THE only difference between some men and many boys is the price of their toys. This observation has much relevance to decisions made over recent years at the Australian Bureau of Meteorology, BOM.
A toy is a game, a gadget, a gizmo that may have no real practical value. But of course it is fun to play with.
They have a lot of fun at the BOM playing with General Circulation Models. These models, run on supercomputers, are essentially about attempting to forecast future and past climate including cyclones, droughts and ice sheets crashing into the Southern Ocean. The super computers are worth tens of millions of dollars and there are over 130, mostly big boys, playing with them from 9am until 5pm Monday to Friday at the BOM in Melbourne.
It all happens at considerable expense to the Australian taxpayer and I say they do little more than play, because the output from their favorite model generates mostly nonsense at least when it comes to that most important of all climate variable for Australians – rainfall.
The BOM has directed most of its research efforts over recent decades towards modeling climate systems as part of a global effort that began back in the 1950s when a small team of American scientists set out to model the atmosphere as an array of thousands of numbers.
By the mid-1970s computing power was catching up with their ambition and by the 1980s there was a growing confidence in the models and, in particular, their ability to forecast the impact of increasing levels of carbon dioxide on climate. Forecasting of rainfall, was considered less important and indeed is less fun than forecasting climate catastrophes.
Australia’s version of this global effort, POAMA, the Predictive Ocean Atmosphere Model for Australia, is promoted as a state-of-the art seasonal to inter-annual forecast system. However, despite such a claim being made for many years now, until very recently, POAMA had not been used for the official seasonal rainfall forecasts issued by the BOM.
This is because rainfall forecasts from POAMA for anything more than a week in advance are not very good. In fact, they are consistently worse than the forecast a schoolgirl could generate based on simply calculating the monthly mean rainfall for a particular locality with a pencil and pad. Such an average value is known as climatology.
But so enamoured have the big boys become with their expensive toys, that this last winter, the BOM discarded the old statistical models relied on for the past 20 years and adopt POAMA as the basis for all climate forecasting.
There is no peer-review paper that indicates POAMA can reliably produce a seasonal or monthly forecast with more skill than the old statistical method – because it can’t.
Indeed the peer-reviewed literature, including a new paper by John Abbot and me (see the journal Atmospheric Research, Volume 138, Pages 166-178), shows that statistical models that rely on pattern analysis will consistently outperform General Circulation Models such as POAMA.
Ignoring this reality, rural industry groups are consistently briefed by the BOM about how good POAMA is. And the misguided enthusiasm for the hi tech computers and massive databases is such that most industry groups including the Grains Research and Development Corporation and Queensland Canegrowers Ltd, have been persuaded to invest heavily in POAMA.
The commitment from the industry leaders to POAMA reduces the opportunity for dissent. But the real loser is every farmer who could benefit from a half reliable monthly or seasonal rainfall forecast.
This article was first published by The Land newspaper on December 12, 2013. Go and buy a copy at your local news agency… to show some support.
More information on alternative methods for medium-term rainfall forecasting, including a powerpoint presentation by Jennifer Marohasy are available on line at
http://www.sciencedirect.com/science/article/pii/S0169809513003141 [see RHS panel for the powerpoint after the page has completely downloaded]
Congratulations on your paper Jennifer.
I read your article in the Land today.
After sitting through a long presentation from BoM very recently, I am much inclined to agree with you.
I do hope that these super duper expensive systems will eventually live up to expectations, especially in the area of regional seasonal forecasting… but so far… in my region… they have NOT performed any better than previous methods. I note in Autumn 2013, their prediction for Winter/ Spring 2013 in Eastern Australia was 80% certainty, wetter than average. Instead, large swathes of QLD and NSW experienced seasonal drought conditions. Some third party impacts have occurred here in the MIA after what’s been called an ‘imbalance’ (but is really an imbalanced DRAWDOWN) of our major storages… partly because of those forecasts.
What gobsmacked me in the presentation that I attended, was there was virtually no recognition or understanding of what the users of this information might need. They were indeed soooooo enamoured with their hi tech computers and massive data bases that they did not even cover how users like farmers, miners etc could benefit from these new approaches. They were clearly uncomfortable when they were asked some very obvious but simple questions about presentation and regional forecasting.
Congrats Jennifer, I shall buy two copies, one fer a friend.
Great job, Jennifer!
Larry Fields says
Speaking of the devil, a popular toy category for males (not counting yours truly) is the computer game. These range from shoot-em-ups to role-playing games. And of course there’s Tetris, which can be useful in monitoring one’s progress in a cognitive enhancement regime, to see if the healthful exercise and improved nutrition makes a measurable difference in quick thinking and mental endurance.
Unfortunately, we climate realists have been neglecting a golden opportunity to spead our Bah-humbug message. Some enterprising programmer should create a climatology-related computer game.
A talented, young, university-based meteorologist is in over his head. He’s become painfully aware of the AGW BS. He’s also aware that if he speaks out, or even reaches an un-PC conclusion in his research, he won’t get published, he won’t get tenure, he’ll lose half of his friends, and his wife will divorce him. The game is all about finding a safe middle ground.
Obvious example: Carve out a niche in an obscure sub-specialty that nobody else is interested in, and which does not involve Global Warming. The upside: He’ll maintain his integrity. The downside: He’ll need to hustle twice as hard for grant money, and won’t have as much time to spend with his family.
The game would provide many hard choices in his career path. Through trial and error, he’ll need to find a linear combination of options that maximizes his job satisfaction, while maintaining his career and his marriage. He need to swim with sharks, without becoming a shark.
The game is a kind of Black Box Problem that they don’t teach you about in school. You do not know what’s inside the box. But for every input you make, the box will give an output.
Here are some Black Box metaphors. Talk to the box. Shake the box. Tap it with a spoon. Run it under cold water. Make detailed notes about what happens. Form tentative hypotheses. Test these hypotheses. Learn from your experiences.
Our game could be modeled after Sim City, a city managers’ computer game. Sorry, my programming competence is limited to number crunching and to interfacing computers with chemical instruments. Any takers?
It has been argued that in science, theories and their overarching paradigms gain their status because they are successful in solving important problems.
AGW theory, with its focus on carbon dioxide emissions, solves a problem that has preoccupied activists for years: it provides ‘proof’ that industrial activity is despoiling the earth. At least this is the case when general circulation models are used according to AGW theory’s particular assumptions.
An alternative theory of climate, for example, one that is useful beyond politics, for example, to industry, might make as its focus detailed and accurate climate and weather forecasts.
General circulation models, however, are not much good when it comes to forecasting weather, particularly not when it comes to forecasting snowfall and rainfall.
Of course there are accurate and objectively agreed measures of rainfall.
Statistical models, based on pattern analysis are much better at forecasting monthly and seasonal rainfall than the expensive general circulation models that underpin AGW theory.
These statistical models work because there are recurrent patterns in the historical rainfall data.
And that the historical rainfall data with some local variables (e.g. minimum temp) and climate indices (e.g. SOI) is useful in forecasting rainfall suggests, contrary to AGW theory, that the climate is not on a new trajectory.
PS I guess what I should add is that the “talented young meteorologist” you refer to, is probably going to make most progress if they move away from the computer games that are general circulation models, and towards forms of artificial intelligence e.g. if they learn about artificial neural networks for pattern analysis.
Dr Norman Page says
I am one of the blogosphere’s old contrarian Realists or as I prefer to think Baconian empiricists. From the start the whole CAGW meme seemed really incredible to me based on common sense thought and simple observations. Also because of the complexities of the climate system it seemed obvious that because of the uncertainty in parameterising the initial conditions of the multiple variables involved and having had some experience in basin modelling there was no way that the model outputs would have any actionable credibility. In particular the IPCC- Met Office models in addition to this inherent weakness were also structured incorrectly. They are beautifully circular in that , in the inputs, the CO2 contribution is assumed to be dominant so that is what comes out at the other end. The result of this is that all the science and the vast amount of computer time used is actually irrelevant and the outputs merely reflect the various RCPs. Future temperatures from the models therefore simply depend linearly on the total amount of CO2 emission implied at any future date by each RCP.
The RCPs themselves are merely imaginary scenarios based on speculations on such things as population and the energy mix at various times in the future – in short pipe dreams.
There has now been no net warming for 16 years with CO2 up 8%. The earth has been cooling slightly since about 2003 and this cooling trend is likely to steepen until 2035 and perhaps continue for hundreds of year beyond that.
By now it is obvious that reality is diverging sharply from the models. Another forecasting method is required. In a series of posts on my blog at http://climatesense-norpag.blogspot.com I have made forecasts of the coming cooling based on quasi periodic- quasi repetitive patterns in the temperature and 10Be data
Bob Tisdale says
Thanks, Jennifer. Well done.
These people are seriously mad, bad and sad. Gawwwdd help us, we’re dealing with imbeciles.
Ken Ring says
Great article! For thousands of years there has been consensus across all cultures except the western that it is the moon and sun that drive weather. To this day those cultures have retained the unversal moon calendar for that very purpose – to get a handle on the weather. Solunar cycles are still known and used by aborigines in the north. However there is little or no funding available to study this because cycles do not advance elitism. If knowledge of workable patterning became widespread farmers could easily do it themselves, without the BoM and without climatologists, computer developers, and the whole self perpetuating and self-funding earth science industry. The problem is not scientific, it is political and therefore pecuniary. Meteorologists get more funding when they are consistently wrong. No one is going to throw buckets of money, for the multimillion dollar upgrading of supercomputers, for the deployment of extra teams of staff and for underwriting junket international conferences at exotic locations, at forecasters that get the weather right all the time.
The talented young metorologist who wants to beat “The team” could take time to look out the window too. Mind you, you can beat a GCM at predicting the future using a random number generator. Apparently some GCMs have been shown to have skill at predicting what won’t happen, that is there is a better chance the outcome will be opposite to the prediction of a GCM than in accordance to it, interestingly the US EPA ignored these facts and used then anyway….
It is clear that climate models that claim to be able to forecast temp change until 2100 over the tropics HAVE FAILED.
Just read Christy’s address to Congress on extreme events plus weather/climate and modeling etc. Particularly look at the graph of all the models and actual observations on page six.
These super duper models are just a super expensive joke and waste of time. Christy has said recently that the models are an EPIC FAILURE.
Jen has done us a great service over the years by encouraging us to be sceptical and evidence based.
So it’s pleasing to see that Robert and Debbie have given the new scheme the big tick on faith without even reading – excellent unswerving blind support has to be commended.
And Jen’s concern for the poor brow-beaten dissent-diverted landholders is admirable. So may we be pointed to the web site where next month’s Queensland forecast is and the track record of how the last 10 years have gone (given as Jen says BoM has made a hash of it themselves). So I’m sure new standards will be set. So if Jen could point us to the web site that would be great.
And as sceptics always lament – those darned papers behind paywalls. I’m sure Cohenite would be advocating free-range open publication. But you have to admire anyone going the yards with the peer reviewed establishment journals.
So it would be good to know what was compared with what. BoM being an impenetrable fortress of satanic secrecy and all that – what version of POAMA with exactly GCM output was used to compare what against what. It’s all so confusing so maybe Jen could tell us exactly how POAMA 2.4 (the current system) was ass-whopped would be sciencey.
Of course POAMA is an ambitious critter than goes across the nation and into the South Pacific Islands (not being as we now know at risk from CO2 induced sea level rise) so I guess we’d need to know how ANN POAMA-slayer V2.1 goes in the regional context.
Debs yourself and Robert can go an have a cup of tea and a lammie and check out Robby’s kayak as you’ll only be interested in the broad social policy aspects and Robert will only want to anecdote mine the data.
2,000 more of those cold weather extremes over the past week in the USA. And Cairo has had it’s first snow in Dec in 112 years. Must be the revolution that caused it?
Remember just 13 years ago they told us that snowfalls were a thing of the past? Have these people any shame at all?
[Neville, I do very much appreciate your updates on general climate related issues at the ‘Open Thread’. But if you keep posting general information at this thread it will be deleted. And I’ve just deleted your last two. Cheers, Jen ]
There is no tick of faith from me. Whichever system starts delivering useful results will get my tick then.
I congratulated Jen for having her work published and agreed with her that BoM’s GCMs is not performing as well as expected… and then also agreed that BoM’s staff appear more enamoured with their super duper technology rather than actually providing a useful public service.
I’m unclear why you’re apparently annoyed that I would be interested in what you call ‘the broad social aspects’ and that you have repeated that as a criticism several times.
Leaving aside the fact that I’m interested in all sorts of things (but not lammies… I have never liked lammies…have never made one etc) …What…according to you… am I supposed to be interested in Luke?
That we can produce a forecast of monthly rainfall with a correlation coefficients above 0.7, would indicate that the system is not chaotic.
That the BOM can not produce a forecast with a correlation coefficient above 0.5 for the same localities and time periods, doesn’t mean that the system is inherently chaotic, but rather that their model/their tool doesn’t work.
I agree with you: the future is in better pattern analysis.
I also agree that the moon is important. Of course it creates the sea tides, and also the atmospheric tides. We do not input the moon’s influence directly into our ANN models because we can’t work out how to quantify its direct effect in terms of an array of numbers that will give a meaningful output.
That is not to say that it can’t be done. Rather we just don’t know how.
As regards versions of POAMA, we compare against the last two available versions in our paper.
But as with the forecasts from the BOM, despite promises that they will improve in skill, there is no evidence to support the same.
That is, after two decades, and hundreds of millions of dollars spent, even the most recent versions of POAMA can’t reliably produce a rainfall forecast better than climatology (i.e. a forecast that is better than simply calculating the long term average).
To reiterate, there is no version of POAMA that can reliably produce a seasonal forecast better than a school girl could by just calculating the long term average for a particular locality.
A useful system of forecasting and/or a system of forecasting that shows promise should be able to produce a forecast with a greater level of skill than computing the long term average. We can do this with our ANNs (Artificial neural networks are a form of artificial intelligence).
Beth et al.
Yes Jen all very rhetorically interesting and fascinating personal opinions but this is science surely – but is there a web site where we can follow your forecast and its results historically. What about your hard done by oppressed landholders – the readers of your “The Land” article – surely you guys are the liberators?
And what did you compare with what exactly. Which versions of POAMA? With what output? Otherwise we really have no idea what you’re talking about.
You may have an R of 0.7 but any of us could do that. But history tells us many schemes come a cropper from simply overfitting or mapping the seasonal cycle. What’s does the cross validation statistics and skill test results say?
Debbie you’ve had POAMA since May 2013 – hardly time to know anything unless you like sample sizes of ONE. What are you interested in Debs? Well we don’t know – you’ve never been able to us despite many attempts. You have said you’re weren’t interested in the technicalities but moreover the broad social aspects. Forecasts tend to be numeric though and not anecdotal.
“That is, after two decades, and hundreds of millions of dollars spent, even the most recent versions of POAMA can’t reliably produce a rainfall forecast better than climatology”
do we have a broad Australian study that shows this? a citation perhaps ?
Not that I’m buying anyone’s long term outlook yet, I’m just sure I won’t be buying the BoM’s. David Jones has promise as a colourist, but I’m more concerned about next year’s spring rainfall, not the Spring Collection.
For those confused over terms, Luke’s notion of “anecdote” is “stuff that actually happened”. His idea of “science” is best described as “stuff that hasn’t happened yet”. I congratulate Jennifer on her initiative, not on things that can only be confirmed in the remote future by actual results. (See: “Stuff That Actually Happened”.)
By way of nuance, when an “anecdote” is accompanied by detailed and compelling evidence, that evidence is referred to by such terms as “kayaks” and “lammies”. As to why Luke sees things this way, I guess he must be one of those DEEP people. (Psst…he sees dead supercells.)
There is a classic POAMA-2 study in Weather and Forecasting Volume 28 page 668 (2013). The BoM got a heap of cash from AusAID to provide rainfall predictions for a set of poor counties in the Pacific.
Well, the POAMA boys did their best, and one can say that they did give the poor folk forecasts as good as they provide for the True Blue Ones back home. Certainly no skimping because they are getting a pro bono job.
Yes, the POAMA-2 study gave correlations of between 0.4 and 0.5 as usual. Everyone gets the same standard of forecast – noise!!!!!
If we takes Luke’s advice, we all just need to be patient and maybe, just one day, when POAMA-99
is unveiled, then we will get better forecasts than climatology from POAMA. Climatology is a long term average – this is achievable with a pencil and paper by a 7 year old schoolgirl if you give her a list of 30 numbers to average and may take 10 minutes, if she is interrupted by Twitter or Facebook.
I suspect many bears could do this too.
Johnathan Wilkes says
“We do not input the moon’s influence directly into our ANN models because we can’t work out how to quantify its direct effect in terms of an array of numbers that will give a meaningful output. ”
In cases like that it is customary to compare a period without the influence with the period where the influence is expected.
If the difference is inside a margin of error and constant over a lengthy period of time, then you can easily calculate the impact or discard it altogether.
@Comment from: Luke December 14th, 2013 at 10:23 am:
“do we have a broad Australian study that shows this? a citation perhaps ?”
Be pro-active Luke. Or find contrary evidence. The science is settled we are told. http://www.theguardian.com/environment/2013/nov/21/warsaw-climate-talks-the-worlds-poorest-cannot-wait-for-a-2015-deal?CMP=twt_gu
Should be easy when the consensus science is on your side.
Yes all very fascinating Mr Bear – but I’m interested in getting some answers to my questions not more advertising. At this stage this post is 100% bluff and rhetoric and 0% evidence of anything.
Handjive – keep on topic pls.
Too easy Luke.
Simple forecast of Cairns Aero monthly rainfall for the 12 months using only the monthly mean gives an R of 0.91 for 2008. Who needs a neural net .So this patch up rerun is a travesty.
I was still getting a laugh from the first paper which used Sydney as a proxy for Qld temperatures. As I recall Qld split from NSW in 1859. But you had to go to Sydney to find a site without a trend.
I have just purchased the neural network paper form the publisher.
If you all do this, you will be able to participate intelligently, rather than talk about something you have not looked at.
Yes- this method definitely produces much better rainfall forecasts for Queensland than POAMA-2.
The science is settled.
Now have a nice weekend, as no rain is predicted for coastal Queensland.
OK Bazza – you win the $10 – stonewalling it is.
Ken Ring says
Koala bear: I disagree. The next rain for coastal Queensland may be around 21st-22nd, give or take a day. This is derivable from lunar patterns. The apogee moon will be trekking south on its declination cycle, inducing a surface trough to move through southern QLD. Rain should affect the SE QLD coast to Capricornia, Central Highlands to the Granite Belt and Darling Downs, patchy rain Tropical North Coast, Peninsular etc but not much in the Northwest. I expect Brisbane to get a fair sized dump then. Then another lot between Boxing Day and New Years Eve. We can quantify the lunar effect because it is the nature of all tides, including the tide of the atmosphere, to be quantifiable. POAMA is what funding buys. You can get the moon for nothing.
We are still laughing at the dopey Queensland government and their very successful attempts to flood Brisbane in 2011.
Rubbish rainfall forecasts coupled with totally incompetent management of Wivenhoe dam led to billions of dollars in unnecessary property damage and many deaths.
These incompetent government people need to be replaced by those that can do the job much better.
Hopefully the upcoming class action lawsuit will reveal just how bad it all is.
Bring on the neural networks.
Yes Luke we have had POAMA since AUTUMN 2013…it came with all the fanfare of its 80% certainty of a wetter than average WINTER/SPRING 2013 and lots of advertising about how much more reliable this system is.
And yes. . . its PUBLIC face is only a sample size of one, but that is NOT the case for the back room boys and girls and all the work that has gone on prior to the release in MAY 2013.
I will also repeat. . .as you seem to have missed this point every single time. . . I sincerely hope that the new super duper system will start delivering better results. It has cost a lot of money and a lot of time.
I will also repeat. . .for emphasis. . .that the BoM staff appear to have lost their way. . .they’re way too enamoured with the hi tech system and all the different ways they can present the information from the huge data bases and seem to have completely missed the reasons why people and organisations would use all this information and what they would expect BoM to focus on.
Concluding things like 2012 was 30% wetter than average in Australia and the latest announcement… that 2013 is shaping up to be one of the warmest on average in Australia. . .while an interesting stat. . . isn’t particularly useful for the people who would like to use info from BoM in their daily lives because they work with the climate/weather/environment 24/7.
Deb,source pls for 2012 30 percent wetter than average. Tell me too how you would summarise Australian rainfall for a year short of listing the stations wetter than average etc etc. maybe bom has that info too and they actually do an excellent job catering for different scales and perspectives.
BoM presentation @ Menzies Hotel Sydney NSW Nov 7th 2013.
Asked to present re BoM brief via Water Act 2007 and the Federal Water Resource Plans.
I don’t doubt that BoM has the info Bazza. . .never have.
Larry Fields says
Comment from: jennifer December 13th, 2013 at 8:20 pm
“PS I guess what I should add is that the “talented young meteorologist” you refer to, is probably going to make most progress if they move away from the computer games that are general circulation models, and towards forms of artificial intelligence e.g. if they learn about artificial neural networks for pattern analysis.”
I take it that you don’t like the idea of the climate ‘science’ office politics computer game. What about making a movie about the “talented young meteorologist?” Of course, we could throw in a few zingers, like no trend of increasing global temperature for more than 15 years. But we shouldn’t make it too didactic.
If you have room in the script for a red-neck sheriff, I volunteer for the part. 🙂
Jen stated at 8:15pm Dec 13 : “And that the historical rainfall data with some local variables (e.g. minimum temp) and climate indices (e.g. SOI) is useful in forecasting rainfall suggests, contrary to AGW theory, that the climate is not on a new trajectory.” The suggestion is simply incorrect. The climate trend as evidenced by the temperature trend is obviously small in comparison to natural variability. Best stay up the shallow end!
Which climate trend is that Bazza?
You say ‘as evidenced by the temperature trend’. . . Not another one of those Australian (as in WHOLE of Australia) between Jan 1st and Dec 30th by any chance?
In what way do you think Jen’s comment is ‘simply incorrect’? Wasn’t she referring to regional, seasonal forecasting and particularly rainfall?
I have no brief for POAMA, indeed don’t know anything about it. But I just want to pick up on a basic point.
The implication in the article and comments is that the model got the 2013 rainfall wrong. To quote: “in Autumn 2013, their prediction for Winter/ Spring 2013 in Eastern Australia was 80% certainty, wetter than average. Instead, large swathes of QLD and NSW experienced seasonal drought conditions.”
But an 80% prediction is not ‘certainty’; there was a 20% chance of not wetter than average conditions. As a farmer, it was probably worth taking a risk on an 80% evaluation, but no grounds for being surprised (disappointed maybe) when the 20% chance eventuated. So the model wasn’t ‘wrong’ – it identified the result, just with a lower probability.
However good, bad, complex the model or the pattern method, it is always a matter of probabilities. High probabilities are not ‘certainties’ – even 99.9% ones.
It is the new trajectory for the globe as evidenced by the last several decades each being warmer than the previous.
But that’s a GLOBAL decadal average/ mean/ median temp thing Bazza. That’s even MORE generalised than using the WHOLE of Australia averages.
Wasn’t Jen referring to regional forecasting and particularly rainfall?
It ended up that over winter/spring there was a SEASONAL DROUGHT! That is NOT a 20% probability of LESS than average rainfall.
The result for winter/spring 2013 in large areas of QLD and NSW was a seasonal DROUGHT.
Your 20% probability of LESS than AVERAGE rainfall is NOT a 20% probability of a seasonal DROUGHT.
There was no stated % of probability for DROUGHT for eastern Australia in Winter/Spring.
Less than average rainfall is not the same as drought. Depending on when that ‘less than average’ rainfall occurs in Winter/Spring, it is still possible to grow decent crops and livestock.
Quite obviously your questions are for Jen and Abbot. I’m sure they can let you know if their conclusions would apply to other regions if you ask them.
I have to confess I am unclear about your meaning re ‘the shallow end’. I could take a reasonable guess. . . but can’t be bothered.
Would you care to elaborate?
I hope this doesn’t eventually appear several times as my original comments went missing in the spam attack.
That is a remarkably similar comment to Luke’s.
Here is my perspective:
The result (as in what actually happened) was there was a seasonal DROUGHT in large areas of QLD and NSW in Winter/Spring 2013. That is not the same as a 20% probability of LESS than AVERAGE rainfall. There was NO stated % probability of seasonal DROUGHT for Eastern Australia.
There may have been one somewhere, but it certainly would not have been 20% probability of DROUGHT.
It is still possible to grow crops and raise livestock when there is a less than average rainfall in Winter/Spring. It is actually also possible to lose crops and livestock if it is wetter than average in Winter/Spring.
It’s OK, please don’t tie yourself up in knots because I used ‘certain’ instead of ‘probable’. . . we’re a long way from expecting ‘certainty’ but we do expect some improvement in the efficacy of seasonal forecasting. . . considering the size of the investment.
As I have commented to Luke on several occasions. . .I sincerely hope that there is some decent focus in this area and some measureable improvements.
If Jen and Abbot’s work helps progress in this area I am all for it.
Johnathan Wilkes says
probability or certainty when it comes to an 80/20 prediction is mere semantics.
Some of my clients would give their eye tooth for a predicted 80% chance.
It is indeed semantics.
But Jay (and Luke) seemed to think it was so very, very important and also somehow negated the substance of my comment.
I was happy to concede that probability was the ‘right’ word because as you point out, when we’re talking figs like 80/20 that isn’t the actual issue.
Gotta love this precious little lecture though:
“However good, bad, complex the model or the pattern method, it is always a matter of probabilities. High probabilities are not ‘certainties’ – even 99.9% ones.”
So even if the BoM gave a probability of 99.9% wetter than average Winter/Spring and it turned out to be a DROUGHT in Winter/Spring that would still mean they were right? (or maybe not wrong) ? ????????
Johnathan Wilkes says
yup that is a good question Debbie.
What’s the excuse when predicting 99.9% prob. and proven wrong?
To my mind it just playing on words.
I can take 70/30 as a borderline ‘prediction’ but anything above that conveys an amount of certainty as far as I’m concerned.
After all they must have done some serious research to come to that conclusion, not just reading tea-leaves.
Otherwise as you said even a near 100% is disputed. I know nothing is certain in this world even tossing a coin sometimes produces an unexpected ‘edge’ ie. the coin refuses to lay flat but that is by chance, we are talking here of grown serious people doing serious research where thousands rely on their advice.