Display Consensus rating versus initial post rating

Routes do change over time and when holds break or get worn down the grade could go up or down. Just display both 5.8 (5.7) one for FA or original grade and one for consensus, heck you could even have an account setting to set which one you personally want displayed.

Problems I see are grade creep. Old routes are graded stiffer than new ones. Another problem is routes are subjective by area. Most Red Rocks routes are soft. Many Yosemite cracks are stiff. Many Gunks roofs are stiff. So for example, a Gunks climber, like myself, would consider a Red Rocks roof (e.g. Solar Flare, 5.10c) to be a 5.9 dragging down the rating. Someone who doesn't frequently climb routes may find that to be a sandbag. When I put up a FA I usually have a few of my friends onsite it before I give it a grade.

A new route given a grade by the first ascentionist will always differ some from the consensus grades. Rock breakage, grit on holds, and differing crux sequences will always play a part. Perhaps a note should be made as to the "age" of a route, and subsequent ascents can judge accordingly..

For established routes that can be found in a guide there already is a consensus. Authors get input from local and frequently visiting climbers who tend to have some experience under their belt.

A potential problem I see with displaying the MP consensus rating is that I think it will tend toward grade bloating. I don't think all opinions have equal weight based on experience, skill, etc. In the MP consensus context unfortunately they do.

Another issue is that similar to online product or service customer reviews many people will only suggest a grade when they disagree with it. As a result, the "consensus" can get skewed.

Lastly, the consensus rating is already displayed a mere two lines below the listed rating.

psakievich wrote:people would be more likely to leave their own rating if they felt like it was making more of an impact.

+1 for this. Great point.

Nah, do yall know how much extra work/processing it would take to get the consensus to show up everywhere? Lets say you want to display a list of 10 routes and their grades on the page. Right now the server just has to go to the database, pull back all of the routes including the first suggested grade and populate them. Instead, you're proposing looping through each retrieved route, finding all suggested grades, and returning the average. And then do that everywhere on the site where grades are displayed. That's a lot of work for not a lot of benefit, especially when it's pretty easy to see the consensus grade on the route page itself.

Drew Hayes wrote:Nah, do yall know how much extra work/processing it would take to get the consensus to show up everywhere? Lets say you want to display a list of 10 routes and their grades on the page. Right now the server just has to go to the database, pull back all of the routes including the first suggested grade and populate them. Instead, you're proposing looping through each retrieved route, finding all suggested grades, and returning the average. And then do that everywhere on the site where grades are displayed. That's a lot of work for not a lot of benefit, especially when it's pretty easy to see the consensus grade on the route page itself.

It's only "a lot of work" if you calculate the consensus grade on every page load. If you calculate a new consensus grade only when someone supplies a new rating and then store the new grade in the database, it takes very little more work than present (or perhaps less work or perhaps exactly the same, but not a lot more).

Nathan D Johnson wrote: It's only "a lot of work" if you calculate the consensus grade on every page load. If you calculate a new consensus grade only when someone supplies a new rating and then store the new grade in the database, it takes very little more work than present (or perhaps less work or perhaps exactly the same, but not a lot more).

That's poor design. A SQL view that calculates average on the fly would be the way to go. Still a lot of work for not a lot of benefit.

Is this really an issue? Are there lots of routes that have a grade that is significantly different than the consensus grade? I just looked at i. High E in the Gunks. Of the 26 routes, only 4 were different at all.

Route name : Original Grade : Consensus Grade
Ridicullissima : 5.10C : 5.10+
Lichen Forty Winks : 5.8- : 5.7+
Ursula : 5.5 : 5.6
Space Invaders : 5.10d : 5.10+

Can someone name a route where there are more than ~5 people who graded the route where the original grade is significantly different than the consensus grade?

Drew Hayes wrote: That's poor design. A SQL view that calculates average on the fly would be the way to go. Still a lot of work for not a lot of benefit.

Presently, do you think the consensus rating is calculated every time a route page is loaded?

Drew Hayes wrote: That's poor design. A SQL view that calculates average on the fly would be the way to go. Still a lot of work for not a lot of benefit.

Can't imagine how this would be true. Regardless, in MP's case, the consensus is recalculated any time a new rating is added to a given route, and then that consensus is stored as a separate field.

So with no technological excuse... it's definitely up for debate, and has been discussed many times, but it's never become an obviously better choice.

I would agree that most routes don't vary much from original to consensus, but an example of no agreement is Honey bucket at Pipeline in Maple. It is listed as 12a in the DK guide, I posted it at 11b.
Here are the consensus ratings. The problems with consensus is lots of people want the higher grade.

Tryhard 5.12a Oliver Chong 5.12a
kennoyce 5.12a Tyson Taylor 5.12a
Mark Lewis 5.12a Chris Horton 5.11d
Jesse James 5.11d David Mitchell 5.11+
Sam Perkins 5.11+ Brejcha 5.11c
wasatch boy 5.11c Doug K 5.11c
CHopwood 5.11c Derrick W 5.11c
dnoB ekiM 5.11c Luke Bertelsen 5.11c
Thomas Holmes 5.11c Danie White 5.11c
Jeremy Steck 5.11b/c Jake Croft 5.11b/c
Spencer Weiler 5.11b/c Emerson Takahashi 5.11
Skat B 5.11 Brennan Crellin 5.11b
SMR 5.11b bheller 5.11a/b
Murphski 5.11a/b KipHenrie 5.12a

Nathan D Johnson wrote: Presently, do you think the consensus rating is calculated every time a route page is loaded?

That's how I would've built it.

Nick Wilder wrote: Can't imagine how this would be true. Regardless, in MP's case, the consensus is recalculated any time a new rating is added to a given route, and then that consensus is stored as a separate field. So with no technological excuse... it's definitely up for debate, and has been discussed many times, but it's never become an obviously better choice.

I stand corrected! I've always been a fan of calculating on the fly rather than storing calculated values, but I think it's been a habit from building applications with sensitive data. Don't rely on a calculated column (what if the transaction failed? what if the calculation needs to change from arithmetic mean to geometric mean?) - calculate what you need when you need it. I guess it's overkill for this case though!

Jon Clark wrote:For established routes that can be found in a guide there already is a consensus. Authors get input from local and frequently visiting climbers who tend to have some experience under their belt. A potential problem I see with displaying the MP consensus rating is that I think it will tend toward grade bloating. I don't think all opinions have equal weight based on experience, skill, etc. In the MP consensus context unfortunately they do. Another issue is that similar to online product or service customer reviews many people will only suggest a grade when they disagree with it. As a result, the "consensus" can get skewed. Lastly, the consensus rating is already displayed a mere two lines below the listed rating.

+1.

I agree with OP.

Too many "original" ratings are taken from guidebook authors, who as often as not, have not climbed the route, and consulted with maybe a few of their friends. Add the fact that routes change..things break, etc and the historical rating can be meaningless. Grit Roof in JT is an example, the lip broke off, which is the crux, so it's not even the same climb in many respects. Add typos in printed material and it's even more reason for consensus...I remember BITD "accidentally" sending my first 5.11 trad pitch , because the print guidebook rated it 10a as a typo. I thought it was pretty damn hard for 10a, until I spoke with the author who told me "oh yeah, that was a misprint, it's 11a".

I've also seen "initial post ratings" where the person entering it intentionally sandbagged the rating, or used something way out of line with even an "original" rating.

I think it would be nice and maybe even have the consensus calculate only the number of ratings in the past 2 years or less because routes do change over time.

I remember after failing to climb a 5.10 at red river one day I was back at miguel's talking with a local who was saying the route was alot harder now due to one of the overhanging jugs being worn down to a slopper.

SMR wrote:I would agree that most routes don't vary much from original to consensus, but an example of no agreement is Honey bucket at Pipeline in Maple. It is listed as 12a in the DK guide, I posted it at 11b. .

This is a great example as to why consensus makes more sense than the current method. People like to refer to an "original" rating but that is not what MP displays. It just displays whatever the person who posted it put in. Some initial posters use the guidebook rating (which people put WAY too much faith in as being a consensus versus solely the author's opinion) and some use their own opinion.

In the case of Honey Bucket, DK fluffed {probably not intentionally} the rating of this super fun route up big time in the guide.
Clearly DK did not get a consensus on this before calling it 12a in the guide. DK's rating on this disproves the theory that the "original" or FA's rating is somehow more relevant than consensus.

SMR noted this crazy miss-rating in the guide and gave his own opinion as the sacred "original" rating in MP.com. If someone else posted it, they would likely have listed it as 12a. This shows exactly why consensus makes more sense. MP.com's current method gives the initial poster all the power/responsibility...as a wiki style information source...consensus yields a better result. {edit...after writing the above...I took a look and Honey Bucket is actually listed as 12A in MP.com...ha!...this actually adds further to the case for consensus...the "initial" rating is completely arbitrary based on who ever post it first or...apparently repost or has it edited...}

{none of the above is meant to say anything negative of DK or SMR; the route is rad and I do think 11b is closer than is 12a; however, I know rating one's FA is hard}

And as a comment to SMR, I think your example further supports the consensus benefit as most people do not take the higher grade...look at those entries...most people downrate Honey Bucket (as they do it's softy neighbor...Neptune Cocktail).

One more vote for displaying consensus rating as default (especially if there are more than say 5 votes). And I think it would be useful if you could post a guidebook/FA rating, and your own opinion when a route is posted.

I'd like to see the consensus rating too. But is it possible to have both?

I feel like all the fluctuations in a routes grade may not be so much a route changing in difficulty (due to holds breaking or whatever) but how it feels for different body types and skill sets. In this case it doesn't really matter knowing what someone else thought the grade of a route was if they are 6'3" and you're 5'6"just by looking at the ticks. One hold breaking probably won't bump a route up (or down) in difficulty more than one letter but having an extra 6" wingspan, or being really good on your feet can dramatically change how hard a climb is. I would be more psyched with an option to upload pictures directly from your phone to make it easier to tell if an hour approach straight up hill is gonna be worth it or not. And maybe adding a fifth star so the quality can get a little more specific :) grades suck anyway and everything is hard