Backtest History

267 graded games

Every graded prediction from the frozen walk-forward backtest, next to the closing spread and the final margin. Lines are quoted for the home team.

NFL backtest history
Season	Wk	Matchup	Score	Model	Close	Result	Model err
2016	15	MIA @ NYJ	34–13	+1.6	+2.5	+21.0	19.4
2016	15	NE @ DEN	16–3	+2.3	+3.0	+13.0	10.7
2016	15	NO @ ARI	48–41	-3.4	-3.0	+7.0	10.4
2016	15	LV @ LAC	19–16	+2.4	+2.5	+3.0	0.6
2016	15	PHI @ BAL	26–27	-8.0	-5.0	-1.0	7.0
2016	15	PIT @ CIN	24–20	+2.5	+3.0	+4.0	1.5
2016	15	SF @ ATL	13–41	-14.5	-13.5	-28.0	13.5 (model closer than the closing line)
2016	15	TB @ DAL	20–26	-4.2	-7.0	-6.0	1.8
2016	15	TEN @ KC	19–17	-4.7	-6.0	+2.0	6.7 (model closer than the closing line)
2016	14	ARI @ MIA	23–26	+2.0	+2.0	-3.0	5.0 (model closer than the closing line)
2016	14	ATL @ LA	42–14	+7.6	+4.5	+28.0	20.4 (model closer than the closing line)
2016	14	BAL @ NE	23–30	-5.6	-6.0	-7.0	1.4
2016	14	CHI @ DET	17–20	-5.3	-7.5	-3.0	2.3 (model closer than the closing line)
2016	14	CIN @ CLE	23–10	+10.3	+4.0	+13.0	2.7 (model closer than the closing line)
2016	14	DAL @ NYG	7–10	+5.5	+3.5	-3.0	8.5
2016	14	DEN @ TEN	10–13	-1.6	-2.0	-3.0	1.4
2016	14	HOU @ IND	22–17	-6.2	-6.5	+5.0	11.2 (model closer than the closing line)
2016	14	MIN @ JAX	25–16	+4.9	+3.0	+9.0	4.1 (model closer than the closing line)
2016	14	NO @ TB	11–16	-5.3	-2.0	-5.0	0.3 (model closer than the closing line)
2016	14	NYJ @ SF	23–17	-2.6	-3.0	+6.0	8.6 (model closer than the closing line)
2016	14	LV @ KC	13–21	-2.8	-3.5	-8.0	5.2
2016	14	PIT @ BUF	27–20	+4.1	+1.0	+7.0	2.9 (model closer than the closing line)
2016	14	LAC @ CAR	16–28	-0.5	-1.0	-12.0	11.5
2016	14	SEA @ GB	10–38	+0.3	+3.0	-28.0	28.3 (model closer than the closing line)
2016	14	WAS @ PHI	27–22	+4.4	+1.5	+5.0	0.6 (model closer than the closing line)
2016	13	BUF @ LV	24–38	-4.2	-3.0	-14.0	9.8 (model closer than the closing line)
2016	13	CAR @ SEA	7–40	-5.6	-8.5	-33.0	27.4
2016	13	DAL @ MIN	17–15	+5.5	+3.0	+2.0	3.5
2016	13	DEN @ JAX	20–10	+6.0	+3.5	+10.0	4.0 (model closer than the closing line)
2016	13	DET @ NO	28–13	-6.9	-6.5	+15.0	21.9
2016	13	HOU @ GB	13–21	-7.7	-7.0	-8.0	0.3 (model closer than the closing line)
2016	13	IND @ NYJ	41–10	-0.3	-1.0	+31.0	31.3 (model closer than the closing line)
2016	13	KC @ ATL	29–28	-5.2	-5.0	+1.0	6.2
2016	13	LA @ NE	10–26	-14.5	-13.0	-16.0	1.5 (model closer than the closing line)
2016	13	MIA @ BAL	6–38	-1.1	-3.5	-32.0	30.9
2016	13	NYG @ PIT	14–24	-7.7	-7.0	-10.0	2.3 (model closer than the closing line)
2016	13	PHI @ CIN	14–32	-1.4	-1.0	-18.0	16.6 (model closer than the closing line)
2016	13	SF @ CHI	6–26	-2.3	0.0	-20.0	17.7 (model closer than the closing line)
2016	13	TB @ LAC	28–21	-1.5	-3.5	+7.0	8.5 (model closer than the closing line)
2016	13	WAS @ ARI	23–31	+2.4	-2.5	-8.0	10.4
2016	12	ARI @ ATL	19–38	-6.8	-3.5	-19.0	12.2 (model closer than the closing line)
2016	12	CAR @ LV	32–35	-4.2	-3.5	-3.0	1.2
2016	12	CIN @ BAL	14–19	-3.1	-3.5	-5.0	1.9
2016	12	GB @ PHI	27–13	-2.8	-4.0	+14.0	16.8 (model closer than the closing line)
2016	12	JAX @ BUF	21–28	-9.6	-8.5	-7.0	2.6
2016	12	KC @ DEN	30–27	-2.2	-3.5	+3.0	5.2 (model closer than the closing line)
2016	12	LA @ NO	21–49	-7.7	-7.5	-28.0	20.3 (model closer than the closing line)
2016	12	MIN @ DET	13–16	+0.5	-2.0	-3.0	3.5
2016	12	NE @ NYJ	22–17	+9.7	+9.0	+5.0	4.7
2016	12	NYG @ CLE	27–13	+8.1	+6.5	+14.0	5.9 (model closer than the closing line)

Model and Close are home lines; Result is the final margin on the same axis. Model err is the absolute miss of the model's margin, shown green when the model was closer than the closing line.