Some ‘scientific commonsense’ to clarify a few misunderstandings that seem to have arisen since John Campbell’s interview with astrologer Ken Ring.

(If you came here looking for my ’continuously updated’ comments on the 2011 earthquakes the aftershocks go here – older comments are appended to this post.)

I am writing this in response to commentary I have seen since John Campbell’s interview of astrologer Ken Ring and the articles by my colleagues. It’s now very late in the act I know, but I hope these clarifications might still be of help for a few readers and provide a (partial) round-up of sorts.

We’ll see what happens on March 20th

On a number of forums I’m seen people writing ‘wait for March 20th’ or the like as if this were the arbiter of if Ken Ring is right or not.

What happens on March 20th will not determine if his method works.

What matters is if the ‘predictions’ are meaningful and if he is able to reliably predict earthquakes.

You can test if Ken Ring’s ‘predictions’ are not worthy of the word prediction before March 20th by comparing them with what you’d get by ‘dumb luck’ using the typical rate earthquakes occur.

Let’s leave aside the (lack of) plausibility of his methods and that he makes multiple predictions–as vital as these are to judging a method*–and look at the prediction he made for March 20th. (David has covered this.)

You don’t need to be a geologist or use ‘real maths’ to see this prediction is not meaningful. (I’m not a geologist by the way.)

On a colleague’s blog Ken Ring softened his March 20th prediction to a magnitude 4 to 6 earthquake within 500km of Christchurch:

I do not hold that 20 March WILL bring a severe earthquake to Canterbury, but an extreme weather event is possible that day worldwide, and an earthquake within 500kms of the Alpine Fault is a risk on that date. More likely to be a 4-6mag.

Let’s also overlook the vagueness of this prediction and the weather ‘prediction’ too.**

New Zealand gets roughly 333 magnitude 4 or greater earthquakes a year. This makes ‘predicting’ a magnitude 4 to 6 earthquake on any one day somewhere in New Zealand pretty much meaningless.

In addition to that, there will be aftershocks near Christchurch following the February 22nd earthquake, including a M=4.8 late last week. GeoNet has a guideline forecast of these aftershocks, scroll down past the map to the bottom of the table.

The ‘prediction’ isn’t meaningful because the chance of it happening ‘by dumb luck’ is too high.

Similarly, if Ken Ring makes multiple ‘predictions’ in a way that covers most outcomes, then the predictions wouldn’t be meaningful either. David and Alison’s articles cover this well.

The key point I wish to add to what they have written is that whatever happens on March 20th is not really very important. What is important is how meaningful the predictions are, and their accuracy. (Once you have predictions that are meaningful you still need to test if they are accurate enough to be useful.) These are essentially the key points David and Alison were making.

This research area has been dismissed by scientists 100 years ago

Quite a few people have said that scientist in the interview by John Campbell said that how the moon might affect earthquakes was last studied 100 years ago, and not since.

What he actually said, but perhaps didn’t make clear enough for many viewers, was that the classic studies of the effect of the moon on earthquakes were done over a hundred years ago, and then went on to summarise the current understanding from more recent work. There is more recent work relevant to this. He didn’t say this explicitly; it was left implied in his description of the current understanding.

It seems an innocent enough misunderstanding and just a bit unfortunate. Unfortunately, too, the internet and gossip is good at spreading these sorts of misunderstandings.

A few days ago posted a Radio Wammo interview of an author of a research paper published in 2004 investigating possible effects of large tides on earthquakes. (Large tides are of course associated with close points in the lunar cycle.)

Ken Ring predicted the February 22nd earthquake

This has already been covered by my colleague, Alison Campbell, so I’ll keep this brief. Ken Ring claimed ’Over the next 10 days a 7+ earthquake somewhere is very likely.’ Not New Zealand or Canterbury, but anywhere on earth. No magnitude 7 or greater earthquake anywhere occurred during that time. Simply put, his prediction failed.

He picked Christchurch’s lesser magnitude 6.3 as ‘success’ after the fact.

Good prediction does work not by picking ‘second-best’ results after the fact. You design the experiment, and the ‘rules’ it works under, before you start and you stick to them.

(There was also a magnitude 6.6 south of Fiji during his prediction period. I haven’t seen him mention this despite that it occurred before the lesser magnitude 6.3 earthquake at Christchurch so he should have been aware of it at the time of calling the 6.3 ‘success’.)

Interview tidbit

One of the things I wrote in my notes while listening to the interview ‘live’ was:

Ring: “Quakes that scare people the most”

I thought it very telling that Ken Ring let slip that his target was the scared, rather than the vulnerable. It is a shame that John Campbell hadn’t picked up on Ring’s Freudian slip.

Convincing people that his predictions work

In a Science Media Centre Alert, scientists Matt Gertenberger and David Rhoades offered this website as a service that Ken Ring might use to have his method tested. (A copy of this report can be found at WeatherWatch with comments from others, including Ken Ring.)

If Ken Ring is genuinely sincere about his predictions, he ought to publish his methodology and have them tested by the scientific community. The correct ‘media’ for resolving scientific matters is ultimately the scientific literature, not TV or radio.

Follow-ons

17th March, 7:50pm: I’ve written an update in the comments; see the second comment, below. You may also wish to read the press conference about the science behind the earthquake.

20th March: Gareth Renowden, who knows of Ring well from his weather ‘predictions’, has added his thoughts. (For comments you’ll need to read his original site.) It’s interesting to learn from Gareth’s article that the revised ‘prediction’ I present in the second comment of the comments below is in fact itself a weakened revision of an earlier ‘prediction’ that specifically gives a location for the epicentre (‘some geographical point between Hanmer and Amberley’) that he subsequently removed. (Compare the prediction for March19-21 I quote and the one Gareth quotes.)

Footnotes

One way to get an anecdotal (that is, non-scientific) feel for how often magnitude 4 or greater earthquakes occurs in New Zealand is follow GeoNet’s @geonet_above4 twitter stream.

* You can’t ignore these really, but even putting these aside his ‘prediction’ is meaningless.

** I strongly suspect that if you were to look up the frequency of severe weather events worldwide, this ‘prediction’ suffers exactly the same problem as his earthquake prediction. Those with more time and with an interest in weather can fill that one in. Note how he doesn’t give the likelihood that these things will occur only a vague ‘probable’ (how probable?) or ‘at risk’ (to what extent?).

*** Here I’m trying to deal with that Ken Ring’s ‘predictions’ are vague and prone to ‘re-interpreting’ after the fact, just as he did for the February 22nd earthquake.

More on Code for Life:

More on Code for Life about the February 22nd 2011 Lyytelton earthquake (updated 17-April-2011 to include a few written since this article):

Christchurch/Lyttelton Feb 22nd 2011 earthquake – questions and answers

Ken Ring & March 20th – let’s get back to science

Christchurch (Lyttelton) earthquake ground movement captured by satellite imagery

High tides and earthquakes — an interview (This also contains a list of the articles on sciblogs.co.nz on ‘the Ken Ring fiasco’, if we can call it that.)

Liquefaction in a barrow

Lyttelton earthquake peak ground acceleration

6.3 earthquake in Christchurch