Introduction

In this post I will be analyzing the writing levels of the top 10 currently actively whales on Steemit, I will be using

@blueorgy's Catchawhale tool to get a sample set of whales, and then I will be running each of those users into my tool SteemIQ which calculates the Flesch Kincaid grade level of the user and uses a number of other writing metrics to calculate SteemIQ. Using this data, I will be able to get a grasp on what level the whales write at. Let's get to it.

The Data

Whale	Flesch Kincaid Grade Level	SteemIQ	Writes at the level of
@ned	9.5	94.6	Mitt Romney, Carly Fiorina
@dan	8.9	119.2	Tim Ferris, Hillary Clinton, Leo Tolstoy
@blocktrades	12.5	111.6	Soren Kaplan, Joseph Micheli
@smooth	7.2	103.0	J.K. Rowling, Thomas Pynchon
@tombstone	8.1	97.7	Tim Ferris, Hillary Clinton, Leo Tolstoy
@summon	4.2	86.4	Ernest Hemmingway, Donald Trump
@pharesim	8.8	102.8	Tim Ferris, Hillary Clinton, Leo Tolstoy
@complexring	7.2	117.1	J.K. Rowling, Thomas Pynchon
@riverhead	7.0	95.9	J.K. Rowling, Thomas Pynchon
@roadscape	6.5	94.6	Stephanie Meyer, J.R.R Tolkein

Whale

Flesch Kincaid Grade Level

SteemIQ

Writes at the level of

@ned

9.5

94.6

Mitt Romney, Carly Fiorina

@dan

8.9

119.2

Tim Ferris, Hillary Clinton, Leo Tolstoy

@blocktrades

12.5

111.6

Soren Kaplan, Joseph Micheli

@smooth

7.2

103.0

J.K. Rowling, Thomas Pynchon

@tombstone

8.1

97.7

Tim Ferris, Hillary Clinton, Leo Tolstoy

@summon

4.2

86.4

Ernest Hemmingway, Donald Trump

@pharesim

8.8

102.8

Tim Ferris, Hillary Clinton, Leo Tolstoy

@complexring

7.2

117.1

J.K. Rowling, Thomas Pynchon

@riverhead

7.0

95.9

J.K. Rowling, Thomas Pynchon

@roadscape

6.5

94.6

Stephanie Meyer, J.R.R Tolkein

@dan takes the win for the highest SteemIQ out of the whales sampled

@blocktrades takes the win for the highest average Flesch Kincaid grade level out of the whales sampled

Conclusion

Data reports like this one will ultimately be able to give a better insight on how to improve the accuracy of calculating SteemIQ, and also shows that quality content does not have to compromise readability. A high SteemIQ can indeed be achieved with readable content at the 6-8 Flesch Kincaid readability range. In the future I hope to enhance SteemIQ so that it is able to utilize machine learning to look for grammatical structure and content of posts to accurately give a better picture of quality. Currently the accuracy of SteemIQ maybe skewed by lists of data that do not have ending punctuation. Subsequent version of SteemIQ will account for this and the metric will be adjusted accordingly.