Should sandbags/soft climbs exist?

We live in an age of data, it wouldn’t be the most impossible task to regrade everything (I know people have tried). Would you adopt a new system based on accuracy or preserve the old ways?

For me I would adopt a new approach. While I’m sure the new approach wouldn’t be pin point accurate, it would probably fluctuate like gps for running.

We already have this with MP consensus grades versus more static guidebook grades. Wisdom of the masses versus expert opinion

B Gwrote:

We already have this with MP consensus grades versus more static guidebook grades. Wisdom of the masses versus expert opinion

I mean kinda, some people rate climbs based on their home area, some on climbs relative area, some just like to sandbag and sometimes there isn’t enough ratings.

Hobo Greg so you like the adventure of not knowing how hard a climb is?

Edit: to Eric. Elo rating exists, it’s used in chess and I think sports fans use it as well. I would contend that basketball is a pretty sophisticated thing to quantify, and it’s definitely not perfect. While the warriors in their 70 win pursuit were favored to win every game they didn’t not (like how we would expect a 5.13 climber to onsight 5.12). The rating would be a proxy for probablility and while not perfectly accurate it’s been used in more sophisticated applications.

 There is a lot of math I have learned to quantify the “unquantifiable.” Amazon figures out if you are gonna need to buy diapers, which diapers to show you, where the diapers should display, how long to display a diaper ad, and what else to display. This is just figuring out the route grade, if anything is not that complex.

Soft climbs definitely need to exist. Stop downgrading the one soft 12a or V10 at the local area. Yeah it’s a little soft but it helps people break into the grade and break that mental barrier. Everybody knows 11d is harder anyway. Sandbags should exist too to keep people in check. One of my favorite climbs from my trip to Chatt was getting spit off Sandbox at HP40 a couple times to start the day. Not a fan of the safety sandbags though. Just difficulty. 

You are depending on 2 false assumptions:

1. That you can accurately measure the difficulty and assign a grade.  There is little about a route that you can quantify and measure. 

2. That one grade would apply for all people.

Eric Engbergwrote:

You are depending on 2 false assumptions:

1. That you can accurately measure the difficulty and assign a grade.  There is little about a route that you can quantify and measure. 

2. That one grade would apply for all people.

Exactly, sandbag/soft... that's like, your opinion, man.

Grades aren't facts, they're subjective and no one will ever be able to measure them accurately because there is nothing concrete to measure. 

yes

Princess Puppy Lovrwrote:

[...]

 There is a lot of math I have learned to quantify the “unquantifiable.” Amazon figures out if you are gonna need to buy diapers, which diapers to show you, where the diapers should display, how long to display a diaper ad, and what else to display. This is just figuring out the route grade, if anything is not that complex.

I guess the analogy to this is some sort of algorithm to figure out the "right" grade for an individual person for an individual route based on their strengths/weaknesses and familiarity with the particular rock/style. For example I'm much better at technical vertical granite than steep limestone, while I bet those who climb a lot in a gym's lead cave have the opposite experience. If people more consistently ticked routes on MP saying how many tries it took before they sent a route, it would be a fun little regression task to figure out the likelihood of a given person sending a specific route. 

Sometimes I think it would be interesting to not necessarily re-grade things but add additional attributes to go along with the grade of the climb.    For example, we already have danger ratings (PG13,R,X etc) and some grades take the danger into account (ice climbing, the E scale), but what if we had more quantitative factors?

We could have attributes for how sustained a climb is, the angle of the wall, if there were slopers, if there were crimps, how big are the edges in mm, what's the maximum reach between holds, what are the feet like, are there pockets, knee bars, a crack?      I think there are more quantifiable attributes you could associate directly with a grade that would provide a clearer picture around difficulty. Pair that with data from the climber (e.g. hand size, reach, etc) and the implied effort and skill associated with an ascent might be more clear.   

That being said, I think adding all those attributes would take a lot of fun out of things...

This is slowly turning into that thread about how climbing needs to be more like golf.

Not Hobo Greg wrote:

You’re describing a guide book. 

RockFax

Yes, let's take this new accurate grading system. All we need is more data. How many subjective measurements does it take till they become objective? I feel like 7. Seven subjective measurements makes them objective measurements. If six more people agree with me, that proves we're right!

Also, we should make the new system a ratio scale. 

Jordan Wilsonwrote:

This is slowly turning into that thread about how climbing needs to be more like golf.

All those bags of sand would certainly be handy for topping up the bunkers.

B Gwrote:

We already have this with MP consensus grades versus more static guidebook grades. Wisdom of the masses versus expert opinion

I personally stick with the guidebook grade, especially if it was graded before the 1990s or 2000s. I'm not a good/old enough climber to be considered crusty, but grade creep is lame. 

On the flipside, I'm fine with consensus 'sandbagging' like this route that went from 5.12a by the FA party to a consensus 5.11b: mountainproject.com/route/1…

most people didn’t  respond to the prompt. We are assuming some ai system was close to perfect, would you use it?

Princess Puppy Lovrwrote:

Should sandbags/soft climbs exist?

Yes. 

Princess Puppy Lovr wrote:

most people didn’t  respond to the prompt. We are assuming some ai system was close to perfect, would you use it

I would guess no one responded to the AI part of the post because it isn't something anyone is interested in. But arguing about sandbags or soft climbs existing is interesting. 

Getting sandbagged is fun. And I'm tired of pretending it's not.

I remember getting to Yosemite and feeling like the grades were stiff, and realizing it's the 'Yosemite Decimal System'; everywhere I had been climbing was just kinda soft. 

Just (redacted) all the grades. Problem solved.

H.

A quick google search, known for its unquestionable accuracy, says there are about 2 million established routes on earth and about 35 million climbers (around 2015).

To achieve a margin of error per-opinion of 5% or less and a confidence interval of 95% this means one person needs a good 380-420 routes under their belt, of varying regions and styles, for their opinion of a routes to meet this statistical standard. That isn’t actually too hard to achieve, lots of people climb over 500 routes in their life.

Find about 400 of these people to climb a single route and grade it, and their aggregate grade would actually also be within a 5% margin of error and 95% confidence interval.

This is also a reasonable number, lots of routes have had over 400 climbs by people with over 380 climbs.

HOWEVER; the uncontrollable variable is what those hypothetical 400 people would grade a given route.  That standard deviation is very large and really can’t be mitigated.  These subjectivities are the cause for the wide interpretation of climbing grades.  The system/method of experience-based initial grading followed by consensus grading itself is actually not that bad of a system, mathematically.  Far more important things are decided by far less valid sample analysis than climbing grades.