The Oil Drum

171 comments on DrumBeat: March 13, 2007

Comments can no longer be added to this story.

memmel on March 14, 2007 - 2:52am

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.

Comments can no longer be added to this story.

Pitt the Elder on March 14, 2007 - 7:03am

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.

Comments can no longer be added to this story.

JoulesBurn on March 14, 2007 - 8:14am

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".

Comments can no longer be added to this story.

step back on March 14, 2007 - 3:24pm

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. ;-)