Inspired by
@Hyde, I decided to do something very similar with my own rankings, plotting the correlation between Rankings and Height, Speed, Length, Inversions, Total Rides and Year Opened. Total Rides might seem like an odd metric to test, but I added it in because I wanted to figure out whether that good old chestnut nostalgia has much of an effect on how highly I rate a coaster. Will I rate something that I have an enmeshed relationship with and have ridden numerous times more highly than something I've only ridden once?
When I tested this out, the ranking of how highly correlated certain attributes were with ranking were as follows (I've reversed the direction of the correlations to clear things up; the raw correlations given by Python were negative because ranking gets lower as a number as you rate something more highly):
| Ranking | Attribute | Pearson Correlation Coefficient (2dp) | Spearman Correlation Coefficient (2dp) | Average Correlation Coefficient (2dp) | Strength of Correlation based on Average |
| 1 | Speed | 0.57 | 0.57 | 0.57 | Moderate Positive Correlation |
| 2 | Height | 0.52 | 0.53 | 0.52 | Moderate Positive Correlation |
| 3 | Length | 0.47 | 0.48 | 0.47 | Weak Positive Correlation |
| 4 | Total Rides | 0.37 | 0.52 | 0.44 | Weak Positive Correlation |
| 5 | Opening Year | 0.33 | 0.39 | 0.36 | Weak Positive Correlation |
| 6 | Inversions | 0.18 | 0.21 | 0.20 | No Significant Correlation |
So my rankings would suggest that the attribute I favour most strongly in a coaster is
speed, with there being moderate evidence in favour of me generally favouring greater speed. There is similarly moderate evidence in favour of me generally favouring greater
height, there is weak evidence in favour of me generally favouring greater
length, a higher number of
total rides and a newer
opening year, and there is insufficient evidence in favour of me generally favouring
inversions.
To be honest, I would have instinctively said that speed was my favourite of the main 4 statistics before even doing that analysis, so that tallies up quite well, really!
If you were asking me more instinctively what I enjoy in a coaster, though; I would have to say that simply put, I really enjoy coasters that are fun and thrilling, with good smoothness, comfort and rerideability. The main fundamental criteria for me to rate a coaster highly are simply that it's fun, thrilling and rerideable!
There are specific elements I do enjoy, however. In terms of specific elements I enjoy; I absolutely relish the feeling of airtime on a coaster, and 9 times out of 10, a coaster in the upper echelons of my rankings needs to have at least some notable degree of airtime or negative g-forces. That's definitely something I specifically look for and rate highly in a coaster. I also really relish a coaster with a great sense of speed, and to be honest, I do often tend to gravitate towards taller coasters; for instance, a 200ft+ coaster is nearly always a winner for me, with 4 of the 6 200ft+ coasters I've ridden being in my top 10 and even the 5th being only just outside it at #12!
Super Dark and Scary Forest of Statistical Regressions 
