Evolving Thoughts on Deep Decompression Stops

[dmaziuk]

This statement ignores the additional gas loading the diver incurs if they lower the low GF.

I very deliberately never said anything about keeping the total ascent time constant, and the author of gradient factors very clearly stated that lower GFs will increase overall decompression time. The only thing that statement ignores is your unstated assumption that GF low will be increased without padding the shallow stops accordingly -- and the point is that is not how the numbers were designed to work. Wheh used as designed, the model will account for additional on-gassing, and factor it in in subsequent -- longer -- stops.

Where Spisni study is relevant is their use of RANTES and lower score on that test for shorter decompression times. Hopefully they'll be using it in all future studies together with Doppler ultrasound and we'll see more verifiable data points and work out reproducible trends and all that.

 

[dmaziuk]

So if, for example, a 70/90 profile has probability of DCS = p, it does not follow that 10/90 or 20/90 will have probability of DCS <= p. I would expect, as in the Spinsi trials, that there could be areas where the additional time given to the 10/90 profile (for example) would not be enough to counter the detrimental effects of the additional gas loading.

If the given set of M-values is unsafe at any point at <100% of the value, it is also unsafe at 100%. Therefore this set of M-values is unsafe and should never have been used in the first place. I.e. Buhlmann's model is wrong by of reductio ad impossible.

 

[tursiops]

If the given set of M-values is unsafe at any point at <100% of the value, it is also unsafe at 100%. Therefore this set of M-values is unsafe and should never have been used in the first place. I.e. Buhlmann's model is wrong by of reductio ad impossible.

You, sir, live in a very strange purely black-and-white world, and an unusual logic system.
Most of what you post is either spurious, specious or wrong. The post quoted is another good example.

 

[rjack321]

If the given set of M-values is unsafe at any point at <100% of the value, it is also unsafe at 100%. Therefore this set of M-values is unsafe and should never have been used in the first place. I.e. Buhlmann's model is wrong by of reductio ad impossible.

You must like to hear yourself talk. Instead of your dubious logic, why don't you try testing an actual hypothesis?

Does 10/90 have less prevalence of DCS risk than 100/100? Your logic cannot answer this question. If you think it can, you don't yet understand integral supersaturation.

 

[dmaziuk]

Does 10/90 have less prevalence of DCS risk than 100/100? Your logic cannot answer this question. If you think it can, you don't yet understand integral supersaturation.

Again, ISS was not a part of ZHL+GF model as designed. If ISS shows that the model is incorrect, than the model is incorrect, it's as simple as that. Many people agree incl. esteemed Drs Yount and Wienke.

 

[Storker]

I have the impression that someone is unable - or unwilling - to understand that risk/probability is a continuous function, not a digital yes/no function

 

[rjack321]

Again, ISS was not a part of ZHL+GF model as designed. If ISS shows that the model is incorrect, than the model is incorrect, it's as simple as that. Many people agree incl. esteemed Drs Yount and Wienke.

Cite your source for that, I'd love to see where Yount and Wienke discuss ISS.

And ISS doesn't have anything to do with risk, ISS is a logical explanation of why 10/90 can have a high DCS risk than 70/90.

 

[dmaziuk]

Yes, yes, a line drawn through grey area. BTDT, got the t-shirt.

I use a couple cheminformatics toolkits at work, about 10% of their code are covering the basic principles. The other 90% are covering the "special cases" like metal-metal bonds and clathrates and whatever else you people come up with... and that's at the basic chemistry level. Don't get me started on molecules that decided to link into long-chain polymers and to crawl forth and multiply.

PS where do you think my .sig came from

 

[1atm]

Well, even if you leave the ISS aside for a moment, it’s still important to keep an eye on the efficiency of the deco. Total deco time does matter, and does affect the risk profile. Sure, not much will happen if you put the first stop a little deeper if this results in +5min deco. But once you start adding significant time, in cold water, ... the risk does go up. If you dont pad the time, then even worse.

 

[dmaziuk]

Cite your source for that, I'd love to see where Yount and Wienke discuss ISS.

So replace M-values with ISS. Or use the smaller of (M-value at 50/75, ISS). A modern computer would have more than enough cycles to run this in real time. Either way, what you are saying is the basic premise of the dissolved gas model: that you're "safe enough" as long as you're under M-values, is wrong.

Yount and Wienke thought a basic premise of dissolved gas model was wrong because bubbles are always present. You think a basic premise of dissolved gas model is wrong because ISS. Notice the similarity?