Data Comparisons

Bet on Annual Average UAH TLT for 2012

58 Comments

This is a biggie. You are betting for the Jan-Dec average based on values available when Roy posts the December value for 2012. As usual, if he changes the algorithm, the winner is based on whatever algorithm or baseline is used when the UAH TLT value is published in December. This isn’t science, it’s gambling. The uncertainty is part of the game!
[sockulator(../musings/wp-content/uploads/2011/UAHBets5.php?Metric=UAH TTL?Units=C?cutOffMonth=1?cutOffDay=21?cutOffYear=2012?DateMetric=Jan-Dec ave 2012?)sockulator]
Cut off Day: Jan 20.

Update:If you can’t see the image of the quatloo, let me know. Thanks.

58 thoughts on “Bet on Annual Average UAH TLT for 2012”

  1. Is the number we are betting on the average of the 12 months, equalling weighted? That is, 31-day January is the same weight as leap year February? Or does Roy have an annual number which is different than that?

    Just curious. Believe it or not, the accuracy of my monthly guesses does not justify these questions. 🙂

  2. DonB–
    Yes. I will take the 12 months posted by roy and take the simple average.

    This is betting: Points of clarification need to be raised. 🙂

    BTW: If you can’t see a quatloo image in this post, I need to know that. I’m trying to do something to catch and ban image scrapers. I’m sure I’m catching them. (I am not yet banning.) But I think I might be “catching” normal people. (That’s why I’m not banning yet!)

    If you can’t see the image of the quatloo with kirk, a dam etc. in the post, let me know.

  3. I can see a quatloo — but I believe it’s a counterfeit! This strongly suggests that the site has been hacked.

    BTW, aspiring gamblers might want to refer to Roy Spencer’s last update on the UAH Global Temperature estimate, here.

    Remember to add (warmists) or subtract (coolists) 2 C before betting*.

    * that was a sneaky joke, designed to improve my own chances of glorious victory.

  5. I can see the Quatloojust fine — as if it were mine already. Anyone who beats my guess (err, I mean number derived via sophisticated technique) is worse than a Denebian Slime Devil in my book!

  6. wonmders about the possibility of creating a Euro-quatloo market for hedging and tax avoidance purposes

  7. Lucia –
    I can see the Quatloo perfectly well – both on the screen and on the special paper I used for the photocopy and subsequent printing. Just out of interest, and for the sake of comparison, do you have a good quality picture of a 1,000 Quatloo note?

  8. Lucia –

    For the sake of some of us high-rollers, is there any chance of raising the betting limit for the annual anomaly?

  9. Anteros…. hmmm… I’ll think about it. The difficulty is that would involve programming. As many have noticed, I’ve been busy– and some of that has involved programming for higher priority things! (Specifically, the ‘bot attacks. Fortunately, the combined advice of ZBblock and cloudflare has helped a lot and *today* I think I have things more or less under control on that front. I’m going to have to blog the “solution”. But even with those two, the solution involved programming!! )

  10. Spencer is planning to switch to a new 6.0 version shortly – is this bet for the current 5.3 version or whatever the version is as of the end of the year?

  11. Arthur–
    Whatever version is as of the end of the year. After all, I can’t be sure that 5.3 will provide data through the end of the year.

    The switch will randomize things! (I should start one for GISTemp– which as far as I know isn’t planning a new version this year. I think HadCrut might be?)

    Can you suggest which metric is least likely to have a version switch? It would be nice to have one with one of these weird unpredictable version switches and one with out a version switch.

  12. Hi Lucia –
    Interesting bet… Though I don’t understand what the effective difference will be from the single month bet opportunity you posted a couple days ago, unless you increase the bet limit.
    Also- I have to ask “ground rules” questions – since Ray has made noises about changing his baseline.
    First- if Ray comes out with an “annual average” himself, and it differs from the the average of each month already stated, is the betting target the simple average, or Ray’s number?
    Second- if he takes this opportunity to change his baseline, and gives two values – one compared to each baseline – which value will be deemed our betting target?
    I’m sorry to get you into the weeds on these issues, but as one of those who has rather handily been increasing their quatloos, I am getting rather covetous of retaining them and I’m concerned that we have no ambiguity and argument next month 🙂

    My opinion is that you ought to use Ray’s number whatever it is, and if he gives two, that the betting target be deemed to be the previous baseline not the “new” one.

    The new baseline should be specified as our target the month after he makes his change, even if he gives both values for a while.

    I also must confess continuing sadness that you do not avail yourself of the opportunity to bet in these. If you are concerned that people will accuse you of fudging the data base retroactively to give you advantage, I suggest a simple possibility would be you publicly posting your value and bet on the betting thread during the last day or hours before the betting window is closed. You would have a disadvantage in case of a tie. Also, some unscrupulous and unsavory characters might take advantage of your work, however I really think you ought to be able to bet. Hmmm… another possibility would be that you post your entry on the thread the day AFTER betting closes, which would alleviate the unsavory problem, give you no significant advantage, and allow you to participate.

    AFPhys

    PS: what is an “image scraper”? I just grabbed one of those quatloos to use as wallpaper to inspire me to greater efforts while grinding through my extensive calculations prior to betting. Does that qualify me as an image scraper?

  15. Ray – LOL – yup, I meant Roy, not Ray the many times I typed that. No one has ever thought I was good with names…

  16. AFPhys: I would not think it would encourage bettors to place bets if the owner of the betting emporium were know to bet in her own establishment. A hard earned quatloo is not something anyone wants to easily part with.

    As an aside, would I be correct that betting is mainly undertaken by the younger people posting here. The betting on the Arctic ice, or better the interest generating by the betting, interests me and especially if it were to attract more tales of baby ice.

  17. Kenneth– I sometimes bet. I think I will bet for the annual average. After that, I will encourage people to take screenshots so they can satisfy themselves I didn’t change my bet! (It is the case that I have access to the database and can change bets. So… I could fix things. Sometimes…. it’s tempting.

    Who doesn’t want to look smarter than they are?

    AFPhys–
    It’s for the glory!!!

    Actually, I agree with you. But it really is a matter of time to program the change. This morning…. I thought I wouldn’t have time. But oddly, I may have managed to gain the advantage over the bots in the past few days! Having watched the ‘patterns’ and learned cloudflare’s API for banning them at cloudflare, I seem to now be auto-banning the more dangerous bots at cloudflare. (Oddly, while I have detected a “forum/comment spambot signature”, I’m leaving spambots mostly alone. So, visitors may not notice the difference.

    But then, comment spam was never the problem. It was the crackers loading 10…0 pages in an attempt to find an exploit and the, loading 10…0 more pages to figure out if they succeeded and them….. ) Even though I’m pretty sure none ever found a vulnerability, it was bringing the blog to it’s knees. But even worse, it bugged the c**p out of me!!!

    But, I think I thought up the ‘solution’ on Dec. 31, and have implemented it. (I sent a question to Brandon and he told me that my worrying about this on New Years Eve was a very bad sign…. I did, however, drink Champagne after implementing the first phase which permitted “data gathering”! Heh!)

  19. MarchH–
    The entities on Triskellion just make piles of 1 quatloo bills and wrap blue rubber bands around piles of 20s and red around piles of 20*20. They all trust each other not to put only 19 quatloos in some of the piles of 20s. That’s what happens when you are just energy and have no real body.

  20. Hey that’s not fair the guys with their climate models like Hanson etc will all be spot on to the tenth of a degree.

  21. My guess was anomaly and as Neils Bohr said “Prediction is difficult especially if it is about the future”.

  22. “My guess was anomaly and as Neils Bohr said “Prediction is difficult especially if it is about the future”.”

    At CA when tropical storm frequencies and intensities were controversial topics, David Smith used to run an annual contest on predicting the tropical storms in the North Atlantic area. The predictions were supposed to be accompanied with the methods employed to arrive at the prediction. I would suppose at Lucia’s betting emporium that revealing methods would not be advisable as it could reveal an advantage in acquiring quatloos. Anyway we used to have some laughs about a group in Europe (I recall it was Meteo France) who used to acclaim their prediction skills on tropical storms in white papers and then announce their prediction results after the fact. Judith Curry indicated that the reasons were that the information was paid for by organizations that used the information to their competitive advantage. I am not sure I bought that reason.

    Prediction skills for forecasting tropical storms, like investment strategies, can look good over a period of a few years but then fail over the longer run.

    It might be interesting if Lucia’s betting emporium were to establish a criteria against which betting skills of the participating bettors could be statistically assessed.

  24. OT, but WUWT has a post by Scafetta. I wonder if you could analyze his model like you’ve done for the GCMs?

  26. I’m in at 0.015C,

    Based on my current model of UAH using Nino 3.4, the AMO, the volcano influences and a trend (0.043C per decade – the GHG-based trend does not work nearly as well as simple linear and polynomial trends – in addition, once the volcanic influence is added, the trend line for UAH drops considerably since the volcanoes are weighted to the beginning of the dataset).

    http://img254.imageshack.us/img254/2320/uahmodeldec11.png

    This is also based on the following forecasts for Nino 3.4 and the AMO (and no stratospheric volcanoes).

    http://img201.imageshack.us/img201/6308/weeklyensoamoglsstsjan4.png

  27. Bill Illis –
    I was quietly studying your graph in the hope of gleaning some last minute extra info’, and I had a sneaking suspicion you might be perpetrating an elaborate hoax on unsuspecting quatloo hunters such as myself.
    Did you really mean 0.015? More than a tenth colder than 2011? So the second coldest year of the new century?
    If so, my initial projections are completely up the spout!
    *
    Hm, perhaps its time to re-hire the tipster I unceremoniously sacked after last months prediction debacle…

  28. “when the UAH TLT value is published in December”

    Repeated, ” Roy posts the December value in 2012.”

    So this is Jan through November common average?

  30. Tim C-
    All 12 months of 2012 methinks. Lucia wrote when Roy posts the December value for 2012. I.e in the first week of 2013.
    *
    Minor typo for the other bit 🙂

  31. Interesting, I can see the Quatloo from my home (iMac/Firefox) but not at work (PC/Firefox).

  33. Was being a little wicked there, omitted a grin. I do “cannot have written that” too.

    Not sure what they Portaloos are, never mind, took a guess at a figure anyway. is a difficult dataset, too noisy, however, on withholding one year a forecast was 0.01C out, fluke I reckon, so I ran on the lot, not that it makes much difference.

  36. eli you are betting against others not the house. so u just need to handicap properly. like with sensitivity. the ipcc has it down for 3c so u know its less than that

  41. I will take the liberty of doing a quick analysis of the bets.
    Total bets (final bets only) = 44
    Overall max. = 0.91c
    Overall min. = -0.042c
    Overall mean = 0.209c
    Standard Dev. = 0.180c
    Overall median = 0.169c
    Mean bets 1-22 = 0.213c
    Mean bets 23-44 = 0.204c
    The overall mean was heavily influenced by the two bets around 0.9c, and without those, the mean would have been only 0.175c.
    Generally speaking the temperatures declined as the period of betting progressed.
    Lucia, I was disappointed to see that you didn’t enter a bet, as I was interested in your opinion.
    Could you let us know what you think, even if you don’t actually enter a bet?
    I would also be personally interested in hearing the reasoning behind the two highest bets, assuming that they are not typo’s.

  42. Ray –
    I agree that it’s pragmatic to look at the mean without the two bets around 0.9C.
    Can I be greedy and ask what the standard deviation would be without those two?

  45. Ray – thanks.
    Quite a difference.

    I was wondering how close to the mean was my 0.177.. and I started wondering [having not noticed that you put the ‘pragmatic mean’ of 0.175 in as well]

    I agree that it would be nice to see Lucia’s ‘calculation’. I think we should extend the deadline for our host and twist her arm to put a hatful of quatloos on the line…

  46. I just looked at AMSU channel 5, discover.itsc.uah.edu/amsutemps/ from UAH. On Jan 19, the temperature crossed over the line into “coldest” territory. Unfortunately, that graphing tool allows graphing only temperatures only since 2002, so that the current “coldest temperatures” means only for the last 10 years.

    Perhaps I will look later on at temperatures prior to 2002 later on today or tomorrow. I am a bit peeved that the script for this tool a year or so ago removed the ability to graph min, max and average temperatures from the satellite era.

  47. AFPhys,
    Are you sure that there are data available prior to July 2002?
    The “show data as text” option only goes back as far as that.
    The cumulative figure for January is now below all years except 2008, but it seems unlikely to go below that year, unless it gets *really* cold!

  49. Thanks Ray and DeWitt – I didn’t realize that the data presented there was exclusively AQUA. Still, that does not explain why the average, max and min were removed from the script.

  50. Nearly all AMSU channels now are colder than any others since 2002. From channel 5 to 8, right now is the coldest for this date. In channels 9 and above, there is a year or two that read slightly cooler than the current temperature, though other channels are now at record lows.

  51. AFPhys,
    While the individual daily temp. for AQUA CH5 and the other channels may be below previous years, the critical thing as far as the likely UAH temp. for January is concerned is the cumulative figure for this point.
    I don’t know about the other channels, but even though daily CH5 is currently well below the previous temps. for this point in the month, (since 2003), the cumulative figure for January is still about 0.07c higher than that for 2008.
    There was a slight rise in the CH5 temp. on Jan. 24, although it remains to be seen whether or not that is the start of a sustained period of warming. However, it would take daily falls of about 0.03c for the remainder of the month for this January to fall below the cumulative figure for January 2008, which seems very unlikely, if not impossible.
    That said, the CH5 data files don’t contain figure for the first two days of January 2008, so we can’t be 100% sure of the cumulative figure for that month.

  52. Ray: Pointing out that the satellite temps are coldest on record at all altitudes, nearly, for the current date (late January) was the point of my post. I agree with all you wrote in response above.

  53. AFPhys,
    Sorry, it wasn’t my intention to contradict what you said in your post in any way.
    I notice that there has been a surge in CH5 and other lower channel temperatures over the last couple of days, which means it is rapidly catching up on 2008, which will create some uncertainty over the final UAH figure.
    It does puzzle me how the temperature can apparently increase so rapidly over such a short space of time.

  54. Ray or/-

    A layman’s observation – the temperature profile appears to be either heading upwards or downwards. Staying the same seems out of its range of abilities. I guess this could be for many reasons including just a visual impression..

    Is there any physical basis/mechanism for the way heat transfers at particular layers in the atmosphere? Such that it is always moving in one direction or another?

  55. Ray: I took no offense … no apology needed. I appreciate your own observations.
    Anteros: One thing seems clear — the cause of the bulk of the movement is not mere “random noise”, or “faulty readings”. On all channels the change more resembles a random walk superimposed on the underlying (approximately) sinusoidal temperature movements. It seems that one possibility is that there exist convection-like mass movements of the atmosphere at all levels that transport heat. It would be interesting to attempt to correlate this movement with such measurable quantities such as immediate solar activity, cyclonic activity, etc. and attempt to eliminate those reasons. One would have to understand intimately the measurement details by the satellite.

  57. Anteros/AFPhys,
    My “impression” is that the temp. trends one way for a while, and then turns the other way for a while. I haven’t done any detailed calculations but these “cycles” normally seem to last about 7-10 days. During the “transition” between one phase and another, the rate of change seems to be lower. It is very unusual to see such a long period in one direction as we have seen this January. There have been periods where there hasn’t been much of a rise or fall (e.g. March 5-17th 2011).
    I have started to look at the relationship between daily changes in UAH and AQUA CH5 and while the trends over a month are usually similar, the changes on individual days can be in opposite directions.
    I see that CH5 is on it’s way up again, quite rapidly and if things revert to their normal pattern, that should continue until at least the end of the month.

  58. Ah – woe is me!

    My fears and premonitions were all true! The January anomaly is indeed lower than all our guesstimates and because i had the fifth lowest guess I was always going to be out of the money!!

    FWIW (and I can’t say I care overy much 🙁 )Roy has posted the January anomaly as minus 0.09.

Comments are closed.