PLEASURE
PLEASURE
GLOSSARY
AI (artificial intelligence) Often used interchangeably with ‘machine learning’. The process of programming a computer to find patterns or anomalies in large data sets, or to find the mathematical relationship between some input variables and an output. AI algorithms have applications in a range of fields including healthcare, self-driving cars and image recognition.
algorithm Set of instructions or calculations designed for a computer to follow. Writing algorithms is called ‘coding’ or ‘computer programming’. The result of an algorithm could be anything from the sum of two numbers to the movement of a self-driving car.
analytical engine Mechanical computer, designed by Charles Babbage in the early 1800s, intended to carry out arithmetic and logical operations, taking instructions or inputs via hole-punched cards. The machine was not constructed during Babbage’s lifetime, but a modified version was built by the London Science Museum in 1991.
cookies Pieces of information from a website, stored by a person’s web browser, which may help the website remember information specific to that person, such as items added to an online shopping cart or login details.
correlated risk Multiple negative outcomes or losses, caused by a single event. For example, many homes are likely to be damaged and people injured as the result of a single hurricane.
data analytics Obtaining, cleaning and analysing data to gain useful insights, answer research questions or inform decision making.
digital age Time period beginning in the 1970s and stretching to the present day, characterized by rapid technological advances, including the introduction of the personal computer and the rise of the internet.
digital library Large repository or archive of data, sometimes available to access or download through the internet, for commercial or research purposes. Digital libraries may include images, text or numerical data.
esports Electronic sports in which individuals or teams of players compete in international tournaments and for monetary prizes, to win video games.
geolocated franchising model Teams of competitive video-game players, based in a specific city, can form a franchise to compete in international or national esports tournaments for a particular game.
live streaming The live broadcast of video or audio content, via the internet. Esports are usually watched through live streaming.
machine learning Finding a mathematical relationship between input variables and an output. This ‘learned’ relationship can then be used to output predictions, forecasts or classifications given an input.
metrics Quantitative measure of performance. For example, it is important to assess accuracy metrics for automated decision-making algorithms. Similarly, measures such as inflation or the FTSE 100 index could be seen as a performance metrics for the economy.
model/modelling Real world processes or problems in mathematical terms; can be simple or very complex, and are often used to make predictions or forecasts.
STEM The fields of science, technology, engineering and mathematics.
swipe The act of swiping a finger across a smartphone screen, to interact with an app. Swiping is widely used in dating apps, where users often swipe right or left on a photograph of a potential romantic partner, to signal interest or disinterest.
wearable technology Electronic devices that can be worn on the body including activity monitors and smart watches.
SHOPPING
the 30-second data
With the internet giving a home to a variety of retailers, the consumer can now buy almost anything from the comfort of their own home. The consequence of this is that retailers have been able to harvest extensive and accurate data relating to customers, which means they are better able to target shoppers based on their habits. An example of this can be seen on Amazon – the biggest online retailer in the world – with its ability to recommend items based on your previous purchases, ratings and wish lists. However, the ability to perform this type of activity is not only the realm of companies the size of Amazon. Services now exist offering artificial intelligence (AI) solutions that allow retailers of many sizes to be able to harness the power of these types of algorithms to drive business, which means that the next time an online retailer suggests a T-shirt to go with your jeans, it could be via AI. Data science isn’t restricted to shopping suggestions: it also applies to how goods are purchased. Facial recognition technology combined with smart devices allows payments to be authenticated without the use of credit cards.
3-SECOND SAMPLE
Shopping online has changed shopping as we know it, but how is it possible that websites seem to know what we want before we do?
3-MINUTE ANALYSIS
Ever wondered how a website knows the shoes you were looking at the other day? Well, the answer is cookies. These are small pieces of data that come from a website and are stored in the web browser, allowing websites to remember various nuggets of information including past activity or items in a shopping cart, which explains why that pair of shoes just keeps coming back.
RELATED TOPICS
See also
3-SECOND BIOGRAPHY
JEFF BEZOS
1964–
Tech entrepreneur who is the founder, CEO and president of Amazon.
30-SECOND TEXT
Robert Mastrodomenico
Online shopping and new payment methods combine to create a high-tech consumerism that’s feeding valuable data to retailers.
DATING
the 30-second data
Sign up to a dating site and you’re presented with a number of questions that you have to complete which will define you and find you your perfect match – how is this possible? The questions are weighted based on their importance, and using these as an input to the algorithms used allows a score to be calculated that shows your satisfaction with other potential matches. It’s not all about you, though – the best match also takes into account how well your answers mesh with the potential matches. So the stats behind your match doesn’t assume love is a one-way street, which seems sensible. Online dating also includes the generation of ‘swipers’ who use dating apps. Here, you are able to see potential matches based on fixed data such as location, age preference and so on. The application then shows you individuals and you register your thoughts on the individuals by swiping left or right. Who appears on your screen is not just based on your fixed preferences; instead, complex algorithms learn from how you and others have used the app and sends you individuals who you would most likely respond to positively.
3-SECOND SAMPLE
Online dating has changed the game of finding love to the extent that finding ‘the one’ is more statistical than you may think.
3-MINUTE ANALYSIS
The ‘swipe right’ paradox: should dating app users just keep swiping to see everyone on the application? Given that the best selections will come first, every subsequent swipe should give a worse selection, and eventually you will see recycled selections as someone you said no to is better than a very bad match, at least from a mathematical point of view.
RELATED TOPICS
See also
3-SECOND BIOGRAPHIES
DAVID SPIEGELHALTER
1953–
Statistician, Professor of the Public Understanding of Risk and author of Sex by Numbers.
HANNAH FRY
1984–
Mathematician, lecturer, writer and TV presenter who studies patterns of human behaviour in relation to dating and relationships.
30-SECOND TEXT
Robert Mastrodomenico
Love at first swipe? Data scientists are working to make this more likely by matching personal data.
MUSIC
the 30-second data
The movement of music from physical libraries to digital libraries has changed the way we consume music. By having a digital music library, we have access to millions of songs by a variety of artists at the touch of a button. Given this volume of music, how are providers able to give us recommendations and custom playlists based upon our listening habits? Taking Spotify as an example, which is one of the most popular music streaming services in the world, it harnesses the power of data by adopting a three-pronged approach to determining what you might like. The first approach comes up with suggestions by comparing your listening habits to other users. The second approach applies machine-learning techniques to textual data such as news articles, blogs or even the text data stored within the digital music files themselves to find music you may like. The third approach analyses the raw audio content of the songs to classify similarity. Combining the results of these approaches allows music streaming services to come up with custom playlists for each and every user on the platform which can include a variety of genres and eras. Such streaming services are constantly evolving to harness new technologies.
3-SECOND SAMPLE
The digital age has opened up the world of music to us, but with so much choice, how can we find new music we like?
3-MINUTE ANALYSIS
If we consider two users: one may like songs a, b, c, d; the other a, b, c, e. Based on this, it could be suggested that the second user try song d and the first user song e, because they both like a, b and c. This is what is done on a much larger level for all music streaming service users.
RELATED TOPICS
See also
3-SECOND BIOGRAPHIES
MARTIN LORENTZON & DANIEL EK
1969– & 1983–
Swedish entrepreneurs who co-founded the popular music streaming service Spotify.
30-SECOND TEXT
Robert Mastrodomenico
Music streaming services can harness your listening data to introduce you to new music you might not have heard otherwise.
SPORTS
the 30-second data
Fields, courts and pitches have always welcomed professional and amateur statisticians alike to measure team and player performance. Common baseball metrics like Runs Batted In (RBI) and Earned Run Average (ERA) have been reliably recorded since the nineteenth century. Recent advancements in technology, however, have helped to launch a data science explosion felt by both participants and spectators. The invention of wearable technology has allowed data scientists to track athletes and activities. In tennis, for example, many professionals have turned to using racquets with embedded sensors, which allow them to track speed, rotation and ball hit location in real time. Other advancements include the expanded use of cameras and radar devices. In most major sports, the universal use of sophisticated cameras has provided participants and fans access to insights that were previously unimaginable. Using the new metrics provided by high-accuracy camera technology in Major League Baseball, for instance, team quantitative analysts have demonstrated that performance improves when batters correct their ‘launch angle’. As athletes continue to use data science to improve their games, there is no sign that the ongoing arms race in sports metrics will get thrown out any time soon.
3-SECOND SAMPLE
While Moneyball is the most famous example, it pales in comparison to the advancements facilitated by the data revolution in sports, affecting the experiences of players, managers, referees and fans.
3-MINUTE ANALYSIS
Commentary surrounding the increase in data science’s influence in sports often presents two competing factions: the purists and the nerds. The film Moneyball presented a heterodox, quantitatively-inclined baseball manager upending a system designed by experienced scouts. Recently, some athletes have criticized statisticians for their lack of experience. However, the most successful data teams leverage expert insights to complement their analytics.
RELATED TOPICS
See also
3-SECOND BIOGRAPHIES
BILL JAMES
1949–
Baseball writer and statistician who created a wide array of player evaluation formulas.
BILLY BEANE
1962–
Pioneered the use of non-traditional metrics for scouting undervalued baseball players.
30-SECOND TEXT
Scott Tranter
Sports data science works best when integrating athlete experience and scientists’ numbers.
SOCIAL MEDIA
the 30-second data
In just a few years, companies like Facebook, Snapchat and Twitter went from small internet start-ups to multibillion-pound tech giants with priceless quantities of influence. Facebook is now reaching nearly 90 per cent of all mobile users in the US, while Twitter boasts 100 million daily active users, enough to become the fifteenth most populous country in the world. With such a massive user outreach, companies can leverage the enormous amounts of data generated on these platforms to discover new trends and insights about their audience, then apply this knowledge to make smarter, more reliable business decisions. By tracking users and implementing algorithms to learn their interests, social media companies have been able to deliver highly targeted adverts and generate billions of pounds in ad revenue every year. These same machine-learning algorithms can be used to tailor the content each user sees on their screen. From a timeline to suggested friends, social media companies play a prominent role in how users interact with their apps and, subsequently, the world around them. What once started as a way to update friends on one’s status has evolved into a public forum, marketplace and news outlet all rolled into one.
3-SECOND SAMPLE
Since the beginning of the twenty-first century, social media has taken over how humans interact, access news and discover new trends, driven, in large part, by data science.
3-MINUTE ANALYSIS
TV shows such as Black Mirror provide us with a different view towards continuing advancements in social media. The episode ‘Nosedive’ depicts a world where ‘social credit’, deriving from a mixture of in-person and online interactions, dictates where a person can live, what they can buy, who they can talk to and more. China has begun implementing a Social Credit System to determine the trustworthiness of individuals and accept or deny individuals for functions such as receiving loans and travelling.
RELATED TOPICS
See also
3-SECOND BIOGRAPHIES
JACK DORSEY
1976–
Co-founder and CEO of Twitter.
MARK ZUCKERBERG
1984–
Co-founder and CEO of Facebook, and the youngest self-made billionaire, at 23.
30-SECOND TEXT
Scott Tranter
The rapid growth of social media has seen it infiltrate everyday life, with data capture capabilities on an unprecedented scale.
GAMING
the 30-second data
Competitive video gaming, known as esports, is an emerging global phenomenon in which professional players compete in packed stadiums for prize pools reaching millions of pounds. Unlike with traditional sporting events, esports fans engage more directly with content via online live streaming technology on platforms such as Twitch. Esports consumers largely consist of males in the 20 to 30 age range, a prime demographic that companies wish to target. By tracking the fan base’s habits and interests using analytical tools and survey methods, companies have been able to tailor content based on the audience they wish to target. However, because of the esports audience’s reduced television consumption and tendency to block internet adverts using browser-based ad-blocking technology, companies are looking into non-traditional methods to reach this demographic. For example, due to the digital nature of esports, brands have the ability to display their products directly in the video games, avoiding ad-blockers altogether. Additionally, professional esports players have a large influence on how their fans may view certain products. To take advantage of this, companies often partner with these influencers and utilize their popularity in order to reach target audiences for their products.
3-SECOND SAMPLE
Esports is engaging its young, digital-savvy fans through non-traditional online media, paving the way for data science into the recreational trends of young generations.
3-MINUTE ANALYSIS
Although esports thrived off the back of online live streaming technology, it has also begun broadcasting esports on television, with ads displaying during commercial breaks, akin to traditional sports. Esports companies are adopting the geolocated franchising model, which looks to take advantage of television advertising and sponsorship deals for its revenue. With this move, esports has an opportunity to expand its reach, opening up the door for mainstream popularity.
RELATED TOPICS
See also
3-SECOND BIOGRAPHIES
JUSTIN KAN
1983–
American internet entrepreneur who co-founded Twitch, formerly Justin.tv, the most popular streaming platform for esports content.
TYLER ‘NINJA’ BLEVINS
1991–
American Twitch streamer, internet personality and former professional gamer who helped bring mainstream attention to the world of esports.
30-SECOND TEXT
Scott Tranter
As the esports industry grows, top players may soon be able to sign endorsement deals in the ballpark of professional athletes.
GAMBLING
the 30-second data
In gambling, everything from the likelihood that the dealer will bust in blackjack to the placement of specific slot machines at key locations are driven by statistics. And, in the evolving world of data science, those with greater access to it can find themselves at a huge advantage over others. This ranges from the simple approach of an experienced poker player understanding the odds of turning his straight-draw into a winning hand – and the correlated risk of pursuing that potential hand – to the more advanced techniques casinos use to turn vast amounts of unstructured data into predictions on the best way to entice players to bet, and to bet more, on lower-odd payouts. Resources exist for both the house and for the player, and they extend well beyond card games and slot machines. Statistical models can impact the payout of sports events – oftentimes adjusting odds in real time and based on the direction that money is moving – in a way that can minimize the risk of the sportsbook (the part of casinos that manages sports betting). By the same token, some gamblers use or create statistical models to make educated decisions on outcomes that are data-driven rather than narrative-driven, giving them an edge on those following their instinct.
3-SECOND SAMPLE
Data science and gambling can blend together with devastating effect – and has made the adage ‘the house always wins’ even more true.
3-MINUTE ANALYSIS
There have been reports on the ways in which casinos are utilizing decades’ worth of player data (tied back to individual players through their rewards cards), while plenty of ‘expert’ gamblers have written books designed to ‘beat the house’. Those with designs on gambling based on luck are simply playing the wrong game – they should be playing the stats – while hoping that Lady Luck still shines on them.
RELATED TOPICS
See also
3-SECOND BIOGRAPHIES
RICHARD EPSTEIN
1927–
Game theorist who has served as an influential statistical consultant for casinos.
EDWARD O. THORP
1932–
Mathematician who pioneered successful models used on Wall Street and in casinos.
30-SECOND TEXT
Scott Tranter
Move over Lady Luck: professional gamblers now pit their data skills against those of the house.
