Skip to content

Advice: positive-sum, zero-sum, or negative-sum

There’s a lot of free advice out there. I offer some of it myself! As I’ve written before (see this post from 2008 reacting to this advice from Dan Goldstein for business school students, and this post from 2010 reacting to some general advice from Nassim Taleb), what we see is typically presented as advice to individuals, but it’s also interesting to consider the possible total effects if the advice is taken.

It’s time to play the game again. This time it’s advice from sociologist Fabio Rojas for Ph.D. students. I’ll copy his eight points of advice, then, for each, evaluate whether I think it is positive or negative sum:

1. Show up. Even if you feel horrible, show up. No matter what. Period. Unless someone died in your family, show up.
2. Do your job. Grade the papers. Do the lab work. Unless the work is extreme, take it in stride.
3. Be completely realistic about how you will be evaluated from day #1 – acquire a teaching record and a record of publication. Don’t have the fantasy that you will magically get the job of your dreams sans publications. Time spent on other issues is “out of pocket” – do it because you care, not because it will help you.
4. Hang out with winners. These people are actually pretty easy to identify – they do well in teaching and publication and they have a track record of placement. Also, ask around to see if people are nice. Where there is smoke, there’s fire.
5. Be constructive. It is easy to criticize people, but it really doesn’t accomplish much. Instead, if you actually offer to help and present a solution, then you’ll make a difference and people will appreciate it.
6. Say yes (unless it is a crazy person). In other words, join teams and accept projects, and say yes to grad school buddies. Once you get a few projects going, then you can say no.
7. No excuses: the only thing that matters is task completion. It may be long or short, but everyday should involve a core task.
8. Submit, submit, submit. Got rejected? No problem – just resubmit tomorrow. If you thought the reviewers were right, take a week and then resubmit. The key to success isn’t submission – it’s resubmission.

OK, now the evaluation:

1. Probably positive sum. If you don’t show up, this inconveniences others; whereas if you’re distracted and show up and do a crappy job, this still is likely to be better than people having to cover for you or reschedule the meeting.

2. Positive sum. I’m assuming there are enough entry and exit points so you can modulate your effort to match the time and energy you have available.

3. Positive sum. Information is good. But I’d like to emphasize the last sentence of Rojas’s advice #3, and point out that, if you do something un-work-related because you care, that’s cool too.

4. At first this might sound negative-sum in that it would seem to promote a judgmental air (evaluating “winners” and thus, implicitly, “losers”), I actually think it’s positive sum, in that, to the extent more people follow this advice, it motivates potential collaborators and research supervisors to up their game.

Similarly, I generally recommend that students taking classes shop around and, to the extent possible, choose instructors with excellent teaching evaluations. The signal isn’t perfect but it’s there (in my impression), and it also has generally positive effects in the sense of encouraging instructors to try harder and take their teaching more seriously.

5. This advice could be positive or negative sum, depending on how it’s taken. If someone reads this advice and is encouraged to be constructive and help people, this is positive. But if someone reads this advice and is discouraged from offering frank criticism, I think it could be negative. Criticism can be useful, and I’m sure I’m not the only one who doesn’t get enough of it! I disagree with Rojas’s statement that criticism alone “really doesn’t accomplish much.” It might not seem to be accomplishing much, but if you can save someone a year of effort by pointing out they’re on the wrong track, that can be a big deal! Some of the biggest gains in life are the avoidances of big losses.

6. Positive sum. Collaboration is good, and if it doesn’t work, it typically dissipates anyway. I’ve seen so many students say they’re too busy to take on a new project, and what does it get them? They’re just sitting in some room getting stuck on page 34 of some thesis that nobody will ever read anyway.

7. Mix of positive and negative sum. On one hand, everyone benefits when a task is completed. On the other hand, we can all benefit from when we help each other understand things, so if people literally follow the advice that “the only thing that matters is task completion,” this could make like more difficult for everyone.

8. Negative sum, I’d say. If everyone submits and publishes everything to refereed journals, this just overwhelms the system. Yes, I realize that by publishing dozens of papers a year myself, I have an outsized “carbon footprint” myself, so maybe I’m not one to talk. . . .

Ok, in sum, that’s 6 points on the positive side and 2 on the negative side. Overall, the advice is positive sum. Excellent!

P.S. I emailed Rojas to let him know of this post, and he should feel free to respond to any of the comments below.


  1. Entsophy says:

    Here is some positive sum advice for ya:

    (1) Quit academia and get a real job. One that produces something others want. Stop being a parasite on society and earn an honest living. Get some self respect in the process. Time spent on writing papers is “out of pocket” – do it because you care, not because it will help you.

    (2) Think about something important and hard for a long time until it cracks wide open.

    (3) Sit on those insights until you’ve developed them into something deep and simple.

    (4) Tell others about it as clearly as you can.

    (5) Write a letter to whatever grant agencies primarily fund your field and tell them you’d rather have hot pokers stuck in your eyeballs than write a grant proposal begging for money like a loser.

    • Rahul says:

      Students pay (quite dearly too!) for four years of university education. I imagine there’s a product they want?

      • What Economists call “signalling”.

      • Entsophy says:


        I’ve been around Academia enough to have every last self serving justification for taking taxpayers money and producing unread papers from it. I’ve heard them all a thousand times.

        The funny thing is that even though Academics say this stuff to each other constantly, most of them still internally feel like frauds or have “impostor syndrome”.

        Well there’s a good reason for that. Most are imposters.

        • Phil says:

          Allow me to point out the irony of someone reading a blog every day, and frequently commenting on it, and then obnoxiously claiming that it is not worth anything.

          Also allow me to point out that Columbia is a private university.

          Entsophy, I have to say I’m disappointed by your attitude. You’ve always seemed thoughtful, seemed to value the discussion here, and I’ve felt that you contribute to it. If someone is forcing you to participate against your will, call 911. If you are a voluntary participant then at best your attitude on this thread is inconsistent with your behavior.

          • Entsophy says:

            Who said anything bad about blogs? Or that they’re not worth anything? I love blogs.

            • Phil says:

              This is an blog written by an academic. Many of its discussions are centered on papers Andrew or others have written, and almost all of them have an academic bent. If Andrew were not an academic this blog would not exist, or at least not in anything like this form. Your item (1) says that people like Andrew are parasites, that they are not earning an honest living, and they do not produce anything people want.

              And yet, here you are.

              • George says:

                I think you are missing Entsophy’s point. That a large part of academia resembles monasticism. So, instead of living off donations and be submissive to the hierarchy as most monks and academics do pretending you produce something extremely valuable snap out of it and do something of real impact in society. I take this advice to be more sincere than Rojas’ Canon Law…

          • It’s possible to be against the current status of Academia as a social institution, and it’s average productivity, without hating individual academics. In my opinion Entsophy has been pretty consistent on this blog that he appreciates what Andrew does. in fact, in many ways by setting up this blog Andrew has done something very unlike the modern Academic system and directly engaged the public instead of sitting in an Ivory Tower publishing only paywalled articles with no feedback other than a few randomly chosen “peer reviewers” who you can generally circumvent using the advice given in the article “resubmit often”.

            What’s wrong with Academia is not that there are people like Andrew organizing the development of Stan and publishing useful stuff on blogs… what’s wrong with Academia is that it’s in a state where advice like the stuff in the original article is considered useful and maybe necessary career advice

            • Phil says:

              I agree with your first sentence, maybe even your whole first paragraph.

              I don’t quite understand your second paragraph…maybe I don’t disagree with it either. Are you saying the advice “show up, do your job, be constructive, say yes to projects” shows that there is something wrong with Academia because there is something wrong with any institution where this is good advice, or are you saying that the fact that this advice isn’t totally obvious indicates that there’s a problem? If you’re saying the former then I disagree with you; if you’re saying the latter then I probably agree. But I don’t think any of that has to do with Enstophy’s objections.

              Enstophy, do you really think most professors are parasites who contribute nothing? You think the world would be a better place without colleges and universities? I really just can’t make your item (1) seem consistent with your behavior.

              • Entsophy says:

                Yes Phil, I do think most professors are parasites, including most of the famous ones and most of the ones who work hard.

                And yes, I do think our nation, most professors, and the fields the professors claim to champion, would be better off if they followed my advice given in the first comment above. Is there really any doubt that’s true?

              • K? O'Rourke says:

                Phil, about 20 years ago the Dean of Medicine at the University of Toronto (tied with Harvard on many measures of publication impact) obtained software to track citations and I was involved in his efforts to evaluate his faculty’s contributions to the literature. Setting aside self-citations, it was rather bleak. There were those rogues, for example a Rhode scholar well into their career with 100 plus citations listed on their CV of which only 5 or 6 could be verified. But setting those things aside, from memory about 50% had less than 100 and the 90% percentile was around 1000 (it was 20 years ago). Today pressures to publish are much more severe and outlets have expanded including predatory journals that essentially exist to simply publish papers in an apparently peer reviewed journal for a fee.

                Citations are a very faulty measure and there are many other contributions than publication, but I think there are reasons to expect many academics do not contribute much beyond an appearance of doing something meaningful.

                This raises red flags “Submit, submit, submit. Got rejected? No problem – just resubmit tomorrow” and maybe even more this “If you thought the reviewers were right, take a _week_ and then resubmit” generalizing the concept of least publishable unit to least adjustment/response to enable publication.

                Here is one real example. Reviewer B noted reference 2 in the paper actually provided a counter example to the main claim of the paper. Least adjustment to enable publication was to remove reference 2 and submit to another journal. What is known, is the same paper except without that reference2 was published a year or two later.

              • Phil, it’s a combination. Much of the advice given is of the form “this should be obvious” (like: show up?? I mean really). What isn’t obvious is more or less “gaming the system” (resubmit immediately, take at most a week to consider any serious criticism…) or “status chasing” (hang out with “winners”??)

                None of the advice was of the form “work on problems that make a difference in the world” or “just because it’s fancy sounding doesn’t mean it furthers any real goals”, or “don’t fool yourself by pretending you’ve made a discovery based on a few “volunteer” undergraduates who came to your study because they need extra credit”… yet those are some of the serious things that I think are wrong about modern Academia.

                This is advice about how to further your career in the modern academic system, but while much of it is probably “positive sum” for academics, it is probably negative sum for society.

              • Enstophy says:

                And one other thing Phil: please not that I put my money were my mouth is on this one. I follow my own advice.

              • Phil says:

                I think “academia” means different things to Enstophy and perhaps to Daniel L than it does to me.

                I went to a good private college and I think the professors there were generally good. Whether I got my money’s worth (or, rather, my parents’ money’s worth) is of course arguable, but I do not think they were “parasites.” What they were “producing” was my education and I did in fact learn from them.

                I had a pretty good grad school experience too; learned a lot; published exactly two papers, both of them worthwhile although not earth-shatttering…basically I don’t know what Enstophy is on about. If you want to claim there are academics who don’t contribute, well, obviously that’s true. But there are also people in the private sector who don’t contribute or who actively do harm: how about those people who assured investors that tranches of worthless loans were as good as gold?

                Enstophy, if it’s true that you have pursued multiple advanced degrees then I’m unable to see how your behavior is consistent with your rhetoric. Either you are getting something out of the Academy or you aren’t. If you are then shut up. If you aren’t then why do you keep going back?

        • Rahul says:

          Your blog says you are planning on getting a PhD? Are you? What gives? A non-academic PhD?

    • lowend1hz says:

      (1) Either you’re a Neo-Luddite or you have an unrealistic perspective of the role of academia in society. Transistors were developed by academics working at Bell Labs and generally teaching at the same time; although not involved in the development of the transistor, John Tukey is a great example of someone working in academia that produced work on a non-trivial scale. Academia is a society’s research and development arm; if you cut it off because you adopted the philosophy of a group of manufactures from pre-industrialized England, then I’d venture to argue that your actions are orders of magnitude worse for society.

      • Pretty sure Entsophy has a different view of what Academia was in say the first half of the 20th century compared to the second half. Furthermore, the existence of exceptionally productive people who were in Academia does not preclude his overall thesis from being basically correct on average. He has a lot of valid points about the current state of academia. As a marine, his presentation is admittedly rough around the edges. This is one of the reasons we keep him around ;-)

      • Entsophy says:


        -Andrew’s post was about sociologists today not physicists 80 years ago.

        -By any measure you care to use (other than unread and worthless papers) 1850-1950 produced far more results at a fraction of the cost to society.

        -Not only did they do that, but they did it without most of trappings Academics consider “essential” today for science.

        -As it happens, I laid out my views on physics in detail here:

        There isn’t even a hint of luddite in any of it.

        • bxg says:

          The original advice is clearly not aimed at people for whom knowledge production is their sole utility function (the “resubmit until accepted” item alone disproves this.)
          Whatever one thinks of (or is true of) academia, perhaps the advice is net positive for the utilities of his audience. And your advice might be better for someone whose utility function is “be like Fermat or Einstein” or “help society as much as possible, not privileging your own well-being in any particular way”. I took it as read that nothing like this was in the mind of the original advice giver.

          TL;DR version: how can we argue over whether advice is good or bad without being explicit about the utility?

  2. Tom says:

    This type of advice from professors is perhaps the worst sort of just world hypothesis when I was a kid uphill in the snow both ways pap.

  3. Phillip M. says:

    On #4. Coming in as a latent grad student, choosing winners among your grad colleagues is something in my experience thus far that a number of grad students do naturally. Some of their picks seem to be semi-altruistic (an exciting collaboration between equally dedicated individuals) while more often, their picks are not (the radar dodgers, the skaters, and the free riders). This has a negative sum outcome within any number of working groups I’ve witnessed. When the goal of another’s doctorate is simply the diploma, and not the skill set, then that person will make the rational set of choices (ie ‘leeching’) which minimize the complicated road of achieving it.

    On #5. The value of a criticism doesn’t lie necessarily in immediate redirection or the solution one presents. All too often egos stand in the way of objective thinking (in my work life, I find this to be especially true). In such cases, I find criticisms most often a ‘seed’. While immediate dismissal or some conniption fit may happen at the time of even the fairest of criticisms (as little more than ego driven opposition to them), they may eventually bloom, though you may not be around to have been vindicated. So I see this perhaps over-optimistically as a positive sum game, which is the latent value created by he criticism, not so much the initial subsequent action to it.

  4. question says:


    “Ok, in sum, that’s 6 points on the positive side and 2 on the negative side. Overall, the advice is positive sum. Excellent!”
    This is exactly like strawman NHST. All you know at the end is + and – direction of each factor. We need to know the magnitude to which each point of advice affects others to know whether this list is excellent or not.

    I would say the negative effect of encouraging the production of an insane amount of low quality publications probably outweighs the rest of the advice.

    • Andrew says:


      1. In adding pluses and minuses I am following Robyn Dawes in his classic paper on improper linear models.

      2. To be fair, Rojas does not “encourage the production of an insane amount of low quality publications.” He says “submit submit submit” but it looks like he’s talking about submitting one paper at a time, not getting discouraged by rejection conditional on you thinking the paper is worth submitting in the first place.

      • question says:

        This one?
        The Robust Beauty of Improper Linear Models in Decision Making

          • question says:

            From that paper it appears that the accuracy of expert opinion is inversely related to the number of predictor weights they are asked to estimate. This makes some sense, since each estimation is another chance to be way off thus totally screwing up the model. A similar effect occurs when the weights are estimated from small samples.

            When choosing which predictor variables to include, the experts are essentially estimating the order of magnitude for many weights and including only those at the top level in making their final decision. Perhaps instead they should instead focus on only the top two or three. I don’t know if anyone has studied this aspect, but it seems related to the n/k dependance seen for small samples.

            Using weights between -1 and 1 then dividing by number of weights:
            5) +0.5. Coming up with constructive advice is much more effective than simply critiquing. Beyond the social aspect, improved methods also advance science. However, lack of criticism fosters incompetence.

            7) -0.8. I have seen so many hours wasted on “task completion” when if someone had taken the time to just sat down and think about it they could have come up with a way to automate the task. This saves many hours in the future along with allowing collection of more/richer data. There really is no way to estimate the time it will take to complete the task of figuring out how to automate.

            8) -0.9. The actual problems with most papers I see would take much longer than a week to fix. Also, I have never seen evidence in favor of peer review. I have my doubts that in these days of the internet that even the first “submit” to a journal is the correct thing to be doing.

            Estimated utility= -1.2/3 = -0.4

            Because “utility” is so poorly defined here I think we need to settle for net negative.

            • Andrew says:


              If you look at my post, I was implicitly giving each piece of advice a score from -1 to 1. My scores were 1,1,1,1,0,1,0,-1, for a total of 4 (or, as I put it, 6 points on the positive side and 2 on the negative side, with my -2 coming from half of item 5, half of item 7, and all of item 8). Your ratings don’t seem so different from mine, you’re just giving zero weight to items 1-4 and 6, all of which I rated as positive.

              • question says:

                You are right, interesting it worked out that way. I was trying to explore this unit weight vs differential weight concept. It is still odd to me that the relative advantage of unit weights under small sample size conditions can be decreased by simply reducing the number of predictors. I’m not sure I grasp the implication that ignoring any evidence (even if weak) can be preferable. This was the first I came across that info though so thanks for mentioning it.

        • bxg says:

          I think I am taking you too seriously here, but is this really an application of Dawes? My recollection of Dawes is that his results assumes the predictors are normalized in magnitude (in addition, of course, have the “right” sign); you don’t just give weight 1 to un-normalized variables or get to just make up a scale. So there _is_ a magnitude question to be answered. Now what the actual model you are trying to invoke here is not at all clear, so I don’t know what you want to take the magnitude _of_, which tends to confirm I’m taking this too seriously.

  5. Blaise F Egan says:

    I just noticed this advice in the comments to the 2010 post about Taleb:

    >There is literature suggesting thst short, intense workouts are the best way to do this

    This is really dangerous. Last year the BBC journalist Andrew Marr gave himself a major stroke doing exactly that.

  6. Ruben Arslan says:

    > 1. Probably positive sum. If you don’t show up, this inconveniences others; whereas if you’re distracted and show up and do a crappy job, this still is likely to be better than people having to cover for you or reschedule the meeting.

    This raises expectations and if all students treat academia as being more important than anything but death in the family and if total research output is the criterion, you lose people who do better work in less time (but cannot compensate for having to work part-time/being a dad/having a healthy social life/whatever). At the extreme there’s the many who never take any vacation or weekends and you compete with these people. I know it’s hard to force these people to take time out and it’s hard to judge by a metric other than total output, but it still sucks.

  7. Anonymous says:

    Epsistasis between 2. 6. and 8. rules out any ‘sum’ I can think of

  8. Anonymous says:

    Maybe this is good advice. But for me it shows a lot of what is wrong with academia.

    • Rahul says:

      Fabio Rojas & Andrew are both professors with the view from the other side. Here’s my ( facetious but not entirely) annotated view from the other side:

      1. Show up. Even if you feel horrible, show up. [ You are, essentially, indentured labor. My slave. So you better always show up. So screw those frivolous excuses & pastimes. Your child is sick? Bah! Who told you to have a child! ]

      2. Do your job. Grade the papers. Do the lab work. Unless the work is extreme, take it in stride. [ Refer to #1. You are my slave. Don’t go complaining to the Chair. So what if you graded quizzes all week for no pay? Face shields are for sissies. Shield or no shield you are going to complete that reaction in the lab. Go pay for your own face shield! ]

      3. Be completely realistic about how you will be evaluated from day #1 Time spent on other issues is “out of pocket” – do it because you care, not because it will help you. [i.e. Yes I made you spend time preparing my slides & planning that conference. Not that you had much choice about it. Too bad! Now start working on your thesis on your own time please! ]

      4. Hang out with winners. [Winner = Me. Be loyal to me. And my cronies. At Dept. social hour, don’t spend too much time with that colleague I hate. I’m paranoid. Live with it. One can never be too careful. I scoop peoples work too. ]

      5. Be constructive. It is easy to criticize people, but it really doesn’t accomplish much. [Yeah, quit reminding me that I’m dropping inconvenient data points from that regression line. It’s for your own good. If you had been constructive those outliers would never have reached my eyes! ]

      8. Submit, submit, submit. Got rejected? No problem – just resubmit tomorrow. […and this time remember to cite those Professors I told you to cite. Doesn’t matter if their work is unrelated. And if you still get rejected try this journal. The Senior Editor was my Grad School buddy. ]

      • Andrew says:


        Regarding item #2, my biggest problem with people in academia not “doing their job” is professors. There was a guy in our department who would regularly cancel his class or show up a half hour late. He’d never get a substitute, he’d just not show up. The students were happy to not have to go to class that day, of course, so it looked like a win-win situation. It really irritated me. It’s not hard to get a sub, and often that can make the class better, to mix in another perspective from time to time.

        • Rahul says:

          Substitutes might help but I remember a certain Professor who used substitutes ( one of his grad students & a postdoc) to teach ~60% of his classes in a certain semester because he was either away at conferences or delivering weekly seminars at other universities.

          I think that’s certainly annoying (unethical?) behavior!

  9. Louis says:

    Agree on that… also somewhat related to advice 3 of Rojas. I have gotten remarks and weird looks because of spending way too much time on teaching (in the eyes of some).

    After all you get paid, to a substantial extent, for your teaching. I do not buy the idea of not caring about teaching until tenured and then investing time in teaching. Maybe it holds for some, but if you get in the habit of neglecting teaching then it is hard to lose that habit.

  10. […] to activists anywhere on the political spectrum. That said, I’d like to think of this as positive-sum advice in that (a) I hope that if activists on both sides are involved in redistricting, this will help […]

Leave a Reply