[abarro81]

Pullups: don't know
1-armers: 0 from straight, 1 with a cheaty range of motion
Grade: the above 2 answers apply from when I was climbing ~7C to now (8B max, 8A/+ more normally)... so take your pick Somedays doing 'most' of a 1-armer (but not the hard bit at the bottom) feels easier than others, but never have I been able to go up, down and then even make any headway on going back up again.

[SamT]

A mate years ago could crank one armers like no tomorrow, straight out the pub door onto some scaff.  Was probably only climbing e1, e2 at a push at the time. perhaps 6c on sport.

That always convinced me that it was pretty irrelevant with no correlation. 

I concede that some certain problems over ~8A are going to require one arm type lockability. But not if they're font slabs.

[Danny]

Based on the currently-tiny sample size of 12 (ish), grade linearly correlates with pullups. Any stats pedants out there can save the beration for a bigger sample...I'm just cocking about.

data:  pulls$boulder and pulls$pulls
t = 2.4437, df = 11, p-value = 0.03261
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 0.06268649 0.86232918
sample estimates:
      cor
0.5931836


A saturating hyperbolic type relation might be a better fit. A base of strength (in the 1-10 pullups range) might relate to the lower grades
more appropriately. It's telling that a few of the better climbers have no idea how many they can do. i.e. it obviously doesn't matter at *some* point. Though I suspect those climbers can do quite a lot.

[mr chaz]

For the bouldering grade, I'd think it should be something like your *general* max grade. i.e. a grade you've climbed a good handful of, not your probably-soft best effort, with one move that totally suited you, 5 years ago.

Pull-ups: ~25
One-armers: 1
Max grade: 7B+ 7A

Fixed that for you

4 or 5 not that? 

Do dynos count? 

10
1
7B

[yetix]

pullups: 20 something last time I tried
1 armers: just about 1 with my left 3 with my right
max grade: 7B+

[petejh]

Pull-ups: 35 underarm, 25 overarm. Used to be able to do a few more.
One-armers: 0. Have never committed any time to doing one, seems to be some myelination required as well as a base level of strength.
Grade: 7C. Hardly spend any time bouldering. Should be 8A.

'Could do better'

[andy popp]

15ish
A definite big fat Zero
7C+

[teestub]

2 arm: about 10 proper ones, although why a boulderer would ever try to train more is beyond me.
1 arm: 0, although I have considered trying to train to address this
Grade 7C+

[joeisidle]

Pull-ups: I once did 10 six years ago. Sporadic attempts since haven't yielded much above 5-6, so I suspect it's somewhere in the region of 8 now if I really applied myself

One-armers: as if.....

General max grade: 7A/+

Once every couple of years max grade: 7B

[peewee]

25
1
8B

[jamesturnbull97]

9
0
8a

[stokesy]

Pull ups = about 12?, a few more back in the day
One armers = never been able to do a proper one
Max grade = 7B, not done many recently but quite a few in the past

[HaeMeS]

Pull-ups: 15 underarm, 12 overarm. Been the same since my teens but wasn’t climbing back then.
One-armers: 0. Not even close to moving upwards.
Grade: 7A+. Pretty solid. 7B impossible.

[Probes]

15 yr ago..
45
9 on right 4 on left
8a

now.. 
25
2 on right 0 on left
7c+/8a

[Danny]

9! Chuff me.

[webbo]

35 years ago
30 on a regular basis,I may have done 50 once.
3 on right 2 on left
MXS sometimes XS
now none without something exploding.

[Danny]

Approaching a semi-useable sample size and the correlation between pulls and boulder grade has strengthened a little. Probes, you're quite the outlier on the number of pulls and one-armers. I've also used you as a pseudo-replicate. If oldmanmatt could define 'lots' he might join you.

data:  pulls$boulder and pulls$pulls
t = 2.8255, df = 22, p-value = 0.009847
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 0.1422133 0.7609981
sample estimates:
      cor
0.5160075

Anything with the one-armers is zero-inflated data. With that caveat, ones correlate with twos. Grade and ones don't correlate...although there's a superficial trend but again, zero-inflated data.

[GazM]

1 to add to your n:

Pull-ups: 12
One-armers: 0
Max grade: 7B

[remus]

We collect some similar data as part of the Lattice Training assessments, though it's not a good indicator of climbing ability in our experience.

I threw together a quick scatter plot of a subset of data that happened to be at hand. n=58.



We don't collect data on one armers because the vast majority of people can't do them.

My opinion (not lattice endorsed) is that if you're in the 15-20 bracket there's very little point smashing out loads of pull ups. Maybe if you're significantly under that there's a case to be made for doing pull ups as a conditioning sorta exercise.

[Stu Littlefair]

Since this is turning into a nerdy stats conversation -

Remus, you measure so many variables in your tests, have you every tried a PCA analysis?

[remus]

Remus, you measure so many variables in your tests, have you every tried a PCA analysis?

Yep In the models I've put together it wasn't particularly helpful, though, as it basically just picks out the factors that we already know are useful (finger strength, max moves etc.) and pretty much ignores lots of the other factors. My personal opinion is that there's so much variability in the underlying measure of climbing ability (i.e. grades) that it makes it very difficult to investigate effects that are relatively subtle. That's not to say those other factors aren't interesting to consider of course, but we're not at the stage where we can say "You could climb 9b, you just need to do an extra 10 moves on the lattice board, another 6 pull ups and hold a front lever for more than 10 secs."

Maybe one day everyone will start using UKC/8a.nu and we'll be able to use something other than 'best redpoint' as a measure of climbing ability!

[duncan]

I have been a 20 pull-up Severe (1A?) wad and 3 pull-up E6/7b/6C punter. I concluded pull-ups were not that helpful for me.

It seems to me the two groups likely to derive greatest performance benefit from pull-ups (and weight lifting in general) are women and old men. The two groups least likely to do such a thing! Since I now find myself in one of these categories I've started doing pull-ups and pumping iron a little. 

Remus, I expect you've done regression analyses, how much of the total variation in RP grade can your models describe?

[Stu Littlefair]

Remus, you measure so many variables in your tests, have you every tried a PCA analysis?

Yep In the models I've put together it wasn't particularly helpful, though, as it basically just picks out the factors that we already know are useful (finger strength, max moves etc.) and pretty much ignores lots of the other factors. My personal opinion is that there's so much variability in the underlying measure of climbing ability (i.e. grades) that it makes it very difficult to investigate effects that are relatively subtle. That's not to say those other factors aren't interesting to consider of course, but we're not at the stage where we can say "You could climb 9b, you just need to do an extra 10 moves on the lattice board, another 6 pull ups and hold a front lever for more than 10 secs."

Maybe one day everyone will start using UKC/8a.nu and we'll be able to use something other than 'best redpoint' as a measure of climbing ability!

Fair enough, and I agree about the scatter in your measure. There's also the issue of features that will be important but are harder to quantify, such as flexibility and climbing efficiency!

[Danny]

<shameless screengrab and datathief> JK.
Given that 5 is your minimum the questions about what happens at the bottom end remain. I know a few people who climb who can't do one.

Stats nerdery:

Never been a fan of PCA myself, and with tools like NLMEEs about these days you can handle quite complex datasets and make meaningful predictions. Although I admit that I find such models deeply frustrating more often than not. 

We collect some similar data as part of the Lattice Training assessments, though it's not a good indicator of climbing ability in our experience.

I threw together a quick scatter plot of a subset of data that happened to be at hand. n=58.



We don't collect data on one armers because the vast majority of people can't do them.

My opinion (not lattice endorsed) is that if you're in the 15-20 bracket there's very little point smashing out loads of pull ups. Maybe if you're significantly under that there's a case to be made for doing pull ups as a conditioning sorta exercise.

[Danny]

More stats. The one-armers are a classic eg of a zero-inflated Poisson distribution by the looks of it. If I was working this kind of data I'd collect that info despite the pile of zeros (which are pretty common in the likes of ecological survey data, for example) because the tools are available to handle that kind of distribution.