>At surface level, the adage emphasizes the consistency that arises when information comes from a single source and points out the potential pitfalls of having too much conflicting information. However, the underlying message is to question the apparent certainty of anyone who only has one source of information. The man with one watch has no way to identify error or uncertainty.
This ambiguity between endorsing one or two is why I kind of prefer the version of this law Fred Brooks cited:
>An ancient adage warns, "Never go to sea with two chronometers; take one or three." The same thing clearly applies to prose and formal definitions. If one has both, one must be the standard, and the other must be a derivative description, clearly labeled as such. Either can be the primary standard. Algol 68 has a formal definition as standard and a prose definition as descriptive. PL/I has the prose as standard and the formal description as derivative. System/360 also has prose as standard with a derived formal description.
In this context, the reader can be much more trusted to see that one chrono beats out two only by being cheaper (which might be worth it for short voyages), and one description beats out two only by being unambiguously the standard (which might be worth it for small or exploratory projects).
Now do the same with models of reality, but extending into the realm where millions of variables are involved, many of them downstream from consciousness, and culture (cultural norms of "logic", etc). To handle this one will need a lot more than a few simple methods. Alternatively, faith in almost any ideology (religion & science are most powerful) can make things appear to line up perfectly. If you don't believe me, just listen to how people talk, there is a near infinite supply of evidence, and more every day.
When one is not able to understand what another says, assuming the other person is necessarily incorrect/silly may not be optimal gameplay (depending upon what variable(s)[1] one is optimizing for of course, or if one never even thought about that aspect of it in the first place).
Less snarkily: the 1/2/3 watches example is a single, ~deterministic, ~knowable data point; it is roughly the simplest type of problem space you can deal with, and even here there is uncertainty. Now consider a much more complex problem space, reality (better known as "the reality"), and things get much worse. But paradoxically, with all the added complexity, "correct" answers often become simpler, not more difficult. (Here I am referring to the conversations you may find on social media (or anywhere) about complex matters such as war.)
Why? It's neither the subject at hand nor have you expressed it meaningfully. I dismissed you because I'm optimizing for productive concrete discussion in the face of your social gaming, and you argued against me on the same grounds only because you lack the self-awareness to see that. Assuming conclusively someone is necessarily incorrect only because they were snarky when they ignored you is not an optimum strategy.
In the text of your comment. You treat your interpretation and perception of what is going on as being equal(!) to what is going on, despite I'm quite confident possessing knowledge of how unreliable perception is, according to both science and common experience.
Why do You People do this? Why do you not behave more logically?
Larcum Kendall's K1, a copy of John Harrison's H4,
H4 had cost over £20,000 (inflation adjusted £3,260,000) to develop. Kendall's K1 cost £500 (now £74,000) and his cheap model, K3, cost £100 (now £13,400), but by the time the Beagle voyages were over the cost of a good chronometer had fallen to under £40 (now £4,200).
> but by the time the Beagle voyages were over the cost of a good chronometer had fallen to under £40 (now £4,200)
Reminds me of the "wait calculation" paradox, where it's always better to wait to start building your interstellar spaceship, because if you wait another 50 years, you will be building a better spaceship that will shave off more than 50 years of your flight time. [1] (Ok, in the original research paper it's not a paradox, but I like to think of it as one.)
It's awesome that we're seeing the same thing happen to spaceflight now with SpaceX developing fully reusable spacecraft, with simple resources (methane instead of hydrogen or hypergolics) and stainless steel.
I was once told by an older and wiser scientist that you should never measure the same phenomenon with two instruments. If you do, your entire study turns into an inter-instrument comparison.
This is also the reason why I avoid using analogies when making an argument. Usually, instead of talking about the argument itself we end up discussing under what circumstances the analogy works and when it doesn't.
Interesting point (arguments do frequently devolve into pointless meta-arguments), but I'd be surprised if you can totally avoid analogies. Analogies are extremely powerful. They're arguably all we have when it comes to reasoning.
Ironically, Segal's Law, which you seem to approve of, is itself an analogy. I don't think anyone here is really concerned with knowing what time it is or whether or not it's a good idea to wear two watches.
No, not really. It's like, if you go to the store and realize you forgot your wallet, would you say we're talking about a wallet? No, we're talking about a trip to the store.
I think the commenter did fully grasp the meaning. Segal's law doesn't give you any clearly-defined system for grappling with uncertainty, it merely points out the existence of uncertainty and one way that we turn a blind eye to it.
The best approach in some cases is to ignore uncertainty and move forward - carrying one watch, as it were.
Two other valid approaches are voting and medians. These are both accomplished with 3 watches.
The crux of Segal's law is that all measuring instruments are imperfect. We could bicker endlessly about the best way to use our tools, but at some point someone's just going to need to know what the damn time is.
Or, use Bayes theorem for an even stronger belief estimate of the 'true' local time relative to a prior-established 'universal' time. Perhaps I dropped one of my watches in the water last week or I notice the second hand on one of the watches is sticky. Bayes can account for that anecdotal information.
I don't understand a mathematical part of your message. Suppose you have 2 watches, does Bayes theorem allows to make a better judgement about time than calculating an average time between 2 devices?
Depends if your measurements are discrete or continuous. In a continuous domain with arbitrarily long precision, your voting system's failover triggers 100% of the time.
Your version is basically an analogue of explaining a joke. Figuring it yourself is the point.
Besides that, I believe your version explains only one facet of the quote. Another interpretation could be analogous to paralysis of choice - in practice both watches are precise enough, focusing on the difference is just misleading you.
I mean, how is this true if he doesn't know which watch is "wronger" and by what margin of error? If it is 2pm and one watch shows 2pm and another 1:30pm the average is worse off than having one watch that shows 1:46pm.
If you have 2 watches with let's say 5-minute difference, I have a pretty high confidence that the indicated hour is correct. The chance that both watches e.g. broke in the same way is small.
I don't think you will necessarily get a higher precision, but your confidence of the indicated time is higher.
The saying makes less sense to a modern audience. We can all pull out our phones and likely show the same time due to network time sync. BUT, what if we have 1000 phones all showing the same time, and they are ALL wrong? New layer to the sentiment....
This ambiguity between endorsing one or two is why I kind of prefer the version of this law Fred Brooks cited:
>An ancient adage warns, "Never go to sea with two chronometers; take one or three." The same thing clearly applies to prose and formal definitions. If one has both, one must be the standard, and the other must be a derivative description, clearly labeled as such. Either can be the primary standard. Algol 68 has a formal definition as standard and a prose definition as descriptive. PL/I has the prose as standard and the formal description as derivative. System/360 also has prose as standard with a derived formal description.
In this context, the reader can be much more trusted to see that one chrono beats out two only by being cheaper (which might be worth it for short voyages), and one description beats out two only by being unambiguously the standard (which might be worth it for small or exploratory projects).
reply