Results 1 to 15 of 15
  1. #1
    parenting for gnarness
    Reputation: chollaball's Avatar
    Join Date
    Apr 2005
    Posts
    5,667

    OT: ride difficulty ranking (code red geek alert)

    Warning: major nerd warning regarding this post

    DurtGurl posted in the Quad Bypass review about comparing ride difficulties, which is something I was thinking as well. Here is a formula I think works:

    mileage * ((avg grade+1) * 100) * terrain ranking

    1. mileage - self explanatory

    2. avg grade - this is uphill only. Yes, downs are effort, but I think the terrain ranking accounts for the effort of hard technical downs. So, in the case of the quad bypass, let's assume 6900 climbing, divided by (33.3 miles * 5280 feet) = 3.9% avg grade. Add 1 to the total, multiply by 100, for a figure of 4.9. I added 1 so the formula works. Assuming a flat road had 0% avg grade, you cant multiply by 0. And giving it a value of 1 makes it the same as a 1% grade. So (n+1) * 100.

    3. Terrain ranking. I came up with a scale 1-5 for terrain. Its certainly open to interpretation, but I think is a good start.

    terrain 1: road
    terrain 2: packed trail or jeep road
    terrain 3: looser, sandier, wetter, rockier
    terrain 4: like terrain 3 but more
    terrain 5: more of everything, genuinely technical obstacles, deep sand, etc. National, Porcupine Rim, etc.

    You could include DH runs in this as a 5 (or even 6) terrain, but it pushes the ability of the model since DH is just a different world.

    The quad bypass terrain probably averaged ~3, for the sake of argument.

    So to come up with the ride ranking, we do:
    mileage * ((avg grade+1) * 100) * terrain ranking

    Quad Bypass
    33.3 * 4.9 * 3 = 489.51

    For the Whiskey 50 (assumed 47 miles and 5k climbing), I came up with
    47 * 3 * 2.5 = 352.5

    Tour of Whites (6k climbing):
    102*2.1*2=428

    4 Peaks Road (5600 climbing):
    37*3.8*2 = 281

    Soul Ride 66 (6k climbing?)
    66*2.7*2.5 = 445.5

    You could obviously add more variables like wind, heat\cold, night, effect of being chased by wild dogs, mp3 player failing... And having accurate mileage and elevation is key. The terrain ranking works as long as you are consistent.

    Love to hear what others think, particularly those who actually know about complex math and could map this to a 1-100 scale.

    PS: yes I am a geek

  2. #2
    mtbr member
    Reputation: brianc's Avatar
    Join Date
    Jan 2004
    Posts
    3,041
    you should factor in the AZ board official "GNAR" scale in you formula. (just devide it by 2 since that is 1-10 scale.

    the obvious short coming of your formula is rides that lack elevation. I've been on rides that I would call agressive XC. which accoring to your scale would be nothing b/c the average grade factor would be 1. but the physical effort needed for such a XC ride is emense because one is constantly going up and down short steep hills (30 feet max) coupled with step ups, downs, "bony" sections of roots and rock.
    b

  3. #3
    Just another half mile...
    Reputation: Epicrider's Avatar
    Join Date
    Jan 2004
    Posts
    890

    SDMB rating system.

    Besides IMBA trail rating system: http://www.imba.com/resources/trail_...ifficulty.html that I use to rate the trail onmy maps, Sonoran Desert Mountain Bicyclists has a decent rating system that's based on technical and aerobic difficulty: Based on this rating system, the Quad bypass would be a 4B ride.

    HILL RATING
    1: Slight rolling terrain
    2: Rolling hills, no long climbs, low elevation
    3: Hilly, longer/more climbs, possible higher elevation
    4: Sustained steep climbs and/or higher elevation, advanced riders only

    DIFFICULTY RATING
    A: Smooth trail/jeep road
    B: Some technical challenges but mostly nontechnical
    C: Lots of technical challenges (e.g. big/loose rocks, ruts)
    D: Highly technical, advanced riders only




    Quote Originally Posted by chollaball
    Warning: major nerd warning regarding this post

    DurtGurl posted in the Quad Bypass review about comparing ride difficulties, which is something I was thinking as well. Here is a formula I think works:

    mileage * ((avg grade+1) * 100) * terrain ranking

    1. mileage - self explanatory

    2. avg grade - this is uphill only. Yes, downs are effort, but I think the terrain ranking accounts for the effort of hard technical downs. So, in the case of the quad bypass, let's assume 6900 climbing, divided by (33.3 miles * 5280 feet) = 3.9% avg grade. Add 1 to the total, multiply by 100, for a figure of 4.9. I added 1 so the formula works. Assuming a flat road had 0% avg grade, you cant multiply by 0. And giving it a value of 1 makes it the same as a 1% grade. So (n+1) * 100.

    3. Terrain ranking. I came up with a scale 1-5 for terrain. Its certainly open to interpretation, but I think is a good start.

    terrain 1: road
    terrain 2: packed trail or jeep road
    terrain 3: looser, sandier, wetter, rockier
    terrain 4: like terrain 3 but more
    terrain 5: more of everything, genuinely technical obstacles, deep sand, etc. National, Porcupine Rim, etc.

    You could include DH runs in this as a 5 (or even 6) terrain, but it pushes the ability of the model since DH is just a different world.

    The quad bypass terrain probably averaged ~3, for the sake of argument.

    So to come up with the ride ranking, we do:
    mileage * ((avg grade+1) * 100) * terrain ranking

    Quad Bypass
    33.3 * 4.9 * 3 = 489.51

    For the Whiskey 50 (assumed 47 miles and 5k climbing), I came up with
    47 * 3 * 2.5 = 352.5

    Tour of Whites (6k climbing):
    102*2.1*2=428

    4 Peaks Road (5600 climbing):
    37*3.8*2 = 281

    Soul Ride 66 (6k climbing?)
    66*2.7*2.5 = 445.5

    You could obviously add more variables like wind, heat\cold, night, effect of being chased by wild dogs, mp3 player failing... And having accurate mileage and elevation is key. The terrain ranking works as long as you are consistent.

    Love to hear what others think, particularly those who actually know about complex math and could map this to a 1-100 scale.

    PS: yes I am a geek

  4. #4
    igoslo
    Reputation: SoloRider's Avatar
    Join Date
    Aug 2004
    Posts
    971
    Scott Morris came up a system that I'm rather fond of even though it means I'm the subject of ridicule from my wife.

    Difficulty Index and Effort Index

  5. #5
    I am Walt
    Reputation: waltaz's Avatar
    Join Date
    Oct 2004
    Posts
    4,555
    Quote Originally Posted by SoloRider
    Scott Morris came up a system that I'm rather fond of even though it means I'm the subject of ridicule from my wife.

    Difficulty Index and Effort Index
    Interesting...so what does the Quad compute to under that system?
    Ride more; post less...

  6. #6
    igoslo
    Reputation: SoloRider's Avatar
    Join Date
    Aug 2004
    Posts
    971
    Quote Originally Posted by waltaz
    Interesting...so what does the Quad compute to under that system?
    Quad comes out as difficulty of 41.421

    to compare

    AZT South to North is 606.876(info stolen from Scott Morris)

    ToWM 42 mile is 18.955

    Goat Camp XC is 14.013

    BCT LPL is 8.297

    Kathleen's to the Helipad and back is 7.391

    24HOP is 6.335

    Pemberton is 4.597

    and the short preserves loop I do with my wife is 1.007
    Last edited by SoloRider; 01-30-2007 at 03:29 PM.

  7. #7
    Ouch, I am hot!
    Reputation: Dirdir's Avatar
    Join Date
    Jan 2004
    Posts
    4,732
    I like the simpler systems, such as the one Epicrider uses. Chollaball's system gave me a headache. Sorry dude.
    I AM JUST A JERK

  8. #8
    "Yabut"
    Reputation: MyBike'sBroken's Avatar
    Join Date
    Jun 2006
    Posts
    837
    Quote Originally Posted by Epicrider
    Besides IMBA trail rating system: http://www.imba.com/resources/trail_...ifficulty.html that I use to rate the trail onmy maps, Sonoran Desert Mountain Bicyclists has a decent rating system that's based on technical and aerobic difficulty: Based on this rating system, the Quad bypass would be a 4B ride.

    HILL RATING
    1: Slight rolling terrain
    2: Rolling hills, no long climbs, low elevation
    3: Hilly, longer/more climbs, possible higher elevation
    4: Sustained steep climbs and/or higher elevation, advanced riders only

    DIFFICULTY RATING
    A: Smooth trail/jeep road
    B: Some technical challenges but mostly nontechnical
    C: Lots of technical challenges (e.g. big/loose rocks, ruts)
    D: Highly technical, advanced riders only



    Yeah, I'm used to the SDMB scale, which always screws me up during Spring Fling. Because during Spring Fling, the HARD ride is the "A" ride, and the easy one is the "C" ride. Imagine my disapointment when we headed down DC, instead of up National. hehehe
    ~Aaron~
    "I like the night life. I like to boogie."
    ____________

    www.CosmicRaces.com

  9. #9
    Elitest thrill junkie
    Reputation: Jayem's Avatar
    Join Date
    Jul 2005
    Posts
    19,220
    nm....
    "It's only when you stand over it, you know, when you physically stand over the bike, that then you say 'hey, I don't have much stand over height', you know"-T. Ellsworth

    You're turning black metallic.

  10. #10
    .......................
    Reputation: ionsmuse's Avatar
    Join Date
    Jul 2005
    Posts
    3,045
    Quote Originally Posted by chollaball
    Warning: major nerd warning regarding this post [a bunch of damn numbers] PS: yes I am a geek
    This is one of the things I always hated about climbing. Difficulty ratings are pointless. Adjectival description, less so. What's the worse that'll happen? Die and get lost?

  11. #11
    parenting for gnarness
    Reputation: chollaball's Avatar
    Join Date
    Apr 2005
    Posts
    5,667

    interesting to see the other systems

    Thanks Epicrider and Solorider for info on the other systems, and to the other posters for comments, likes\dislikes etc.

    GEEK ALERT II - DONT SAY YOU WERENT WARNED

    I think the SDMB system has a lot going for it: simple, intuitive and gives someone a very easy and mostly accurate overview of a ride. It falls short as a comparative system, which is not a criticism its clearly not intended to be comparative. Ex: How do you look at a 2C vs a 3B ride and know which is harder? Or, does a 50 mile ride with 2 big climbs get the same rating as a 10 mile ride with 3 big climbs? But, if you go with the ranking and a 30 word description, its very useful and effective to let you know what you are riding.

    Epicrider suggested the Quad Bypass was a 4B, but I noticed SDMB ranked Fantasy Island a B and Chiva a C. Its been a few years for me on either, but I think the Quad Bypass was closer to Chiva terrain than FI.


    The "Scott Morris" system seems really good for comparing rides, and the number of rides, data, and personal massaging he put into it I'll bet makes it really effective. He also accounts for mileage as part of his evaluation, which the SDMB system does not. The accumulation of ride points is a nice feature, as it makes the ranking based on granular details and not 1 broad classification for hill-type or terrain type. Obviously, its only effective if you have some raw data. However, the description on his web page leaves some questions about the modelling. Did the model come first, or was the model adjusted to fit the data? Professional modellers will allow for some back-and-forth like this, but its the riskiest part of modelling because you might push the data towards a conclusion. Scott acknowledges some of this on his web page where he talks about deriving his constants. Couple other questions about Scott's system: he seems to have a very strict correlation between grade and technical difficulty, which may be too rigid? And, he does not break out the terrain from the grade. How does the model handle "flat and sandy" or "mild grade with baby heads"? Finally, what type of rides and the rider-style were used as the data? Is the data all based on "big climbs, moderate tech." Does it include "flat rides, very tech" like BrianC suggests? Are the personal opinions of the ride all based on one rider's strengths and styles? Would really enjoy hearing more about this system and its derivation, there is probably a ton of thought that went into it that the web page does not convey!!


    The system I suggested is kinda a cross between the 2. Its not real different than the SDMB system, but is comparative rather than simply descriptive. The place it is really weak is in accounting for climbing - I tried to get all mathy with % grade, but I think spreading it over the whole ride can miss the mark in either direction. 1 monster climb is a lot harder than a gradual grade, but would have the same effect on the total. Rolling hills are a lot easier than isolated hills, but would also have the same effect on the total. You could have further modifiers within the elevation category to account for types of hills during portions of the ride, but now you are getting into a level of granularity that you need all the GPS\Scott Morris data to use correctly.

    I'm going to try for a while using the SDMB rankings, but instead of letters for terrain use numbers, and multiply both terrain and hills by mileage. [terrain * hill type * mileage]. This will provide the comparative basis the SDMB lacks, but keep it simple enough.

  12. #12
    mtbr member
    Reputation: KRob's Avatar
    Join Date
    Jan 2004
    Posts
    9,908

    Hey good idea.

    Quote Originally Posted by chollaball

    So to come up with the ride ranking, we do:
    mileage * ((avg grade+1) * 100) * terrain ranking

    Quad Bypass
    33.3 * 4.9 * 3 = 489.51

    For the Whiskey 50 (assumed 47 miles and 5k climbing), I came up with
    47 * 3 * 2.5 = 352.5

    Tour of Whites (6k climbing):
    102*2.1*2=428

    4 Peaks Road (5600 climbing):
    37*3.8*2 = 281

    Soul Ride 66 (6k climbing?)
    66*2.7*2.5 = 445.5
    Wow. The one ride/race I have accurate numbers for scores a 884.4!! Yes it was tough. Was it almost twice as tough as the Quadruple Bypass? I doubt it. I'll have to try the quad and let you know.

  13. #13
    Saucy Size Moderator
    Reputation: Paul B's Avatar
    Join Date
    Jan 2004
    Posts
    4,483
    Awesome geekage! You could also include a handicapping system for individual riders. Your technical multiplier could change according to your skill level, and your climb multiplier could certainly change according to your weight/fitness.

    p.
    Don't be that guy! Read the forum guidelines.

  14. #14
    Scott in Tucson
    Reputation: Krein's Avatar
    Join Date
    Jul 2004
    Posts
    1,313
    Quote Originally Posted by chollaball
    The "Scott Morris" system seems really good for comparing rides, and the number of rides, data, and personal massaging he put into it I'll bet makes it really effective.
    Hi chollaball, Scott Morris here. Feel free to drop me an email (smorris AT topofusion.com) if you want to go into more detail, but here's a few answers.

    Did the model come first, or was the model adjusted to fit the data?
    Both. I came up with the basis for the model without looking at any data, and then adjusted it to fit the data I had. As for pushing the data towards a conclusion, there may be some of that here, yes.

    Couple other questions about Scott's system: he seems to have a very strict correlation between grade and technical difficulty, which may be too rigid?
    Yes, it is too rigid. This is a flaw in the system. But I wanted a system that did not rely on subjective rankings. First, because they are subjective and vary between people. Second, I didn't want to require any additional input -- only the GPS data.

    So you might have a REALLY steep road ride that gets too high of a ranking. Similarly, relatively flat technical rides might not be getting enough credit, as you noted.

    You can imagine being able to select a portion of your track and rating it moderate technical or sandy, or whatever. Or better yet, getting that info from a large scale trail network where people have already rated each trail segment. Would be pretty cool.

    Finally, what type of rides and the rider-style were used as the data? Is the data all based on "big climbs, moderate tech." Does it include "flat rides, very tech" like BrianC suggests? Are the personal opinions of the ride all based on one rider's strengths and styles?
    I loaded up 40 of my tracks, trying to cover as many types of rides as possible. I think I saved the list of what those rides were, but I know I had quite a few ridiculous ones in there (hike-a-bikes). I know I had some road rides, as well as a few half road / half trail rides. I definitely had some sandy ones in there.

    Only one rider style -- mine. Not sure what that style is. I like long rides and technical rides, if that helps. But, sad as it may be, I can't always do long/techy rides. So I end up doing a lot easier riding, just because I love getting out.

    The type of riding BrianC is suggesting is covered by the difficulty index. Even though the hills are short, if they are steep (and show up on the GPS), they'll be rated high. I can't think of many trails that are truly flat and "very tech." I've ridden across some lava rock sections that would qualify, but generally you need a good grade to expose large rocks.

    In the end, it is a perceptual model. Since it's only based on what I think, it's highly sensitive to things like how I was feeling the day I did a particular ride. It would be interesting to do a study and collect a sample of other people's rankings so it's not just based on me. I'm sure it would tweak things somewhat.

    But I have had a fair amount of feedback on it over the last couple years. No one has yet to say they think the rating is completely wacked or even wrong. Experienced mountain bikers, newbies, roadies and even hikers have chimed in to say they love the rating. Most recent guy I remember was an old hiker who has set a limit on the total difficulty rating he can do per week. He says it's been keeping him from over-doing it. Pretty cool.

    Anyway, thanks for the thoughts. If you have any suggestions for improving it, or other ideas, I'm all ears.
    Author of TopoFusion GPS Software. MTB+backpacking = bikepacking.net. Ride Diary.

  15. #15
    mtbr member
    Reputation: Evil Patrick's Avatar
    Join Date
    Sep 2004
    Posts
    2,610
    I think a great way to rate the difficulty of trails is via a DB of averaged HRM statistics
    mashed up with GPS data, ride time, rider's age-group and self-assessed riding level.

    The more data, the more accurate the mash-up, the better the trail rating.
    -- Evil Patrick

    My Music

    My Videos

    Ride everything! Remember, Elvis died pushing.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •