NBA Predictions With Efficiency Ratings

Note: This is from the current book on basketball betting I’m working on. A bit of a long read, but some decent information.

Efficiency ratings have become a bit of craze when it comes to advanced statistics in the NBA. But all they really are is Points Per Possession multiplied by 100. That’s it. After seeing all the calculations that had to be done before efficiency ratings became so widely available, we should be grateful that the basic concept has caught on. Now we can do game projections in just a couple of minutes by using those numbers.

While there are plenty of different sites which list efficiency ratings, they are not all the same. There is a bit of disagreement when it comes to the formula and free throws. Some sites use the .436 and others use .44, which will throw the numbers off slightly. There is also a bit of a difference in the pace ratings between sites, which is also dependent on the formula used, among other things.

I prefer to use the official NBA stats at, which lists both pace and efficiency ratings on the same page. The primary reason I use the NBA site will become apparent in a little bit, as using the efficiency ratings for projections is essentially a two-part process. For the 2019-20 season through the shutdown, the average pace was 100.5 possessions, while the average Offensive Efficiency was 110.4 and the number for Defensive Efficiency was 109.9. The NBA calls the two Offensive Rating and Defensive Rating.

Creating a Projected Score

The first step in creating a projected score is to come up with the pace rating for the game and we’ll use a game between Memphis and Philadelphia to work through. The Grizzlies averaged 103.3 possessions per game, while Philadelphia had an average of 99.4 possessions. Memphis sees 2.8 more possessions than average, and Philadelphia sees 1.1 fewer possessions than average. Memphis will get a +2.8 for pace and the 76ers receive a -1.1, so you’re left with a +1.7 for pace. Adding 1.7 to the average pace of 100.5 gives you 102.2, so we can expect 102 possessions per team in the game.

Memphis has an Offensive Efficiency rating of 108.9, which is -1.5 compared to the league average and Philadelphia has a Defensive Efficiency rating of 107.6, which is -2.3 less than the league average, so Memphis will get a -3.8 for this game. When you take 110.4 and subtract 3.8, you’re left with a total of 106.6, which is the final adjusted rating for Memphis.

Philadelphia has a 109.7 for its Offensive Efficiency, which is -.7 from the league average and Memphis has a 109.9 for defense, which is the same as the league average, so no adjustment needs to be made there. Philadelphia’s only adjustment is a -.7 from the league average of 110.4, which leaves you with 109.7, which also happens to be Philadelphia’s Offensive Efficiency rating. That makes sense since Memphis is your average defensive team.

Now we’re left with adjusted numbers of 102 pace possessions per game and ratings of 106.6 for Memphis and 109.7 for Philadelphia. The last step before being able to calculate a projected score is to turn the efficiency rating for each team into a points per possession rating, which is simply moving the decimal point two places to the left. All we’re doing is dividing by 100, so moving the decimal point is all that’s needed.

You’ll simply multiply the expected number of possessions of 102 by 1.066 for Memphis and by 1.097 for Philadelphia to get a projected score, which in this case is 111.89 for the 76ers and 108.73 for the Grizzlies, so we have Philadelphia 3.16 points better and a projected total score of 220.62 on a neutral court.

The home court advantage isn’t as great as some people tend to believe, as home teams had a scoring margin of +2.1 in the 2019-20 season, so add one point to the home team’s projected score and subtract a point from the visiting team’s projected score to account for home court. Simply adding two points to the home team will throw off your projected total, which is why the addition and subtraction are done.

If you have a little bit of Excel knowledge, you can easily create a spreadsheet where all you do is enter the league average for pace and Offensive and Defensive Efficiency, along with the numbers of each team to get your projections and it will save a little bit of time.

New York at Atlanta Example

Let’s work through one more just to make sure you have the process down and this time we’ll look at the game between New York and Atlanta, which was one of the four games played on March 11, 2020, the last day of games before the league shutdown.

New York entered the game averaging 99.1 possessions, along with an Offensive Efficiency of 105.9 and a Defensive Efficiency of 112.4. The Hawks averaged 103.3 possessions and had an Offensive Efficiency of 107 and a defensive rating of 114.4.

For possessions, New York receives a -1.4, while Atlanta receives a +2.8, giving a result of +1.4, which will become 102 possessions per team, since the +1.4 is added to the league average of 100.5, for a total of 101.9 and we’ll simply round up.

New York’s offensive rating of 105.9 is -4.5 compared to the league average, but Atlanta’s defensive number is 4.5 points worse than the league average, so New York receives a -4.5 and a +4.5 to give them a 0, meaning they will get the league average of 110.4.

Atlanta’s offensive number is -3.4 compared to the league average, but New York is 2.5 points worse on defense, so the Hawks receive a -3.4 and a +2.5 for the Knicks’ defense, which results in a -.9 rating for Atlanta in this one, which is a 109.5.

Multiply the 102 possessions by 1.104 to come up with New York’s projected score and multiply 102 by Atlanta’s 1.095 to come up with the Hawks’ projection. New York’s projection is 112.6, while the Hawks receive a 111.7. On a neutral court, we’d have New York winning by one, but with the Hawks being at home, Atlanta would be projected to win by a point and a total of 224.3.

The Hawks were favored by 4.5 and the total was 232.5, so we’d be looking at the Knicks and the under and would have split on this game, as the teams were tied 118-118 at the end of regulation and Atlanta went on to win in overtime.

Seeing the game go over the total wasn’t really a surprise, as the Hawks had seen a big increase in scoring recently. For the season, the Hawks averaged 111.8 points per game and allowed 119.7, but if you look at their games from February and March, you see Atlanta averaged 118.7 and allowed 124.3.
The Knicks also had been on a bit of a scoring increase, as New York averaged 105.8 points and allowed 112.3 for the season, but in February and March the Knicks were scoring 110.8 points and allowing 113.6.

Last 10 Games Check

That happens more frequently than expected, and the only way to deal with it is to look at a team’s most recent performances. What we do is run a second set of numbers, but instead of using season-to-date averages, we will use numbers from a team’s last 10 games, which is easily available on the NBA site with a click of a button and is the primary reason I use the website. Under the “Season Segment” just click last 10 games and you get an entirely different group of numbers, which consists of the last 10 games for each team.

You do have to start from scratch, so for the last 10 games, we’ll have an average pace of 100.6, and an average offensive number of 111.5 and a 112.9 for defense.

In the last 10 games, New York has an average pace of 98.7 and the Hawks are seeing 102.1 possessions, so New York receives a -1.9 for pace and Atlanta gets a +1.5, so we’re left with a -.4. If you take 100.6 and subtract .4, you’re left with a 100.2, which we’ll round down to 100. Our expected possessions are now 100 based on the last 10 games.

But we see both teams have really picked up the offense, as New York has an offensive rating of 112.4 and the Hawks have a 113.4. On defense, the Knicks have a Defensive Efficiency of 116.3 and the Hawks receive a 117.
New York’s 112.4 is +.9 better than the league average and Atlanta’s defensive number is 4.1 points worse than average, so the Knicks receive a +.9 and a +4.1 for a total of +5. Atlanta’s offense gets a +1.9 and New York’s defense receives a +3.4 for a total of +5.3.

For the Knicks, 111.5+5=116.5, which is New York’s projected score, and the Hawks end up with 111.5 plus 5.3 for a total of 116.8. We now have Atlanta winning by .3 points on a neutral court and a total of 233.3. Accounting for home court, our line on this one would then be Atlanta -2.3 and a total of 233.5.

Redoing the numbers a second time isn’t too time consuming, as it’s a relatively easy process. But it is something that should be done to enable you to keep up with changes in style. If you only have time to run one set of numbers, I’d be more inclined to use the most recent numbers. Ideally, you’ll be able to run both sets to get the best projections. The key is to find games where both the season and the recent numbers agree.

Recent Trends

Looking at a team’s numbers for the most recent 10 games allows you to pick up on trends that the general betting public may have overlooked. That was the case with the 2019-20 Milwaukee Bucks, who were one of the best offensive teams in the league. But in the final 10 games before the league shutdown, Milwaukee ranked dead last in Offensive Rating, but was easily No. 1 in defense. Despite being tied for No. 2 in pace, Milwaukee was 2-8 in totals over that span.

At the other end of the spectrum was the Orlando Magic. Orlando was primarily a defensive team and scored and allowed fewer points than the league average. But the Magic flipped the switch shortly before the shutdown and were No. 1 in offense and No. 25 defense for the last 10 games. As a result, the Magic finished with 12 straight overs before the shutdown occurred.

Leave a Reply