Table 4. Kruskal-Wallis H tests of product preferences between clusters.

	Cluster 1 ‘Conscious tilters’ (n = 20)	Cluster 2 ‘Unconscious tilters’ (n = 52)	Cluster 3 ‘Non-tilters’ (n = 93)	n a	dif	H	η2H
Frequency of use	Mean ranking	Mean ranking	Mean ranking
Embedded livestream	103.353	87.51	76.101	165	2	6.52*	.03
Virtual live updates	94.70	91.10	75.96	165	2	5.01
Statistics board	103.903	93.873	72.431,2	165	2	11.69*	.06
Cash-out feature	92.50	83.76	80.53	165	2	1.16
Instant deposit	104.803	98.653	69.561,2	165	2	18.39**	.10
Concurrent bets	91.38	87.00	78.96	165	2	1.75
High-odds / microevents	99.48	89.09	76.05	165	2	5.60
Importance	Mean ranking	Mean ranking	Mean ranking
Embedded livestream	73.47	74.48	63.44	136	2	2.70
Virtual live updates	93.093	79.38	61.821	142	2	11.38*	.07
Statistics board	91.18	81.74	69.73	152	2	5.12
Cash-out feature	74.42	75.19	78.66	153	2	.29
Instant deposit	94.753	86.373	67.671,2	154	2	9.63*	.05
Concurrent bets	76.88	64.96	65.14	132	2	1.41
High-odds / microevents	66.75	69.22	63.75	131	2	.60

^aParticipants who reported ‘never’ using a feature did not rate its importance.

^{1, 2, 3}Representative of cluster that is significantly different (p < .05).

* p < .05,

**p < .001.