Grantsmanship Is A Stochastic Process

March 13, 2011

The unpredictability of grant review is exactly why you need to have multiple grant applications in the system all the time, targeting different study sections for review. Seeing young investigators moaning and groaning on the Internet about the fate of “my R01 application” is heartbreaking, because it is so foolish. And anyone who complains that they “can’t write multiple R01s” because they lack the resources, ideas, preliminary data, whatever are simply not thinking creatively. The same exact preliminary data can form the basis for an infinite number of different grant applications.

If your career hinges on the review of a single grant application, you are doing it wrong. Even the most successful grant writers submit multiple competing applications (including resubmissions) for every one that gets funded. Over the years, I have averaged two competing NIH submissions (including resubmissions) per year, with only about 1/3 ultimately getting funded.

Grantsmanship is a stochastic process. Deny this fact at your peril.

35 Responses to “Grantsmanship Is A Stochastic Process”

  1. namnezia Says:

    How big is your lab CPP?

  2. Dr Becca Says:

    What would you say is the ideal time for a brand new faculty person to begin submitting grant applications? Should I be working on one (or more?!) over the summer to submit in Oct, or work through my 1st semester to submit for Feb next year?

  3. drugmonkey Says:

    Mine is way smaller than his and I totally endorse this position, namnezia. Even when my lab was me and one tech I understood this. It is reality. And you deny it at your peril.

  4. BikeMonkey Says:

    I don’t know when you start Dr Becca, but if you think you are going to hit Oct 5 after a month on the job, yes, you damn well better have the science written in advance. Learning a new institution’s submission process takes time and effort.

    CPP’s hit rate is admirable but nobody new should bank on this right now. I’d say you need at least 5 apps (2 can be R21s) including one revision in the first year to have much confidence…

  5. New PIs should *not* be writing R21s, at least until you are awarded your first R01. The time, effort, data, and verbiage should be devoted 100% to nailing an R01.

    This is for a number of reasons:

    (1) Only R01s have a relaxed new PI payline.

    (2) The application is over half the length of an R01, but the maximum total budget is less than one quarter.

    (3) R21s are not renewable; R01s are, and there is a *massive* scoring bonus for competing renewals during peer review.

  6. juniorprof Says:

    In my first 3.5 years on the TT I averaged 4 NIH applications (including resubs) per year. I’ve gone after 5 mechanisms: R01, R21, U01, Blueprint specific stuff (R01 level) and that crazy competitive new investigator award. That doesn’t include numerous foundation grants. I’ve hit on 1 R01 and think I’m dangerously close on another. I also have a foundation grant. Since starting, my lab has consistently been made up of 4 students plus my occasional presence at the bench when needed.

    In terms of preliminary data, I’ve used essentially the same prelim data for several different grants without any problem. I’ve also spun off collaboration prelim data (data from the collaborating lab) for my own grants. I’ve been able to do this largely because I am collaborating with very well funded senior investigator labs who have a vested interest in my success.

    While I’ve been able to get my grants into a variety of study sections this has largely occurred because I have gone after some RFA with special study sections. For normal R01 apps I have struggled to find study sections that fit the science but we recently generated some prelim data that can be combined with some other published work to move into a different area and a new study section. I’m hopeful that this will be a big help for the lab and I’m going to get ahold of the SRO to make sure this will fly before putting that app in for the June deadline.

  7. cackleofradness Says:

    What has been the % chance of funding over the course of your career? Does your strategy of 2 RO1’s, on average, per year mean that you spend a great deal of time tweaking each one? I ask since my contemporaries are sending in ~4-8 RO1’s per year with the chance of funding ~10% or less.

  8. drugmonkey Says:

    Well cackle, you just start from here and work backwards. Stop when you think you’ve sufficiently guessed his age….even in the most conservative estimate, well we’ve seen quite a range of success rates from NIH…

  9. Namnezia Says:

    DM – but what I was getting at is, currently we have 2 grants with about 2-3 years left on each of them plus a foundation grant. My lab is smallish so we have more than sufficient funding, so the question is should I be writing new grants like crazy, preparing for my grant renewals, starting new pilot projects for future grants, or what?

  10. You’re damn right you should be!

  11. cackleofradness Says:

    DM – I believe what I’m asking, in more general terms, is does one submit more grants when paylines are low, as my friends seem to be doing – or – does one submit fewer, better grants when the money is thin?

  12. DrugMonkey Says:

    The notion that one can write grants “better” in a way that beats the odds of the stochastic process is a fallacy. Common enough, you hear it from POs frequently, but totally wrong. Many of my funded proposals have glaring “flaws”. Many of my unfunded ones are damn close to

  13. GMP Says:

    (Trying to learn more about the NIH) Is there a limit in the number of grants you can submit to the NIH per year (without counting specific solicitations)? I submit to the NSF, and you are allowed only one proposal as a PI or co-PI per division per proposal window — my division is generous as we get two proposal windows per year, many divisions have only once. There’s no way I could submit 8 proposals per year (which I close to what I do) if I stuck with the NSF unsolicited proposals alone…

  14. arrzey Says:

    As someone who (I think) is significantly older than DM & PP, I can only concur with them. Becca, start writing now. Save bits & pieces to be used elsewhere.

    There are times when I hit 1 in 2 being funded (in the heyday)and times when I did not. Did not even come close.

    The best piece of advice from someone on my thesis committee (back in the Paleocene), “papers (and grants) are never finished, they are merely abandoned to publication (and study sections)”. Do not get hung up on perfection, do not get hung up on “if I wait till the next cycle it will be perfect”. Submit early and submit often.

  15. […] -CPP has a great post up on why you should always have multiple grants in the hopper. […]

  16. Pregnant Elephant Says:


    “What has been the % chance of funding over the course of your career? Does your strategy of 2 RO1′s, on average, per year mean that you spend a great deal of time tweaking each one? I ask since my contemporaries are sending in ~4-8 RO1′s per year with the chance of funding ~10% or less.”

    It seems that, indeed, grantsmanship is a stochastic process. The question arises as to how much time do your contemporaries devote to write the science involved in 4-8 R01’s per year, as opposed to the time devoted to do science and make science happens in their labs… It may well be that science is becoming a stochastic process as well. How wonderful!.

  17. DrugMonkey Says:

    If you are not eventually using text and figures generated for grant applications in your manuscripts, something is wrong PE.

  18. Pregnant Elephant Says:

    Sorry DM. I don’t seem to get it. It is not unusual though. Being pregnant at an old age is taking its toll. That’s the consequence of too much love and little, if any, precautions. My only wish, right now, is that the little one is ok and healthy!. I will take care of my grantsmanship when my maternity leave is over. Good luck to you, your grant (s) and paper(s).

  19. […] gadfly and MFJ connoisseur CPP has a post titled “Grantsmanship is a Stochastic Process.” If you are, or will be, involved in grants in any way, even peripherally, you should read […]

  20. lylebot Says:

    Science has been a stochastic process at least since Fisher, Pearson, and Gosset invented statistical hypothesis testing.

  21. Pregnant Elephant Says:

    Thanks Lylebot for your input.

    Is science a stochastic process? I understand that science has a stochastic component in that its “products” ought to be analyzed and validated in terms of probability. Modeling scientific questions are also stochastic processes but defining science as a stochastic process would seem restrictive to me.

    I’d appreciate if you would elaborate a little more. Maybe, I am too much attached to the meaning of the words.

  22. andrewD Says:

    Grantsmanship is a stochastic process. Deny this fact at your peril

    I read this as grants are allocated by a process with a high degree of Randomness rather purely than on the basis of the best submissions.

    I have a suggestion that would give more money for grants and cut out wasteful administration,-
    Once a year there is a competition for funds, the NIH allocates a total sum and each person wanting a grant sends in a Card (say 8”x4”) with their name and address, project title and sum of money requested. These cards are all put in drum and the drum shaken; cards are drawn at random (by a celebrity Nobel Prize winner?) And the requested funds are allocated to the submitters of the cards drawn; a running total of the sum allocated is kept, when the allocated sum equals the declared amount available the draw stops and any remaining card destroyed. It might be sensible to cap the amount allocated to each applicant to prevent someone scooping the whole pool
    This is a purely stochastic process, NIH admin is reduced and CPP and others would not need to write grants but could do research just think of the extra real work that could funded and done .

    Please note I have no irons in this fire: for 2 reasons,1) I am in the UK and 2) I am an industrial chemist
    Andrew D

  23. […] query at PhysioProf’s blog from cackleofrad asks: What has been the % chance of funding over the course of your career? Does your strategy of 2 […]

  24. DrugMonkey Says:

    there is no limit on the number of NIH apps a PI can submit, no. The only related rule is that one PI cannot have two substantially similar apps under consideration at the same time. Oh, and the occasional special funding opportunity when an *institution* is limited to one proposal.


    2-3 yes out should you be doing pilot work- yes, this shouldn’t even be a question.

    -readying for renewal? Heck yes although you need to work backward from the end of your existing awards to figure out when the A0 has to go in to give the expected A1 a chance of funding when you need it.

    The question of throwing down new projects now, and risking becoming a 3 R01 lab when you feel most comfortable as a 2 R01 lab is an interesting one. Depends on your risk tolerance, I guess. At current paylines I think you have to be trying all the time to advance the date at which your last bird-in-the-hand expires..

  25. DrugMonkey Says:

    One thing that would relieve pressure on labs to become a N 1 R01 lab (just to ensure continuity, not because the PI wishes to expand) would be the ability to delay the start of an award for a year.

  26. […] Flexibility: The ability to change directions as necessary. Follow where the data is taking you. Start up the new projects (or new branches of existing projects) you’re going to need for the 2-3 R01 (or equivalent) applications you’re going to be submitting each year.* […]

  27. GMP Says:

    DM, thanks for answering my question!

  28. anonymous Says:

    The part that is messed up to me is the idea that the exact same preliminary data should be used for multiple grants. This just seems like a giant waste of everybody’s time.

    I think NSF has it right TBH – one grant at a time. If you want multiple grants, write to multiple agencies, or allow for people to ask for more money (that is, enough to support what would previously have been “two projects”) on a single grant.

  29. The part that is messed up to me is the idea that the exact same preliminary data should be used for multiple grants. This just seems like a giant waste of everybody’s time.

    Either you don’t understand the role that preliminary data play in a grant application, or you have no scientific imagination.

  30. drugmonkey Says:

    Preliminary Data often functions to show you have the chops to get done what you say you can get done. Technical feasibility, so to speak. It can be sufficient even with no direct test of the hypotheses you are advancing being presented. Or even if 50/50 with direct Preliminary data, the technical stuff can make a huge difference.

    Particularly for newbies.

  31. juniorprof Says:

    yes, particularly for newbies

  32. anonymous Says:

    I’m not saying it’s impossible or even difficult to come up with innumerable “separate” grant applications supported by a given set of preliminary data.

    I’m instead saying that it’s a waste of everyone’s time (PI’s, reviewers, panelists, etc) that the system is set up so that you have to do this in order to get a single NIH grant. No matter which one of your submissions eventually gets funded you’re ultimately going to do the same research anyway. Some of it might by chance be the proposed research in the funded grant but it probably will be a mix of stuff from the various grants or whatever is most relevant by the time the money comes and everything has changed.

  33. No matter which one of your submissions eventually gets funded you’re ultimately going to do the same research anyway.

    You are still thinking in a very short-sighted and uncreative fashion, essentially like a post-doc and not a leader. A lot of the time, the same preliminary data supports completely different lines of research, and if you are creative and an effective grant writer, you can actually end up with more than one grant funded and be able to pursue those multiple lines.

  34. drugmonkey Says:

    If you think you will literally be doing the exact same research than you are going to have difficulty writing an application that has a high probability of succeeding.

    It only works if you can genuinely see a line of productive work ahead of the Aims you propose….

  35. Dr. O Says:

    Anon, I’m agreed with DM and CPP on this point. I’m always keeping an eye out for alternative directions my research could go, if only because I don’t know when my *main* route will hit a dead end. If you’re not thinking bigger picture, you’re bound to end up hitting a wall. And if you don’t see how your research could apply to multiple aspects of biology, therefore leading in many different directions, you’re either 1) not working on a high-impact enough project or 2) are very short-sighted, as CPP indicates.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: