[slackline]

Updated Report

There have been about another 20 or so people completing the survey so I've re-generated the graphs and recompiled the report.  I'm too lazy to upload all of the individual graphs to Imgur and then embed them here again, so if you want the most recent/complete overview see this PDF*.

Might get round to doing some more of the planned analysis today, but further work may be put on hold as I'm likely to have to reinstall Gentoo on my laptop in the coming week as its having its hard-drive replaced under warranty.


* An archive of versions is maintained on the UKB Wiki.

[andy_e]

Is there a facility by which to test hardest grade climbed against hardest grade flashed? I'd be interested to find out if there are as many people as me who have flashed a whole number less than they've climbed or if I'm just truly terrible at reading sequences...

[slackline]

Yes I can probably find time to generate a scatter plot of that.

I did sit down over xmas and start playing at modelling relationships but the data is actually quite sparse in so much as very few people filled out all fields.  I could test each training variable to see how it predicts route/bouldering performance but I don't think such an approach is useful because for starters the factors influencing performance don't work in isolation and are going to be correlated with each other in some way.  Its also prone to problems associated with multiple testing.  Instead I prefer multivariate approaches, but for each predictor variable you generally need 10-20 observations to detect a resonable size effect.  Thus I'd be looking to have around 200 complete observations and there weren't more than a dozen.

[Sasquatch]

Is there a facility by which to test hardest grade climbed against hardest grade flashed? I'd be interested to find out if there are as many people as me who have flashed a whole number less than they've climbed or if I'm just truly terrible at reading sequences...

I'm interested as well, for both bouldering and routes. 

My hardest flash is 6 number grades lower than my hardest RP   (7c vs .13d)

[andy_e]

Was that data recorded in the original survey? I don't remember mentioning hardest ever grade or onsight.

[andy_e]

6 number grades... (7c vs .13d)

Har har har.

[slackline]

Was that data recorded in the original survey? I don't remember mentioning hardest ever grade or onsight.

Good point, I can't provide you with a plot of hardest flash v's hardest worked grade for bouldering as I assume you're interested in (get the impression you boulder more than sports/trad) because I only requested indoor/outdoor bouldering grades.

I did however request the on-sight and red-point grades for routes so can do a plot for those.

Or is it that you'd like to know how your hardest boulder grade stacks up against your hardest onsight of routes?

[andy_e]

Yes, I boulder more than sport or trad (just like one is more than zero  )

I'd like to see max flash vs. max workedfor bouldering but the other permutations could be interesting too.

[slackline]

Yes, I boulder more than sport or trad (just like one is more than zero  )

I'd like to see max flash vs. max workedfor bouldering but the other permutations could be interesting too.

Can't do that for bouldering, only have onsight/redpoint for routes and for bouldering, well see my last post. 

[andy_e]

collect more data perhaps?

[abarro81]

You don't need data for this, I can confidently inform you that if your flash/onsight grade is a number grade below your worked limit you are undoubtedly relatively bad at flashing/onsighting.

[slackline]

collect more data perhaps?

Go for it, the survey is still there.  How you encourage more people to complete the all fields of the form I don't know (I did link to it on Reddit Climbing which got about five more people filling it in).  If someone fills it in again but more complete then please ask them to record the same name as the original so as to facilitate removal of duplicates.

I have thought about a second survey in light of comments made in this thread and including flashed/worked boulder problems is included.

[Rocksteady]

Very interesting report.

I think something that would be interesting and significant to track would be hours climbing/training per week vs grade. I suspect this would have one of the biggest correlations.

I may be reading this box plots wrong but it looks to me like someone on here can bench press 200kg, do 40 pull ups, and climb Fb7C! That's some serious all round strength! 

[andy_e]

You don't need data for this, I can confidently inform you that if your flash/onsight grade is a number grade below your worked limit you are undoubtedly relatively bad at flashing/onsighting.

Thanks!  It'd be interesting to know if I'm an anomaly though or if other people are affected by the same tardiness as me.

Slackers, is it easy to set up a new survey just asking those questions I want?

[slackline]

Slackers, is it easy to set up a new survey just asking those questions I want?

Yes, I used Google Docs, but you could use Survey Monkey or similar, there are lots of options.  The problem is (always) getting enough people to complete the surveys.

I may be reading this box plots wrong but it looks to me like someone on here can bench press 200kg, do 40 pull ups, and climb Fb7C! That's some serious all round strength! 

Well the person who can do 40 pull-ups climbs F7b+ but if they were in the F7c category you can't tell from box-plots that the data points relate to the same observation (unless they were labelled).

[highrepute]

Only justnoticed this thread. I'm getting a 404 on the pdf download any chance of re-uploading slackline?

[slackline]

Only justnoticed this thread. I'm getting a 404 on the pdf download any chance of re-uploading slackline?

Not at present, it appears that when the UKB Wiki was upgraded the file permissions for uploading were reset to allow only png, gif, jpg, jpeg.  The section on the Training : The Science (UKBenchmarking) still links to Imgur where I hosted the  individual  box-plotfiles.

This page suggests that the file should be there, but as you say returns a 404.

@Bubba : Has the location of uploaded files changed between wiki upgrades?


Note also that jwi is doing a benchmarking survey too.

[slackline]

P.S. : Check your PMs.

[Sasquatch]

Any chance of getting a copy of this or getting it to be available online again?  I was curious about checking it out a couple of years later

[Bencil]

I'd be interested to see this too. Are there any obvious correlations?