NBA Preseason Rankings & Season Projections For 2014-15

With the NBA season kicking off Tuesday, here is our annual post outlining our NBA preseason ratings, with season projections added for good measure, for every NBA team in 2014-15. Below the data table is more information on how we rank and project teams.

Preseason NBA Rankings Highlights

When we make comparisons to “Vegas” below, we are comparing where a team ranks in our NBA championship probabilities with where they rank in the futures odds listed at Vegas Insider. Note that we are NOT comparing our probabilities with the payout odds, in order to find value. We are comparing the rankings, in order to illuminate which teams are ranked higher or lower than expected.

  • Agreement with Vegas regarding top teams. There is close agreement between us and Vegas on the five teams most likely to win the NBA title. We both project Cleveland as the most likely, followed by San Antonio, then Oklahoma City. We then have the Clippers 4th and the Bulls 5th, while Vegas has those two teams reversed.

  • San Antonio and Cleveland cut above the rest. From a ratings perspective, last year’s NBA Finals teams (and by that we mean the Spurs and Whoever LeBron Is Playing For) also project as the clear top two teams in 2014, and are only separated by one-third of a point. There’s a bit of a drop to Oklahoma City at #3 (aggravated by Kevin Durant’s injury woes) and the Clippers at #4, then an even larger drop to the Bulls and Warriors.
  • Our three most underrated teams. While it can be misleading to compare teams only by rankings, compared to Vegas, we’re most optimistic on the following three teams: Raptors, Suns, and Hawks.
  • Our three most overrated teams. On the flip side, compared to Vegas we’re the most pessimistic on the following three teams: Mavericks, Celtics, and Lakers.

Glossary

  • Team ratings (“TR Rating”) are expressed as points better (a positive rating) or worse (a negative rating) than the average team in the NBA, when playing on a neutral court
  • “Playoffs” is odds to make the playoffs; “Division” is odds to win the division; “1 Seed” is odds to be the #1 seed in the conference; “NBA Champs” is odds to win the 2014 NBA title

 2014 NBA Preseason Rankings & Projections

TR RankAtlanticTR RatingWinsLossesPlayoffsDivision1 SeedNBA Champs
1San Antonio6.357.324.793%53%27%16.1%
2Cleveland6.057.624.497%62%40%17.8%
3Okla City5.355.126.990%63%19%11.1%
4LA Clippers5.154.327.788%53%18%10.4%
5Chicago3.651.530.589%30%16%6.9%
6Golden State3.550.131.979%33%9%5.3%
7Houston3.048.533.574%17%7%4.1%
8Toronto2.849.332.785%56%12%4.6%
9Portland2.747.734.371%26%6%3.1%
10Dallas2.647.234.869%13%5%3.1%
11Washington2.347.834.280%35%9%3.8%
12Memphis2.246.235.865%13%5%2.7%
13Miami1.044.637.472%25%6%2.2%
14Charlotte0.944.137.970%23%5%2.0%
15Phoenix0.842.239.852%12%2%1.3%
16Atlanta0.041.840.261%16%3%1.2%
17Denver-0.339.642.442%10%1%0.7%
18New Orleans-0.339.442.641%4%1%0.7%
19New York-0.441.140.959%22%3%0.9%
20Brooklyn-0.639.942.156%19%2%0.9%
21Indiana-1.437.744.347%4%2%0.6%
22Detroit-1.836.645.442%3%1%0.4%
23Sacramento-3.729.852.213%1%0%0.1%
24Minnesota-4.328.453.610%1%0%0.0%
25Orlando-4.728.553.517%2%0%0.1%
26LA Lakers-4.727.454.68%1%0%0.0%
27Utah-5.225.956.16%1%0%0.0%
28Boston-5.426.955.113%3%0%0.0%
29Milwaukee-5.825.956.111%1%0%0.0%
30Philadelphia-9.417.564.52%0%0%0.0%

How We Rank & Project NBA Teams

Our preseason process has two steps:

  • Generate preseason team ratings. First we gather a bunch of data on things like historical team performance, offseason roster changes, and coaching changes. We feed this data into a model we built to generate our preseason team ratings and rankings. Every year, we re-calibrate this model based on having a new season’s worth of data.
  • Run simulations of the 2014-15 season. Second, we feed our preseason ratings into our season simulator (internally known as the “Awww-simulator” because of its awesome computational power), which plays out the 2014-15 season thousands of times.

Thanks to randomness, every individual season simulation ends up looking differently. But once we run lots of simulations, the outcomes — such as the projected final win-loss record for each team — start to converge on the numbers that end up being our official 20142015 preseason projections. The simulator is sophisticated, accounting for the fact that our preseason rating for any given team may end up being wrong, and for the randomness and variance inherent in any given game.

Unlike the vast majority of pre-season rankers out there, we do very little subjective evaluation of teams. This approach generally serves us well. Inevitably some years turn out better than others (occasionally for factors out of our control, like mid-season injuries to key players), but we beat Vegas last season and haven’t changed the method since then.

Here are our previous preseason projections:

Three Things To Keep In Mind

Before you make a comment about where we’ve ranked your favorite team and call us a bunch of no good bleepety-bleeps, please keep a couple things in mind:

  • We’re using a systematic approach to rank all 30 teams. Because our approach generalizes predictive factors, it’s going to get plenty of individual teams wrong, and a few of them very wrong, for many different reasons. The goal here is overall accuracy across the entire system of 30 teams, and to be right more than we are wrong when a projection of ours deviates the most from consensus opinion. If we wanted to maximize our odds of making the most accurate projection for only your specific favorite team, we’d likely take a very different approach.
  • Look at ratings, not just rankings. For example, less than a quarter of a point separates the #8 team (Toronto) from the #10 team (Dallas). Those three teams are all close to being equals. Often times, the distinction between two closely ranked teams is essentially negligible from a ratings perspective, so it’s not even worth debating.
  • Making futures bets requires more information. Just because our models think that a team has a higher or lower expected 2014-15 win total than the current Vegas line does NOT mean that it’s a good bet. Among other things, payout odds for futures bets make a huge impact on your expected returns. In short, hunting for value in current futures odds based on our projections is a more involved process than is covered in this post (here’s an NFL example from 2012).