With the season starting in just under a month, and media day for most teams only a few days from now, it is time to unveil the final version of my 2017-18 projected NBA win totals for each team — unless, of course, a team makes a major move or a coach says “we are going to start X player instead of Y.”
The first step in projecting wins for the 2017-18 season is creating a baseline expectation for how each player will perform. I have a database of every season played going back to 1951-52, the first in which minutes were tracked. Using this database, I can compute a similarity score to current players. After finding the most similar players to everyone who played last year, I use how those similar players changed to project how the current players will perform in 2017-18.
Each of these player projections take place in a vacuum, without context of how other teammates change. After I have a projection for every player who I expect to play in 2017-18, I throw them all into one shared universe. In this digital world, every missed shot must result in a rebound, every steal must result in a turnover. Certain interactions, like increased spacing from 3-point shooting and playing with high usage players, lead to adjustments in an individual’s expected production.
After making all of these adjustments, I add every player on a team’s contributions together to get a win projection. Each of these win projections are in a vacuum and do not actually add up to 1230 wins — the total number in an NBA season — so these totals are further adjusted accordingly.
Even at this point, the projections are not quite done. The last step is taking this digital world and putting it in to the real NBA schedule. And with that last adjustment, I arrive at my NBA win totals.
The process is obviously a bit more complicated than what I have described, but the nitty-gritty details are not exactly exciting to read about.
2017-18 NBA Projected Win Totals
My model is higher on the Nuggets and Clippers than I am personally, but none of the win totals really jump out and shock me.
The West seems to have clear tiers to me:
- The Warriors are clearly the best team.
- The Rockets were really good last year and added a top-15, at least, player in the league.
- The Spurs and Thunder are going to be a really fun match up when they’re fighting for the No. 3 and No. 4 seed.
- The Timberwolves, Nuggets and Clippers are all solid playoff teams with maybe a little upside to get home court in the playoffs if a team above them falters.
- The race for the No. 8 seed has the Trailblazers, Pelicans, Jazz and Grizzlies in what is sure to be a down-to-the-wire finish.
- The other bottom feeders are teams in various stages of rebuild. I would be surprised to see any in the playoffs this season.
The Eastern Conference does not quite break down as cleanly as the Western Conference. With no team projected for 50 wins, I would put the tiers as follows:
- Cavaliers. Until LeBron James does not play for them (so next year), they should be considered the best team in the East.
- Raptors, Celtics and Bucks(!) all project to be around 48-49 wins and make for a really exciting group that can challenge the Cavaliers.
- The Wizards and the Hornets are built very differently but project to be similar in actual wins totals. The Hornets wildly under-performed their expected wins last season, and should rebound nicely. The Wizards are being pegged as a lock for home court by some fans, and the evidence simply does not back that up.
- The race for the bottom of the East playoffs includes: The Pistons, Heat, 76ers, Knicks(!!) and Magic(!!!!!!!). I’m not sure I would actually consider the Knicks or Magic as actual playoff-level teams, but I will be sure to change that tone if I turn out to be right at the end of the year.
- The bottom of the East is a whole lot of blah: The Nets, Pacers, Bulls and Hawks. One of these teams will win less than 20 games, one will win more than 30. Do not expect any of them to jump in to the playoff hunt (unless Myles Turner is even more amazing than I think he is).