<Predicting the Past: The Hubbert LinearizationThe Oil DrumSo will it be the Emperor Coal?>

171 comments on DrumBeat: March 13, 2007

Comments can no longer be added to this story.

JoulesBurn on March 13, 2007 - 4:19pm Permalink | Subthread | Parent | Comments top

Let me see if I follow this logic:

Observation: Texas production will seemingly max out at about 62 Gb. The 1958-2004 Texas HL analysis gives a bit more than that, but there is a linear region prior to peak (6-7 data points) which seems to give an intercept closer to 62Gb.

Conclusion: This short linear region is more "Hubbertian" than the inflection (never mind the error bars on the value of Qt from so few data points, or what it means for that line to have a different slope than the latter 25-year linear region).

Observation: Texas was the swing producer. KSA is now the swing producer.

Conclusion: The inflection period reflected Texas' efforts to maximize production prior to the peak. Thus, the HL for KSA should also show a "Hubbertian" region followed by an inflection as they try to maximize production prior to the peak. Indeed, HL for KSA has a linear region from 1991-2002 followed by an inflection, which can be ignored in estimating Qt.

My Observation: To me, this underscores RR's point that concluding anything definitive from HL of KSA's production data is very problematic. WRT Texas, it seems more likely that, if production for any time period follows a geology-constrained model (Hubbert), it would be right before and over the peak--before they knew what was hitting them. Before then, production was constrained by the TRC. With KSA, you seem to want to focus on a decade when oil prices were crashing through the floor, call that the Hubbert-constrained region, and write off the upswing that happened as prices were climbing. In effect, you're saying that ramping up production after years of steady production, for whatever reason, is a sign of desperation. I would call it wanting to take advantage of increased demand.

Because of the economic and political factors that it cannot account for, the Hubbert model clearly cannot explain production trends over certain time periods, and there is no justification for excluding data which doesn't fit your expectations. Just because it's linear on an HL plot doesn't mean that it is "Hubbertian".

0
Comments can no longer be added to this story.
memmel on March 14, 2007 - 2:52am Permalink | Subthread | Parent | Comments top

You should read my posts on Roberts thread. This is exactly how HL should be used. I.e only use the points that fit the underlying assumptions not try to fit all the data.

You need reasonable and defensible criteria for excluding points. But the is the right way to do HL.

If you want to use all the data even if it does not fit the criteria for HL then use a different model.

0
Comments can no longer be added to this story.
Pitt the Elder on March 14, 2007 - 7:03am Permalink | Subthread | Parent | Comments top

This is exactly how HL should be used. I.e only use the points that fit the underlying assumptions

Assumption: KSA is peaking.
Test: use only data points that show KSA is peaking.
Conclusion: KSA is peaking.

Brilliant.

This is a combination of the logical fallacy of Begging the Question (assuming what you're trying to show) and of the data-analysis fallacy of Cherry-Picking (only using data that shows the conclusion you want).

The result is a completely invalid method of analyzing data. Let the data tell its own story.

0
Comments can no longer be added to this story.
JoulesBurn on March 14, 2007 - 8:14am Permalink | Subthread | Parent | Comments top

This is exactly how HL should be used. I.e only use the points that fit the underlying assumptions not try to fit all the data.

Well, you can do that...but it ain't science, pal. In the words of a 70's glam-rock band, "You're fooling yourself and you don't believe it".

0
Comments can no longer be added to this story.
step back on March 14, 2007 - 3:24pm Permalink | Subthread | Parent | Comments top

Wasn't it John F. Kennedy who said:

Ask not what your Delusion can do for you.
Ask what you can do to perpetuate your Delusion.

Well, it was something like that. :-)

Let's all do our part to perpetuate what ever delusions we have, lest they be forgotten and fall away into the dust bins of history.

Praised be the giver of all BS. Amen.

(BTW, there is overwhelming scientific evidence that my deity, the Holy BS Giver exists. Just look all around you. The proof is there. The Flying Spaghetti Monster, on the other hand, is a fraud. ;-)

0
Comments can no longer be added to this story.