documented info on why gyms require you to back up a figure eight using an overhand

There is no physical/mechanical reason why a figure eight tie-in knot would need an overhand backup. Testing shows that the rewoven figure eight or a figure eight on a bight is one of those knots that can't simply slip apart because you pull on it or put it through a lot of cycles of tension. It can't turn inside out, for example, like an offset figure eight (not the same as a Flemish bend), which should never be used as a bend. From the physical theory of knots, I think the only parameter that could affect this would be the coefficient of static friction of the rope on the rope, which is plenty high enough, although conceivably it could become too low if the rope was icy.

The textbook used in the AMGA SPI class says simply, "The figure eight knot does not require a backup knot."

However, I think it's pretty common at gyms, including the one I climb at, to require people to tie an overhand backup. If you take their class, they teach you to tie the backup. If you take their test to get a belay card, and you don't tie a backup knot, they tell you that you did it wrong.

I'm sure we could generate endless debate about hypothetical scenarios where the climber messes up in some way (e.g., the figure eight really isn't a figure eight), and then the backup saves you. But what I'm more interested in is whether there is any reliable information from a documented source on what actually causes gyms to say this.

My guess has always been that it's a way for them to make sure you have enough of a tail. If a gym employee is walking around and looking up at people climbing, the existence of the backup knot automatically shows that the figure eight was tied with plenty of tail -- enough tail to make the backup knot. They can easily see this from far away.

Another possibility would be that it has no objective justification but only a legal one. If they or their insurer gets sued, they simply don't want to have to convince the jury that they were correct not to require the backup. They think the jury will be impressed that they required the backup, and if they don't equire it, they're afraid that the plaintiff could accuse them of not following standard practices. On this theory it would be sort of like the tail of a peacock. It's a random useless ornament, but once it starts to evolve, you have to go along with what everyone else is doing.

I've taken a lot of belay tests while traveling the last couple years, mostly in California and Nevada but also a couple in Oregon and Arizona, and Minnesota, and Colorado. Never tied a backup knot, and noone has ever mentioned it. ¯\_(ツ)_/¯

Yukon Cornelius wrote:

I've taken a lot of belay tests while traveling the last couple years, mostly in California and Nevada but also a couple in Oregon and Arizona, and Minnesota, and Colorado. Never tied a backup knot, and noone has ever mentioned it. ¯\_(ツ)_/¯

Interesting. Maybe it's not as widespread as I thought. I climb at Sender One in Santa Ana, CA. Here's an example of someone else who perceives this as a standard thing to require at gyms: https://outdoors.stackexchange.com/q/27863 The fact that the AMGA single pitch manual specifically says it's *not* necessary would seem to indicate that it's common to believe that it is.

It's ridiculous but yes, most gyms require it. Perhaps the rationale is that if the figure 8 is tied incorrectly, at least the backup knot will work. Or maybe it's just the insurance companies thinking that two knots are better than one.

i think it has more to do with insurance. I got into a debate with a guy giving me a belay test at PG in SF because he wanted me to have like 3 ft of tail and tie a back up knot. He said their insurance required it.

I imagine that it’s incase the knot is tied wrong and a stopper knot would hopefully clog it

I've been to gyms that require it, and gyms that do not. It's one of those things that I ask about before I tie in, and produce whichever knot the gym wants to see.

There's a lot of variation among gyms. We stopped randomly at Gravity Vault gym in NJ on a drive back from NYC recently. The staff member who did the belay test was ok with my husband tying in on a bowline+re-thread /backup knot, which was very nice. Most gyms just require figure 8, and do not allow other options. It was also the first time ever that I was in a gym that didn't require a lead fall during the lead belay test. Weirdly, they also completely neglected to test my friends, who showed up there, filled the waiver, and went on to climb, without anyone checking that they are toprope-belay-proficient. I suspect that if I didn't ask about borrowing the rope, and doing the lead test, they wouldn't have tested us, either,

My understanding (likely misinformed) is that the most common potential danger with a figure 8 is that you have too short of a tail. (<6 inches) If you can tie a backup knot then you basically prove that you have enough tail. So it’s not because the backup is necessary but that it shows the knot was tied correctly. 

The common mistake is to assume the goal of a gym is to teach perfect, by-the-book techniques that are perfectly vetted to reduce objective risk.

This is incorrect.

The goal of a gym is to move hundreds of people of a variety of skill sets and a huge spectrum of intelligence and attention span through a high-risk, high-consequence, human-error-prone activity by adopting easily-repeatable safety practices that give a handful of barely-trained minimum wage employees the best odds to notice and respond to human errors before they produce a catastrophic result. There's also the insurance angle, but it's basically the same story with the addition of unspeakable amounts of money and imperfect regulatory oversight.

Don't look to gyms to have perfectly-dialed safety protocols that immediately make sense to you. That's not what they're after. They're chasing a different animal, notably, "It's easy-ish for us to operate at scale, nobody dies, and we don't get sued into the center of the earth." These rules produce very different results than "What's the current best practice according to pull tests?" The gym doesn't care that the AMGA says a safety knot isn't necessary; the gym cares that requiring a figure 8 with a stopper knot (or whatever similar protocol: assisted-braking devices, locking carabiners, only using the gym's lead ropes, tethering or not tethering for certain tasks, failing you for screwing up a specific part of the lead test, etc.) makes it easier and more repeatable for Joseph Jeremiah McEightDollarsAnHour to wander through a group of forty people and check that gravity isn't going to kill any of them in the next five minutes.

Th "back up" knot is simply to keep all of the extra rope out of the way, since gym climbers tend to use way to much rope to tie in. I used to use the Yosemite finish, but I don't even do that anymore, I simply let the small tail flop around. 

Prior to a few months ago, I would have agreed with OP but I’ve actually seen loops develop on some friends poorly tied figure 8s. Rather scary.  Granted, it was on 70m dynamic lines that were fixed on a slab route and being jugged up using a figure 8 on a bight, connected to the belay with 2 locking carabiners.  4” - 6” loop developed on two different lines and could have been catastrophic.  No stopper knots were put in.  I was at the belay as they were coming up and watched it develop, I then pulled the tails tight!! getting the open loops out.   So yeah, I would say it is best practice for the masses to back up! 

Jake S wrote:

i think it has more to do with insurance. I got into a debate with a guy giving me a belay test at PG in SF because he wanted me to have like 3 ft of tail and tie a back up knot. He said their insurance required it.

That's funny, cuz with that much of a loop at the end I'd be worried about it catching during a fall.

Don't expect any common knot strategy to ever be 100% adopted. It's an unrealistic expectation. To me the backup knot is about as necessary as a cordellete, ie zero necessary, Yet people still use cordelettes, defying MY common sense goddamnit! And you know what? I still tie that backup overhand knot above my fig 8 and sure it sops up excessive tail but I tie with long tail precisely so I can tie the useless backup knot to begin with, thereby defying YOUR common sense goddamnit!

But you know what, hombre? I find the sight of it reassuring, when I'm climbing and I look down at my package. And for me, that alone is enough. Sure I've climbed without it, and can tie in with a several unnecessary variations of a fig 8, or go the bowline route too (where a backup knot may be more well-advised). Doesn't matter, I'm putting a backup knot on it. 

I go to a gym every once in a while that doesn't allow you to use an ATC at all, and on some climbs you have to use their gri gri, and on other climbs you can use your own. I've never been belay tested or anything there, they never ask. The gri gris that they supply are already attached with a bowline with a zip tie on it? And the gri gris move from climb to climb, seemingly randomly. It makes absolutely no sense, and the employees can't answer anything as to why or the reasoning or why the pre attached gri gris move to different routes, seemingly day to day. 

Also not allowed to use any gri gri alternative. None of it makes any sense. None of the employees seem to know why or even question it. Climbing gyms are kinda weird. 

Erik Strand wrote:

I imagine that it’s incase the knot is tied wrong and a stopper knot would hopefully clog it

I know someone who is alive and well having been lowered on the stopper knot. 

Any extra safe procedure can have a downside. A few years back, an experienced climber in the southeast arrived at the anchors of a sport climb. Clipped into the anchors, untied, brought the end of the rope through the anchors, and then tied back into the rope. He thought he had tied in using a figure 8 but had only tied in the backup knot. He fell and was fatally injured. There was discussion at that time that the backup knot being unnecessary could lead to danger, especially in adverse conditions such as dim light.

Ben Crowell wrote:

There is no physical/mechanical reason why a figure eight tie-in knot would need an overhand backup. Testing shows that the rewoven figure eight or a figure eight on a bight is one of those knots that can't simply slip apart because you pull on it or put it through a lot of cycles of tension. It can't turn inside out, for example, like an offset figure eight (not the same as a Flemish bend), which should never be used as a bend. From the physical theory of knots, I think the only parameter that could affect this would be the coefficient of static friction of the rope on the rope, which is plenty high enough, although conceivably it could become too low if the rope was icy.

The textbook used in the AMGA SPI class says simply, "The figure eight knot does not require a backup knot."

However, I think it's pretty common at gyms, including the one I climb at, to require people to tie an overhand backup. If you take their class, they teach you to tie the backup. If you take their test to get a belay card, and you don't tie a backup knot, they tell you that you did it wrong.

I'm sure we could generate endless debate about hypothetical scenarios where the climber messes up in some way (e.g., the figure eight really isn't a figure eight), and then the backup saves you. But what I'm more interested in is whether there is any reliable information from a documented source on what actually causes gyms to say this.

My guess has always been that it's a way for them to make sure you have enough of a tail. If a gym employee is walking around and looking up at people climbing, the existence of the backup knot automatically shows that the figure eight was tied with plenty of tail -- enough tail to make the backup knot. They can easily see this from far away.

Another possibility would be that it has no objective justification but only a legal one. If they or their insurer gets sued, they simply don't want to have to convince the jury that they were correct not to require the backup. They think the jury will be impressed that they required the backup, and if they don't equire it, they're afraid that the plaintiff could accuse them of not following standard practices. On this theory it would be sort of like the tail of a peacock. It's a random useless ornament, but once it starts to evolve, you have to go along with what everyone else is doing.

I can tell you exactly why it's required at S1 as I had a hand in the policy.

The backup knot allows the knot to be confirmed "safe" from a distance. If the fig 8 was tied incorrectly the backup would stop a total failure. So, staff looks for that backup because it's easy to see and with 100+ people in the gym that helps lifeguarding move faster.

Tradiban wrote:

I can tell you exactly why it's required at S1 as I had a hand in the policy.

The backup knot allows the knot to be confirmed "safe" from a distance. If the fig 8 was tied incorrectly the backup would stop a total failure. So, staff looks for that backup because it's easy to see and with 100+ people in the gym that helps lifeguarding move faster.

Cool, thanks for posting. This was exactly what I was hoping for -- actual evidence about why gyms require it, as opposed to speculation. So it's actually kind of a hybrid of two of the hypotheses earlier discussed. It's about visibility, but not visibility of a long enough tail.

Were you influenced by having worked at some other gym that did this with the same rationale?

This would explain the inconsistency with the AMGA SPI manual. That book is written for guides who are going to be closely supervising a small number of clients, not for gym employees supervising a mob of climbers.

I'm trying to imagine what knot someone could tie that they thought was a rewoven figure eight but actually wasn't a knot. I did hear a story about someone at SO who simply didn't tie in before he started a lead, and he had earbuds in, so he couldn't hear his belayer yelling at him that the rope had fallen out of his tie-in loops and was on the floor. The story was that once he realized it, he started downclimbing, then inexplicably jumped off from high up rather than finishing the downclimb, and got hurt.

dave custer wrote:

I know someone who is alive and well having been lowered on the stopper knot. 

Did he just think he'd tied the figure eight, but hadn't really tied a knot at all? And then he tied the overhand to back up the nonexistent knot?

Or was he at a gym, where someone had left a figure eight tied on a toprope setup, and he never did the step of completing the figure eight, but he tied the overhand?

dave custer wrote:

I know someone who is alive and well having been lowered on the stopper knot. 

If the figure 8 wasn't there it would be a Scaffold knot (or similar knot) which is commonly used to make work positioning lanyards or cow tails. 

Opening this thread I was hoping to see replies with insurance or climbing gym industry literature explaining why the "safety knot" is so ubiquitous. One can dream, I guess. 

We in AZ and the university of Canada documented how you don’t need to back up an 8.

Essentially, the final follow through is the back up.

Same testing groups also found the overhand was beyond adequate, but should have a long enough tail.  And that an overhand on a bight with long tails is an ideal knot for rappel ropes. 

Ben Crowell wrote:

Cool, thanks for posting. This was exactly what I was hoping for -- actual evidence about why gyms require it, as opposed to speculation. So it's actually kind of a hybrid of two of the hypotheses earlier discussed. It's about visibility, but not visibility of a long enough tail.

Were you influenced by having worked at some other gym that did this with the same rationale?

This would explain the inconsistency with the AMGA SPI manual. That book is written for guides who are going to be closely supervising a small number of clients, not for gym employees supervising a mob of climbers.

I'm trying to imagine what knot someone could tie that they thought was a rewoven figure eight but actually wasn't a knot. I did hear a story about someone at SO who simply didn't tie in before he started a lead, and he had earbuds in, so he couldn't hear his belayer yelling at him that the rope had fallen out of his tie-in loops and was on the floor. The story was that once he realized it, he started downclimbing, then inexplicably jumped off from high up rather than finishing the downclimb, and got hurt.

Did he just think he'd tied the figure eight, but hadn't really tied a knot at all? And then he tied the overhand to back up the nonexistent knot?

Or was he at a gym, where someone had left a figure eight tied on a toprope setup, and he never did the step of completing the figure eight, but he tied the overhand?

The rationale was developed from talking to other gyms. With a gym that big you need an efficient way to check all climbers. AMGA means nothing to gym managers.