Random Independent Events And Probability
In the Probability of Independent Random Events, we look at the probability of multiple random events and we step through three different problems as practice. We'll be increasing the difficulty of problems, so practicing these simple exercises will help us later. Stay tuned for future videos on SQL In Six Minutes.
Speaking of statistics, if I pull data from my recent seven posts (where 1 is most recent), we see that my posts receive more views than upvotes. which after doing a quick random sample is outside the bell curve of normal (still within 1 standard deviation). The data I pulled recently (and added this post):
| Post | Votes | Views |
|---|---|---|
| This post | 14 | 53 |
| 1 | 6 | 16 |
| 2 | 8 | 29 |
| 3 | 5 | 18 |
| 4 | 5 | 9 |
| 5 | 4 | 10 |
| 6 | 5 | 13 |
| 7 | 5 | 14 |
Some random posts to contrast (ceteris paribus measurement period lengths):
| Post | Votes | Views | Rating |
|---|---|---|---|
| 1 | 135 | 145 | More views than votes, but oddly close after a short period and with many votes |
| 2 | 56 | 51 | More votes than views; very common pattern that we see on SteemIt |
| 3 | 96 | 16 | Much more votes than views; these are more common than you think, but not as common as the above example |
Consider the numbers: my average post exceeds 3 multiples as many views as votes after a recent period. The best performing post (from a marketing standpoint) for these posts is 1.07 multiples as many views as votes. In one case - which isn't as rare as we'd think - the ratio is less than 0.20 multiples as many views as votes.
While it's true that some posts do get more views than votes, the pattern of being consistent is incredibly rare with the exception of a few members on this platform. The big difference is these members tend to have tons of followers. Consider this assessment from a marketing perspective: this means that highly rated posts do not translate into actual views or people considering what's written. This also matches some of the random data I've pulled which show users with large Steem Power balances and quite a few followers, yet have never posted or interacted with others on this platform (but vote for specific users). This is one of the many advantages with learning statistics; we can uncover patterns which the average person may not see or recognize.
For me, this is a great accomplishment for myself, as few members consistently achieve obtaining more views than votes within a short post period. I consistently do, unlike many members here. This just demonstrates one of the useful features of Elementary Statistics and how we can apply it to research we may be doing.
If you're interested data, I teach the course Automating ETL, which teaches you how to get data easily so that you can quick analyze data in meaningful ways. If you're looking for a short introduction to T-SQL (a SQL language), check out fast-paced course T-SQL In Two Hours.