So you want to get a PhD?

I am Batu, a 2017 NYU Abu Dhabi graduate. I am in the second year of my Ph.D. program in Computational Media at the University of California, Santa Cruz. There are several things I wish I had known before I started my Ph.D. This piece of writing hopes to give my perspective and also serve as a compilation of other useful resources.

First off, stop reading this and go absorb the following two documents. I have scoured the internet for advice over the past two years. The ones linked here are the ones that I personally find the most useful and unlike me these people actually finished their PhDs. (Unfortunately most of the links I give and my own experience present a very narrow idea of what a Ph.D. is, mostly from a technical Computer Sciency way. So take it with a grain of salt.)

1. Andrew Karpathy, A Survival Guide to a Phd
2. Philip Guo, PhD Grind

Bonus: Phd Comics

The Decision

Now that you are back let’s get to things that I add to the conversation. A bit about myself. I decided to get a Ph.D. after grabbing lunch with my introduction to Computer Science professor, Michael Paik. He basically said: “If you want to contribute to the overall knowledge of humankind, Ph.D. is the way to go.” That idea resonated with me. I really wanted to help humans make “better” games! Matt Might has a beautiful illustration that gets this idea across succinctly (and a similar version for art, thanks Max). Orthogonal to these two, in the past two years I reached my own formulation of the Ph.D. bug, which is a bit more visceral:

“One day you ask a question to the all-knowing Google.

Your search query results in zero answers. You decide to spend 6 years answering that question. Figure out there are much more interesting questions to answer. End up with a different answer to a different question but now humanity as a whole knows a little bit more, or at least Google does. “

If either my or Matt Mights descriptions resonate with you, you might be in the right track. (Or you might want to become a professor and then this is your only way there. If this is the case please be aware of your future prospects, the academic job market is tough.) That being said there are several reasons not to get a Ph.D., and you can just find the listing online, I do not want to preach to the choir. However, a more important question that I invite you to ask is “Do I need to get the Ph.D. NOW?”

I went into my Ph.D. program straight after getting my undergraduate degree, without a masters degree. Whether that was smart or not is debatable. Regardless, there are a few things I observed related to this fact:

The cohort is much more diverse than any of my earlier experiences. No, I am not talking about cultural diversity, but rather, talking about diversity of experience. There are a few students who are similar to me, fresh college graduates. There are students who finished their masters and decided to get a Ph.D. There are people who worked in the industry for several years and came back to school. There are people with families, some with kids etc.

This is very different than most people’s college experience. Most freshmen have similar backgrounds, of having finished high school. However, this experience diversity in Ph.D. level means that some people will just know much more than you do. It is really important not to use others as a measure of your own value or capacity. I struggled with this in my first year, all around me I saw experts, people who had published research, people who had worked jobs, and knew… like stuff and things… I had a similar feeling when I started college as I didn’t know how to code before and other people had been coding for several years! In both cases, it balances out. If you know less than others, you also learn faster thanks to all of the smart people around you that can answer your questions.

This diversity brought in a lot of specialized knowledge. I got into grad school thinking that I knew what I wanted to specialize in: “Games.” Eaaasy! No, it was not. There were several different labs, some working in AI, some working in narrative, some working in the humanities aspect of it. I was really stressing about finding what I really wanted to do, and comparing myself to people who had a laser focus, honed with time in their masters or their industry work, and some just knew what they wanted right off the bat. Once again, it became important for me to not compare myself to others, as everyone was at a different point in their unique journey.

These observations pushed me to ask not whether or not I wanted to get a PhD, but rather whether if I wanted to get a Ph.D. NOW. After middle school obviously (obvious for those who were lucky), we went into high school. After high school most of us went directly into college. From then onwards, it is easy to think that the easy next step is just getting a Ph.D. It does not have to be. If you do something else for a few years you will lose nothing and might gain a lot: You might dive deeper into one area of study, you might face real-life problems that you get excited to solve, you might get industry experience which you can use to compare your Ph.D. to, you might save some money which will come in handy.  Or maybe get a masters if you can fund it easily which might teach you if you like the field and give you more of a perspective before jumping in the deep side.

So you decided you want to get a Ph.D., for whatever reason. And then you decided to get a Ph.D. NOW, for another reason. It is time to pick a place and then apply.

The Search

The first step is to figure out where to apply. While it is not feasible to totally disregard the overall ranking of the university (because usually higher ranking universities have an easier time getting funding) it is definitely not the main thing that you should be considering.

What matters the most is who you will be working with, your adviser. If you don’t believe me, let the rabbits tell you. You will be working with them for several years and it is very important to have a good match. A questionable report showed that the average marriage length to be 8.2 years long in the U.S.. And a decade ago an NY times article claimed that the average Ph.D. is also 8.2 years long. Are these number debatable? Yes. But is the average marriage and the average Ph.D. on the same order of magnitude when it comes to length? Also, probably yes.

So how do you go about finding the magical fit? I don’t know. But here are two extra tips that seem to work:

First, start by gathering a list of topic that you like. For me it was Game AI, for example. If you go to google scholar (Google Scholar is a service that helps you find papers.  You will get very familiar with it soon.) and type label:<Your Label> into the search bar, you will see the researchers who publish academic papers under the given label, ranked by their citation counts. This technique is definitely flawed and shouldn’t be your only source when it comes to finding researchers but it is a quick and dirty way to know who is publishing in the field. More importantly, this gives you an idea of where they are right now, at what institutions. Take a look at their papers and see if you resonate with any of them.

Second, figure out a conference you like. If you do not know how to find them check out this source, but probably just ask the faculty that you are already working with. When you find those inspiring conferences, look at who is publishing in them. What institutions? What labs? You don’t even have to fully understand the papers but the title and introduction should give you a good idea.

With these two additional approaches, you can start visualizing a network of labs and people who tackle questions that you are interested in. If you are lucky some has already compiled a research ranking for you but for most fields this is not that common. When you have a list of places picked, the time comes for the application.

The Application

There are whole books written on this topic and career centers focusing exclusively on this. Instead of treading the same ground I will quickly mention what I believe is very important and can get you ahead of others.

First of references are incredibly important when it comes to applications. I believe I lucked out and some of my referenced actually carried me to wherever I am today. Try to treat the faculty you are interacting with like they are actual human beings and hopefully the rest comes naturally. Give them enough time to write the reference letter (at least a few weeks.) and giving them a bullet point list of your interactions is usually appreciated.

If you can DO reach out to the faculty that you plan on working with. Hopefully, by now you found one of their papers that you liked and actually have some questions about it. Most faculty will be glad to answer them for you, unless they are overworked and straight up ignore you. That is also very possible, don’t get discouraged! But regardless of the response, reaching out to them shows that you really care about the field enough to seek answers to your questions. And of course they will hopefully recognize your name when the actual application opens. Finally, if you show that you are familiar with their work in the application that will definitely give you another leg up over other candidates. (NOTE: I spoke to some faculty about this. BE CAREFUL NOT TO SEND BASELESS EMAILS. They are busy and only reach out if you have something of substance to say.)

Go forth and Ph.D., Maybe

Remember, the Ph.D. is a long marathon and it is essential to have a solid reason to dive into the Ph.D. at this moment in time. But it seems not horrible, at least in my humble experience so far. There is free pizza sometimes and good conversations. Hopefully, this write up has been useful in contextualizing and helping the pre-application phase of the Ph.D. journey.

Please reach out to me with your comments, questions, and feedback.

Thank you to my peers at the Computational Media Grad Salon for their valuable feedback and tips.

Best of luck and enjoy life,