The Typology of Curation: What Kinds of Curators We Have Now?

Two days ago, I posted How Much Do Whales Influences on Rewards?. After posting it and having discussion with others, I found this study can provide more deeper understanding of curation patterns. As a follow-up study, this post is focusing on typology of curations: grouping curators by certain patterns.

Method

I used k-means clustering to classify the top 100 curators (by influence) into four groups. The sample's total influence is 99.5%, which means almost all of curation is done by the top 100 curators.

Data and Variables

The data is from Steem blockchain from 2016/8/16 to 2016/9/10.

• Influence: Percentage of payout a curator used
• Power: Percentage of Steem Power a curator owns
• IPR: Ratio between influence and power (Utilization of Steem Power)
• N: Number of votes a curator casted during the study period
• Range: Number of unique writers a curator voted for
• Mean: Average number of votes per writer
• Max: Maximum number of votes on a writer repeatedly
• StDev: Standard deviation among a curator's votes on each writer

The following table presents some descriptive statistics about the four groups.

Group	Sum_Influence	Avr_Influence	Sum_Power	Avr_Power	Avr_IPR	Avr_N	Avr_Range	Avr_Mean	Avr_Max	Avr_StDev
1	15.86	0.51	30.33	0.98	0.52	120.51	123.71	0.97	5.97	0.90
2	46.99	1.20	18.11	0.46	2.59	582.55	329.72	1.77	24.09	2.65
3	31.83	1.38	10.82	0.47	2.94	802.70	260.96	3.08	61.93	6.58
4	5.27	0.75	1.35	0.19	3.90	1410.57	164.57	8.57	70.00	12.08
Total	99.95	-	60.61	-	1.65	547.91	238.48	2.51	30.39	3.67

1. Lazy Lions

Curator in this group show low utilization of Steem Power on average (0.52), but they still have a great potential to become influential curators since they have more power than the other groups (0.98). The list of curators is as follow.

Account	Influence	Power	IPR	N	Range	Mean	Max	StDev
@dantheman	3.41	2.54	1.34	222	340	0.65	10	1.24
@jamesc	2.72	3.19	0.85	150	135	1.11	14	1.86
@rainman	1.80	1.49	1.21	203	171	1.19	10	1.31
@ned	1.56	5.72	0.27	45	254	0.18	3	0.37
@val-a	0.88	2.80	0.32	26	77	0.33	2	0.49
@fuzzyvest	0.57	0.58	0.97	306	214	1.43	12	1.93
@dan	0.44	4.49	0.10	1	12	0.09	1	0.36
@witness.svk	0.42	0.54	0.78	169	130	1.30	10	1.24
@wackou	0.34	0.62	0.55	92	57	1.61	10	1.59
@skywalker	0.28	0.63	0.44	18	13	1.38	3	0.62
@laonie	0.27	0.14	1.87	1230	1227	1.00	17	1.32
@blackjack	0.26	1.00	0.26	15	12	1.25	2	0.43
@roadscape	0.26	0.65	0.40	54	60	0.91	4	0.77
@lafona-miner	0.25	0.28	0.91	201	148	1.36	8	1.00
@firstclass	0.21	0.66	0.31	17	12	1.42	4	0.86
@arhag	0.19	0.61	0.31	30	40	0.75	3	0.60
@freedom	0.18	0.53	0.34	17	12	1.42	4	0.86
@cryptoctopus	0.17	0.15	1.13	147	111	1.33	9	1.17
@proskynneo	0.16	0.54	0.30	59	82	0.72	2	0.44
@datasecuritynode	0.16	0.44	0.36	21	18	1.17	3	0.50
@books	0.14	0.16	0.85	17	13	1.31	4	0.82
@disney	0.14	0.15	0.92	16	13	1.23	2	0.42
@bmw	0.14	0.18	0.74	17	13	1.31	4	0.82
@samupaha	0.14	0.20	0.67	-2	241	-0.01	17	2.24
@alphabet	0.13	0.33	0.40	15	14	1.07	2	0.26
@excalibur	0.11	0.32	0.36	15	14	1.07	2	0.26
@analisa	0.11	0.40	0.27	3	3	1.00	1	0.00
@arsahk	0.11	0.39	0.28	4	3	1.33	2	0.47
@bitcoin2016	0.11	0.42	0.25	17	12	1.42	3	0.76
@delegate.lafona	0.10	0.10	1.03	201	148	1.36	8	1.00
@cryptogee	0.10	0.08	1.25	411	236	1.74	12	1.79

2. Hidden Gem Explorers

This group consists of the biggest part of the total influence (46.99%) and they have the highest range and the second lowest Max and StDev. This implies that their votes are dispersed to various authors.

Account	Influence	Power	IPR	N	Range	Mean	Max	StDev
@blocktrades	11.29	3.71	3.04	630	486	1.30	22	1.51
@berniesanders	10.41	2.97	3.50	997	787	1.27	15	1.71
@tombstone	3.62	1.87	1.94	321	196	1.64	12	2.18
@summon	3.00	1.29	2.32	364	221	1.65	14	1.99
@complexring	2.49	0.67	3.69	958	544	1.76	40	3.08
@nextgencrypto	1.75	0.51	3.41	1061	759	1.40	14	1.51
@pharesim	1.64	0.87	1.88	837	406	2.06	54	3.93
@silversteem	1.55	0.46	3.37	1381	760	1.82	15	2.08
@riverhead	1.55	0.69	2.25	518	336	1.54	13	1.41
@steemit200	1.54	0.56	2.74	438	178	2.46	14	2.35
@satoshifund	0.97	0.35	2.78	329	130	2.53	21	3.61
@au1nethyb1	0.78	0.43	1.84	417	209	2.00	11	1.81
@silver	0.71	0.19	3.76	1403	763	1.84	15	2.13
@badassmother	0.69	0.26	2.68	662	237	2.79	16	2.83
@donkeypong	0.47	0.25	1.90	481	286	1.68	15	2.02
@joseph	0.45	0.19	2.37	374	155	2.41	30	3.63
@xeroc	0.35	0.18	1.93	478	240	1.99	45	3.48
@rossco99	0.30	0.15	1.99	210	119	1.76	13	1.56
@justin	0.25	0.07	3.60	1097	774	1.42	14	1.50
@clayop	0.25	0.56	0.44	209	78	2.67	27	4.53
@slowwalker	0.24	0.08	3.19	895	437	2.05	28	3.61
@cass	0.21	0.10	2.04	454	219	2.07	20	2.53
@boatymcboatface	0.20	0.06	3.44	185	111	1.67	11	1.37
@james212	0.20	0.09	2.20	570	231	2.47	29	4.21
@steempower	0.19	0.12	1.66	495	215	2.30	37	3.62
@onceuponatime	0.19	0.28	0.68	380	198	1.92	43	3.88
@sean-king	0.16	0.13	1.22	399	228	1.75	32	2.68
@dollarvigilante	0.16	0.07	2.29	622	239	2.60	46	4.23
@steemservices	0.16	0.05	3.13	921	780	1.18	14	2.66
@stellabelle	0.15	0.21	0.74	477	243	1.96	44	4.27
@renohq	0.14	0.13	1.07	270	113	2.39	15	2.51
@anwenbaumeister	0.13	0.07	1.79	771	425	1.82	28	2.40
@toxic	0.13	0.05	2.40	173	107	1.62	12	1.48
@kevinwong	0.13	0.06	2.20	973	472	2.06	42	3.08
@steemychicken1	0.12	0.15	0.81	259	149	1.74	19	2.63
@the-alien	0.11	0.04	2.54	247	248	1.00	8	0.77
@steemrollin	0.11	0.07	1.51	332	231	1.44	14	1.81
@inboundinken	0.10	0.03	2.96	539	257	2.10	20	2.20
@liondani	0.10	0.09	1.11	594	292	2.03	60	4.44

3. Village Builders

While this group has a significant influence on payout, they votes for small groups of writers repeatedly, probably due to build their own communities (they have very high Max and StDev). A good example is @abit and @boombastic who support the Chinese community. They have relatively high possibility to be developed into their own DCGs suggested by @ned

Account	Influence	Power	IPR	N	Range	Mean	Max	StDev
@smooth	13.19	2.86	4.62	791	343	2.31	65	7.50
@steemed	4.07	1.60	2.54	298	80	3.72	45	6.54
@itsascam	3.94	1.53	2.57	297	79	3.76	45	6.60
@smooth.witness	2.29	0.53	4.30	782	327	2.39	65	7.63
@steempty	1.28	0.70	1.84	478	165	2.90	34	4.83
@kushed	1.17	0.50	2.34	657	251	2.62	85	7.40
@xeldal	1.15	0.65	1.78	597	204	2.92	48	5.65
@enki	0.91	0.53	1.72	560	174	3.22	46	5.58
@steemroller	0.46	0.18	2.57	295	73	4.04	44	6.69
@abit	0.45	0.64	0.71	783	201	3.90	60	7.48
@cyber	0.43	0.11	4.02	2058	450	4.57	66	8.08
@boombastic	0.33	0.10	3.41	870	261	3.33	44	5.71
@anonymous	0.32	0.13	2.43	937	685	1.37	114	5.50
@nanzo-scoop	0.29	0.07	4.14	859	196	4.38	57	7.23
@eeks	0.24	0.07	3.51	1064	231	4.61	43	5.85
@barrie	0.23	0.06	3.63	832	220	3.78	44	6.02
@tuck-fheman	0.20	0.11	1.91	953	350	2.72	76	5.99
@juneaugoldbuyer	0.18	0.06	2.99	1230	303	4.06	54	6.42
@gavvet	0.18	0.11	1.66	687	274	2.51	50	6.18
@ihashfury	0.16	0.10	1.57	848	203	4.18	55	7.75
@benjojo	0.14	0.13	1.06	870	233	3.73	28	5.44
@mummyimperfect	0.11	0.02	5.35	621	154	4.03	45	6.16
@nabilov	0.11	0.03	3.27	1098	545	2.01	211	9.11

4. Superhumans

They are possibly bots. They have very high N on average (1410.57) and the highest IPR (3.90). They may have very high curation rewards profitability.

Account	Influence	Power	IPR	N	Range	Mean	Max	StDev
@wang	1.93	0.66	2.91	1778	213	8.35	82	10.51
@recursive	1.60	0.31	5.22	1198	123	9.74	51	11.81
@hr1	1.02	0.21	4.80	1086	104	10.44	52	12.00
@jl777	0.19	0.05	3.90	2434	393	6.19	78	9.80
@recursive2	0.18	0.04	4.07	1173	102	11.50	80	13.88
@taoteh1221	0.18	0.04	4.67	1010	116	8.71	67	11.57
@recursive3	0.17	0.04	3.99	1195	101	11.83	80	14.98

Conclusion

Curation is a very important task in Steemit. More understanding of curation will provide constructive foundations for further growth of Steemit.