What's new

What makes a coaster good?

Archie

Mega Poster
Something I have been thinking about recently is what makes a roller coaster good. The philosophy of coaster design has defiantly changed in the last few years, in the 90s up to the mid 2000s it seemed for a coaster to be considered good it had to be the fastest or tallest or just be the biggest in some form, TTD and Millennium force for example. Today it seems more like quality over quantity, we have coasters such as Helix and Toutatis, not the fastest or tallest by any means but considered some of the best. But what is it about them that makes them the good?
 
What makes a coaster good is truly up to the rider. Some people like rides with lots of airtime, some like rides with strong positive Gs, some like rides with good inversions, some like rides with launches, some like long rides, some like unique rides, some like their rides butter smooth, some like janky rides, some like rides with good theming, but most land with a mix of several.

It really depends on the rider. In most cases there is no objectivity in coasters.
 
Positive and negative forces. Length is a consideration also when ranking. 3 great airtime moments on a short ride arent as good as 3 on a longer ride, because there is more experience ot enjoy
 
Great question actually. I think when we're young we assume the higher the better, but actually there are so many factors e.g theming etc....

Certainly not number one for me but I do think one thing that features higher than me for most is actually the restraints. Nothing like getting into one of your favourite coasters and looking forward to the pure comfort of a smooth and fun ride.
 
This is why it's so hard for me to rank the coasters I've ridden--there are so many different things that go into making a coaster "good," and a lot of them can't be quantified. Beyond just G-forces and height and speed stats, a lot of coasters have uniquely cool moments that you don't find on many or any other rides.

Speaking for myself, one thing I've found I like that I haven't seen other people talk about is just, how far away from the station the ride takes you. That's one of my favorite things about Fiesta Texas, that two of their coasters sprawl out against the cliff wall and take you on a journey across the back end of the park. Feels like I'm on an adventure!
 
Today it seems more like quality over quantity, we have coasters such as Helix and Toutatis, not the fastest or tallest by any means but considered some of the best. But what is it about them that makes them the good?
In the plainest terms, variety.

If I'm talking to someone about the hobby, the inevitable usually comes up as to asking what's my favourite? And then why? Suddenly it gets very hard to answer that question to anyone with little context (and without boring them with all the subtleties) and I'll settle on something like 'it does a bit of everything'.

As alluded to though, there's a hundred different factors at play and each of us has different preferences, priorities and emphases on each, which also differ over time as we ourselves change and/or experience more.
 
For me it depends on what the coaster is going for. You wouldn't slate something like the Big One for not having inversions because it's not an inversion focused coaster, you would slate it for not using its height and speed well. Similarly, Nemesis doesn't really have much airtime but that's fine because it's not the focus, it's the speed, intensity and inversions. I judge a coaster on how well it executes its core concept and ideas - in this example Nemesis does a much better job at this than the Big One.

Rides with variety are obviously really good like Helix, but I judge based on execution and intent more than just variety. It would be "how does this stack up as a hypercoaster, or an airtime centric wooden coaster, or a multi-inversion coaster or whatever else". That's why I like how many different types of coasters there are. If they were all everything at the same time it would get samey. I suppose I haven't ridden enough coasters to get a general trend, especially since in the UK most of the best coasters are inversion based, and so are most of the foreign coasters I have been on.

The remaining part is how comfortable it is - this is why Infusion is awful and another knock against the Big One and basically everything at Blackpool minus Revolution and Icon. Roughness cripples a coaster for me these days as I get older.
 
So this has actually been the core question I have built my own rankings tracking around: "What specific elements/features are most correlated with higher-ranked roller coasters?" I've focused on easiest characteristics to track (I went down a rabbit hole of trying to classify/track specific inversion types - A LOT of time would be needed to properly code out!) and landed on:

  • Top Speed
  • Height
  • Length
  • Drop Height
  • Drop Angle
  • Inversions
  • Year Opened
  • Duration (Time)
  • Cost
By working through many, many RCDB pages and correlating these stats with respective roller coaster rankings (that is, which of these stats, if any, is closest linked to changes in rankings, such as shorter coasters being ranked lower), we arrive at the following table:

CategoryR Value Correlation (Higher is Better)Rank
Top Speed0.691
Height0.642
Length0.613
Drop Height0.574
Drop Angle0.485
Inversions0.376
Year Opened0.327
Duration0.318
Cost0.239

All of which is to say: a good roller coaster is a relatively fast, tall design with good to decent length. Inversions, angles of first drop, and relative age (Year Opened) do not appear to play much of a factor.

If you're interested to dig in more, including many, many charts that show respective correlations, click my Top 10 in signature to go straight to the Google Sheet. :)
 
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):
RankingAttributePearson Correlation Coefficient (2dp)Spearman Correlation Coefficient (2dp)Average Correlation Coefficient (2dp)Strength of Correlation based on Average
1Speed0.570.570.57Moderate Positive Correlation
2Height0.520.530.52Moderate Positive Correlation
3Length0.470.480.47Weak Positive Correlation
4Total Rides0.370.520.44Weak Positive Correlation
5Opening Year0.330.390.36Weak Positive Correlation
6Inversions0.180.210.20No 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!
 
A
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):
RankingAttributePearson Correlation Coefficient (2dp)Spearman Correlation Coefficient (2dp)Average Correlation Coefficient (2dp)Strength of Correlation based on Average
1Speed0.570.570.57Moderate Positive Correlation
2Height0.520.530.52Moderate Positive Correlation
3Length0.470.480.47Weak Positive Correlation
4Total Rides0.370.520.44Weak Positive Correlation
5Opening Year0.330.390.36Weak Positive Correlation
6Inversions0.180.210.20No 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!
Always love a good correlative study! <3

As a quick aside, appreciate your trying Pearson vs. Spearman Correlation; the dataset is most likely too basic (aka simple, number inputs versus larger dataset or complex data) to have an appreciative difference. I played around with similar stat comparisons in my own list (table below) - basically the data is too simple to have any additional, fun insights from running TTest or SteyX Test lol

🌲 Super Dark and Scary Forest of Statistical Regressions 🌲
VarianceTTest
Length2,091,158.5Length0
Height4,244.6Height0
Drop4,525.2Drop0.00000004080046273
Inversions4.5Inversions0
Top Speed331.9Top Speed0
Duration2,016.9Duration0.000000006108816823
Drop Angle364.7Drop Angle0
Year Opened360.3Year Opened0
Cost$644,905,912,413,449.00Cost0.00000000151203259
Pearson TestSteyX
Length-0.61Length69.43
Height-0.64Height67.09
Drop-0.58Drop65.74
Inversions-0.37Inversions81.56
Top Speed-0.69Top Speed62.19
Duration-0.31Duration83.38
Drop Angle-0.48Drop Angle72.70
Year Opened-0.32Year Opened83.06
Cost-0.23Cost75.27
 
Top