UTD Decompression profile study results published

[Patoux01]

One reason could be that I'm a cheapskate and don't want to spend 2k$ on computers.
Another reason could be that I don't bring my laptop on the boat, so that if there's a change in the dive, I can still dive and not go "Oh but my plans are at home"?.
Sure, I could carry every possible table with me... That takes forever to make, takes a bit of space, and quite some fiddling when I look through them. I could only keep "some of them", and go for a "well, my profile is close enough to this one, so that deco will do as well" approach and call it "tables-ratio", which is what RD does.

Personally, even under very high stress, I'm able to compute 15*2. Now others may or may not be able to do that, but that is irrelevant to whether RD can be used or not. Not everyone can dive a rebreather, that doesn't mean no-one should use them. The "maths" in RD are a middle school level at best.

 

[huwporter]

Edit: I'll explain a bit more what I mean with "does it matter how we compute an ascent plan?"" I could compute a deco schedule by hand using Buhlmann algorithm, then I could implement it on Matlab on my 64-bits computer, and then I can do it on an 8-bit mcu. Do you think the end result are gonna be dramatically different? The errors are gonna be in the "close enough" range. What RD (the methodology) tries to do is to make something that lands in the "close enough" range. What UTD-thing tried to do is "Lol, physics? Who needs that. I'm just gonna put some here, and there, and some more here, all good that's the best we can do".

Exactly. As soon as you get in the water, you are 'cutting with an axe' any deco schedule anyway. Doesn't matter whether you are diving tables, or a dive computer or whatever, generating a platonic ideal profile with floating point arithmetic, your depth is never going to be exactly at the planned depth to arbitrary decimal places, your stop times and ascent rates are never going to be split second accurate, and your body isn't a perfect suite of Buhlmann compartments, precisely exponentially on- and off-gassing. A couple of minutes here and there really makes no difference at all when compared to all that messy physiology and real world conditions. Your personal variability from day to day will have more of an impact.

However the study does clearly conclude that the UTD Ratio Deco(TM) profile is NOT better than straight GF30/80 - and there is some evidence that it is worse - in spite of being 11 minutes longer. The UTD profile moves significant amounts of time around compared to GF - it arrives at 15m a whole 11 minutes later than the GF profile, more than a third of the entire GF deco time - in pursuit of a range of pseudo-scientific principles (deep stops, oxygen window).

The study doesn't narrow down why, or which of the lucky-rabbits-foot inclusions in the UTD profile might be responsible, or whether there might be other profiles that are even better - but given a choice between these two profiles you can at least conclude that you might as well get out of the water 11 minutes earlier with some indication of lower decompression stress by following the standard GF profile, and all those pseudo-scientific additions were at best pointless, at worst counter-productive.

The study's main flaw is that if the UTD profile HAD been found superior, you wouldn't know if that was because it was intrinsically superior or because it was 11 minutes longer. But that didn't happen.

 

[UWSojourner]

Personally, even under very high stress, I'm able to compute ...

Kudos to you. But most aren't

 

[Kevrumbo]

I disagree: the study proved that "AG's lucky rabbits feet UTD Ratio Deco(TM)" schedule is not as efficient as a GF 30/80 schedule. It sheds no light on ratio deco as a methodology. . .

And there's no need to rhetorically denigrate UTD Ratio Deco as "lucky rabbits feet" either, because that implies that GUE's version is just as illegitimately "voodoo" as well.

GUE's Basic Cascading Ratio Set Points for the O2 segments are the same or nearly similar as UTD's:
150’/45m O2 segment = 1/2 * Bottom Time;
200’/60m O2 segment = 1 * Bottom Time:
250’/75m O2 segment = 1.2 * Bottom Time;
300’/90m O2 Segment = 1.5 * Bottom Time;
350’/105m O2 segment = 2.2 * Bottom Time;
400’/120m O2 segment = 3 * Bottom Time.

The tenets behind Ratio Deco are perhaps "mis-taken" now in hindsight with regard to bubble models and the NEDU Deepstops Study, but in its time, the Ratio Deco method and the explanation of "stopping the diver much deeper than in traditional decompression algorithms, bringing them up slower without penalizing for the deeper stops, and then making the shallow stops much shorter with the decompression effectively completed at depth and not at 10’(3m)" -it all seemed attractive and authentic given AG's pride in his accomplishments and confidence (or now cynically, AG's hubristic pride with self-aggrandizement and arrogance).

 

[UWSojourner]

The tenets behind Ratio Deco are perhaps "mis-taken" now in hindsight with regard to bubble models and the NEDU Deepstops Study, but in its time...

Mistaken is a good way to describe Ratio Deco.

 

[Patoux01]

And there's no need to rhetorically denigrate UTD Ratio Deco as "lucky rabbits feet" either, because that implies that GUE's version is just as illegitimately "voodoo" as well.

Sorry what?

(from Huwporter UTD Decompression profile study results published )

 

[steady1570]

Does anyone have the "third" graph from above - ie UTD-RD(TM) vs GUE? And perhaps even a fourth with all 3 profiles on one? Would be interesting to view but I'm not infront of a computer to do it myself...

-Mark

 

[huwporter]

To be clear - the 'GUE' profile isn't anything official from GUE; GUE does not dictate official profiles. It is a ratio style profile consistent with GUEs current SOPs that I'd happily dive myself and propose to my team. And it started with... wait for it... the GF 30/80 profile from the test.

 

[steady1570]

Yes, that makes sense... so perhaps more accurately combining the graphs of the UTD-RD(TM) algorithm vs the ratio profile that uses your view of the methodology discussed by GUE to generate a 30/80 profile "on the fly"?

From my interpretation of his post, Kevrumbo is suggesting that the "GUE-style" profile is basically the same as the UTD one, which from the graphs posted doesnt initially appear to be the case... and so putting both on the same graph would allow a closer comparison (for my poor eyes at least).

But I take your point about cutting with an axe etc, and perhaps its not a particularly useful comparison in the context of this thread anyway - as I understand it the UTD-RD(TM) is treated as a decompression algorithm, which was compared to the Buhlmann 30/80 algorithm. "GUE-style" ratio deco methodology has been brought in slightly erroneously, as its a methodology used to generate on the fly plans that mimic your algorithm of choice within certain conditions (and here the example is the 30/80 profile). I have no particular knowledge of the history of DIR / GUE / UTD etc so its all interesting to me, particularly how the different terms have become conflated.

-Mark

 

[huwporter]

Yeah, you have that pretty straight.

There is no claim here that the 'GUE'-style profile is better than the GF one - just that it is close enough in the real world, and that
3-3-3-3-3-15
is easier to remember than
1-1-2-3-5-16