Backtest Results

Thursday, February 26, 2026

Generated at 1:09 PM ET

Backtest Results - Thursday, February 26, 2026

SPREAD PICKS SCORECARD

Pick	Result	Outcome
INDIANA PACERS +12.5	Hornets 133 - Pacers 109	Failed to cover by 11.5

Spread Record: 0-1 (-1.1u)

OVER/UNDER PICKS SCORECARD

Pick	Result	Outcome
POR @ CHI UNDER 234.5	Trail Blazers 121 - Bulls 112 (Total: 233)	Under by 1.5
SAC @ DAL UNDER 234.5	Kings 130 - Mavericks 121 (Total: 251)	Over by 16.5
WAS @ ATL UNDER 235.5	Wizards 96 - Hawks 126 (Total: 222)	Under by 13.5

O/U Record: 2-1 (+0.9u)

PARLAY RESULT

Leg	Pick	Outcome
1	POR @ CHI UNDER 234.5	WIN
2	WAS @ ATL UNDER 235.5	WIN

Parlay Result: WIN (+2.6u)

DAILY SUMMARY

Category	Record	Units
Spreads	0-1	-1.1
Over/Under	2-1	+0.9
Parlay	WIN	+2.6
Daily Total		+2.4

QUALITATIVE ANALYSIS

Model Performance Review - February 26, 2026

WHAT WORKED

Under totals strategy was solid - Hit 2/3 under bets with strong reasoning on pace and defensive matchups
Washington/Atlanta under crushed by 13.5 points - correctly identified two inconsistent offenses in a potential blowout scenario

WHAT MISSED

Indiana spread catastrophically wrong - Predicted CHA -2.6, actual line was CHA -12.5, and Pacers still lost by 24. Model severely underestimated injury impact compounding effect
Sacramento/Dallas over missed by 16.5 points - Failed to account for pace increase when bad teams play each other in high-scoring environment

MODEL CALIBRATION NOTES

Injury adjustments appear flawed - Adding individual injury values (+4.5 + 3.0 + 3.0) linearly doesn't capture how multiple key players out creates exponential weakness
Totals analysis stronger than spread work - 2/3 success rate on unders suggests better calibration for pace/defense factors than team strength

METHODOLOGY ADJUSTMENTS TO CONSIDER

Implement non-linear injury scaling - When 3+ rotation players are out, apply 1.5x multiplier to account for roster depth collapse
Add "bad team pace-up" factor - Sub-.400 teams often play faster in desperation, inflating totals beyond BPI-based projections

Bottom Line: Solid totals methodology, but spread model needs major recalibration for injury impact and team strength differentials.

CUMULATIVE SEASON RECORD

Category	Record	Win %	Units
ATS	40-35	53.3%	+1.6
O/U	40-40	50.0%
Parlays	6-23
Season Total			+1.6u

View original analysis