[Muenchener]

Awesome I'm the very bottom of one of those little sticks!  Does this mean I should do more pull ups?!

Me too!

Whereas pull-up strength is one thing I don't have to worry about to get from bouldering 6B to 7B.

I probably need to work on my bench pressing instead.

[moose]

Years ago, I used to do sets of 30+ pull-ups off a single-joint deep door frame edge... I could boulder around f6c+. 

I then moved to Yorkshire, I couldn't do any pull-ups (parents' house: weak door-frames and no chin-up bar) but I was unemployed and spent a year or two solidly grit bouldering - did lots of f7b+'s and the odd soft 7c.  After that spell, I had a go on a chin-up bar and found I could only manage around 10 pull-ups. By my understanding of science, if I chop off my arms and get down to zero pull-ups, I'll have a bouldering grade of font-infinity!

[slackline]

Years ago, I used to do sets of 30+ pull-ups off a single-joint deep door frame edge... I could boulder around f6c+. 

I then moved to Yorkshire, I couldn't do any pull-ups (parents' house: weak door-frames and no chin-up bar) but I was unemployed and spent a year or two solidly grit bouldering - did lots of f7b+'s and the odd soft 7c.  After that spell, I had a go on a chin-up bar and found I could only manage around 10 pull-ups. By my understanding of science, if I chop off my arms and get down to zero pull-ups, I'll have a bouldering grade of font-infinity!

presumably you mean negative infinity?

see the text above that precedes the figure about how correlation can be misleading.

[moose]

Positive font-infinity based on my own two, longitudinal data points (30+ chinups = f6c+; 10 chinups = f7b+/c)!  And yes, I am aware that incomplete statistics are misleading and, even where the numbers are accurate, correlation doesn't determine causation.... that was kind of the point I was making, albeit in a ham-fisted way.  All other things being equal, more arm and finger strength is good, but if technique / core are terrible they'll be the limiting factor. 

I wonder if only studying people who've climbed a good few years makes any difference to the results.  If limiting the data set to those for whom "rookie" technique is unlikely to be limiting, makes strength based correlations clearer?

[Sasquatch]

Moose may have a good point, and # of years climbing may be a good addition to the survey (If its not there now, I can't remember).

[petejh]

Slackers have you ever thought about turning your statistical-skills to fixing the broken trad grading system?

[slackline]

Positive font-infinity based on my own two, longitudinal data points (30+ chinups = f6c+; 10 chinups = f7b+/c)!  And yes, I am aware that incomplete statistics are misleading and, even where the numbers are accurate, correlation doesn't determine causation.... that was kind of the point I was making, albeit in a ham-fisted way.  All other things being equal, more arm and finger strength is good, but if technique / core are terrible they'll be the limiting factor. 

I'm crap at picking up sarcasm on the net.

I've absolutely no idea how to capture or quantify technique accurately, but figured the distinction between on-sight/red-point sports routes might be one method.  My reasoning being  that technique is of far greater utility to on-sighting and the strength of any relationship between feats of training strength would be weaker when that is the outcome.

I wonder if only studying people who've climbed a good few years makes any difference to the results.  If limiting the data set to those for whom "rookie" technique is unlikely to be limiting, makes strength based correlations clearer?

Moose may have a good point, and # of years climbing may be a good addition to the survey (If its not there now, I can't remember).

Duly noted and will be incorporated into the summary and where to take things next I'm compiling, cheers.

Could be confounded by indoors/outdoors though where the above "technique" is better learnt outdoors, yet many people these days start indoors and tend to develop strength over technique.

[tomtom]

Time climbing is a red herring IMHO, I've been climbing for 20 years (a week ago!) but half of that time has been bugger all/no climbing..

The biggest proxy for technique is probably average highest bouldering grade.. Which I suspect is more technique than strength controlled...

[moose]

The way this is going, it'll soon be demanding more computational time than SETI and the Human Genome Project combined. 

By the way did you ever see the height versus sport grade study that someone on 8a produced.  The analysis method was rather odd, as I recall, the average height for "top" climbers was around 173cm and the average increased as more non-"top" climbers were included.  However, this suggested that the most elite group of climbers would be a tribe of pygmies..... that said, I am sure we all remember the rapid rise in standards that accompanied the cast of Time Bandits visiting Malham.

[slackline]

I guess it depends on how honest people are and how the question is phrased....

When did you start climbing?   <--- Could easily lead to mis-leading results.

v's

How long have you been training/climbing in your current cycle?  <--- Might be more appropriate.
How long have you been training/climbing without a significant break?  <--- Might be more appropriate.


Its a major drawback of these sorts of "epidemiological" surveys, you can't control things*.  Which is why people do experiments where they can control things (viz. everyone hanging on the same hold and timed as discussed above).



* You can't even control how people respond to questions...

• One person has entered their height in centimetres as 1.7 when presumably they meant 170.
• A number of people seem able to do weighted pull-ups with between x1.5-2 times their bodyweight which makes me suspicious that this is perhaps their body weight + additional weight.
• Others include the units when there is no need because they've been stated in the question and it only means the data needs purging of these so that variables are numeric and not forced into strings by the presence of "20 secs"
• Some people have entered V-grades when Font grades have been requested.

...all of which needs 'cleaning' before you have a dataset you can use.  This isn't a problem though, and highlighting these shouldn't discourage or prevent people from completing the survey, the more complete data the better.

[abarro81]

  My reasoning being  that technique is of far greater utility to on-sighting and the strength of any relationship between feats of training strength would be weaker when that is the outcome.

Actually I find that my onsighting relies just has heavily on my raw power as my redpointing, probably even more so - when redpointing I can trick my up stuff more by finding sneaky beta whereas on onsights there's more reliance on being overstrong and being able to pull through moves in inefficient ways.

[slackline]

After having written that I gave it more consideration and thought it might be the reverse situation you describe, but then countered myself thinking perhaps technique is just as important when on-sighting as it means you are more efficient in your climbing meaning you don't pump out hanging around placing gear etc.

I don't think there will ever be an easy way to quantify someones technique.

[Grubes]

Slackers do you have anymore graphs for us showing their is or isn't a coloration between two answers?



I like graphs 

[slackline]

I've the same graphs for each of the questions for each of the bouldering/route grades but am holding back on putting anything else up as I'm writing it up as a PDF report (when I get round to having time to spend on coding the analysis and writing the text to go with the figures).

I'll put everything up on the wiki when done and link it from here, including a copy of the cleaned data (anonymised by the removal of names that people have entered) along with the scripts I write should anyone wish to do any further work.

[tomtom]

I've the same graphs for each of the questions for each of the bouldering/route grades but am holding back on putting anything else up as I'm writing it up as a PDF report (when I get round to having time to spend on coding the analysis and writing the text to go with the figures).

I'll put everything up on the wiki when done and link it from here, including a copy of the cleaned data (anonymised by the removal of names that people have entered) along with the scripts I write should anyone wish to do any further work.

Any chance of putting the anonymised data (give em names A, B c etc..) on a google docs spreadsheet we can copy from?

[slackline]

I've the same graphs for each of the questions for each of the bouldering/route grades but am holding back on putting anything else up as I'm writing it up as a PDF report (when I get round to having time to spend on coding the analysis and writing the text to go with the figures).

I'll put everything up on the wiki when done and link it from here, including a copy of the cleaned data (anonymised by the removal of names that people have entered) along with the scripts I write should anyone wish to do any further work.

Any chance of putting the anonymised data (give em names A, B c etc..) on a google docs spreadsheet we can copy from?

That (or similar, most probably a CSV file) was the intention.

Or would you like a raw uncleaned copy now?

The data set does keep on getting new entries, currently upto 80 entries now, and I'm a bit lazy to keep on reposting it whenever its updated and would rather share it when my analysis is complete (although the survey will remain open and it will be straight-forward to re-run things at a later date).

[tomtom]

Maybe wait until its been up a week?

[slackline]

Maybe wait until its been up a week?

Probably a good idea rather than waiting for me to get my arse in gear!

[slackline]

Well, I've not had masses of time to spend on summarising results, but as I'm away this weekend, busy with work next week, then off to Spain the week after for some bolt clipping I thought I should share what I have done so far.

I've added a section to the Training : The Science page of the wiki that links to the PDF report ukbenchmarking.pdf that I've written.  I did a short literature search and have hopefully summarised some of the papers I found to give some background on similar work that has already been done in this area.  There is then a description of the summaries and the intended analysis (its the analysis that I'm yet to get round to doing).  Lots of pretty graphs are then presented, there's an empty section on the modelling results and a virtually non-existant discussion and future work before the references that are cited are listed.

All analyses were done using R, the graphs were created with the R package ggplot2 and the PDF was written and prepared in LaTeX.

I would warn against over-interpreting these graphs, they're showing the spread of things at the moment, the planned statistical analysis will then quantify if any of the patterns seen are of any significance.  I've tried to include links to web-pages in the PDF that describe the graphs and statistics that are intended to be used (likely send most people to sleep, but they're there for those who are interested).

I've uploaded all of the graphs I've generated so far to a gallery and added brief titles and descriptions (hopefully I've got them all right, but if in doubt they are correct in the PDF, labels on imgur or below are wrong if there is any discrepancy).


For simplicity they're embedded below....

Weight and Height



Boulder Grades...


Bench-Press by Boulder (Outdoors)


Bench-Press by Boulder (Indoors)


Dead-Lift by Boulder (Outdoors)


Dead-Lift by Boulder (Indoors)


Pull-ups by Boulder (Outdoors)


Pull-ups by Boulder (Indoors)


Hang-time by Boulder (Outdoors)


Hang-time by Boulder (Indoors)


Hang-size by Boulder (Outdoors)


Hang-size by Boulder (Indoors)



Route-grades climbed....


Bench-Press by Route (On-Sight)


Bench-Press by Route (Red-Point)


Dead-Lift by Route (On-Sight)


Dead-Lift by Route (Red-Point)


Pull-ups by Route (On-Sight)


Pull-ups by Route (Red-Point)


Hang-time by Route (On-Sight)


Hang-time by Route (Red-Point)


Hang-size by Route (On-Sight)


Hang-size by Route (Red-Point)



For anyone wondering the data on front-levers was so sparse I didn't think it worthwhile graphing.

Yes I could derive correlations for each of these but thats not overly useful in isolation and would give you lots of correlations whereas the intended generalised linear model will allow the derivation of partial pseudo-R² statistics for each dependent variable and some ROC analysis to determine the sensitivity and specificity of the model (hopefully!).

The survey remains open so if you've not submitted any data yet then feel free to do so, but please bear in mind the point is to include some training metrics as well as your height, weight and bouldering/sports route grades otherwise you won't be providing particularly useful information.  And please do not bother entering units, the questions indicate the units you should enter them in (cm/kg/mm/seconds) you don't need to write them (nor should you enter in other units, e.g. don't put your height in as 1.8 rather enter it as 180 as the former is in metres the later is in the requested centimetres).  Its a bit of fun, nothing too serious, but I can see it being improved on based on suggestions here and taking a little more time to design the survey (I did it in about five minutes after people started posting their training feats and climbing grades to the forum).

I'll try and get round to doing the intended modelling (was planning on doing some whilst on holiday, but my laptop is going away today for repair and may not be back in time).

[JohnM]

Interesting to see the anomaly to the trend of number of pull up to boulder grade outdoors.  So there are some 6B+ climbers doing 20+ pull ups! 

[andy_e]

Interesting to note the users on here have a modal indoor grade of 7A and outdoor grade of 7C! It would seem that those who get the opportunity to climb outdoors climb harder, or is that an invalid observation for some reason?

[slackline]

Interesting to note the users on here have a modal indoor grade of 7A and outdoor grade of 7C! It would seem that those who get the opportunity to climb outdoors climb harder, or is that an invalid observation for some reason?

Check how many NA's there are for each.

[Stubbs]

Interesting to note the users on here have a modal indoor grade of 7A and outdoor grade of 7C! It would seem that those who get the opportunity to climb outdoors climb harder, or is that an invalid observation for some reason?

You could also come to the conclusion that indoor grades are off by 2

[tomtom]

Interesting to note the users on here have a modal indoor grade of 7A and outdoor grade of 7C! It would seem that those who get the opportunity to climb outdoors climb harder, or is that an invalid observation for some reason?

You could also come to the conclusion that indoor grades are off by 2

It could also be due to longevity of problems... you may try that 7C for 5 years and eventually get it, but depending on how often you go to the wall (and how often they re-set) you'll only get so many chances before its gone.....

[rodma]

Interesting to see the anomaly to the trend of number of pull up to boulder grade outdoors.  So there are some 6B+ climbers doing 20+ pull ups!

It may be because 6b+climbers kip 20+ times in a row and still count those as pullups. Everyone counts differently after all, so maybe even those doing only 10 pullups are kipping too 

Anyway, well done slackers. It looks like it took a lot more work to do this than i had certainly imagined