June has busted out all over! It’s been rainy and the grass needs constant mowing. And it’s time for you to bet on the UAH temperature anomalies. You know the drill.
[sockulator(../musings/wp-content/plugins/BettingScripts/UAHBets5.php?Metric=UAH TTL?Units=C?cutOffMonth=6?cutOffDay=23?cutOffYear=2013?DateMetric=June, 2013?)sockulator]
Cut off day for bets is midnight June 22, 2013. Good luck!
+0.3 based on this predictor.
http://s1291.photobucket.com/user/RichardLH/media/uahtrendsinflectionfuture_zps7451ccf9.png.html
Lucia: You could always turn that predictor into a bet of “where and at what value to place the next nodes”? 🙂
Richard LH–
Only those interested in “nodes” would want to predict such a thing. Beyond that: is there an established body reporting on the placement of said “nodes”? If no one has been reporting these, then we have no ones future posted values to bet on.
Ok. Pin the chart to the wall and see how good it turns out. I am only observing a pattern and calling out that it exists. Time alone will show if the human pattern recognication end of statitsical science is useful.
Lucia: Obvoulsy I being my usual opaque self when trying to communicate. The ‘nodes’ are the inflection points through which the various monthly running average filters will traverse. (see data summary ref)
The UAH data itself will produce them in time. By the end of 2014 we will be able to see if my Dec 2013 values of 0.13 is right or wrong and by how much.
RichardLH
This is an overly specialized idiosyncratic thing to bet on. I think betting on temperatures reported by Roy makes much more sense.
With respect to your ‘node’ issue, people with other notions would have to familiarize themselves with “RichardLH’s ideas about averaging and so on” and then based on the idea of the specific blog reader “RichardLH”, insert their notions about what temperatures are going to do and so on. Meanwhile, each of them have their own theory or idea of how to gain insight, but for some reason, we aren’t setting up bets based on that.
I mean… I regularly show trends since 2001. I could set up a bet where people predict “The difference between the reported June 2013 anomaly and the value predicted based on the Jan 2001- June 2013 trend”.
Or I could observe that, currently, the monthly value of GISS falls below the trend line computed since 2001 and ask people to predict when the monthly value will exceed the trend line.
Or I could set up all sorts of complicated bets based on this method of plotting data. Predict the next date for the crossover between the annual average temperature and the ols trend? Whatever you want. One could bet on that.
But… those bets are just too complicated and require people to not only think about what UAH does, but what trend lines, effects of start points, focus on “my” way of displaying data etc.
In contrast, betting on UAH is just betting on UAH. I think that’s easier for me and others.
But this is betting on the UAH. It is just predicting where the running averages of the values will fall rather than the values themselves.
As I say, I am quite happy to pin this on the wall and see how good it turns out.
Richard LH
Are you the Climate Grinch? Allowing for some argument, the UAH is something quasi-physical to bet on. No only that, I suspect that in addition to myself, there are those that have their own little physical-statistical ‘models’ to get at the number. It is a fun, good-natured, quiet corner where we can play.
Lucia, I like the idea of introducing spread betting on temperature, then we could hand over the temperature network to the Mafia in the knowledge that they would keep the system uncorrupted; if you are going to have crime you might as well have organized crime.
mwgrant (Comment #115643)
June 12th, 2013 at 11:25 am
Richard LH
“Are you the Climate Grinch?”
I do hope not. I was just pointing out that there appears to be a pattern present that would allow for more accurate prediction of the series into the future. Unfortunately the ‘noise’ present means that only averages can be predicted with any reasonable chance of success.
Richard LH
Thanks, you look safe…but now, this Martyn fellow…this is beginning to really look unsavory.
Richard LH
I’m not sure how you get 0.3c from that graph.
Ray (Comment #115682)
June 12th, 2013 at 4:00 pm
Richard LH
“I’m not sure how you get 0.3c from that graph.”
It is about trying to place where the next falling node is (Dec2013 looks good from the history so far). The value is where I am currently expecting the 12 month moving average will ‘cross’ the local wavefrom center line in the future and its value and that is probably more difficult to get right this far out. All the other longer filters output will follow through this point later in time.
Should be good for verification by end 2014. 🙂
Sorry – missed the typo – that should be 0.13 for end Dec 2013 as the graph shows.
Here is a 6/12 snapshot of Aqua Ch6 (7.5km) adjusted down by Roy’s suggested 0.2degC correction for instrument drift identified around August 2012
It’s pretty clear that the first 12 days of June remain significantly colder than June 2012
Richard LH,
Thanks.
You mean the ? in the black diamond?
I thought so but wasn’t sure.
let’s see. wrong the last 5 times so must be wrong this time.
Last 2 years dropped below 0.0 a couple of months before this and I so wanted it to be negative. Then shoots up early this year before teasing again. I will not go negative.
Best betting method is reversion to the mean as in a random walk. So I should bet 0.000, all the time to win but I won’t.
I want it to be less than 0.075 but it won’t.
Using great will power I will demand that it does not shoot up to far so 0.1 this time. Note last 2 years suggest 0.37 would be right.
Ray (Comment #115818)
June 13th, 2013 at 2:47 pm
“Richard LH,
Thanks.
You mean the ? in the black diamond?
I thought so but wasn’t sure.”
Yes. From the data so far that looks to be the most likely value (if the patterns continues).
The data set is so short though.
I will say that it means that 2014 will, on average, be colder (UAH Global wise) than 2013.
Just wait and see I suspect.
Lucia
Bet entry was unstable again.
FYI
OS-X 10.6.8/Firefox 15.0.1
Behavior – submittal (without clicking ‘dubmit…’ button, immediately after [rtn] on math proof entry.
Also for the past week or so going to Blackboard plays hobb with my scrolling when I go to this betting page page. OS is getting long in tooth and Firebox is becoming an increasing problem.
Please scrub any entry for mwgrant. I will not re-submit. Such an isolated issue is just not worth your time or my time. As suggested above, I do suspect my firefox does not like something about this page’s script and after the last time I don’t care to try re-entries. Thanks for all the previous efforts. I’ll send a runner for the balance of my quatloos.
hmmm… I wonder if it’s the ads. Those are recent.
They don’t pay much. Maybe they are more trouble than they are worth.
Lucia,
Really don’t worry about it as I am the only one having problems. Seems to me that you got a lot more interesting irons in the fire and the culprit probably lies on my end. I just passed on info in case you have other problems.
Best regards and give the cats a scratch under the chin.
lucia (Comment #116324) ,
I had a problem with your pages when you first put the ads in. I can’t remember exactly what the problem was… but the problem went away when I blocked the ads. Running XP with IE8.