Shearwater's take on the helium penalty

[RainPilot]

---

---

A ScubaBoard Staff Message...

This thread will not be allowed to turn into another bashing location. Any posts that are not directly about the topic will be summarily deleted, as will threads containing ANY form of personal attack.

---

 

[Storker]

On the deco stop, we can write the truth, and expose all the facts... without interference or fear of having your post deleted. Such freedoms are not permitted here.

Scubaboard does not censor facts. Scubaboard doesn't even censor wildly erroneous claims and leaves it up to the board members to refute those. That is a clear instruction to the moderators.

The board does, however, have Terms of Service that prohibit - among other things - profanity, off topic posts and personal attacks. Posts that stay within the ToS are not moderated, but posts which violate the ToS will be moderated.

And most people agree that for a statement to be accepted as fact, it has to be substantiated. Unsubstantiated claims without documentation cannot reasonably be regarded as facts.

 

[cathal]

On the deco stop, we can write the truth, and expose all the facts... without interference or fear of having your post deleted. Such freedoms are not permitted here.

Feel free to join in the discussion over there......

.

Truth = peer reviewed evidence. No need to leave this forum to provide the truth.

 

[Beau640]

Some people on here really need to take a statistics class before they attempt to criticize someone who's life and career revolves around this research...some of the arguments here are covered in stats 101

 

[stuartv]

In any event, the imprecise risk prediction by the model used to select the profiles does not invalidate the results of the study. Identical decompression profiles were performed with either heliox or trimix. The trimix dives resulted in two cases of DCS and the heliox dives resulted in none. This allowed the authors to conclude that the use of heliox was not inferior in terms of decompression efficiency to use of trimix.

Simon M

Hmmmm.... I guess there's a difference between knowing how to write software and knowing how to perform statistical analyses... How about that!

I am reminded of an anecdotal observation I have made in the past. It seems there are a lot of divers who claim to use XYZ algorithm for their diving and that XYZ is safe because they have done many dives using that algorithm without getting bent. But, when you drill into exactly what they do, they ultimately reveal that they use XYZ algorithm as a baseline and then they add extra deco time to it, usually at the shallowest deco stop depths, and in amounts that are determined by very precise "feelings". Which is a long way of saying that they don't actually dive XYZ algorithm at all and their claims regarding the safety of XYZ are almost negligent in their meaninglessness.

Thank you Dr. Mitchell for the work that you do and for continuing to be willing to share your insights here on ScubaBoard.

 

[rossh]

I am reminded of an anecdotal observation I have made in the past. It seems there are a lot of divers who claim to use XYZ algorithm for their diving and that XYZ is safe because they have done many dives using that algorithm without getting bent. But, when you drill into exactly what they do, they ultimately reveal that they use XYZ algorithm as a baseline and then they add extra deco time to it, usually at the shallowest deco stop depths, and in amounts that are determined by very precise "feelings". Which is a long way of saying that they don't actually dive XYZ algorithm at all and their claims regarding the safety of XYZ are almost negligent in their meaninglessness.

This is one of the main reasons we put our database together (150,000 records), and the priorities in the way it collects data. We wanted to really see what the world was up too, and see just how close people followed the plan.

Our desktop programs have tools to work the problem backwards as well. Load a graph image from a divelog, or even a picture of one from the dive computer, and use it as a backdrop image, planning over the top of it, to work out what kind of a plan one really did. (press the red/blue graph button, and add image). There is a pinning feature in the next version.

.

 

[stuartv]

This is one of the main reasons we put our database together (150,000 records), and the priorities in the way it collects data. We wanted to really see what the world was up too, and see just how close people followed the plan.

How do you know what their plan was, in order to know how closely they followed it?

Do you have data from those dives on what, if any, symptoms of DCS were presented following each dive?

Being anonymized, do you have data to know if any given dive was the first dive after a multi-day surface interval, versus a repetitive dive after a short surface interval?

Do you have any data on the exposure protection used during the dive? The diver's personal biometric data (height, weight, age)? How about if they were diving Nitrox but had their computer set to Air, for conservatism?

Most importantly, what conclusions have you been able to draw from the 150,000 dives in the database?

 

[rossh]

..answered below...

this is way off topic.....

How do you know what their plan was, in order to know how closely they followed it?

People use computers so the plan can be made on the computer, on the fly, which they follow.
It's a dive computer --- the diver enters settings and mixes - the dive computer re-calculates the ascent every few seconds based on progress..... It creates a ceiling. How close did the diver follow his new ceiling? Check the log. There is your answer. You can also back check the log and computer ceiling, as described above.


Do you have data from those dives on what, if any, symptoms of DCS were presented following each dive?

Yes.. that is part of the DAN and the PDE system, 5% (7000) records have been sent to DAN by the users through our divelog manager program integration to PDE (note: we have a copy).


Being anonymized, do you have data to know if any given dive was the first dive after a multi-day surface interval, versus a repetitive dive after a short surface interval?

The unit serial number links the dives in sequences. All dive records have time stamps. Surface interval is required for PDE... the PDE collects a full set of data, as does the dive log and the divelog manager.

Do you have any data on the exposure protection used during the dive? The diver's personal biometric data (height, weight, age)? How about if they were diving Nitrox but had their computer set to Air, for conservatism?

Yes - DAN has this as part of PDE reporting.

Most importantly, what conclusions have you been able to draw from the 150,000 dives in the database?

Plenty: Things that we wanted to know about techniques, rates, styles, mixes, common profiles and settings, etc, etc. The big question you are trying to ask - injury rate - we know it too (very low), and you will have to wait for DAN to publish official results.

************

This is the biggest and most comprehensive set of information on current tech practices available.... public, searchable, free: database.

.

 

[stuartv]

this is way off topic.....

How do you know what their plan was, in order to know how closely they followed it?

People use computers so the plan can be made on the computer, on the fly, which they follow.
It's a dive computer --- the diver enters settings and mixes - the dive computer re-calculates the ascent every few seconds based on progress..... It creates a ceiling. How close did the diver follow his new ceiling? Check the log. There is your answer. You can also back check the log and computer ceiling, as described above.

Ah. We are not speaking the same language. What my dive computer tells me on the fly is not my "plan". My plan is what I work out in Multi Deco or PastoDeco or Subsurface before I get in the water.

How closely did the diver follow the on-the-fly calculated ceiling is (to ME) a different question than "how closely did the diver follow their plan".

Thank you for clarifying.

Do you have data from those dives on what, if any, symptoms of DCS were presented following each dive?

Yes.. that is part of the DAN and the PDE system, 5% (7000) records have been sent to DAN by the users through our divelog manager program integration to PDE (note: we have a copy).

Cool. So, what percentage of those 7000 dives had symptoms or suspected symptoms of DCS? Roughly speaking? Distribution by algorithm of the computer they were using?

Most importantly, what conclusions have you been able to draw from the 150,000 dives in the database?

Plenty: Things that we wanted to know about techniques, rates, styles, mixes, common profiles and settings, etc, etc. The big question you are trying to ask - injury rate - we know it too (very low), and you will have to wait for DAN to publish official results.

These are conclusions based on the 150,000 dives in your database? The 7000 you transmitted to DAN? Or the 200,000 dives in DAN's database, of which 7000 came via your software?

I'm not asking about injury rates necessarily. I don't want to presuppose what kinds of conclusions you might have been able to draw. As far as I know, you are not a representative of DAN, so I'm not asking for conclusions based on their data. I'm specifically asking you about your data. What conclusions have you been able to draw from the 150,000 dives in your database? If there are many, how about a couple of highlights? I'm just curious where the state of the art in dive research is heading.

 

[Dr Simon Mitchell]

The big question you are trying to ask - injury rate - we know it too (very low), and you will have to wait for DAN to publish official results.

Cool. So, what percentage of those 7000 dives had symptoms or suspected symptoms of DCS? Roughly speaking? Distribution by algorithm of the computer they were using?

Stuart,

Ross has claimed to have outcome data for a very small proportion of the dives on his database for a long time, but has never revealed it. I don't know what to make of that. If the outcomes are in his database it is his data, and there is no reason at all why he could not reveal it here. I will contact Petar Denoble at DAN and ask about Ross's data - see if we can get it that way.

It has been pointed out to Ross before that his database (particularly any outcome data) has to be interpreted cautiously because the data are recorded retrospectively and voluntarily. Thus, there is the potential for a substantial reporting bias in the dives that get up-loaded. Divers who suffer problems may wish to bury all trace of their problem dives and outcomes (in the clinical realm we find this is a frequent problem with technical divers who prefer no one to know about their issues), and would be the least likely to upload dives onto a database and report that they had a related issue. Thus, extrapolating incident rates from data collected this way to the wider population is tenuous. That's why in the hey-day of Project Dive Exploration much of the data was collected by observers in the field who actually watched and talked to divers after dives recorded on data loggers.

Ross's database certainly has its uses, but it is unlikely to reflect an accurate incident rate, even if dive outcomes were always recorded (which they are not in the vast majority of cases).

Simon M