Does GSW Have a Magic Number of Points Scored to Win or Lose? + Poll
INTRODUCTION
In looking at first 22 SAS games in 2017-18, I realized SAS had won every game in which they had scored 95 or more points (14-0) and lost all but one in which they had score less than 95 (1-7). In the context of NBA scoring during 2017-18 this seemed a surprisingly powerful number of points and a low number, so I looked at a few other teams. First, to understand how low the SAS win point level is. Consider that the average (2017-12-02) team had scored 105.4 points per game. Of the 325 games, the median (163rd highest) winning score was 111 and the median losing score was 99. (These medians are far apart, so unsurprisingly only 21 games were decided by 2 or 1 points.)
Breaking this down some, 115 winners scored 115 or more. On the other side, 66 losers scored 110 points or more, 104 scored 106 or more, 145 scored 101 or more, 155 scored 100 or more, and 208 scored 95 or more. Only 117 losers scored less than 95. (The lowest score so far in season (of course, by a loser) was 69 points by CHI in losing to OKC 69-101. The highest number of points scored by a loser was 131 by BRK in losing 131-140 to IND on October 18.)
ANALSIS
To return to the issue of key scores for winning. I decided to eyeball estimate two characterizing scores for each team that I looked at:
• The score needed to win: team wins all or almost all games when score this or more (preferably at most two loses scoring more)
• The score needed to lose: team loses all or almost all games when score this or less
I tried to have an empty middle with few games played having scores between these two scores.
How often do such magic point totals exist? I choose some team because of interest – SAS, GSW, DET (because I thought might resembled SAS in speed of play), HOU, and BOS. I choose 4 that I thought would be hard to establish the two numbers for because IMHO their seasons have been strange in some way – POR, MIN, WAS, and MIL. I felt these four might challenge my ability to establishlevel for needed to win and needed to lose. Here is what I found listed as team, (record above score needed to win), score needed to win, (record between score needed to win and score needed to lose), score needed to lose, (record below and including score needed to lose) and [Pace].
GSW (16-1) 110 (0-0) 107 (1-5) [101.7 PACE (Basketball Reference)]
SAS (14-0) 95 (0-0) 94 (1-7) [94.7]
DET (13-2) 99 (0-0) 95 (1-5) [96.6]
HOU (17-2) 105 (0-0) 90 (0-2) [99.3]
BOS (14-1) 101 (0-0) 100 (5-3) [96.3]
POR (13-3) 99 (0-0) 97 (0-6) [96.8]
MIN (9-0) 111 (1-1) 107 (3-9) [96.7]
MIL (9-3) 98 (0-0) 97 (1-6) [95.8]
While I had to stretch a little in two cases (POR and MIL) allowing 3 losses in group above need-to-win score, it worked well for these teams. [It did not work well for WAS.]
I took a peek back at 2016-17, and GSW had a magic point level at 102-103.
GSW (67-6) 103 (0-0) 102 (0-9)
SAS (50-8) 100 (11-9) 92 (0-8)
In 2016-17, both SAS (92 points) and GSW (102) had clear-cut needed-to-lose values guaranteeing a loss, (0-8) and (0-9). GSW’s needed-to-win situation was anything above its needed-to-lose value had an 84% chance of winning. Thus, GSW did have a magic point total of 103. SAS needed-to-win score was 100 for an 86% chance of winning. However, SAS score to be sure to lose was 92, 8 points lower. This compared to GSW’s 102 that guaranteed a loss. In between SAS’s two values, SAS played .550 ball (111-9). Roughly, SAS’s point levels do not appear to have changed substantially between 2016-17 and 2017-18 despite the absence of Parker and Leonard. Contrastingly, a good chance exists that GSW’s values have changed for the early part of the 2017-18 season compared to all of 2016-17 – values for these possibly mythical numbers.
PACE
The difference in Pace could explain 7 of the 13-15-point differences between SAS and GSW point levels in 2017-18. The other 6 points of difference between GSW’s 107 needed to lose and SAS’s 94 might reflect a difference in defense but it also as a component from GSW’s stronger opponents played (strength of schedule). Teamrankings gives this strength of schedule difference a 1.6-point value, as I understand it. Altogether, this is consistent with other comparisons of early 2017-18, GSW has a better offense than SAS and a worse defense. (In their one 2017-18 game, GSW won 110-98 meeting GSW’s needed-to-win point level.)
Correlation with Pace has one obvious outlier, MIN. MIN has a basic anomality that may be behind this – its difference in offensive prowess (5th in OffRtg by Basketball Reference 110.9) and defensive (26th in DefRtg 110.6) for a NetRtg of 0.3. While it has won only one more game than Basketball Reference estimated (mainly from ratings), it being a combination of top 5 and bottom 5 values is unusual. (My favorite simplification is that Coach Thibs’ young players did not see his success with defenses in BOS and CHI and love it when he urges them to play better offense and ignore him when he hoarsely screams at them the same thing concerning defense.)
CONCLUSION
Magic scores to win and to lose may be a more viable concept than I ever dreamed, especially magic scores to lose. I therefore found these results interesting. However, I will need to explore the concept further before I can establish limits to its viability – and possibly my skepticism might decrease substantially or not.