PDA

View Full Version : How Roller Coaster Ratings Are Calculated


Sum_Zero
04-09-2005, 12:34 PM
This thread is for specific information and theories about how roller coaster ratings are calculated. Because of fewer unit conversions and simpler numbers when calculating things like speed, force, and curve radius Iíll stick with SI measurements. Iíll focus on coasters that meet the minimum drop, speed, and other thresholds (no junior coasters for me.) To keep the length shorter Iíll just edit my existing posts as I learn more

Sum_Zero
04-09-2005, 12:36 PM
Results from a WOODEN ROLLER COASTER:

SPEED RESULTS:

Maximum Speed: Score doubling (e,i,n) threshold exists between 9.14m/s and 10.03m/s.

Ride time/Ride length: At the moment I view these as a medium for creating exposure to speed, force, air time etc. If you make a song out of a single note it will be about as boring if its 2 minutes or 20 minutes long. Similarly, data from my flat rectangular coasters taken through multiple circuits demonstrate how an increase in 1896m led to no change in ratings.


FORCE RESULTS:

Maximum Positive Vertical G: No sign of thresholds as with lateral G, drop height etc. Up to 10.26 g tested. In the real world, long exposures to forces over 6.3 g would have all the passengers riding around unconscious :)

Maximum Lateral G: There is an upper limit between 7.70 and 8.01 g where intensity jumps to 14+ and excitement drops to around 0.30.

I have some partially tested ideas about how forces impact ratings in the theories section below.


DROP RESULTS:

Number Of Drops: A score doubling threshold exists at 2 drops minimum. Even a bump of lift-flat-hill-lift will qualify. Raising the number of drops beyond 2 has no effect on ratings, aside from creating new speeds forces and air times.

Highest Drop Height: Another score doubling threshold, at a minimum height between 5.43m to 7.43m.

Number of Inversions: No special thresholds here as with number of drops.
Donít be fooled by extra gains from setting new levels of maximum vertical g on the first inversion.
0-20 inversions tested. (higher numbers of inversion involved multiple circuits)

More about drop/inversion results in the theory section below.

Reminder: results above come from tests on the WOODEN ROLLER COASTER

Sum_Zero
04-09-2005, 12:37 PM
UNPROVEN theories and early observations based on tests with the WOODEN ROLLER COASTER:

Interaction between the score numbers themselves: itís possible that the intensity rating itself will alter nausea and or excitement, other similar interactions are possible. This will complicate the final goal of determining exactly how all factors come together to generate ratings.

SPEED RESULTS: Some of my older data (possibly from the wooden wild mouse instead) indicate that at speeds around 35m/s and above, the positive vertical g maximum for going from flat to gentle slope was actually less than at lower speeds, and air time increased. Maybe this was because the exposure was too short in time to count. A more interesting possibility is that air flow under the coaster at high speeds is considered to have a cushioning effect : the coaster, literally, begins to fly :)

Average speed: trying to isolate a change for a given average speed is proving difficult, in the next series of tests Iíll try treating speed like a force (see force theories below). So far average speed seems to affect intensity the most.


FORCE RESULTS: I think maximum force levels contribute to ratings.

I also think that, after the minimum thresholds for things like drop height and number of drops are exceeded, the bulk of ratings are determined by the type (e.g. lateral, vertical etc), level (gís) and duration of forces. Speed might also be treated this way. This exposure to force or speed can be visualized by the area under the lines in the test graphs. A large area in a short time should lead to intense coasters, while the same area taken over a long time will decrease intensity for about the same level of excitement. I wonder, after peak forces and ride time are determined, will it matter whether the remaining exposure to force is concentrated (close to the maximum level of g) over a small portion of track, with low force pieces of track as filler to equalize ride time, or distributed at lower levels of g over a large portion? I'm also curious how multiple forces at the same time(e.g. helix, laterally curving hill) are handled.


DROP RESULTS:
The peeps seem to scream during drops, so maximum and total drop heights might be factors in generating ratings.

I think that the change in score for inversions is due to the vertical g involved for making a loop on the wooden coaster. My guess is the actual number of inversions is not a factor.

Again this section is just unproven theory based on WOODEN ROLLER COASTER tests

Sum_Zero
04-09-2005, 12:38 PM
Data:

WOODEN ROLLER COASTER
Iím too lazy to go through all the data in detail. Instead Iíll just do vague descriptions and a few samples of the more interesting tests.

SPEED RESULTS: I think its best to do speed tests on flat pieces of track, using boosters and brakes to prevent changes in vertical or lateral G so you can isolate changes from other factors like lateral G.

Evidence of the minimal effects of ride time/ track length as a thing in itself: 2 flat rectangular coasters, Coaster B is 632 meters longer of flat, chain pulled track than coaster A.
Base ratings (e,i,n) A= (058,044,029) B=(064,045,029)
Difference in length = 632 meters
Difference in ratings : e 0.06, I 0.01, n 0
Taking the 2 coasters through multiple circuits, I found that the ratings for both coasters increased slightly but the DIFFERENCE IN RATINGS (.06, .001,0.00) between the 2 coasters REMAINED THE SAME even when the track length (due to extra circuits) was increased to a difference of 1896m.

DROP RESULTS: I ran a small series of tests involving single lift bumps and larger hills, multiple circuits etc.

For the inversions I tended to go with minimum speeds to keep intensity manageable, I also put a bit of helix track to exceed the maximum vertical gís generated by the first loop. For amusement some stats from the 20 loop/20 inversion (4 circuit) coaster: e: 7.01 i: 8.45 n: 4.73.

Sum_Zero
04-09-2005, 12:40 PM
The next steps will probably be to measure the length and curvature of each type of track piece, then figure out how friction is handled, and use these to get precise measurement of how the different force and speed exposures affect coasters of different ride times and ratings levels. For now Iíll keep going with the wooden roller coaster only.

rizzla
04-12-2005, 03:37 AM
Cheers, Sum_Zero, that is quite interesting. In one of the coaster competition threads I have been experimenting a bit but simply by comparing the same layout for different coaster types. Now I don't want to put you off your experiments with the Wooden Coaster, but I can say that all ratings differ depending on the coaster type.

Well, admittedly, this is quite easy to figure out by simply looking at stats coasters produce in general. As I have argued before, the Giga Coaster for example is too good - if your goal is to have very high excitement, medium to high intensity and low to medium nausea. Other coaster types will generally make it impossible to have an excitement rating higher than intensity and just as difficult to keep the nausea down if you want any semi-decent excitement. The Woody I'd say is an in-between, producing about equal excitement and intensity and manageable nausea.

A few coaster types can't even be classified as "serious" - try what I did, a simple four track pieces climb, then the same drop, then onto a banked 180 degree turn, a back straight, another banked 180 degree turn back to the station. While the "serious" coasters will produce something like 1 excitement, 1 intensity and 0.x nausea, smaller coasters (like, for example, the Water Caoster) will produce ratings of 4, 5 and 4 respectively.

caribcaoster
05-04-2005, 09:23 PM
Are the coasters and aother rides in the game realistic in terms of Speed, G forces and what not? In other words, if I designed a coaster in the game, then construct it in the real world, would it function in "safe" manner as projected in the game?

CoasterExpert
05-04-2005, 09:54 PM
No caribcaoster necause the coasters in rct3 are not realistic and would cause serious injuries because of the small transitions.