Our most prolific early-season system is a strong 12-8-2 ATS so far this year, with a couple of days left still before this angle goes into hibernation. Picks are generated when home teams on ATS losing streaks (of one or more games) are laying more than a longball to non-division opposition during the first ten games of the NBA season. The play is to back the road dogs in such situations.
Adhering to this system has proven good for a (statistically significant at the 95% confidence level) 10% hypothetical return on investment, back to 2005-06. This empirically validated track record conveys high conviction in the notion that the profitability this system has amassed over time reflects a genuine relationship, and is not merely due to chance.
We posit that the success of this approach owes to the casual betting public's (misplaced) faith in better teams' ability to beat the number. Evidencing a bias toward favorites in NBA betting markets, data from the Bet Labs database suggests that of almost 21,000 regular season NBA games since 2004-05, favorites have received the backing of the majority of spread bets tracked by Sports Insights 78% of the time.
But the best players are probably not in full stride during the initial action of the season, because pre-season efforts tends to emphasize evaluating and developing younger players, whose contributions are less certain. As a result, in general, dogs are more likely to cover early in the year, as elite players use the first games of the season to tune up. Supporting this hypothesis, again from Bet Labs, 14,000+ home favorites have covered at about a 49% clip over since the 2004-05 season; this cohort beat the number a statistically significantly lower 47% of the time through the first ten games of the year though.
Additionally, we think that the perceived advantage that home field connotes draws more bettors toward favorites—despite the normalizing effect of the spread, and even when a team's negative momentum (i.e. recent ATS loses) hints that mass support might be unwarranted.
The following graphic indicates each of the plays generated by this system this season. For Saturday, November, 5th, plays are triggered on the Rockets +8.5 at the Timberwolves and the Blazers +11.5 at the Suns (we are passing on the latter contest, as we have observed a conflicting signal).