How I Handicap The NBA

Posting this to hopefully get some feedback.

Step 1: Find each team's relevant (home/away) stats:

- Pace

- Off Eff Rating

- Def Eff Rating

Step 2: Find the following stats:

- League Avg Off Eff Rating (Home)

- League Avg Off Rating (Away)

- League Avg Pace

Step 3: Calculate game pace:

Home Team Pace * Away Team Pace/ League Avg Pace

Step 4: Calculate Expected Points Per Possession (PPP) For Each Team:

Home Team Off Eff Rating * Away Team Def Eff Rating / League Avg Home Team Off Eff Rating

Away Team Off Eff Rating * Home Team Def Eff Rating / League Avg Away Team Off Eff Rating

Step 5: Calculate Expected Points Scored For Each Team

Home PPP * Game Pace / 100

Away PPP * Game Pace / 100

Step 6: Handicapping

Subtract home team's expected points from away team's expecte points to determine pointspread.

Add both expected scores together to determine game total.

****EXAMPLE *****

EG. Tonight Lakers @ Timberwolves

LAL (Away)

****

Off: 103.0

Def: 108.6

Pace: 103.07

MIN (Home)

****

Off: 113.1

Def: 106.8

Pace: 97.14

League Avg 

****

Off: 105.51

Def: 105.51

Pace: 96.27

Game Pace:

103.07 * 97.14 / 96.27 = 104

PPP:

LAL: 103 * 106.8 / 105.51 = 104.25

MIN: 113.1 * 108.6 / 105.51 = 116.41

Exp Points Scored:

LAL: 104.25 * 104 / 100 = 108.42

MIN: 116.41 * 104 / 100 = 121.06

Handicap:

MIN -12

Total: 229

Current Line:

Min - 10

O/U 227

Does anyone have any SQDL info that could help to narrow / remove bad picks and increase ROI with this?

Any improvements or suggestions would be appreciated. 

I calculated:

MIL - 7

Total: 216

So I'd take MIL - 4 and the over on that one too.

Appreciate the suggestion -- hopefully I can take that and more and tighten up my pick to improve the ROI

Im actually putting this method out there because by itself it is insufficient.

I've had it running as high as 58% for short bursts, but generally its season average is 48-52%. Not good enough.

I'm hoping that some suggestions from others can make it even better, and everyone can benefit as a result. For example, are there obvious lines or situations where we should avoid games? indicators to look for that could be used with this method to improve the win rate?

Hoping to do a bit of a knowledge/tech share here.

Interesting.  Where do you obtain your data??  I have a DB going back to 2013 and can run formulas in seconds to do a whole year's worth of data to see how well a method performs.  (Would also like to find historical 1Q and 1H closing odds as well.)  By the way, using data from COVERS NBA printsheets, averaging team A's OFF + team's B DEF and then B's OFF + A's DEF usually hits right around 52-53%.  Not quite good enough for the 52.38% needed to break even on straight bets at -110 long term.

The NBA is all about the 'matchup' and situational handicapping (strategy & logic)

with Sabermetrics absorbing 30% of the cap....the rest is a formulated methodology; initiating a system to a particular 'spot' on game....

May the NBA bring prosperity to those who are interested.

take note: the NBA is a very lazy league during the regular season (last four minutes) until the playoffs begin; then it becomes a whole new season. 48 min  of true championship basketball.

I get my stats from NBA.com, Hollinger on ESPN, and Covers.

This system was crusing along at 58% in Nov/Dec.....and came back down to earth in January. After the trade deadline? It's now positively better to fade the predictions. lol

Appreciate the feedback, looks like I have to add more subjective factors into the analysis.

Can you get enough data, let's say 2 or 3 years worth, to experiment with the formula?  Sports_Network definately has a point about not relying soley on one formula.  There's a myriad of other stats/situations that can help as well.  Am definately interested in your work.  Would also like to find 1Q and 1H historical odds, can't find any as of yet.

Agreed.  Take for example the Cavs.  Favored by 10 or more points 12 times this season, 0-12 ATS.  Wonder why they were favored by 10 or more so many times then?   GS the same.  14-15 ATS (favored10 or more).  So you could say there are quite a number of reviewable stats when capping a game.

Thank you very much guys. I haven't been tracking that stuff. That helps a lot!

I've added in some median data versus the line to my stats analysis of the games.

New Method Picks:

BKY/CHA: Over 215.5 (4 point adv)

PHI/CHI: Under 213.5 (0.5 point adv)

OKC/SAC: Over 214 (3.5 point adv)

LAC/GSW: LAC+10 (3 point adv)

Hey I'm going to post my line plays that I have been testing, so you can compare them with your formula, in the hope that we can work together and figure something out.

I had PHI as a play at -5.5, line is now -6.5 so will still play.

NYK +3.5

PHI -6.5

Thanks for the input rivermonsterz!

My formula made the following picks for tonight:

BKY +8 / Over 215.5

PHI -6.5 / Under 213.5

ORL -3.5 / Under 215.5

OKC -7.5 / Over 214

CLE -5.5 / Over 220.5

LAC +10 / Over 234

********************

Then, I took the calculations and compared them to historical performance afgainst the lines posted. Only the following picks were "double verified" by both parts of the analysis:

BKY/CHA Over 215.5

PHI/CHI: Under 213.5

OKC/SAC: Over 214

LAC/GSW: LAC+10

Cheers

Night's Results

**********************

Handicapping Method Picks: [3-9] 

Handicapping + Historical Data Consensus Picks: [2-2]

Historical Line Data Standalone Picks: [8-3-1]

Looks like I've got more work to do....

I think there is more to capping these games than just the scores.  Checking the ATS, how teams do on the road and the like.  Like today, MIL-TOR game.  TOR is 11-3 ATS playing at home against avg/weak teams.  At home SU, they are quite good.  TOR worth looking at for that reason as well as calculating the score/spread.  Against common opponents this year:  TOR +8.6 pts per game and MIL +0.1.  UNDER might be worth a look too.

Thanks NevadaBet!

Yeah, I'm still trying to find the right mix of factors. I'm testing a theory that the consensus picks between handicapping and the median line history is a long term winner. This is definitely a work in progress lol

Today's Consensus Picks:

ATL +7.5 / Under 211.5

CHA +3.5

BOS -2

MIL +7.5

CLE / MEM Over 214

MIA +2.5

HOU -8.5

UTA -4.5 / Under 207

DAL / LAL Under 223

Night's Results:

*******************

Handicapping Method Picks: 14-7 [17-16 YTD] 51.5% 

Historical Line Data Standalone Picks: 11-9-2 [19-11-2 YTD] 63.3% 

Handicapping + Historical Data Consensus Picks: 7-4 [9-6 YTD] 60.0% 

Better tonight. Changed 2 variables with the handicapping. Will continue to update and track record. Combining with consensus data seems to be a strong improvement.

Today's Consensus Picks:

ORL +9.5 / Under 216.5

MIA -8.5

NYL +5 / Under 211.5

Strongest Play Tonight:

BOS/NYK Under 211.5

4 different methods of analysis all in consensus.

BOL 

1-4 on the day.

Just gotta keep refining this process.....

Have you a method to check you methods using historical data?

No, not at the moment. Any suggestions?