Table 4. Statistical analyses on when to stop.

Number of players and rounds included in the analyses in each group
Group	Player		Round (Mean)		Round (SD)		Round (Min)		Round (Max)
High-Involvement	1679		6001.9		14176.8		12		200706
Low-Involvement	651		440.9		959.3		13		13114
ANOVA
Effect	df		MSE		F		ges		p
Involvement Level (High vs. Low)	1, 2328		0.09		5.36		<.001		.021
Prior Outcome (Loss vs. Win)	1, 2328		0.29		3353.74		.520		<.001
Interaction	1, 2328		0.29		8.95		.003		.003
Pairwise comparisons
Comparison (A vs. B)	A-mean	B-mean	diff	lowerCI	upperCI	df	t	p	lnBF	g
High-Loss vs. High-Win	1.24 (0.15)	0.17 (0.49)	1.07	1.04	1.10	1678.0	68.5	<.001	1115.02	3.297
Low-Loss vs. Low-Win	1.21 (0.22)	0.25 (0.79)	0.96	0.88	1.04	650.0	24.4	<.001	207.52	1.884
High-Loss vs. Low-Loss	1.24 (0.15)	1.21 (0.22)	0.03	0.01	0.05	898.1	3.0	.007	3.33	0.140
High-Win vs. Low-Win	0.17 (0.49)	0.25 (0.79)	-0.08	-0.14	-0.01	850.1	-2.3	0.027	0.83	0.105
(High-Loss—High-Win) vs. (Low-Loss—Low-Win)	1.07 (0.64)	0.96 (1.01)	0.10	0.02	0.19	860.4	2.5	0.027	1.47	0.114

Note: ANOVA: df = degrees of freedom. In a 2 by 2 ANOVA, the dfs for all effects are the same. MSE = mean square of the error. ges = generalized eta squared. Pairwise comparisons: Comparison (A vs. B) = the two variables compared in each comparison. A-mean, B-mean = means of the left (A) and the right (B) variable in a comparison, with standard deviations in parentheses. diff = difference between A and B. lowerCI, upperCI = lower and upper boundary of 95% confidence intervals of the difference. df, t, p = degrees of freedom, t value and p value from the Welch’s t tests (between-subjects comparisons) or paired-samples t tests (within-subjects comparisons). P values were corrected for multiple comparisons using the Holm-Bonferroni method. lnBF = the natural logarithm of Bayes factors. g = Hedges’s average g.