under/over

Hubble proposal deadline tomorrow, lest you forget.
How many proposals do you think we will get?
I think I know. You wanna bet?

The Hubble Space Telescope is a fine telescope.
Currently it has very limited capabilities due to instrument failures, but it is, hopefully, getting one more servicing mission – during which a new Wide Field Camera (3) will be installed – with a moderately wide field, well sampled broadband performance (UV to near-IR – an interesting compromise).
The Cosmic Origins Spectrograph – a fixed single aperture medium resolution blue spectrograph.

In addition there are plans to fix the Space Telescope Imaging Spectrograph and the Advanced Camera for Surveys (yay!) if they can, as well as various housekeeping and upgrades of operational hardware (usual gyro/battery/power/control stuff)

So… this will assuredly be the last servicing mission, but Hubble ought to have full capabilities for a few cycles until it reaches the end, so everyone wants to get their priority science in asap.

So… how many proposals? What oversubscription rate? Some people are getting very ambitious for large programs.

I’m betting over 2000 proposals! – no stakes, just yer NASA Nerd Pride level and community reputation – modulo the dictum that one should never bet on peoples’ behaviour.
I also suspect a record oversuscription rate for science orbits, especially with new GTO teams and need to do some orbit engineering and verification tasks.

So, less than 5% success rate, anyone?

You have 30 hours to stake your claim.
Anything after 4:55 pm EST friday doesn’t count, that is close enough that proposal numbers will be converging on a final count, you must stick your neck out before then or I will ignore you.

PS: starting 4:55, feel free to tell us your highest successfully submitted assigned proposal cycle number…

Comments

  1. #1 mihos
    March 6, 2008

    I think I’m going to sit out this cycle, just so as to give the rest of you schmucks a fighting chance.

  2. #2 Steinn Sigurdsson
    March 6, 2008

    wait! weren’t we writing a proposal together…?

  3. #3 E
    March 6, 2008

    If I recall, the rates last year were 5x oversubscription and ~700 proposals, to use two working instruments (WFPC2 and NICMOS). This year, we’ll nominally have five working instruments (WFC3, ACS, STIS, COS, NICMOS). If both rates are roughly linear with the number of instruments, then this suggests ~12x oversubscription and ~1750 proposals. Of course, astronomers are greedy nonlinear creatures, so these numbers may not be worth much.

    Personally, my gambit is to go with small proposals and try to squeeze in on the margins with high-risk/high-reward ideas. I’m also trying an archival program this year. There are so many new instruments to play with, I wouldn’t be surprised if AR proposals are somewhat neglected in the rush to get new data.

  4. #4 Julianne
    March 6, 2008

    I’m guessing we won’t quite break 2000. There is a limit to how many proposals you can write. On the flipside, though, you have many people (i.e. spectroscopists) who were sitting out the past several cycles. I’ll go with 1900.

    Data point: They’re currently up to 200. I think this is about 2x where there were last year at a similar point (as I recall).

    Next question: Will the oversubscription be worse in the galactic or extragalactic panels?

  5. #5 Karen
    March 6, 2008

    I’ll guess 1820.

    I’m doing archival, too, mainly for the reasons stated above. (and that I got lots of data last cycle that still needs some work)

  6. #6 Ben
    March 6, 2008

    1400, but the oversubscription will be high. People will write big ambitious proposals for new instruments. There are only so many things you can point COS at, and many orbits in the GTO program.

    I thought the oversubscription for the last cryo cycle of Spitzer would be higher than it turned out, about 5.3. I think people scared themselves off.

  7. #7 JohnD
    March 6, 2008

    11:25pm and I’m working on #2 of 3 props that I’m on. Since I put one in last year, I’m going for 2100 (are we playing Price is Right rules?). I think that will be an underestimate. I smell hunger for orbits out there.

  8. #8 Ben
    March 7, 2008

    STScI should run a proposal ID lottery. They pick a random number (I’m sure there’s a procedure in stsdas for it) and whoever had the lucky proposal ID, the Director takes them out for a crab cake sandwich, driving Karl Gordon’s Mustang.

    No, I’m not punchy. Well, maybe a little.

  9. #9 Pat Durrell
    March 7, 2008

    OK…I will go with 1763. Do we need error bars? 1 sigma or 3?
    Just checking.

  10. #10 Steinn Sigurdsson
    March 7, 2008

    Error bars?
    We don’t need no steenkin’ error bars!

    I’m a theorist, dammitt.

    Wait, Karl has a mustang? Classic or one of these new jobs?

  11. #11 Kayhan Gultekin
    March 7, 2008

    Done!

    Now it’s time to work on my Chandra proposals……..

  12. #12 Caryl
    March 7, 2008

    I’m going with 1600, but agree that the oversubscription will be high. There are lots of large orbit WFC3 proposals going in this cycle! New data point, #534 at about 3 pm EST.

  13. #13 JohnD
    March 7, 2008

    Well, one’s in the bag. Onto the other one…

    #630

  14. #14 E
    March 7, 2008

    A couple of data points for compiling dN/dt:

    Submission time: 12:30, Proposals: 550
    Submission time: 2:00, Proposals: 650

    (both time are PST)

  15. #15 Craig Heinke
    March 7, 2008

    Submission time 5:21 pm (EST), proposals: 663.
    It’s all Hail Mary throws this round, but they’re all fun.

  16. #16 Pat Durrell
    March 7, 2008

    More data…time for a spline fit (or some such thing)

    #657 at 5:18pm

  17. #17 Julianne
    March 7, 2008

    Interesting! These are not going up as fast as I had guessed! Given that you can resubmit up to the deadline, there’s not reason not to do an early first submission to grab yourself a not-yet-exhausted-from-reading low proposal number. I’m guessing linear rise until the deadline.

  18. #18 Julianne
    March 7, 2008

    Because I have nothing to do, I posted a plot of the data above. I’ll update if you give me more data points!

  19. #19 Caryl
    March 7, 2008

    #764 at 6:27 pm EST

  20. #20 JohnD
    March 7, 2008

    #779 @ 6:43 EST

  21. #21 Pat Durrell
    March 7, 2008

    Just a thought…can someone get funding for an archival proposal for analysis on the HST Prop Number vs Time plot? Perhaps over several years? Now I *have* lost it…

  22. #22 E
    March 7, 2008

    One more data point:

    Submission time: 4:28 PST, Proposals: 860

  23. #23 Robin
    March 7, 2008

    #870 at 7:38 EST

  24. #24 Jane Charlton
    March 7, 2008

    #928 at about 7:50 EST

  25. #25 Julianne
    March 7, 2008

    Latest updated figure is over at Cosmic Variance. If the extrapolation is valid, we might break 1000, but it’s not guaranteed. This is completely off what I was expecting.

  26. #26 Steinn Sigurdsson
    March 7, 2008

    Ok, Ben wins NASA Uber Nerd of ScienceBlogs award of the week.

    And he was still off by 40%!

    I’m in shock, how wrong can I be.
    ‘course I myself was on fewer proposals this year than last year, so I shoodanode,
    but I figured extrapolating from local biased data was invalid

    What we really need, for archival research, is a proposal meta-analysis.
    Not just numbers of proposals and orbits, but success rate as a function of submission ranking, by proposal class; and then post data impact factor and h-score by proposal number.

    Yeaargh!

  27. #27 Julianne
    March 8, 2008

    Ben used to beat me at everything in high school too ;-)

    I’ve also wondered about success vs. proposal number. The ones I submit early are the ones where (1) I’m resubmitting from last year, so it’s already been a loser once or (2) I’m way on top of my game. From the other side as a panelist, if you get a very nicely written proposal near the bottom of the stack, you feel an extra little gush of gratitude.
    My guess is that it’s a wash.

  28. #28 Brad Holden
    March 8, 2008

    Be interesting to see if there are lot of big orbit programs

    I wonder if a lion’s share of spectroscopists were on the COS GTO team, so are getting their photons for free (so to speak).

  29. #29 Ben
    March 8, 2008

    Hey, I was only 0.15 dex off.

    Don’t be taken in by Julianne’s plea for pity. She used to beat me all the time too.

    A cynical (or practical) person could say that the number of proposals is governed by the amount of funding available as well as the number of instruments.

    Regarding success by proposal number, sometimes I deliberately read them out of order. Plus I read the ones that I’m primary or secondary on first, or more carefully. At the most recent panel I was on, we discussed them in order of subject area rather than number, which turned out well.

  30. #30 Steinn Sigurdsson
    March 8, 2008

    What high school, in these times, produced two astro faculty in a single class?!

    I’d think shuffling proposal order would be a good thing to do, but there may still be systematic differences between proposal qualities based on submission ranking – as Julianne noted, the early propoals tend to be retreads while the later ones may be poorly prepared.

    There were some studies in Europe in the ’90s which effectively said that proposing is limited by perceived success probability, people do give up and stop sending in proposals, they suggested the level was close to where expected return became comparable to the mean effort.

  31. #31 Ben
    March 9, 2008

    http://www.pps.k12.pa.us/allderdice/site/default.asp

    Public schools, baby. Steinn, you probably have its grads in your classes.

  32. #32 Julianne
    March 10, 2008

    It’s also on wikipedia:

    http://en.wikipedia.org/wiki/Taylor_Allderdice_High_School

    There’s a small army of astro folks who hail from the ‘Burgh. Ben tried to run away and be an english major, but it didn’t stick…