An analysis of the difference between men and women when they say they climb ‘5.something’.
Every once in a while I’ll meet a climber and they’ll ask me, “What grade do you climb?” Instantly my mind has flown off into calculation mode.
Well, I just climbed my first 5.12a in Malibu last weekend, so maybe I’m a 5.12a climber. No, I’ve only climbed one, and maybe it was soft. I’ve climbed five high 5.11s, but I don’t feel confident on ~all~ 5.11s. Hell, I couldn’t even hang-dog my way to the top of Little Buckaroo in Wild Iris and that was 5.11b.
“Um, I climb… low 5.11s?” I’ll respond. And then I’ll ask them what they climb.
So how do other people decide what grade they climb? Is it once they’ve climbed one of that grade? Five? A dozen? And does this differ by gender?
Since I’m at the tail end of Metis’ three month intensive data science bootcamp, I decided to go about answering this question by using data science.
I scraped my data from Mountain Project, one of the most popular online guidebooks with access to more than 170,000 routes. In addition to having route information, Mountain Project also has a feature called “Partner Finder” where Mountain Project users can find climbing partners based on their area. Each user lists what grades they’re able to lead and follow for each climbing discipline. Below is a screenshot of what the Partner Finder looks like.
Since only 500 partners show up at a time, I used various zip codes and ages to reload this page. Every time I loaded a page I scraped all of the information on the Partner Finder page. Using Selenium, I also clicked on each user name and downloaded their CSV file that contained a list of every climb they’d ever ticked on Mountain Project. This gave me the ability to compare what they claim to lead and how many times they’ve actually led that grade or higher.
Overall, I was able to scrape data from 3,960 users, with a total of 514,921 rows, where each row represents a Mountain Project tick by a user.
Notable Changes to the Data
Note 1: For this analysis I removed the rows where a user had climbed over a hundred routes that were at or above the grade they claim to climb. Those were outliers, and I’m making the assumption that those might have been mistakes where users improved their climbing abilities without updating their Partner Finder profile.
Note 2: When I’m comparing gender, I’ll only be comparing users that identify as male or female. There were 1,135 users that didn’t specify a gender, and there were 55 users that identified as genderqueer. Since that isn’t enough data to make any conclusions, especially after I separate them by climbing discipline, I’m omitting them from the analysis.
Distribution of Gender and Age
In line with the real life gender disparity in the climbing world, my dataset contains far more men than women. There are 3,206 male users, and 731 female users. This is around 4.4 times more men than women.
Most of the users were in their 30s, followed by 20s, 40s, and over 50. According to Consumer Health News, the average age of an indoor climber is 26. On Mountain Project, all of the climbs are for outdoor climbing. This might represent an age difference for outdoor versus indoor climbing, or be a result of the zip codes I scraped from.
What do you lead?
For the sake of simplicity, I’m going to use the term “Claimed Lead Grade” to refer to the grade people claim to lead. And when I say that somebody has led something, that means that they ticked the route and indicated that they led it on Mountain Project.
The first thing to examine was how a person’s Claimed Lead Grade compares to how many times they’ve led at that grade or higher. After filtering out users who weren’t lead climbers, I was left with 459 female users and 1,865 male users.
The findings showed that the average person has led their Claimed Lead Grade 8 times, with a difference between women and men. Women, on average, have led at their Claimed Lead Grade or higher a total of 9.8 times. Men are different in that, on average, they have led at their Claimed Lead Grade or higher a total of 6.4 times. This means that if you ask the average male and female climber what grade they lead, the female climber will have led 3 more climbs at that grade than the male climber.
The following graph shows how this plays out by grade. I removed grades where less than ten women were represented. There were always more men represented in each grade than women.
For nearly every grade, women have led their Claimed Lead Grade more often than men. This is particularly true for the mid-level climbs: grades 5.10a through 5.11d. Men and women similarly say that they lead 5.7, 5.8, and 5.9 when they’ve led around 6 climbs of that grade. The one instance where the the average male has climbed at or above their Claimed Lead Grade more often than women is for 5.12b.
There’s one particularly interesting trend for men in this graph. Looking within each number grade (5.10, 5.11, 5.12, etc.), there seems to be a downward trend. At the lowest end of each number grade, the average number of climbs a user has led at that grade hovers at around 6. As the climbing gets harder within that number grade, the count gets lower and lower. This seems to mean that men have more inhibitions about stating that their Claimed Lead Grade is a 5.10a or 5.11a than they are about saying that their Claimed Lead Grade is a 5.10b or 5.11b.
One possible explanation for this is that men perceive the difference between letter grades to be smaller than the difference between number grades. It’s possible that claiming to climb 5.10b, when you’ve only climbed a few, isn’t a big deal. But claiming to climb a 5.11a is a bigger deal, since being a ‘5.10’ climber versus a ‘5.11’ climber is a significant difference in the climbing world. So climbers might be thinking, “I better be able to put money where my mouth is” when bumping their Claimed Lead Grade up to a new number grade.
Another possibility is that when men are mentally calculating what their Claimed Lead Grade is, the number of the grade sticks out more than the letter. So a climber who has climbed six 5.10as and four 5.10bs might think to themself, “Well I’ve climbed a lot of 5.10s, and a good amount of them were 5.10bs, so I’m probably a 5.10b climber.”
Given the small sample size of women, especially once they’re divided by Claimed Lead Grade, I don’t think that much weight should be put on the women’s pattern within a number grade. With more data, a similar trend might show up.
Although the trend isn’t apparent for women in this graph, it does show up for women in a graph of the total number of users for each Claimed Lead Grade.
More women say that they climb 5.11a than 5.10d or 5.10c, even though 5.11a is more difficult. This is true for the men as well. Even though the men’s graph doesn’t have a slowly descending pattern within a number grade like the women users, there are peaks for the men at 5.10a and 5.11a.
This could mean several things:
- People tend to give the lower end of a numbered grade as their Claimed Lead Grade until they’ve started climbing a full number grade higher. Climbing six 5.11as might be more of a wakeup that you’ve improved and that you’re now a 5.11a climber than climbing six 5.10ds. If there are a significant amount of users who never define their Claimed Lead Grade as a ‘b’, ‘c’, or ‘d’ grade, with users jumping straight from 5.10a to 5.11a, this would also account for the fact that when men say they climb an ‘a’ grade, they’ve typically climbed quite a few of that grade. They may be climbing ‘b’s or ‘c’s but are waiting until they can climb a full number grade higher before changing their Claimed Lead Grade.
- Another possibility is that people are consciously picking ‘a’ over ‘b’, ‘c’ or ‘d’. By clustering to the low end of the grade they climb, they’re setting themself up for success by lowering the expectations of those around them. If they fall on a ‘c’ grade, no big deal, they already said they climb an ‘a’ grade. If they send a ‘b’, they’ve surpassed the expectations they’d set for those around them.
- People might also consciously low-ball their Claimed Lead Grade since different areas vary in difficulty. A 5.10a in Yosemite is much harder than a 5.10a in Red Rocks. By setting their Claimed Lead Grade low, they might be accounting for some sandbags.
Does this mean that women who say they lead 5.10a are better climbers than men who say they lead 5.10a?
We can’t necessarily say that yet. We have to consider this possibility:
My boyfriend and I are both avid Mountain Project users, but we differ in when we tick a route. I tick all the routes I get on, even if I chopped the sport climb into a series of boulder problems by reaching each bolt, shakily clipping it, and then sitting on it for half an hour. My boyfriend, on the other hand, only ticks a route once he’s sent it. What if this difference is true for men and women in general? Or what if men only attempt higher grades when they’re more confident they could send?
To test these hypotheses, I filtered for only sport climbing ticks that users had both led and sent. This got rid of ticks by users who hadn’t specified their lead style, and it removed ticks where the user said they’d hung/fallen on the route. I was left with ticks where the user had sent, including pink points, red points, onsights, and flashes.
Here’s what I found: Men were decidedly ticking more routes that they were sending proportional to all of their lead routes compared to women. The percent of female leads that were sends was 59.21%. Meanwhile, the percent of male leads that were sends was 68.88%.
But this didn’t mean that men and women send the same number of grades that are at or above their Claimed Lead Grade. Once again, there was a disparity! On average, female climbers had sent their Claimed Lead Grade 5.2 times. Whereas male climbers had, on average, sent their Claimed Lead Grade 3.6 times. The average number of climbs above the Claimed Lead Grade for all users was 4.4 routes.
What about the grades that people say they can follow?
The last analysis I conducted was comparing users’ Claimed Follow Grade to how often they’d followed that grade.
In general, users had very few follow ticks at or above their Claimed Follow Grade. The average for all users was a low .73. The average for women was a somewhat higher .94, and the average for men was .52.
There are a couple ways these low scores could be interpreted. The first is that people are drastically over-estimating their following capabilities. The second, and the one I favor more, is that it’s just hard to find someone capable of leading above a person’s Claimed Follow Grade. If people typically follow grades that are higher than what they’re capable of leading, and if people typically climb with climbers who climb similar grades as them, then the opportunity to follow a climb at a person’s absolute limit is few and far between.
When someone says that they lead ‘5.something’ in sport climbing, how many climbs have they led at that grade or higher?
- Average for All Users: 8.1
- Average for Women: 9.8
- Average for Men: 6.4
When someone says that they lead ‘5.something’ in sport climbing, how many climbs have they sent (defined as not falling or not omitting their lead style, including red points, pink points, onsights, and flashes) on lead at that grade or higher?
- Average for All Users: 4.4
- Average for Women: 5.2
- Average for Men: 3.6
How often do people tick routes that they’ve fallen or hung on?
- Percent of Female Leads that are Sends: 59.21%
- Percent of Male Leads that are Sends: 68.88%
When someone says that they follow ‘5.something’ in sport climbing, how many climbs have they followed at that grade or higher?
- Average for All Users: .73
- Average for Women: .94
- Average for Men: .52
When women claim to lead at a certain grade, they have climbed that grade more often than men who claim to lead the same grade (assuming Mountain Project is a realistic portrayal of how people are climbing in real life).
Although it’s true that men tick more sends proportional to their total leads compared to women, the disparity still shows up.
The difference is also apparent in ‘follow grades’, but in general all users have followed very few of the grades they claim to be capable of following.
This analysis only focused on sport routes but it’d be interesting to analyze trad, boulder, aid, and ice routes as well. I also only focused on the difference between genders in this blog, but I suspect there might also be differences between age groups. There’s so much more to explore!