Skeptics 0, Ken Ring 1
After last week’s nocturnal earthquake, Ken Ring has been on Facebook proving how right he is. He’s pointed out that he predicted the earthquake in his 2021 almanac:
From the “Earthquake diary”, on p339 of the 2021 NZ Weather Almanac, it is written that the days of earthquake activity for March would be the 4th, 9th, 18th-20th, 25th, and 29th. This morning’s shake struck at 2 hours after midnight, close enough to 4th to be counted as a prediction.
I guess it’s time for the NZ Skeptics to throw in the towel and admit defeat, in the face of this uncannily accurate prediction. Except, I can’t help shaking the feeling that maybe, just maybe, by Ken’s own admission, his predictions cast the net wide enough that it’s hard for him to miss…
So, pulling this one apart, let’s have a look at the dates in Ken’s March predictions. If we allow Ken his one day leeway, that would mean that he predicted an earthquake for the 3rd-5th, 8th-10th, 17th-21st, 24th-26th and 28th-30th of March – a grand total of 17 out of 31 days in the month. For any earthquake that happens in March, he appears to have a slightly above even chance of the quake landing on one of his predicted dates, even if his predictions are totally random guesses lacking any kind of scientific rigour.
However, a one day leeway isn’t all Ken has asked for. Last year, talking about his earthquake predictions on Facebook, Ken stated that people need to “Allow 1-2 days error as in all forecasting”. Gosh, if we were to allow him the “2 days error” he asks for, his March predictions would have covered 26 out of 31 days. It seems that, if he’s willing to be this loose in his predictions, it’s going to be hard for him to fail. But maybe that’s the point.
We’re now going to be a little bit cheeky, and borrow the ideas of sensitivity and specificity from the field of medicine. What we need to know in order for us to do this are very rough ideas of Ken’s true and false positive and negative rates. The “true positive” rate for Ken’s predictions is okay (maybe half or more of the major earthquakes we experience fall on one of Ken’s predicted earthquake dates). However his “false positive” rate (the times Ken predicts an earthquake and it doesn’t happen) is much higher – most of the dates Ken throws out there don’t turn out to be days where a sizable earthquake hits New Zealand. His “true negative” rate appears to be fairly good (many of the dates he doesn’t predict an earthquake for, earthquakes don’t happen), but his “false negative” rate (dates for which a major seismic event occurs but he hasn’t predicted it) is not great. The upshot of this is that both the sensitivity and the specificity of Ken’s earthquake predicting ability are pretty abysmal.
The astute among you (and I’m assuming that’s many of you, given that you’re skeptics and prone to questioning what other people say) will have spotted that this method doesn’t translate overly well to Ken’s earthquake predictions – not least because of several major issues with the kinds of predictions Ken makes.
To explore this, let’s play a game of make-believe. Imagine a world where Ken Ring has figured out a way to predict the timing of earthquakes, a method that the world’s leading seismologists have somehow totally missed. Even if that were the case, and Ken’s ability was real, it turns out it would still be almost useless. In our make-believe world, where Ken’s specificity and sensitivity are both 100%, let’s look at his March 2021 predictions. If earthquakes occurred somewhere in New Zealand on the “4th, 9th, 18th-20th, 25th, and 29th” of March and not on any other days, exactly as Ken predicted, what practical steps could Civil Defence, or any branch of government or business, take that would help people? Ken usually offers no useful information beyond listing dates, and occasionally a rough location. He has claimed weak and strong earthquakes, and quakes thousands of kilometres away, as “hits” for his predictions. Without knowing specifics about location, depth, strength, time of day, etc, these dates are not terribly useful. I’m struggling to think of how this level of predictive power would change how we live our lives, or make any difference to our earthquake preparedness. In short, Ken offers too little information in his predictions for them to be of any practical use.
As we said before we know that, in the real world, Ken’s predicted earthquake dates – no matter how vague he makes them – still aren’t very accurate. His specificity and sensitivity are not 100% – in fact, they’re nowhere near. We have to conclude, given his lack of accuracy, lack of detail, and lack of plausibility (using phases of the moon to predict earthquakes seems iffy), that it’s vanishingly unlikely that Ken Ring really has the ability to predict earthquakes at all. It’s much more likely that Ken has found his schtick; a trick that allows him to extract money from the unwary, with which he lines his pockets.
tldr; Ken Ring’s earthquake predictions are not accurate, and even if they were they wouldn’t be very useful.
Now that we’re fairly confident that Ken’s more shyster than visionary, I think that we can safely ignore his recent advice that tsunamis are not a risk for New Zealand. He’s been asserting that most of New Zealand is underwater, i.e. shallow waters surrounding our islands, and that these shallows mean that a tsunami will never affect New Zealand. If the experts say that we are in danger of a tsunami after a large offshore earthquake, as they did last week, and Ken tells us we’re not in any danger, I’m going to follow the advice of the experts with science and computer modelling behind them. Plus, Wikipedia has a thing or two to say about tsunamis affecting New Zealand. Sorry Ken, it’s nothing personal.