HOW THE "SCORE PREDICTION" MODEL WORKS?
The predictive model uses regression analysis to determine coefficients for the factors that might have impact on the game outcomes. As the model continuously improves itself by employing some optimization techniques, it only focuses on predicting the nearest final scores individually for road team and home team.

HOW TO EVALUATE THE RESULTS?
Since the goal is to output nearest score for road team and home team, criterias such as margin of victory, accuracy of picking the winner team should not be primarily considered when evaluating the success. Though, the record for predicting winning teams has been tracked and is 59-23 (.720) since April 17, 2010.
An assigned grade for both road team and home team does a better job of classifying the results by taking points differential between predicted and actual score.

Complete prediction history for the Playoffs 2010 has been documented at the prediction archive.

ROAD TEAMS

HOME TEAMS

HOW TO INTERPRET THE CHARTS? Each score prediction is evaluated by a grade that measures the differential between predicted score and actual score. If this differential is in the "0-3" points range, it gets an "A"; for the "4-7" points range, it gets a "B" and so on.

Subscribe to 2010 NBA Playoffs Score Predictions.

Find out the sortable, filterable Playoffs 2010 schedule and results at prediction archive.

You can always follow current season's NBA rest days stats including back-to-back game stats at NBAstuffer.

After the win over OKC, Portland improved to 14-3 W-L (.823) in second game of back to back sets. This is by far better than any team in the NBA.

They have played with a plus 4.3 efficiency differential, which is already adjusted to opponent strenght and game location.

What's more, the Blazers also lead the NBA in "4th game in 5 nights", having won 3 of 3 (with +9 efficiency differential) so far. In back-to-back situation of "3rd game in 4 days" (3IN4-B2B) they trail San Antonio and Cleveland according to winning percentage.

In the recap McMillan says: "It was a gut check for our guys, we have been talking about winning back-to-back games." It appears that Stats are backing McMillan. Portland Trail Blazers is a good regular season team, and hopefully will be better in Playoffs where there are no back-to-back games!

Following article is a guest contribution by our long time reader Brian Sands, I asked him to tell us how he has been using NBA rest days stats to make better decisions.

Is there any statistical edge to get over the sportsbooks? When NBA betting, closing NBA number by a point can be beaten and it is possible to add extra 4% to winning percentage. Taking an extra half point on 2 or 7 makes much more valuable than taking an extra half point on 8 or 3.

Let’s look deeper into it:

The chart tracks how the league average differs for three different possession based stats: Offensive efficiency, defensive efficiency and pace (the number of possessions per 48 minutes).

And these applies the numbers to seven different conditions; 4th game in 5 nights, 3rd game in 4 nights (played yesterday), 2nd game of a back to back, 3rd game in 4 days (rested yesterday), 1 day rest, 2 days rest and 3 or more days rest.
Offensive Efficiency
-1 Points/Per 100 Possessions – 4th game in 5 nights
-1 Points/Per 100 Possessions – 3rd game in 4 nights (played yesterday)
-0.1 Points/Per 100 Possessions – 2nd game of a back to back
-0.2 Points/Per 100 Possessions – 3rd game in 4 days (rested yesterday)
-0.2 Points/Per 100 Possessions – 1 day rest
+0.3 Points/Per 100 Possessions – 2 days rest
-0.3 Points/Per 100 Possessions – 3 or more days rest

Defensive Efficiency
+1 Points /Per 100 Possessions – 4th game in 5 nights
+1 Points/Per 100 Possessions – 3rd game in 4 nights (played yesterday)
+0.9 Points/Per 100 Possessions – 2nd game of a back to back
- 0.4 Points/Per 100 Possessions - 3rd game in 4 days (rested yesterday)
-0.4 Points/Per 100 Possessions - 1 day rest
-1.1 Points /Per 100 Possessions - 2 days rest
-0.5 Points /Per 100 Possessions - 3 or more days rest

Pace
-0.8 Possessions - 4th game in 5 nights
-0.3 Possessions - 3rd game in 4 nights (played yesterday)
-0.2 Possessions - 2nd game of a back to back
-0.2 Possessions - 3rd game in 4 days (rested yesterday)
-0.1 Possessions - 1 days rest
+0.5 Possessions - 2 days rest
+0.5 Possessions - 3 or more days rest

The numbers for a team playing their 4th game in 5 nights and 3rd game in 4 nights is what we would expect, the offense goes down by a point and the teams give up a point on average throughout the NBA. Remember that the books will know this but they might try to over adjust the NBA betting lines, if the game line is more than 2 points off what it should be, than the bookie is over inflating the line due to the public’s perception of the affect of rest on the team.

The number that stands out for me is teams that are coming into the game with 2 days rest, the NBA average shows teams in this situation are playing at their optimal level. The teams only score 0.3 points more than the league average but the defense tightens up by 1.1 points a game. Another point I found interesting but it makes sense is that teams scoring goes down on 3 or more days rest, the shooters are like the rest of us, a little slow getting back to work after a long weekend.

Of course, those numbers represent the NBA as a whole; the chart can extrapolate which teams are the best under each circumstance. For example two of the worst teams this season have played under some of the worst conditions, the Indiana Pacers and the LA Clippers have both played the killer 4 games in 5 nights, 4 times this season, more than any other team. In addition to those 4 in 5’s they have played 20 and 19 back to back games this season respectively, which is the first and third highest totals in the league.

With the chart, you can use the individual team situations and compare them to the line as a way to help you get an idea of determining the value of the line.

See the NBA Score Prediction Model in action! Check out prediction archive and prediction records for the current season.                    

If you have already subscribed to the premium plan, sign in below to access score predictions.

I'm glad to announce another valuable addition to team stats. As you could recall, NBAstuffer's NBA schedule heat map has done a pretty good job of visualizing road trips & homestands. With the road trip tracker you can follow your team on its road trips and find out the information such as total games played in road trip, efficiency differential averaged during the road trips.

I'm pleased to announce a new feature:
The chart which can be found at the 2009-2010 regular season rest days stats does a good job of understanding what impacts various rest days have on NBA teams' performances.

Since we are just four weeks into the season, the differentials should lead to more meaningful results as the season progresses.

Long awaited NBA 2009-2010 season tips off today, just wanted to share a few things for those NBAstuffer users who are asking for score predictions:

First of all, the prediction model needs some real season data, so it will be starting by December, 1, 2009. In an effort to keep the results consistent, I never feel like to start predictions earlier than December. As usual, only registered members are getting the chance to access daily NBA score predictions. Sign-up NBAstuffer for free, it takes 30 seconds!

Since I've been frequently asked for doing it, this year the prediction content may be offered with a premium plan, but nothing is clear about it. I'm working on technical details and let you know whether a premium plan will be revealed or not.

Until then, stick around here to learn more about basketball analytics or take look at 2009-2010 advanced stats for teams and players. Pay a visit to this season's locked-out referee stats page. Download NBA box score stats or schedules in excel and manipulate the data in your own way! Advanced stats tables for the 2009-2010 season will be up in a few days.

Let's have a good season together.
Any concerns, or suggestions? Your feedback is appreciated.