GFHi - practical meaning?

[njdiverjoe]

Except that the empirical data doesn't really support the "more conservatism: less risk" conjecture. It is probably true, up to a point, but as far as the recorded numbers are concerned, you have X in Y risk at low conservatism, and you still have X in Y risk at low conservatism. Well, maybe X in Y+1.

What exactly are you trying to say? That there is not an increased risk of surfacing with a GF of 100 versus one of 80?

I disagree with what your saying. Would you be able to point me to some data source demonstrating your point?

 

[dmaziuk]

OK. Here's the guys who have 10 divers bent at <= 60 vs 8 divers bent at >= 100.

 

[scubadada]

Except that the empirical data doesn't really support the "more conservatism: less risk" conjecture. It is probably true, up to a point, but as far as the recorded numbers are concerned, you have X in Y risk at low conservatism, and you still have X in Y risk at low (high?) conservatism. Well, maybe X in Y+1.

Check out the SAUL recreational dive planner you cited in another thread Recreational Ascent Rate in the last 15 feet

Do a dive on 32% to 80 feet with Shearwater low conservatism (45/95), 44 min, and your probability of DCS is 0.31%. On medium conservatism (40/85), the dive would be 36 min with a P(DCS) of 0.13%. At high conservatism (35/75), the dive would be 28 min with a P(DCS) of 0.02%. Sure these are all small numbers, these are no stop dives on a commercially available dive computer. These correspond to a risk of 1/323, 1/769. and 1/5,000. This is more than an order of magnitude difference in risk. As you know, the SAUL algorithm includes a routine 3 min SS.

What it really comes down to is one's own risk tolerance within a spectrum something like this. Of course, few to no divers take every dive to the NDL limit. Many dives would carry less risk than the examples.

 

[njdiverjoe]

OK. Here's the guys who have 10 divers bent at <= 60 vs 8 divers bent at >= 100.

Thank you for the link!

Except that the empirical data doesn't really support the "more conservatism: less risk" conjecture. It is probably true, up to a point, but as far as the recorded numbers are concerned, you have X in Y risk at low conservatism, and you still have X in Y risk at low conservatism. Well, maybe X in Y+1.

The data in the paper actually contradicts your claim. The authors say that "Depth, dive time, and GF were statistically higher p < 0.0001, p = 0.001, p < 0.0001" are DCS risk factors.

If you check out Table 3 they specifically call out that "Increase GF increases DCS" with p < 0.0001. That's not to say that there aren't additional DCS risk factors, but GF is definitely one of them.

 

[njdiverjoe]

What it really comes down to is one's own risk tolerance within a spectrum something like this. Of course, few to no divers take every dive to the NDL limit. Many dives would carry less risk than the examples.

I disagree -- there are a lot of tech divers which take almost every dive past NDL.

 

[scubadada]

...When you select low/med/high conservativeness on your computer, you are making a decision on the maximum acceptable risk of DCS that you are willing to tolerate in the event you need to make a direct ascent to the surface. It should not include the (optional but highly recommended) safety stop, if it did then it would be a mandatory stop and now we are quickly moving away from recreational diving.

Except that the empirical data doesn't really support the "more conservatism: less risk" conjecture. It is probably true, up to a point, but as far as the recorded numbers are concerned, you have X in Y risk at low conservatism, and you still have X in Y risk at low conservatism. Well, maybe X in Y+1.

Check out the SAUL recreational dive planner you cited in another thread Recreational Ascent Rate in the last 15 feet

Do a dive on 32% to 80 feet with Shearwater low conservatism (45/95), 44 min, and your probability of DCS is 0.31%. On medium conservatism (40/85), the dive would be 36 min with a P(DCS) of 0.13%. At high conservatism (35/75), the dive would be 28 min with a P(DCS) of 0.02%. Sure these are all small numbers, these are no stop dives on a commercially available dive computer. These correspond to a risk of 1/323, 1/769. and 1/5,000. This is more than an order of magnitude difference in risk. As you know, the SAUL algorithm includes a routine 3 min SS.

What it really comes down to is one's own risk tolerance within a spectrum something like this. Of course, few to no recreational divers take every dive to the NDL limit. Many dives would carry less risk than the examples.

I disagree -- there are a lot of tech divers which take almost every dive past NDL.

Your initial post was about no stop/rec dives. My post was prompted by a comment on the same topic. I thought that my post was also clearly addressing no stop/rec diving. I guess I should have explicitly made it clear, to avoid confusion.

 

[rsingler]

OK. Here's the guys who have 10 divers bent at <= 60 vs 8 divers bent at >= 100.

As noted by @njdiverjoe above, this is incorrect.
Increased GF was associated with incidence of DCS. I believe you have misinterpreted the findings.
Not that a post hoc study with a host of other variables is the way to address this in the first place.

 

[dmaziuk]

As noted by @njdiverjoe above, this is incorrect.
Increased GF was associated with incidence of DCS. I believe you have misinterpreted the findings.

In-Depth Analysis of GF-Value in the 320 DCS Cases (Table 4)
✓ Only eight cases (2.5%) showed a GF > 1
✓ 14 cases had a GF > 0.9 (4.4%)
✓ The majority of cases (236–73.7%) showed GF-values between 0.70 and 0.90
* 37.5% between 0.8 and 0.9
* 36.2% between 0.7 and 0.9
✓ 46 cases (14.4%) had a GF lower than 0.70
✓ 10 cases (3.4%) lower than 0.60
✓ Only 3 cases had a GF lower than 0.50

Clearly, since 73.3% of the cases were in GF range of 0.7..0.9, then GF 0.7..0.9 is the riskiest of 'em all.

Also clearly, since 4.4% cases had GF > 0.9, but 14.4% had GF < 0.7, then GF < 0.7 is three times riskier than GF > 0.9.

Puh-leez.

 

[rsingler]

Fifteen people had fatal accidents walking down the sidewalk.
One person had a fatal accident walking down the street.
Walking on the sidewalk is fifteen times as dangerous as walking in the street.

Puh-leez.

If I am reading Table 3 above correctly,
Median(StDev) GF among DCS victims was 79, versus
Median(StDev) GF in whole database was 66, with a P<.0001
Ergo, there is a statistical association between higher GF and DCS, notwithstanding the distribution of cases within the database.

 

[dmaziuk]

If I am reading Table 3 above correctly,
Median(StDev) GF among DCS victims was 79, versus
Median(StDev) GF in whole database was 66, with a P<.0001
Ergo, there is a statistical association between higher GF and DCS, notwithstanding the distribution of cases within the database.

You're looking at the posterior distribution and you're positing that the likelihood function is linear increase in surfacing GF. It does not follow. It's just as likely that all of the cases were diving at their computers/tables default conservatism and got bent well within the predicted -- using Craig's numbers above for illustration -- 1 in 323 or 1 in 769. It just so happens that in 75% of the cases that default was in the range of GF 0.7..0.9. Correlation is not causation, with a P of "absolutely certain".

If you run from the priors, like SAUL, you'll have to cut your bottom time in half and your GF: by 100%, to shift your chances from 1/500 to 1/5000. And still some 30 minutes after you surface you'll either be 100% bent or 100% not.