The old tables vs computer argument

[gcbryan]

I kind of don't get the on the fly part of it just for the sake of showing that you can do it on the fly. As a backup I can see the usefulness. The "we" speak gets a little old as well.

 

[Rainer]

Talk about too funny ... as I said, there is a certain level in the field (really in any field) where there are people who need do little more than read a bit and perhaps have a conversation, you may not understand that because you are not at that level and do not routinely move in those circles..

Look, the point is neither you nor the person you "learned" RD from has actually heard about it from the horses mouth. God only knows what differing assumptions you've made, especially seeing as the rules have been modified over the years. As such, it's pretty hard to take your assessments of RD seriously. You first talk about "catastrophic holes", but then can't provide any such specific cases, and I suspect you couldn't produce an RD deco schedule for most dives (something anyone who has sat through an RD class could do). If you can't generate the schedules, how can you debate their merits? In the circles in which I and others are expert (mostly related to my MIT Ph.D. work), I'm hard pressed to think of cases where we think we're too good to still learn something new. Then again, I try not to let ego get in the way of my academic education.

 

[Mr Carcharodon]

A note about ratio deco:

Ratio deco is basically a curve fitting exercise. There is NO underlying or unifying theory. Ratio deco is rather untested, when compared to "accepted" tables that were derived from a theoretical construct.

But ... ratio deco appears to work and the profiles that are derived from it usually fit on the "safe" side of those that are derived from more conventional models ... but the science has yet to be done.

The concern about the lack of an underlying theory is a bit odd. What is the underlying theory that established maximum pressure gradients, critical bubble volumes or gradient factors? All the models out there are based in empirical observation at some level. Go way outside of the previous observations and the dive may not turn out as planned. RD is pretty good about setting the boundaries and about setting methods for pushing the boundaries.

Also the concern about lack of validation has been overtaken by events. There are a large number of divers routinely doing deco dives on RD and not getting bent. If you have a system that works repeatedly there comes a point where claiming it might not work becomes unreasonable. I’d say that point has been reached.

 

[Mr Carcharodon]

I kind of don't get the on the fly part of it just for the sake of showing that you can do it on the fly. As a backup I can see the usefulness. The "we" speak gets a little old as well.

The on the fly part allows for the plan to be tailored for the dive real time. So if your planned depth is less than your real depth the dive plan changes rather than following an obsolete plan. Also it allows tailoring the deco for added stress during the dive or lost gas. Those are practical reasons that enhance your safety and go beyond just showing you can do it. I guess you could cut tables that covered a nominal plan and a set of contingencies. But the intent of RD is have a set of simple rules such that there is no need to consult a list.

 

[Thalassamania]

Look, the point is neither you nor the person you "learned" RD from has actually heard about it from the horses mouth. God only knows what differing assumptions you've made, especially seeing as the rules have been modified over the years. As such, it's pretty hard to take your assessments of RD seriously. You first talk about "catastrophic holes", but then can't provide any such specific cases, and I suspect you couldn't produce an RD deco schedule for most dives (something anyone who has sat through an RD class could do). If you can't generate the schedules, how can you debate their merits? In the circles in which I and others are expert (mostly related to my MIT Ph.D. work), I'm hard pressed to think of cases where we think we're too good to still learn something new. Then again, I try not to let ego get in the way of my academic education.

The idea that Hamilton, or any other professional table cutter, would have to sit through a presentation to understand something that simple is an expression of ego on your part. I never said that anyone (especially myself) is too good to learn something, just that when you work in the field it is often not necessary to spend the time to sit in a class that is designed for non-professionals, that's all.

But the intent of RD is have a set of simple rules such that there is no need to consult a list.

So which is it? So complicated that everyone must take a class or simple enough that even a professional decompression schedule designer and researcher with a PhD could grasp without personal coaching?

 

[Eric Sedletzky]

Diligence.

Personally, when I dive N/MDL tables, I use the average depth of my profile, not the maximum. I also know my table without looking at it (it simply consists of the limit at each 10' rather than a huge array of information that is useless to me, and is thus very easy to memorize). If a site put my average depth a little deeper or shallower than I expected, I would adjust either my bottom time or my ascent profile accordingly.

This is good.
I would also add that I would incorporate a series of deep stops into virtually every dive starting at 30 feet and working up to 15 feet.
Combine that and the use of nitrox and you're golden.

 

[Rainer]

The idea that Hamilton, or any other professional table cutter, would have to sit through a presentation to understand something that simple is an expression of ego on your part. I never said that anyone (especially myself) is too good to learn something, just that when you work in the field it is often not necessary to spend the time to sit in a class that is designed for non-professionals, that's all.

The fact that one is a professional table cutter doesn't mean one therefore inherently knows the rules of RD. That's just silly. What it does mean is that Hamilton is a very good person to ask about how well the schedules RD generates match against decompression theories, but one first has to know what schedules RD actually generates (and that does have to be learned, even by a "professional"). How else could one judge the validity of the curve fitting? You have to know what curves are being generated. As for RD, that depends entirely on how you define the rules (which standard gases, for what exposures, for what depths, for what conservatism, for what other assumptions, etc). These all have to be defined by someone. They are not some sort of universal givens. Hamilton cannot deduce what rules GUE/AG have defined based on any sort of first principles or understanding of decompression theory. They aren't one and the same. I don't know how else to make that clear.

 

[Blackwood]

I guess my major issues with RD are first of all the ... lack of substantial testing

Again, I have to disagree that it's not tested.

If I have a bag of M&Ms, all of which have been tested an approved as safe by the FDA, and I choose to only eat the green ones, I'm still eating something that has been tested.

That's as close an analog as my brain is willing to provide at 6:40a.

Ratio Deco is a gross simplification of decoplanner inputs (i.e. only a few gases), and observation about what schedules it produces with both Buhlmann and VPM algorithms using that reduced input set.

It's no less tested than the algorithms upon which it is derived, and between the two of them that represents the vast majority of empirical evidence out there.

(I'm possibly wrong, but I'm not really expressing myself well)

I am in complete agreement that if you are doing a new dive, one that is dramatically different than others you've successfully completed, by all means run them through the software to ensure that your profile is in the ballpark.

 

[Rhone Man]

I have never been trained in RD either, but the thing that seems "obviously wrong" to me about the concept is that there is a linear extrapolation between bottom time and decompression time, whereas just about every other decompression model agrees that deco time is exponential to bottom time (until saturation).

If you put enough padding on your algorithm, you can mask that. But as Mark Powell points out in his book, this leads to two significant disadvantages - firstly, it means that the level of conservatism varies significantly between profiles. Secondly, it only works between a normative range for bottom times - outside of that range it becomes unsafe.

On repetitive dives: to be honest, until I read this thread I wasn't aware that you could even do repetitive dives using RD, but the idea that you just "double the stops" doesn't sound terribly scientific (apologies if I have paraphrased Tracey incorrectly).

 

[rstofer]

Again, I have to disagree that it's not tested.

If I have a bag of M&Ms, all of which have been tested an approved as safe by the FDA, and I choose to only eat the green ones, I'm still eating something that has been tested.

That's as close an analog as my brain is willing to provide at 6:40a.

But do you know whether the FDA approved the green M&M's because they knew that they would only exist as a percentage of total M&M's?

Do you know whether 100% green M&M's was tested?

Before someone accepts test results, they should try to find out what was tested and how. Then they can try to fit there problems to the results of tests.

Richard