PhD

Today, my PhD graduation ceremony took place (no academic dress at NTUA, for good). I feel quite perplexed right now, in between of being relieved that this saga is finally over, humiliated for the time it took and at the same time happy for the outcome. In any case, here’s a long story short.

I graduated from NTUA’s Mechanical Engineering School right before the Athens Olympics, at the age of 22. I spent the next couple of months working for the Games (two full-time jobs in parallel, it felt great!) and right after, while tinkering between my options to set sail for a graduate school abroad, I received an offer to stay at my school, and take the skin-deep research I had started much further. It was a tough decision; I recall running a number of Monte Carlo simulations to figure out which decision would provide for a better future. It turns out though no decision is right or wrong, just your hard work or its lack thereof turns one to be such. And, did I work hard for the next years.

I was enthusiastic about the topic of forecasting, already from my diploma thesis. And I decided to focus my phd research on exactly that. I spent a full year studying forecasting in general, from time series to econometrics to machine learning. But I was also digesting a number of textbooks, not necessarily related to the topic, that for some reason I found of interest. I happily read and learned a lot during this year, and my excitement about research got even stronger.

At the same time though I realised that most fields of forecasting were quite mature for my humble research to bear any significant fruits or have any kind of impact. This understanding came along with myself discovering a pretty new research direction, which kind of brought together a number of interests of mine by that time. It did not take a lot of time to start working on it, until soon it became clear: I would like to focus my thesis on prediction markets.

I devoted the second year of my research in meticulously studying every single reference on the topic I was able to locate. It being a young field, I ended up publishing the first literature review available on the topic and getting a pretty good idea of the promising new directions a naive researcher might have an impact on.

Eventually taking some advantage of my engineering background and having dug deep into the field, I shaped the opinion that what really matters in a market are not the various case studies or descriptive theories, but rather the mechanism facilitating its operation. With this in mind, I spent half a year to set up a framework of mathematical properties for the coherent operation of such a mechanism and to finally figure out a price function that satisfies it.

These two and a half years in total left me with the belief that, if I was excited enough to truly commit myself to something, I could really become world class and get on top of it, no matter the difficulty or my previous experience or its lack there of. This is a way of thinking that wasn’t there before but has been with me since then, and I consider it to be one of a phd’s best pay-offs.

Having contributed what I considered to be the core of my thesis, I was in need of some experiments and real world implementations that would bring the theoretical findings into practice and provide some indications they can actually work. In other words, a construct was required that would enable people to participate in a marketplace, this time utilising my proposed mechanism. Correct, we are talking about a software tool here, yet the graduate courses I had taken were quite limited on that front.

I started learning programming on my own -as I did with everything else during my short career in research- but I never really managed to get on with it. At the same time however, I was not thrilled with the idea to put all this work and create something just for experimental usage in a lab. What fascinated me instead was the opportunity to put a tool I was fantasising as useful on the hands of as many people as possible.

You got that right, I was on the verge of creating a startup, and in the next couple of years my focus gradually switched fully on that. I was happy to cross roads with Thimios Mpothos, who also happened to pursuing a phd in NTUA focusing on prediction markets and brought a strong developer background on the table. We partnered to create a tool and a company.

Since 2007, the time I was devoting on research got limited; I thought I was on track to deliver my thesis on time, but my thoughts have moved elsewhere and my time followed suit. Until 2009, I was spending most of my time on the company, which managed to launch a -quite innovative yet regrettably poor in general- product, run a small number of case studies and eventually fail.

Then, Openfund came into the picture. Truth be told, I started feeling that the PhD was a very selfish way to spend my time, in between of managing other people’s money, and helping unique products, value and jobs get created. No matter how bad it felt, I was also in denial. It took me 4 years to open the latex archive again.

In July 2012, while patiently awaiting for Openfund II’s setup processes to complete, I found the courage to give it a final try. Having already performed all of the underlying research and with about 70% of the final document ready, I was expecting it could take about 4 months to wrap things up. It only took 20 tough days.

And the day I sent the draft to my advisor, I was the happiest, the most relieved man in the world. I could after all those years spend a few days on holidays, without feelings of guilt due to the unfinished business I should get back and work on. I finally defended my Thesis on June 21, 2013 and attended the graduation ceremony today, almost 10 years after I started. To put it kindly, this is clearly not a performance to celebrate. Yet, for some weird reason, it still does feel good.

Looking back, I didn’t have a clear plan or any specific aspirations when I started. I just loved the topic, and thought it would be great to be able to spend time and study it. It was and I did; when you find something that you really love and you truly commit to it, the rest will take care of itself. At least, this is how things have worked out so far for me. I wish you the same.

Tziralis PhD Defense

Tziralis PhD Thesis

Tziralis PhD Thesis Summary in Greek