I am RiotDerivative and I am a Data Scientist at Riot Games. I was hired back in October and have finally gotten around to answering some data related posts. I see some people doing some cool stuff with data on the forums. That is how I got started in my field -- pulling interesting data and doing something with it. In my case, I used to crawl Twitter and Facebook and analyze tweets and FB profiles and the connections they exposed.
A little about my background if anyone is curious. I went to UCLA and majored in Mathematics, Statistics and Computer Science. I continued onto grad school at UCLA and continued studying Statistics and Computer Science (I decided proofs were not for me) and am currently a Ph.D. candidate in Statistics at UCLA. Hoping to be Dr. Derivative this summer.
The title "Data Scientist" is relatively new in industry (about 5 years). It is a mix of a bunch of different disciplines: mostly data analysis, statistics, machine learning and computer science. Where, say, a Data Analyst focuses on analyzing data, and a Statistician focuses on predicting the future, a Data Scientist has experience in the entire "data" process. We write code to extract data in hard to reach places such as by scraping, web crawling, parsing XML/JSON/HTML and by working with "big data" packages like Hadoop. Then we analyze and visualize the data for consumption by others and build models that explain trends in play, player behavior and how to make the player's LoL experience better. Finally, we work with engineers to turn these models into scalable code that actually makes decisions and has effect on the game. In all of this, my specialty is in text mining and training computers to understand written text. The biggest thing that distinguishes a Data Scientist from a Data Analyst or Statistician is extensive programming and algorithms experience.
Before joining Riot, I worked in the web ads industry trying to use data to get people to click on ads. It wasn't for me. I had heard a lot about the work that Riot was doing in the Big Data space and how interesting the data sounded. I contemplated applying online, but was worried that I did not have enough LoL experience. Oddly enough, a day or two after I last looked at the job notice, I was contacted by a recruiter on LinkedIn and the rest is history. In my case, they looked more at my interest and passion of working with the type of data generated by games and their players than my overall gaming experience. Most of my gaming experience was in simulation games such as Sim City, Tropico, etc. I also played a lot of Angry Birds and for a couple of years was majorly addicted to Zynga games like Farmville and CityVille. While playing those games I became intrigued by the type of data produced by players and how it can be used to improve the game.
TL;DR Graduated from UCLA, do cool stuff with data and developing algorithms to make the LoL experience better. :-)
I love it! The past few years I have not been on campus much though, since I was last a TA. UCLA is great, but Riot is better!
Quote:
Originally Posted by MelonBalla
Would you say that as a Data Scientist, you are more interested in what data tells us about large trends and patterns in the behaviour of a large population, or the tiny details/unique links that can be found in data such as unexpected behaviour patterns in a small subset of a population?
We are interested in both. I find that researching small subsets of the population yields more unexpected and interesting findings though. Most of the larger trends across the entire population found in data can also be found by in-person community involvement etc. that other departments at Riot specialize in.
Quote:
Originally Posted by Kziki
Out of curiousity, have you started to play the game a bit more after being hired? I hadn't thought about it till now, but not every Rioter can be a dedicated gaming type, so how many pick up League after starting to work with Riot? I would expect at least an account and a couple games, but are there any employees who just don't play? Could it be considered a gateway game into more serious things like WoW or such to people who've never touched a game before getting acquainted with League upon being hired and finding they like it?
I find the idea of working at a games company fascinating for the sheer diversity involved with the employees. One would expect an employee to be at least somewhat involved with the end product, but in terms of games not everyone who would traditionally work in even a software company would play. As you say the things Riot was doing apart from League were what drew you to them originally, I would assume there are a number of cases like that. I'm not sure why I see it as interesting how working for Riot would affect them differently than if they worked at a non-entertainment software company, but I do.
Oh yes, I play a lot more now. At least 2 games per day. I don't know exact numbers, but in my hiring class only a couple of us did not play religiously. AFAIK, there are no Rioters that never/rarely play LoL.
Quote:
Originally Posted by Zuko Prince
Oh GOD !!!!!!
It sounds like this data scientist will be digging into what we say in text chat in game and seeing how it makes us lose/win. He will come after me for b*tching at my team for face tanking turrets.
I never said that I work on text mining... just that it is my research specialty ;-). Besides, if anything complains, it would be a computer program, not me ;-).
I am applying to Riot's Big Data department and I had a couple of questions about the department/the application process. Since you're relatively new, I figured you'd have some good insight.
I friended you in game, so if you have any time to talk whenever, maybe I could send you a message?
Also, do you have a favorite equation/expression/theorem?
If you catch me on LoL, you can send me a message sure. I am not on under that account much though.
You may want to post your questions on our LinkedIn group. That way, everyone gains from your question and responses, and if I can't answer your question, another Rioter can jump in and respond:
In terms of math, I love anything involving the Fibonacci sequence as it is found in nature so frequently and in some beautiful things. As far as theorems go, the Central Limit Theorem makes my job a lot easier, but my favorite would have to be the Fundamental Theorem of Arithmetic stating that any number is either prime, or has a unique factorization of prime numbers. It is incredibly useful in things like hashing and security. It is mindblowing enough as it is !
@RiotDerivative Introduction?
I am RiotDerivative and I am a Data Scientist at Riot Games. I was hired back in October and have finally gotten around to answering some data related posts. I see some people doing some cool stuff with data on the forums. That is how I got started in my field -- pulling interesting data and doing something with it. In my case, I used to crawl Twitter and Facebook and analyze tweets and FB profiles and the connections they exposed.
A little about my background if anyone is curious. I went to UCLA and majored in Mathematics, Statistics and Computer Science. I continued onto grad school at UCLA and continued studying Statistics and Computer Science (I decided proofs were not for me) and am currently a Ph.D. candidate in Statistics at UCLA. Hoping to be Dr. Derivative this summer.
The title "Data Scientist" is relatively new in industry (about 5 years). It is a mix of a bunch of different disciplines: mostly data analysis, statistics, machine learning and computer science. Where, say, a Data Analyst focuses on analyzing data, and a Statistician focuses on predicting the future, a Data Scientist has experience in the entire "data" process. We write code to extract data in hard to reach places such as by scraping, web crawling, parsing XML/JSON/HTML and by working with "big data" packages like Hadoop. Then we analyze and visualize the data for consumption by others and build models that explain trends in play, player behavior and how to make the player's LoL experience better. Finally, we work with engineers to turn these models into scalable code that actually makes decisions and has effect on the game. In all of this, my specialty is in text mining and training computers to understand written text. The biggest thing that distinguishes a Data Scientist from a Data Analyst or Statistician is extensive programming and algorithms experience.
Before joining Riot, I worked in the web ads industry trying to use data to get people to click on ads. It wasn't for me. I had heard a lot about the work that Riot was doing in the Big Data space and how interesting the data sounded. I contemplated applying online, but was worried that I did not have enough LoL experience. Oddly enough, a day or two after I last looked at the job notice, I was contacted by a recruiter on LinkedIn and the rest is history. In my case, they looked more at my interest and passion of working with the type of data generated by games and their players than my overall gaming experience. Most of my gaming experience was in simulation games such as Sim City, Tropico, etc. I also played a lot of Angry Birds and for a couple of years was majorly addicted to Zynga games like Farmville and CityVille. While playing those games I became intrigued by the type of data produced by players and how it can be used to improve the game.
TL;DR Graduated from UCLA, do cool stuff with data and developing algorithms to make the LoL experience better. :-)
I find the idea of working at a games company fascinating for the sheer diversity involved with the employees. One would expect an employee to be at least somewhat involved with the end product, but in terms of games not everyone who would traditionally work in even a software company would play. As you say the things Riot was doing apart from League were what drew you to them originally, I would assume there are a number of cases like that. I'm not sure why I see it as interesting how working for Riot would affect them differently than if they worked at a non-entertainment software company, but I do.
It sounds like this data scientist will be digging into what we say in text chat in game and seeing how it makes us lose/win. He will come after me for b*tching at my team for face tanking turrets.
I am applying to Riot's Big Data department and I had a couple of questions about the department/the application process. Since you're relatively new, I figured you'd have some good insight.
I friended you in game, so if you have any time to talk whenever, maybe I could send you a message?
Also, do you have a favorite equation/expression/theorem?
You may want to post your questions on our LinkedIn group. That way, everyone gains from your question and responses, and if I can't answer your question, another Rioter can jump in and respond:
http://www.linkedin.com/groups/Insid...ular=&trk=tyah
In terms of math, I love anything involving the Fibonacci sequence as it is found in nature so frequently and in some beautiful things. As far as theorems go, the Central Limit Theorem makes my job a lot easier, but my favorite would have to be the Fundamental Theorem of Arithmetic stating that any number is either prime, or has a unique factorization of prime numbers. It is incredibly useful in things like hashing and security. It is mindblowing enough as it is !