A Real Life Off-by-One Error
leejo.github.io
leejo.github.io
I expected the article to eventually answer this puzzle:
> The competition started and got through a number of rounds. There were some comments about how the climber on the left always won.
Near the end:
> The kicker is that the out of place hold hasn’t been used in a long time. The climbers have optimised their route such that it is skipped. The same happens to the fourth hold from the bottom. So either being in the wrong place is immaterial to the climbers’ technique as long as they don’t get in the way.
So it seems like the error discovered by the article author should not have conferred any advantage to the climber on the left.
Anyone who can shine light on this matter?
> A few of the climbers had said that the automatic belay ropes on the right hand lane did not feel right, so the cherry picker was replacing those and not the hold that I had noticed being out of place. The climbers had noticed something wasn’t quite right, but hadn’t said anything about the out of place hold.
It was probably just two separate problems.
Also the author rules out psychology, but I wouldn't, especially since there were multiple confirmed errors in the route preparation, which I expect could reduce one's trust in the fairness of the competition. In the moment, I might start to wonder, "If one hold was out of place, why not more? Is anyone even checking this?" even if untrue / unlikely.
Certainly, the rope feel is a much more significant factor, since the feel of the rope tugging on your harness is a non visual part of your body position feedback (maybe “I know that I’m going fast enough/pulling hard enough if I’m outracing the rope retraction rate”).
Literally it was an off-by-one error. Literally, literal meaning.
There are other reasons as well. I think the more common causes are inclusive vs exclusive comparison errors and fencepost.
The spelling is off by one. It's "borne"
the spelling was actually off by e , so you're stilll off by 1.71828...
It’s for example called “the fencepost problem”: https://betterexplained.com/articles/learning-how-to-count-a...
I agree subtracting or adding one to any number is not the problem. It has to do with counting.
if you are sprint climbing and you put your hand out to grab and you add or subtract the number 1 from where your hand needs to be, it is a problem or contributes to a speed difference.
They might not use the hold by physically touching it, but they might still use it as a visual indicator of where the other holds are in relation. These competitors are used to the same layout for many years. If there is a slight misrepresentation it can surely put them off.
- either as 11 empty holes between the holds; - as 11 holes, start counting 1 just above hold A; - or as 11 holes, start counting with hold A as number 1.
Another real-life example, is a plumber who tells the construction worker that the distance between the holes for hot and cold water needs to be 15 cm. This was meant to be measured center to center, but the constructor worker interpreted it as the distance from the right side of the first hole to the left side of the second. The result can still be admired in our house, 10 years later.
Of course use of expressions may vary. My in-spirit-meaning of ‘off-bye-one error’ may differ from yours, and that is fine. (Had it really mattered in a discussion then we would simply agree to find a suitable definition of the expression.)
But I think it's possible that 'extra' holds are potentially like 'junk' DNA. People fall into the trap of thinking that DNA is useless if it's never transcribed, but we know that's not actually the case. Non-expressed DNA can do things like alter binding affinity for neighboring sequences, affecting how often those neighboring sequences are expressed. I think it's possible that climbers are taking in a lot of information subconsciously as they sprint through this route in order to mike very small adjustments. The position of surrounding holds, even ones they never touch, could very well be a part of that information stream. They're fighting over hundredths of a second, so even a very small effect could be meaningful.
Here a pyramid of 120 blocks is planned, with one placed every ten years. It started on the 1200th anniversary of the town, and is planned to be complete 1200 years later... and I'm sure you've spotted the problem.
Matt Parker, the 'stand-up Mathematician' has a video on it: https://youtu.be/FAdmpAZTH_M?si=_u8fM-fprUWiEqZ9
This is odd to me since I've successfully used stereoscopy in the past to find small differences. For some reason, with this image, rather than causing a change in perceived z-level, my eyes fight for dominance and my left ends up winning.
Same, I think there are too many other things around it to make it work for me though.
I'm also one of them... actually not, I'm not "in US" :)
There's a challenge where you see words like "RED BLUE YELLOW GREEN PURPLE", but they are written in color different to their name. You are challenged to say the color of the text and not read the word out loud.
Intentionally defocusing your eyes makes that challenge trivial.
Am I the only one?
When autostereograms were all the rage in the late 80's I had a program on my Mac Plus that let me make/edit them and I used to edit for hours WHILE looking at them in 3D. Then one time I was walking down a hallway with a repetitive wallpaper pattern, my eyes did the thing, the entire hallway appeared to shift in front of me, and I stumbled and fell.
Rare that I meet someone else that does this. I learned how to do magic eye puzzles as a young child, I think my first was in a magazine and I ‘solved’ it the standard way of placing it close to your nose then slowly pulling back. Before long I could just do it on command and as an adult I find myself doing it all the time, often unconsciously. Makes spot the difference puzzles trivial, that’s for sure
One comment on what the article says:
> If this were actual code review the correct comment would be something like “this [piece] hasn’t been used for years, it should be deleted”. But this is something in physical space, and there would be arguments that removing it (them) means the route has changed, thus times are no longer comparable.
Hmm, I think the correct analogy is rather a benchmark. Like code in a benchmarking tool or test, the whole climbing course does not serve any purpose, any actual goal, except to be completed as fast as possible.
You wouldn't say "these instructions should be deleted because branch prediction and speculative execution in recent years have made it so that total cycle count is the same without them", for the reason stated ultimately after in the article already: That may not have been true in the past, and may change again in the future.
Then a new CPU architecture becomes popular, and spiders start winning every speed climbing event
This is a very different statement than the original, and furthermore the original might actually trigger performance issues e.g. while the operation might not be used anymore it can have side effects (like triggering a prefetch) which end up affecting downstream code.
The fast path that the speed climbers commonly use doesn't involve the slower path involving the buggy portion.
But if the climber can't get up/further using the fast path then the climber may be forced to take the slower path.
Since the slower path is standardized, the climber knows how much effort/time is needed on the slower path.
Except on the righthand side, the climber's typical effort along the slower "buggy" path is different from the typical slower path and will result in an even slower time aka the execution time in the error-handling path is larger than usual...
I discovered an off-by-one error in college as I was graduating. Apparently the "class plan" I had put together with an advisor during my second year was missing a class, and I discovered it as I petitioned it for graduation. "You're one class short."
(It was for my second major, not the primary one, and the head of the physics department was nice enough to credit a nonlinear optimization course from engineering toward the major, so I earned it.)
The deal with my Math dept head was I had to make it my primary major (no true doubles at my weird undergrad, had to have an "additional"). It was a deal I was happy to make if it meant skipping an intro-to-proofs course after having taken the masters level series in most the of the fields they offered!
For some reason I find this a bit humorous. My dudes, you do not have much leverage.
Also, by the way, where is the photo on your about page: https://images.squarespace-cdn.com/content/v1/60f40aa1942bba...
I feel like it's Crater Lake, Oregon, but considering where you live it's probably somewhere in Switzerland. Looks lovely.
There's been some discussion of having new speed routes every few years. I think this would make the event a thousand times more interesting. That said, I'm still not sure I'd be interested.
So the day of your fifth birthday is the first day of your sixth year alive.
https://www.aljazeera.com/amp/news/2023/6/28/why-are-south-k...
They could even still incorporate the one standardized wall config as a speed round once in a while or integrate it into the competition in some other way.
There are other forms of competition:
- bouldering: how many of 4 short boulder problems can you finish
- lead: how high can you get on a longer, higher route (pictured on the right of the image in the article)
In these ones the problems are switched up every competition.
It's an entirely different discipline. Seeing lead climbers compete in speed climbing is like seeing Eliud Kipchoge matched up with Usain Bolt for sprinting.
Boulder and lead are at least somewhat related, although I still think it would be nice to see separate competitions here, too.
not every climber. Its clearly a different discipline. Thats why boldering/lead is separate from speed climbing.
But then lead is different to top rope and sport different to trad. And indoor is very different to outdoor.
It wasn't separate in the previous Olympics, that's my point.
>But then lead is different to top rope and sport different to trad. And indoor is very different to outdoor.
Except for trad most climbers I know do a mix of all of these.
Speed climbing is just so different from everything else that I don't even know one climber out of the hundred or so that I know that does it regularly.
Plenty of other sport are pretty much the same each time, particularly track and field.
Not to mention other variables outside of the human body. Perhaps the type of rope could matter in your performance. The age of the holds could matter too; even when the governing body standardizes on a replacement period for holds, I'm sure competitors would have strong opinions about the difference between a hold at the start of its service life versus one about to be replaced.
Also, the one thing I love seeing in physical contests is how competitors eke out the last bit of performance advantage with sheer willpower. Muscle memory takes care of the actions but performance and willpower is a conscious effort.
In short, no, I don't think competitors can literally turn off their conscious brain and just let muscle memory take over. If a field has jargon, there's a hell lot to geek over it.
or people who describe an expert as someone who is unconsciously competent, they no longer have to think about what they are doing they just "know".
And as yallpendantools is saying, operating at that level of expertise is much much more than sheer muscle memory, though the muscle memory is critical. My typing skill certainly helps me find flow in my programming work, I don't have to think about how to type each letter, when I'm in flow the ideas "just appear" on the screen.
So rather than saying the competitors turn off their conscious brain, maybe it is better to say that they transcend it?
Are you?
This reminds me a lot of how Brood War meta changes as new 'bugs' are discovered, since the fandom loves the game without it ever being touched so when edge cases are discovered they become part of the game rather than something to be fixed.
https://archive.nytimes.com/www.nytimes.com/library/cyber/we...
I never asked the distributed.net team, but I always suspected this was intentional to maybe thwart people from front-running a submission, so not truly a bug. If Adam, Jeff or David read this, maybe they could chime in.
1: 1, 10 :0
2: 2, 9 :-1
3: 3, 8 :-2
4: 4, 7 :-3
5: 5, 6 :-4
Since the sum on each line is 11, the sum of all the numbers from one to ten is 55.
The cool thing is that this generalizes
1: 1, N : 0
2: 2, N - 1 :-1
3: 3, N - 2 :-2 <-sum is N+1
…
N/2: N/2, N -(N/2)+1 => N/2+1
So the sum of N numbers is N/2 * (N/2+N/2+1) => N/2 * (N+1) if N is even.
It appears to be broken for odd numbers
1: 1, 9 :0
2: 2, 8 :-1
3: 3, 7 :-2
4: 4, 6 :-3
5: 5, 0 :-4 <- can’t reuse 5
But for odds, setting the odd number K equal to N+1, N is an even number so the total sum is sum(N) + N+1. We showed that sum(N) = N/2 *(N+1). So we have N/2 * (N+1) + (N+1).
But that means N/2+1 * (N+1) equals sum(K)
=> (N+2)/2 * (N+1)
=> (K+1)/2 * K
So the formula N/2 * (N+1) computes the sum of N numbers if N is odd too! It works for all numbers. Wow!
[Edit: Formatting]
I’m guessing the belay rope issue was real and the actual cause of the other lane always winning then.
I realize there's an infinite number of ways to accomplish this that would be more complex. What I was stating is the simplest possible way being pure prompt engineering. You could even try with OpenAI right now, I didn't try.
However
Off by one is the error, not the cause. sure, a lot of off by one errors are caused by index starting confusion, but not all. The hold is clearly off by one. thus its a "real life" off by one error.
People can mean something else than you when using the same words, and they might not be wrong as those word definitions might be acceptable too (and they might change over time too).
It is possible to use words incorrectly, but if 90% of people use a word in a certain way, that becomes the definition.
That is, you can use words in such a way that many people misunderstand what you mean. You could then call that usage "incorrect" if you want, and I wouldn't quibble too much with that, but the more important reason not to use those words in that way is that people don't understand you.
But in this case, this usage of "in real life" is not confusing anyone. Even those who pedantically quibble with it being "wrong" actually understood what was meant.