[cheque]

I have no opinion on Egrades but it is brought to you by "Reyt" and can I just say I hate that fucking deliberately faux-spelling of words in a Sheffield accent that you see around these days. Get in the bin on that affront alone.

  About the only thing they could have done to make it less appealing to the average climber would have been to sell it as an NFT.

[Bradders]

Give it time.

[edshakey]

https://www.outside.io/

Close enough for you?

[Nemo]

If you're easily bored, I suggest to look away now.
I gave all this stuff a fair bit of thought a long time ago, when I was vaguely involved in documenting climbing in various ways.
A lot of the following is cut and pasted from things I wrote over a decade ago, but I've not changed my thoughts much since then.

Agree with almost everything Duncan said.
Disagree with most of what Shark said (sorry Shark!).

I've always thought that what we already have - E grades and UK tech grades - is the best grading system for trad routes in the UK up to around E6.
And a combination of E grades and either French or Font grades (for highball type routes) is the best grading system for trad routes in the UK roughly over E6.
(In reality many trad guidbooks in the UK have simply started appending French / Font grades to hard trad routes and this works fine.  Although it does make the UK tech grades of 6c and 7a that those routes typically still get assigned even more pointless than they were in the first place).

So, sure for high level trad routes, the tech grade needs replacing with French / Font grades.
What emphatically does not need changing is the way the E grade works.

To point out the blindingly obvious, as Andy said, all the eGrader system really is - is a simplistic translation of a US style grade to an E grade (which may as well be called an F grade, as it's not an E grade as we currently know it).
ie: it's a French grade plus a danger grade, slightly adjusted around the use of pads - kicking out a calculated E Grade.
In other words, disregarding pads on highballs, the calculated E grade is giving precisely zero additional information to the US system which was the starting point.

This calculation bypasses the entire point of E grades and keeps all the flaws of the US system.  And the US system is a long way from ideal for the vast majority of UK trad climbing.  Of course, it works to some extent, but is in no way whatsoever the best way of grading UK trad routes (hard or otherwise). 

The eGrader also appears to be significantly over emphasising the danger element, which is one of the reasons why it is ballooning E grades at the high end when applied (particularly when applied to not very high roadside type crags).  It's very rare for routes to go up 3 E grades from what they would get for pure physical difficulty (Indian Face goes from E6 to E9, but that level of jump is really very rare).  E grades have significant width - it's pretty common for pretty hairy routes to get the same E grade or only 1 grade higher as they would for pure physical difficulty.

Any kind of algorithm for working out E grades (in their normal sense) is never going to work.
It is and should be the overall grade of the route.  Which takes into account a million details that algorithms can't handle.

Where I disagree with Shark, is that I think that E grades (in their normal sense) are extremely useful.  Both to people trying to work out what routes to climb, and to anyone trying to understand at the top end what is genuinely newsworthy.
And E grades are actually (or at least should be) incredibly simple, both for low grade routes and at the top end.
They answer very successfully one simple question.  How big a deal overall is an ascent of route X in any particular style. 

That is (along with the physical difficulty) ultimately both what aspirant ascentionists, and anyone interested in climbing news wants to know.
The US system does not tell you that. 

It doesn't take too much effort to see the problem with the US system.  Is the genuinely dangerous bit the crux at the top of the route, or on the VS bit after all the hard climbing.  Is there a sustained dangerous section, or is it just a couple of relatively easy moves at the end of a runout to get to some good gear.  Is the dangerous section on positive holds, or is it a sketchy smeary slab move where you're genuinely going to die.  Is the dangerous climbing at an easy standard but with some loose rock, particularly if you have to completely trust a foothold which has a decent chance of breaking.  Etc etc etc. 

There's so many other factors that need to be considered - loose rock, soft rock, insecureness or otherwise of climbing, location and atmosphere of crag, level of commitment required, level of intimidation etc etc.  Of course, the E grade on it's own doesn't tell you any of that detail either.  But what it does and always has done extremely effectively is answer that one simple question - overall, when you take everything into account, how big a deal is climbing this particular route, in any chosen style. 

e.g: If you put the main pitch of Positron on the ground at a roadside crag where you could see and hear your belayer, and didn't have hundreds of feet of exposure dropping into the sea to deal with it would probably be E3.  Where it actually is, on your own on that headwall, round the corner unable to see or hear your belayer?  Completely deserves the E5 it gets.  And it's not because it's particularly dangerous.  It's because of many other things.

And danger is a very hard to pin down thing clearly anyway.  The skill in getting good at trad climbing is to take something that more beginner climbers would find extremely dangerous, and become competent enough to climb it safely.  When ElMo or Caff are onsighting E7s, for the most part I don't think they are in huge amounts of danger.  I don't even think when Alex Honnold is soloing, that for the most part he's in huge amounts of danger.  They've developed over a long time, the skills to do what they are doing reasonably safely.  The E grade is essentially about what overall skill level you need to make the route a reasonably safe proposition (and that applies to the Indian Face as much as to Left Wall).

As an extreme example, take headpointing the Indian Face vs soloing Freerider.  With the US system the first would be 7b+ X, the second would be 7c X.  The first is something that's well within the capabililties (given sufficient work) of hundreds of climbers just in the UK.  The second probably isn't within the capabilities of anyone other than Alex Honnold.  You can't die more than once, so it's not that falling off soloing Freerider is really more dangerous than falling off high on the Indian Face.  It's that the skills required to make soloing Freerider even remotely safe are vastly harder to acquire than the skills required to headpoint the Indian Face.

Of course, no doubt the proponents of the eGrader will say that you can take account of all that by adjusting the danger component before putting it into the calculator.  But then you're really just coming up with a way of cheating the calculator to calculate what you know is actually the right E grade, in which case there was no point in using it in the first place.  ie: you're no longer actually talking about a danger factor.  You're talking about an everything other than physical difficulty factor.  Which is exactly what the original E grade system gives us if used in conjuntion with a French / Font grade for harder routes (as IMO it should be). 

All that said, there's clearly a fair bit of confusion out there amongst hard trad climbers doing first ascents.  But that really doesn't have to be the case, given a little thought.
UK tech grades are completely dead for high end routes.  Just use a French / Font grade.  No problem.
And for them to have any value, E grades need to work in the way they always have - simply compare (taking everything into account) this route with other routes you've climbed in a similar style (e.g in a day, or 10 day battle, or full blown multi year siege).  Of course at the highest level, this is tricky as there isn't that much to compare to, but that's just the same problem that boulderers and sport climbers have at the highest level.  Only once there's a significant number of routes / boulders at a particular level, and typically only when someone like Ondra comes along and climbs lots of them and compares them - do sensible grade boundaries emerge at the top end.  Again, it's no problem, just takes a while for a consensus to emerge, especially when some of these routes don't get a lot of ascents. 

[Nemo]

A better E Grader

If they'd avoided all the fanfare, not pretended it was using algorithms or doing anything clever, not attempted to apply it to anything under E5 where everything works perfectly fine already, this might have been better received.

IMO, a better eGrader is simply this and requires precisely 8 lines of text:

Proposed boundaries for totally safe climbs (ie: U.S P.G rating):

7a+ - 7b routes would be E5
7b+ - 7c routes would be E6
7c+ - 8a  routes would be E7 
8a+ - 8b routes would be E8
8b+ - 8c routes would be E9
8c+ - 9a routes would be E10
9a+ - 9b routes would be E11 (currently aren't any trad routes of this physical difficulty)
9b+ - 9c routes would be E12 (currently aren't any trad routes of this physical difficulty)

Traditionally E6 was a bit wider than that, and E7 a bit narrower - with things like Cave Route Right and Left at 7b+ and 7c+ getting E6, yet Requiem at 8a+ getting E8.  That did seem a bit daft though, so no problem with having things have a bit more of a consistent width. 

That's it.  It ain't rocket science.
But that is a useful little pocket guide for someone trying to grade a new hard route.
You then, obviously, add to those E grades depending on loads of other factors, of which, danger is one, and compare the overall difficulty with other pre existing routes.

It's worth pointing out that the above is a little different to where those boundaries have sometimes been placed, and it's a half a grade different than where the eGrader seems to be placing the boundaries.

It's also why I don't think E12 yet exists. 
Because I simply don't think there's anything in the trad climbing world that is the equivalent in terms of overall difficulty (ie: when you take everything into account, as you should with an E grade), as redpointing 9b+ and 9c.  If you're using E grades consistently, then giving something E12 essentially amounts to saying you think you're in the top 5 of route climbers worldwide.  Steve McClure was 15 years ago.  I don't think any of the other people involved are.

Whenever Steve McClure goes on any trad route in the UK, he typically does it in a day or so.  Clearly Steve has climbed 9a+ and 9b, but only after very prolonged sieges.  If he's climbing things in a day or two, I don't think it's unreasonably to suggest that overall, it's unlikely that they are a bigger deal than redpointing a 9a.  ie: Unless Steve is vastly better at trad climbing than he is at sport climbing (unlikely), then it would suggest that Lexicon and Rhapsody are hard E10.

In other words, rather than ballooning the E grading system, a little thought and consistency, and realising that E grades have always had significant width - suggests that at the top end, far from being undergraded, most things are about right already.  Perhaps I'm wrong and Echo Wall or Bon Voyage really will end up being considered as big a deal as redpointing upwards of 9b+.  Time will tell, but I'm skeptical that's really the case.

If people just asked a couple of simple questions of themselves after doing a new route, this stuff really wouldn't be that complicated.
Is the route you've just headpointed, overall as big a deal as redpointing a Fr8c+ sport route?  Then it is probably at least E10.  Is it as big a deal as redpointing a 9a+ sport route?  I'd suggest there's very few routes in that category.  Perhaps just Echo Wall, Bon Voyage and possibly the odd one or two others.  These probably deserve E11.  Given the level of the trad climbers involved at sport and bouldering, I find it hard to believe that there's anything yet in the trad world that's the equivalent of Fr9b+ sport routes.  The only thing would perhaps be Honnold's solo of Freerider - if that were actually a route, then perhaps that would be the first to justify E12 (it's the only thing I'm similarly impressed by as someone climbing upwards of 9b+).

So a little thought, and assigning E grades to things at all levels really isn't terribly difficult.  If that was what the eGrader was meant to achieve, then fine.  Not sure what all the fanfare and fuss was about though, for what could easily have been 8 lines in a notepad file.

[petejh]

Despite all the fancy talk of algorithms etc. it's basically just an old fashioned grade conversion chart that probably work just as well (i.e. not very) on paper.

The eGrader also doesn't involve any original thinking. It's just the below that I listed two years ago, in a thread about Gresham's ascent of Lexicon. I used 4 danger ratings, they've used 9. Not that I'm claiming any original thinking either as I was just repeating other's ideas that I'd heard over many years:

The process of grading trad routes shouldn't be that difficult.
...
How is the below so complicated to use as a guide:

For averagely safe routes
5 - 5+  E1
6a-6a+ E2 
6b-6b+ E3
6c-6c+ E4
7a-7a+ E5
7b-7b+ E6
7c-7c+ E7
8a-8a+ E8
8b-8b+ E9
8c-8c+ E10
9a-9a+ E11
9b-9b+ E12

Bold/scary/long runouts but not exceptionally dangerous  +1 E-grade
Exceptionally dangerous  +2 E-grades
Exceptionally safe  -1 E-grade

It's just a repackaged idea that's been around for years, a list of sport-with-danger variables-added. Repackaged by, as Shark notes, a certain climbing 'sleb' who seemingly needs to keep themselves in the public eye, and a consultancy business that calls itself reyt and whose strapline is -
''Elevating sports, outdoor, and wellbeing brands to deliver innovative digital solutions.''
 
Get in the nearest body of water.

 

[Nemo]

A few other tangential bits, from previous things I've written ages ago, which seem vaguely relevant:

Are E grades really for the onsight?
The one bit where I disagreed with Duncan is the widespread opinion that E grades are for an onsight ascent.  Of course, that's what everyone says, and it has elements of truth.  But in reality on 99% of routes E grades work just as well for headpoints, flashes, in a day ascents or onsights.  In other words, it's reasonably rare for a route to be massively harder or easier relative to other routes for any particular style.  Left Wall is roughly 3 grades easier to onsight than Right Wall but it would also be roughly 3 grades easier to headpoint than Right Wall.  The same would apply to comparing Silly Arete to Resurrection to Hunger.  Each is going to be one grade harder than the previous one regardless of whether you are onsighting or headpointing. 

The exceptions are usually when there's a particularly desperate to read sequence or some vital obscure gear which wouldn’t normally be carried.  Thus, this is largely an elitist issue as there are very few routes under E5, where crucial sequence beta, or vital obscure gear are a large problem – on easier routes there are nearly always alternative sequences or gear.  How should these routes be dealt with in guidebooks? 

Well there's a bit that everyone typically forgets in trad grade discussions in the UK.  In the UK, trad route guidebooks are usually excellent and have descriptions, not just a topo and a grade as tends to be the case in many places outside the UK.  What do guidebook writers do when there's vital obscure gear, or a crucial unlikely to onsight sequence?  They typically give away the crucial gear or beta in the description.

So in reality, routes are almost always graded for the easiest possible sequence and the best possible use of gear.  The skill involved in onsighting a route, is to use a sequence and gear as close to the optimal as possible.  In the very few cases where onsighters are likely to get into trouble because of very obscure beta (either gear or moves), the description in the guide should give an appropriate warning or give the relevant beta.  If any crucial gear or beta is given in the guide, then you can still pretend that you are “grading for the onsight” if you want, as long as any beta in the guide is considered fair game.  But in reality, you're not really.  (e.g: If someone set off on Master's Edge genuinely onsight without the crucial tricam, I'd suggest it would be a fair bit harder than E7 - routes like this are not really graded for a genuine onsight.)


Why the UK tech grade has died for hard routes and why that's a good thing:
As discussed at length on here and elsewhere previously, the problem with the old UK trad system wasn't the E grade bit.  It was the UK tech grade, that in reality died a long time ago for routes over about E6.  For these routes, replacing (or just adding to) the tech grade with a French grade (or highball Font grade for short routes) solves all the problems.  And many UK trad guidebooks have now realised this and include the French / Font grade for hard routes.

Some people have suggested that instead of doing this, we should extend the UK tech system. But there is a good reason why no one has bothered to do this. The majority of high level trad climbers talk about physical difficulty in terms of French / Font grades. They tend to have climbed lots and lots of Fr8as or Font 7Cs, but only a few routes of comparable physical difficulty as trad routes. Thus it is not surprising, that when it comes to grading a new trad route, that they usually turn to their vast experience of French or Font grades to compare the new route to – rather than trying to extend the UK tech grade system, where they would have done very few trad routes of similar physical difficulty to compare with.

At lower grades, typically the opposite is the case.  Lower grade trad climbers (particularly the older variety) often have way more experience of the UK tech grade than French grades.  And also for lower grade routes in the UK, the tech grade is much more specific than French grades with 4a, 4b, 4c, 5a, 5b, 5c covering roughly the same ground as French 4 – 5+. This is the grade range of the vast majority of UK trad routes – and the very specific nature of the grades is one reason why it is popular with lower grade climbers.


Historical aside:
Where E grade boundaries have been placed until now is something which has evolved over time.  And it has evolved in some pretty odd ways...  Those that have been around a while can think back to the late 80s and early 90s before magazines and guidebooks had started using sport grades for sport routes – remember how they used to be graded…

Obsession         E6 6c   (Now Fr7b+)
New Dawn           E6 6c   (Now Fr7c)
Dominatrix         E6 6c   (Now Fr7c)
Mescalito          E6 6c    (Now Fr7c+)
L’Ob Session      E6 6c   (Now Fr7c+)
Raindogs          E7 6c   (Now Fr8a)
Zoolook            E7 6c   (Now Fr8a)
Overnite Sensation   E8 6c   (Now Fr8a+)
Predator         E8 6c   (Now Fr8b)
Mecca            E8 6c   (Now Fr8b+)
Cry Freedom         E9 7a   (Now Fr8b+)
Liquid Ambar      E9 7a   (Now Fr8c+)
Hubble             E9 7b   (Now Fr8c+)

From that list it doesn’t take much to see the silliness of the UK tech grade for high end routes, but that's been covered already.
It's also clear that E6 and E8 were a lot wider than E7.

Back then there were probably as many sport routes using grades over E7 as there were trad routes, which has meant that the way the E grades were applied to sport routes has had lasting consequences on the way E grades have developed.  Over time as more and more trad routes over E7 have been done, I think things have smoothed out a little.  But I think, currently the state of play for safe trad routes still largely mirrors how the E grade boundaries were laid out for sport routes nearly two decades ago.
So one bit where I'd agree with the eGraders, is that a bit of tidying up to make things a bit more consistent widths isn't a terrible idea, as suggested above.

[andy moles]

I like the majority of what you've said Nemo, but not so sure about this:

And E grades are actually (or at least should be) incredibly simple, both for low grade routes and at the top end.
They answer very successfully one simple question.  How big a deal overall is an ascent of route X in any particular style. 

This suggests that there should never be any difference between how much of a 'big deal' a route is in an on-sight or a headpoint style respectively at its given E grade. I don't think that can be true. If, for example, a route is accurately graded for an on-sight, but has some crucial piece of hidden gear or non-obvious beta that an on-sighter is never going to find, it's likely to be a soft headpoint at that grade. I don't think on British rock this is especially unusual. Another example I guess would be a long but just-about-safe runout. To me that's so much easier to deal with as a headpoint, knowing and possibly having tested the fall, compared to questing close to the limit of safety on-sight.

Edit: I see you've addressed this point in a later post. Your point is fair that it relates mostly only to harder routes, but given it's harder routes that have ignited this debate in the first place, it's relevant. It's not always possible to put the necessary detail of these things across in a description, they're going to require inspection.

[andy moles]

.

[Fultonius]

Somewhat agree with Andy about (in my experience at around the E4-E6 category) that some routes do drop more for the headpoint/repeat than others. Great example is Chemin de Fer at Dumby and Lady Charlotte at Upper Cave. Both E5.

CdF is still nails when you know what to do, as it's sustained and pump; lady charlotte on the other hand is pretty mellow on a top-rope but hard to read, hard to protect and generally a bit scary on the onsight.

Otherwise I pretty much agree with most of what Nemo has said. If I had the choice I would re-scale the tech grades - if you just adopted Font grades, it wouldn't change many routes much at all (max 1 grade) in the VS - E2 category and it could just keep going nicely in the upper grades.

[Kingy]

I agree with Nemo's assessment and don't have the hard trad background to comment further.

One small thing, Raindogs is E7 6b in my 1990 Rockfax  It was always contrasted with Zoolook in that the moves were easier than the crux of that route (which was English 6c) but that it was very sustained at 6b most of the way with no rest whereas Zoolook was furnished with a decent recent not far below the crux   

[ferret]

The exceptions are usually when there's a particularly desperate to read sequence or some vital obscure gear which wouldn’t normally be carried.

That's fine for safe routes but for things with genuine danger I would say there's a big difference between onsighting something with dynamic moves vs static or smears, slopers, pebbles etc vs positive/incut holds.
If you are genuinely grading for an onsight I would say these factors could bump the grade in the right circumstances and perhaps also how exposed something feels.
There's routes of all grades that aren't relatively popular to onsight for these reasons.

[shark]

Hi Nemo 

A great set of posts there. 

Perhaps the E grade shouldn’t be thrown out with the bath water for the very top grades 

To summarise/paraphrase your position you start with the premise that the French grade is generally felt to have established a ‘linear progression’ with a consistent bandwidth of difficulty for each grade and couple that with bolted/safe routes at each E grade occupying two French grades (historical niggles at E6 aside). Using this base info the E grade can be consistently extended as far as the cutting edge climbers take it by adding risk and other factors accordingly with a limit of 3 E grades ie 7b+ can be an E9 if it is something like Indian Face.

So as you say using these premises Lexicon is E10 because at 8b if it was bolted it would be highish E8 but as it is on good rock and demonstrably possible to fall off without dying then bumping it up 2 E grades seems right. Sounds plausible.

Perhaps we are now allowed to say this as the wads have abdicated their cliquey consensus power to a democratic algorithms so we should all feel empowered to exercise our new computational rights to make that sort of call   

I’m still niggled by one thing though.

But in reality on 99% of routes E grades work just as well for headpoints, flashes, in a day ascents or onsights.  In other words, it's reasonably rare for a route to be massively harder or easier relative to other routes for any particular style.  Left Wall is roughly 3 grades easier to onsight than Right Wall but it would also be roughly 3 grades easier to headpoint than Right Wall.  The same would apply to comparing Silly Arete to Resurrection to Hunger.  Each is going to be one grade harder than the previous one regardless of whether you are onsighting or headpointing

Is this actually true? I’ve very little experience of headpointing so can’t really comment very factually but if I compare it to soloing at Stanage for example the grades go out the window where an HVS might be harder than an E2. Not the greatest analogy but I’m sure you get my drift. Perhaps those who regularly onsight and headpoint might comment? ie 1. whether an E grade is consistent when measured against other headpoint routes and 2. Superfluous if a French grade is provided as you can safely establish the other factors when working it

[remus]

A better E Grader

If they'd avoided all the fanfare, not pretended it was using algorithms or doing anything clever, not attempted to apply it to anything under E5 where everything works perfectly fine already, this might have been better received.

IMO, a better eGrader is simply this and requires precisely 8 lines of text:

Proposed boundaries for totally safe climbs (ie: U.S P.G rating):

7a+ - 7b routes would be E5
7b+ - 7c routes would be E6
7c+ - 8a  routes would be E7 
8a+ - 8b routes would be E8
8b+ - 8c routes would be E9
8c+ - 9a routes would be E10
9a+ - 9b routes would be E11 (currently aren't any trad routes of this physical difficulty)
9b+ - 9c routes would be E12 (currently aren't any trad routes of this physical difficulty)

What's the basis for squeezing 2 sport grades in to one trad grade? As far as Im aware there isn't a popular opinion that french sport grades are overly narrow, so seems strange to effectively double the range.

For example, it would make a big difference to me if I was getting on a safe E6 and it was 7b+ (for me there's ~50% chance of onsighting it) or 7c (~10% chance of onsighting it).

[shark]

It’s not what Nemo has declared but an observation on the bandwidth of the E grade which is consistent with that short period of when bolted routes where given E grades in the mags and guides. So yes the bands of physical difficulty are that wide for safe routes (and always have been) in each E grade which is another shortcoming of them.

[andy moles]

1. whether an E grade is consistent when measured against other headpoint routes and 2. Superfluous if a French grade is provided as you can safely establish the other factors when working it

I've given my tuppence already, but to put it another way: this basically claims that there is no difference between routes of a given redpoint standard in terms of ease of reading the rock or of various psychological factors on first encounter.

To me this just isn't true. Take ten routes of different styles, rocktypes and situations, that would have the same input on e-grader, say 7c and run-out. Assume they aren't chalked. Even for a really rounded on-sighter like Caff, is he likely to find them as close to equal in difficulty as they would feel when practiced? I really doubt it. Some things are physical but relatively obvious, some are hard to read or knacky.

[andy moles]

What's the basis for squeezing 2 sport grades in to one trad grade?

Hasn't the e-grader done exactly that too?

If you keep the danger point constant and increase the sport grade, there are two sport grades for each E grade (easy and hard).

The only slight difference is whether you distinguish between shades of 'very safe'.

[Teaboy]

If the two “metrics” of French grade and danger give an accurate indication of the route’s difficulty why obsfuscate that information behind a single grade that gives less detail, why not make 8b, danger 2, or whatever, the grade?

E grade makes sense if you are on-sighting as these were meant to encompass a third subjective component, the “feel” of the route, made up of many different factors that are largely eliminated on a head point but they do exist and can make a difference. Mostly they don’t make too much of a difference which is both why the eGrader works but also why it doesn’t fix the edge cases like it was intended to!
Also, one of the proponents of the eGrader said it was supposed to address the obvious inconsistencies in the higher grades. If there were obvious inconsistencies why not just say what that are and regrade? Surely a bit of beef and armchair downgrading of each other’s routes would be more fun than this?

[andy popp]

Surely a bit of beef and armchair downgrading of each other’s routes would be more fun than this?

That's the true British trad spirit.

[abarro81]

RE onsight Vs headpoint, I can't comment on trad but for sport the gap in 'onsightability' between routes can be enormous. Like 2-3 grades. Onsightable 8b can certainly be easier to onsight than 'unonsightable' 8a. Ditto 8a vs 7c etc. From way back when I did a bit of trad I imagine this to be even more true in trad since you have the same factors as sport and the added issue if how hard gear is to find and place (without getting too pumped)...

[mrjonathanr]

Disclaimer: I don’t climb these hard trad grades, but do have some thoughts about what’s been proposed.

Steve McClure takes a logical stance: "The basic point is, a mathematical model can't be argued with. Because a mathematical model takes a large amount of data, and for a grading system ideally all of the routes will lie on a linear graph, precisely because grading is linear!

Except that anyone who lived through the pandemic (remember that?) and the endless debates about the accuracy of the mathematical models Imperial and Sage churned out, will remember that it isn’t the reliability of mathematics that’s in question, but the assumptions of the model makers. In other words, as we all know, there’s lies, damned lies, and statistics algorithms.

Seems like the algorithm has the effect of deflecting debate about the judgment of the first ascentionist/grader onto a supposedly ‘objective’ grade, despite the fact that the only thing that’s remotely objective is the pitch length if you ab down with a very long tape measure. Every other parameter is a matter of judgment and therefore subjective, surely?

It seems to me that when Ste M goes on to talk about why outliers should be debated and adjusted or not, that’s exactly how the E grade works now, holistically.

[abarro81]

That quote has to be a joke right? 

[petejh]

Perhaps he meant ''a mathematical model that has what turn out in retrospect to be perfectly accurate assumptions that perfectly match everybody's experienced reality, cannot be argued with''. And was just using shorthand because he thought we all knew that's what he meant really...

What struck me about reading the ukc back n forth about 'linear' versus 'normal distribution' is that you could flip the object of the grading system to not grade 'climbs' but instead use it to grade 'climbers'. So that a route would be graded according to where it sat on a standard distribution bell curve of number of ascents. In this world, a route with only 1 ascent would be harder than a route with 10 ascents until it received 11 ascents. It's a stupid idea and would never work, but really that's what a difficulty grade should do - tell you how many people in a population could get up a climb, if everyone in that population tried it. 

[mrjonathanr]

Sorry for the rant, but no, it’s a direct quote from the article, ~ 10th paragraph down under the sub heading Project Concept
https://www.ukclimbing.com/articles/opinions/the_egrader_calculator_-_rebooting_e-grades_with_linear_consistency-15123

The full quote is:

Steve McClure takes a logical stance: "The basic point is, a mathematical model can't be argued with. Because a mathematical model takes a large amount of data, and for a grading system ideally all of the routes will lie on a linear graph, precisely because grading is linear! The eGrader converter has used a large number of known routes with widely agreed grades to create the linear graph - and it works. Routes that now lie outside the linear line can be discussed as to why they lie outside the linear line. [If] there is no reason for them to be outside the line, then the grade needs to be changed (up or down)."

[mrjonathanr]

What struck me about reading the ukc back n forth about 'linear' versus 'normal distribution' is that you could flip the object of the grading system to not grade 'climbs' but instead use it to grade 'climbers'.

You may be onto something here…