not take into account narrative or voter fatigue I think it is too overconfident in some player’s chances of winning. For instance, I think James Harden’s true chances of winning the MVP award are closer to zero percent than 38 percent, especially if Houston complies with his preseason trade demand.

Nevertheless, it is worth pointing out how bullish this model is on Jokic, the player with the second best chances of winning the award, according to my model (he also has the second best chances in Basketball-Reference’s model). His box score stats are eye-popping for a Center and even though his team is 3-4, they’re expected win around 60 percent of their games by the end of the season. That is an important detail since only two players since 1981 have received the MVP award on a team that won fewer than 60 percent of their games — Moses Malone in 1982 (56.1 win percentage) and Russell Westbrook in 2017 (57.3 win percentage).

One player’s who MVP chances jumps out to me further down the table is Paul George. Before the season, George was listed as +10,000 to win the MVP by the consensus on the Action Network. I find that somewhat surprising considering George is only two years removed from finishing third in MVP voting. As of today, George is listed as +7,000 by the consensus and you could argue that those odds are still too high based on his production. If “Propaganda P” can make people forget about his disappointing postseason last year then look for hid odds to drop even further.

I could probably improve this model around the margins by taking into account some additional variables (i.e., whether the player has won the award before), but I think it’s fine to keep things simple given that most MVP voters aren’t considering much beyond box score stats when casting their ballot anyway.

I’ll continue to provide updates on how this model performs throughout the season and how it compares to betting markets and Basketball-Reference’s MVP Tracker.

---

The “haha, j/k …. unless?” Fake Trade Of The Week

Listen, I’m going to experiment with this newsletter and try out some reoccurring gimmicks and see if any of them stick. Inevitably, there will be stinkers. Please bear with me.

Here’s a four-team Harden trade that is highly unlikely, but makes all parties involved instantly more interesting:

Why Houston Does It

• Michael Porter Jr. has the makings of a bona fide offensive star. The list of players his size who can create and make their own shot is not long. Add in Minnesota’s first round pick that Golden State owns and Houston is set up nicely for now and for later.

Why Golden State Does It

• Steph Curry isn’t getting any younger and James Harden is the best player available. Defenses can’t double-team both of them.

Why Toronto Does It

• There’s a hard ceiling on a team when Pascal Siakam is the primary offensive weapon. Swapping Siakam for James Wiseman and Wiggins is a down grade, but the Raptors are still a playoff team (or at least a play-in team) as long as they keep Kyle Lowry. Also, who wouldn’t tune in to watch Nick Nurse teach Maple Jordan how to play defense?

Why Denver Does It

• Pascal Siakam in the Jermai Grant role is worth the juice.

---

Thank you for reading The F5. If you found it refreshing, please consider subscribing.