Advance look - Exploring the Steem blockchain: Who are the best curators?
According to a scoring method that I'm toying with (and assuming no errors in my code), here are yesterday's top-50 curators with more than 10 curation rewards. If I'm thinking about this right, the scoring should be independent of wallet size.
Since this is a first draft of code, please take it with a (very large) grain of salt. ;-)
| Rank | Account | Score | # curation rewards |
|---|---|---|---|
| 1. | curx | 3.448100851990027 | 89 |
| 2. | kevinwong | 3.3527491606275746 | 89 |
| 3. | thranax | 3.2627874077159684 | 13 |
| 4. | rosselena | 3.1552191654670936 | 15 |
| 5. | hive-176147 | 3.115957028494533 | 22 |
| 6. | misterajay | 3.113158538921441 | 89 |
| 7. | suntr | 3.078082605874703 | 71 |
| 8. | remade | 3.0705505774239734 | 119 |
| 9. | kingporos | 3.021754333757083 | 61 |
| 10. | hive-180821 | 3.0204745427892563 | 61 |
| 11. | veryhappyday | 3.0110350496954337 | 107 |
| 12. | harkar | 2.970134240473849 | 126 |
| 13. | steem-d-anlovnit | 2.9514029628077534 | 16 |
| 14. | ducato | 2.9220098770860043 | 40 |
| 15. | karja | 2.8868016330771606 | 168 |
| 16. | moh.arif | 2.8288143568215864 | 30 |
| 17. | alexmove.witness | 2.7873675216387777 | 17 |
| 18. | hafizullah | 2.7369652741319035 | 13 |
| 19. | happy.free | 2.7335559438964854 | 11 |
| 20. | yonaikerurso | 2.722749760504736 | 23 |
| 21. | ydavgonzalez | 2.7103017030721883 | 15 |
| 22. | tingtong12 | 2.7014702092525837 | 36 |
| 23. | alexmove | 2.689933986417561 | 17 |
| 24. | turkcayi | 2.677522536847546 | 37 |
| 25. | steemit-market | 2.6769754898238523 | 33 |
| 26. | apoloo1 | 2.6627333914485054 | 44 |
| 27. | yakspeace | 2.6534343366226465 | 12 |
| 28. | mayepariata | 2.65276663371876 | 20 |
| 29. | enamul17 | 2.6478354800525423 | 12 |
| 30. | gabrielagg | 2.6231312757647727 | 14 |
| 31. | sct.curator | 2.598700022603238 | 11 |
| 32. | bangla.witness | 2.5706947265248457 | 17 |
| 33. | mcgome | 2.5386128796094525 | 11 |
| 34. | uco.bnb-l | 2.5307333424308585 | 11 |
| 35. | nevlu123 | 2.506946007201965 | 17 |
| 36. | realmankwon.scot | 2.5011603049550355 | 11 |
| 37. | crisvera | 2.497439162555394 | 16 |
| 38. | maytte | 2.496834903943723 | 21 |
| 39. | uco.research | 2.4918047901990636 | 11 |
| 40. | oppps | 2.4709790677807217 | 130 |
| 41. | realman0711 | 2.4549583093282226 | 11 |
| 42. | steem-venera | 2.4531241633708953 | 13 |
| 43. | ezzy | 2.437971758459919 | 211 |
| 44. | radard | 2.43687048211729 | 11 |
| 45. | ripon0630 | 2.4271568411578595 | 36 |
| 46. | cub1 | 2.4217650939509627 | 54 |
| 47. | steemit-pak | 2.39911097010915 | 13 |
| 48. | mdkamran99 | 2.395637383479352 | 11 |
| 49. | sct | 2.376352367176782 | 11 |
| 50. | ctime | 2.364418940709981 | 422 |
Eventually, I hope to start publishing regular updates as well as the code I'm using, but it might be a while... Finding time to work on this has been a challenge.
Looking at the SteemDB curation history from a handful of accounts in the list, these accounts seem to be collecting curation rewards at a rate in the neighborhood of 20-26% of their staked SP.
@remlaps-lite,Thank you very much for your interesting research.
I would like to know what indicators did you use to study the effectiveness of curation?
And what do you think is more important: Avg. Efficiency or Weighted Avg.?
It's actually fairly simple. The curation score that I'm assigning on any post is a multiple of the ratio - [VESTS received] divided by [rshares voted]. Then, that number is averaged over the entire day for each voter. (I multiply VESTS times 1 billion, to get "nanovests" just to reduce the decimal places in the final fraction).
This is a simple average, since it's not weighted by rshares voted. That's on the "to-do list".
Excellent work