Skip to main content

Hackathons as a market

A hackathon is an event where people gather to share ideas, collaborate and solve problems, usually within a 2 or 3 day span of time. Participation is generally limited and often application based, making these events more and more competitive between applicants. To run a hackathon, organizers often ask companies to sponsor them, and in return the company gets access to an enthusiastic group of people with particular skill sets. We can consider this event to be several markets.

In this scenario, we can consider the hackathon to be a market where organizers petition companies to sponsor the hackathon and pay the costs in return for exclusive access to the talent it attracts. Organizers get the money to run a successful hackathon and gain prestige while companies get positive publicity and the opportunity to pick out potential new hires from a wide pool of enthusiastic people. Therefore this trade can only be executed between companies and organizers, effectively forming a bipartite set of organizers and companies. Organizers of hackathons will not have money to fund others and will find little meaning in doing so. Often, organizers can run themed hackathons, ones that are focused on a particular subject or problem. This has the upside of attracting specialized and niche talent, therefore becoming an opportunity of great interest to the companies in the field.

We can also consider the relationship between applicants and organizers to be a market. Applicants trade their time and resumes for an opportunity to have fun and meet with companies. In addition due to its growing competitive nature, applicants can gain prestige and prizes for completing a successful project. Since entry is application based, organizers can pick out people to maintain a wide variety of skill levels, which will make their event even more attractive to companies and get more sponsor money, creating a complicated situation involving a tripartite set of agents.

Finally, we can consider the exchange of ideas to be another market. This currency can be shared between anyone, companies, organizers, or participants. A well executed idea will be attractive to many parties, prompting more investment and prestige to the originators. In this example, everyone is nearly an equal and this ‘currency’ can be freely exchanged among any agents.

The Bachelor as a Market

Last Monday, a significant amount of the population tuned in to ABC to watch the season finale of the Bachelor, and determine who this season’s Bachelor would propose to.  While this show provides much entertainment value, it can be looked at through the lens of a market and allocation.  In this case, a proposal of marriage is what is being exchanged and what the buyer is hoping for.  The end goal of the show is for the bachelor to propose to one of the girls that he is simultaneously dating, so that is the allocation, and the girls are hoping to get that proposal.  In this situation, the currency in this market is romantic feelings, and those are exchanged by both the buyer (the girls hoping for the proposal), and the seller (the Bachelor).

One feature of this market that makes it so dramatic and fun to watch is that there is only one seller and multiple buyers, which means that all the buyers are fighting to be the one who receives the allocation of the marriage proposal.  This setup makes the show more interesting because it allows for people to get sent home every week for dramatization.  ABC, however, could have set up the show in a way where there was a market with the same number of Bachelors as Bachelorettes, and then we could see a matching where everyone has the potential to get matched depending on their preferences.  It would still be an interesting show to watch, as I am sure there would still be lots of drama as multiple people fell in love with the same person, or someone was left out of the fray altogether (pulling at the viewers’ heartstrings), but unlike the current set up of the bachelor where there is only one exchange of a proposal, this setup would have the potential for a perfect matching.

Music Taste as a Market

Now we all know the already established markets that revolve around music such as record labels, music streaming devices, music award shows, etc. However, here we will examine people’s music taste on how that is in of itself its own market. First, we must define what we mean by “Music Taste”. It is perhaps more appropriate to label it one’s music choice that is: the songs a person decides to listen to more than once and by their own accord. One way to think about it is the music one might add to their collection whether that be on Spotify, Youtube, or any type of physical or virtual collection.

Now we will mainly try to focus on the music listener and thus this will closely resemble a one sided market where the music listener has certain preferences over what they desire to listen to. The preference list a person will have will depend on how much they like the song given their particular taste. Here it can be very interesting to think about what an agent’s preference might be dependent of since usually the music we like is strongly correlated to what we listened to growing up or perhaps what people we care about listen to. We can also think of matching a song to an agent as bringing the agent some sort of positive utility. Whereas, it will follow that an agent will prefer a song to another if it will bring them more utility than another song. One interesting point to examine is that musical taste is often dynamic and many times the utility of a song can change given if that person has been “matched” (has listened) to other songs are similar. For example, a person who doesn’t really like classical music will have low utilities on classical songs. However, say they take a class that studies classical music and through listening to different pieces and learning about it they find a newfound liking to the genre thus, increasing the utility of classical songs.

Many music streaming services are actually very dependent on how well they navigate this market. Pandora, Spotify, and Soundcloud all have features such as “suggested songs” or “song radios” that try to match agents to songs they might like. Here these services take a different approach on how they handle this dilemma. The approach in many ways can be compared to the model that the service decided to implement. Spotify’s Discover Weekly approach to this market is agent driven. It utilizes information of other agents (other people’s playlists) and information about your taste as well as machine learning to try and recommended new songs for you that you will like. On the other hand, Pandora took a more musically driven approach. Pandora’s approach relies on something called a Music Genome. “Pandora relies on a Music Genome that consists of 400 musical attributes covering the qualities of melody, harmony, rhythm, form, composition and lyrics.” It is important to notice that Pandora has limited Music Genome to contemporary artists and has still yet to approach a model for classical composition. That is they are trying to model musical composition versus people’s musical taste composition.

Which model is more effective? Try both and see which you prefer.


Tinder as a Market

While sitting at the dinner table and wondering what topic to write about, I looked over to my friend next to me and saw him swiping through Tinder. That’s when it hit me that Tinder itself is a type of market. Specifically, it is a type of matching market in which neither side of a potential matching knows the other’s preferences (whether or not the other person has/will swipe left or right). In most markets, there are people who create and sell products to others. In the case of Tinder, people’s profiles are the products that are being “sold” to others. The app serves as a matching platform in which you can display a profile to “advertise” yourself. However, you do not determine an intrinsic value for yourself, rather your value is subjectively determined by other users who are debating whether or not to swipe left or right on you.

There’s no physical currency in this market. Rather the purpose of this market is to facilitate a “contract” between two parties. Swiping left is considered to be declining to a contract, while swiping right is considered to be agreeing to a contract. When both parties agree to a contract with each other, they are given the opportunity to chat with each other (swiping could be considered to be “window-shopping” a product, but chatting could be considered to be quality testing). The ultimate sign of a successful deal in this market is an exchange of contact info between two matched users. Additionally, unlike many typical matching markets in which all items involved are organized into unique pairs, Tinder allows you to be a part of multiple matchings with other users. For some, this encourages the practice of swiping right on as many profiles as possible in order to maximize their success rate.

One of the design choices of Tinder is the ability to “super-like” profiles. Once a day, you are allowed to super-like someone else’s profile. The user will then see that you super-liked them when deciding on whether or not to swipe left or right on you. This essentially gives you a “competitive advantage” with this person because they know that you used up your super-like on them, which makes them more likely to agree to a deal (swipe right) with you because their valuation of your profile has increased. A possible alternative to this would be to allow you to send one “pre-emptive” message to someone you’ve swiped right on if they haven’t swiped on your profile yet (like super-likes, you’d only be able to do this once a day). This may be even more successful in increasing your subjective value to another user because not only will they know that you used your daily pre-emptive message on them, you will be able to create a tailored message for the specific purpose of attracting this user.



BorrowDirect – Interlibrary loaning as a market

On any given day, a Cornell student working on a paper, a research project, or any other assignment may make their way to the library in search of information. At one of Cornell’s libraries, they have access to over eight million volumes; books, manuscripts, old newspapers, musical scores, etc. But even with so many resources, it sometimes happens that students cannot find what they are looking for. To solve this problem, BorrowDirect was created. BorrowDirect is an interlibrary loan system connecting the libraries of al Ivy Leagues schools, MIT, Duke, John Hopkins, Stanford and the University of Chicago. This interlibrary loan system can be viewed as a market.

While this is different from the one sided matching markets we have discussed in class it is very much a market. Here the agents are students of the 13 member universities who need to be allocated books (or any type of document) for their school work. In this market there is not a currency. One of the ways that this market is different is that students will first go to their own library to find information, they will only go to BorrowDirect if they cannot find what they are looking for. It is still a market because it is a place where agents go to be allocated books (although they do have to return them). There is no currency (students do not pay in any way to loan out a book from BorrowDirect), the documents being loaned out really hold information, which is what agents enter the market in search of.

To use BorrowDirect, students request a document, which they then receive within 3-5 business days. The value in this market is that it increases a Cornellians access to information from roughly eight million volumes to about 70 million volumes, a significant increase. This market allows for all agents to have an increased access to information. The agents are allocated “information” when they request it. If there is a document with only one copy, if one agent has loaned it out, another agent must wait if they wish to access the same document. In some sense there is exchange in this market, this comes in the form of the universities who participate. In the market, they are required to loan their documents to (students of) the other schools which participate in the market. However, participating universities never relinquish ownership of anything, they simply lend it out.

One design choice was to only include certain elite universities in this interlibrary loaning system. This design choice seems reasonable as the schools that are included have some of the oldest, largest and most extensive university libraries in the country. With just 13 schools each member has access to over 70 million volumes. The advantage to not adding more schools is that the mechanism for distributing more books can become increasingly complicated, with factors such as shipping and storing and accessing records. Additionally, if schools with less comprehensive libraries joined, they may have less to offer to the market than they have to gain. One way the market may change if more schools are added would be that the size of the market simply increases, with more agents and more information to be distributed all around, although it would increase the logistical complication of the underlying mechanism.




















Searching for a mate

A very well known real-world situation that can be abstractly viewed as a market would be searching for a mate. In a sense, it is not so different from many other things that we do in our life, e.g. looking for a job, applying for school, where to eat for lunch, etc. Prospective partners, like jobs, each have their strengths and weaknesses. This is a two-sided market and the agents are the single people that come into the market and the things on the other side of the market are other single people. These agents don’t really know the other single people that are available to them, but they get to interact with each other and get to know more about each other before deciding anything in the market. The ‘things’ being allocated in this market is other people, more specifically prospective partners of an agent who are also looking for prospective partners. However, these allocations are different from other allocations, in that both parties must be enthusiastic about the match for it to happen.


It is hard to describe the currency in this market completely because it consists of many things and none of these things are materialistic like money. One of the thing the ‘currency’ consists of is definitely “happiness”. People who are in a relationship are usually much more happier with the person they are in the relationship with. So being happy is like receiving currency from the prospective partner. In addition, an “experience” with someone is unique and can not be replaced. This is something a person can get from being in a specific relationship which are the matchings in this market. And this experience can be traded in this market. And also “attention” that an agent has to offer and give is another factor to the currency because you can give and get. In addition, social-psychological rewards can also make up part of the ‘currency’. For example, being with someone can definitely improve a person’s reputation and this is desirable. The currency in this market would be meaningfully exchanged between the pair of agents because both agents in a relationship are in the same position as the other, they exchange all the things that make up the “currency”.


One of the design feature in this market would be new technological advances such as Tinder. These new dating apps has greatly changed the dating market. Since now everyone has a phone and many people has access to the internet, it is so much easier for people to “meet” over the internet since all people have to do is to make a profile and upload some pictures. The chances of you meeting another person has increased greatly. In return, this would add more edges between the nodes on the two sides in our market giving us more choices when making matchings. Another alternative to dating apps is social gatherings to meet other people. This will greatly change the market because of many factors. First, because in order to be invited to a social gathering, an agent  must be socially active and so this requires time and effort. And those who are on top of the social game would be doing very well in this market and there will be many edges coming out of these agents to other people. In addition, doing well in this market would be very hard for people who are shy and uncomfortable in social gatherings. These agents would be left with very little choices, and often the choices they are left with are other people that they are very close with and part of their “friend circle”. Therefore, a market with this design will contain many constricted sets and making a perfect matching very hard. In all, it is very interesting to apply networks principles to our everyday lives.


Negotiating a better job offer through Networks II

One of my most stressful moment at Cornell was full-time recruitment. As if the course load wasn’t heavy already, add filling out and sending out application, interviewing, and traveling only to be brushed aside with an impersonal email that starts with “Thank you so much for your application to _____.” However, hard work pays off and often times you are left with job offers and for the lucky and most successful few, some may have few offers to choose from. This situation creates a matching market where companies and students act as an economic agent who makes self-interested choices and decision based on his preference. The goal of the student is to accept the best job based on his preference. Typically, some factors students might consider are compensation, culture of the firm, location, exit opportunities, job growth, and industry. The goal of the firm is to fill their entry-level class with as many talented, skilled, motivated as they can afford and fill their labor need. Interesting thing about this market is that students can accept a maximum of one job but firms can accept as many as they can afford.

There exists an exchange between the firm and student. When the student and firm is matched, they exchange a contract in which the student promises to exclusively work for the firm while the firm promises to give employment to the student. When the student recruits, he or she applies only to acceptable positions and within acceptable positions may have a preference. For example, a person might apply to both Google and Facebook because they are acceptable companies for him to work at but might prefer Google to Facebook. In this exchange, however, the company has the upper hand in that they are the ones making the offer.

An interesting thing about this market, unlike the ones we studied in class, is that the companies have the opportunity to change their compensation and reveal more information about their company that could change the student’s preferences. For example, say student A is first priority for all companies that exist in the job market (company A, company B, and company C). Student A finds all companies acceptable and receives offers from all companies (say they want to higher at minimum one student). Student A prefers Company A over Company B and company B over company C at the current offer. However, company C can increase their compensation, offer him better location, and reveal more information about future prospects to lure student A to taking the job offer. While at initial offers, student A preferred company C the least, after C’s renegotiation, student A could end up accepting company C’s offer.

We find that in the loosest definition this job market is not strategyproof. Because company’s can change their offers and use other tactics to luring their desired students to accepting the job offer, it is in the best interest for students to lie about their preference. Say student A prefers company A>B>C. He could lie and say to each company that he prefers them the most in order to entice a better offer that could be strictly better than all the offers he has. The student would be incentivized to lie so that he or she could encourage a better offer that could Pareto dominate his current offers.

Job search is always stressful but with the knowledge of networked markets, we could all learn to negotiate a better offer.

The NBA Draft

As college and professional basketball head towards their respective postseasons, the NBA Draft looms large as teams evaluate their strengths and weaknesses and decide which players are best suited to improve their on-court product. The Draft can be viewed as a market where players are allocated to teams based on the teams’ preferences.

In its most simple form, a sports draft can be represented as a one-sided market in which the agents (the teams) have strict and complete preferences. A serial dictatorship mechanism is used to allocate players to teams, where the team with the worst record from the previous season is given first choice, the team with the second worst record is given second choice, and so on.

However, the NBA draft is more complicated than this. For starters, this is not exactly how the serial dictatorship is determined. To discourage teams from tanking (in other words, attempting to “win” the worst record in the league and thus be entitled to the first pick), the NBA uses a lottery system where the first three picks of the draft are awarded at random to one of the fourteen teams that did not make the playoffs the previous season. Teams with worse records have a higher chance of receiving one of these top three picks, but the worst team does not necessarily receive the first pick. After the first three picks are awarded, the remaining teams are allocated picks in reverse order of wins.

Teams may also choose to trade picks (either before or after players are chosen) for cash, other players, or other draft picks. Hence, teams can alter the matching mechanism if they deem it useful. Because of this market feature, modeling teams’ preferences on rank order alone is futile; these types of preferences provide no indication of a team’s willingness to trade a pick that the serial dictatorship would have given them. Rather, preferences may be modeled using the utility teams assign to players, thereby allowing players, draft picks (which are used to acquire players), and cash to be compared directly against one another. For example, suppose Team A has pick 6 and plans on using its pick on Player X while Team B has pick 10 and would like to draft Player Y. Perhaps Team B suspects Player Y may be drafted before its turn and values the player highly enough, and Team A is confident Player X will still be available by pick 10. Teams A and B may choose to trade picks, with Team B also giving Team A something of value to make up for the worse draft slot.

The tradability of draft picks alleviates potential frictions in the market. Namely, although the draft itself is not competitive, NBA teams compete against one another in zero-sum games using the players who are drafted. Without the ability to maximize utility by trading for better or worse picks, a team might find it beneficial to draft a player not because he maximizes the team’s utility given the remaining options, but because drafting that player hurts a competitor. While this type of sabotage might not be inefficient per se, it would be contrary to the good-natured spirit of the draft, which exists to improve one’s own team rather than hurt other teams.



The Market for Parking Spaces

In this post I will be exploring the possibility of viewing parking spots as a market. Anyone who has tried to park on the street in a city will recognize that by leaving a spot to use their vehicle, they are forfeiting their spot, and will likely have to search for a while to find another. In essence, they are exchanging their good (the spot) for the utility of the use of their vehicle. In most cases, they are not exchanging that space with a particular individual, but rather the pool of people without a space. I would argue that the two pools of people, (those enjoying the utility of their vehicle, and those currently occupying a space) can be viewed as two “agents” in a parking space trading market. Of course, not everyone driving is looking for a space, or already has a designated space for them, and thus these people are not participating in the market. People could also give their spot to a friend, but we will ignore that possibility for the sake of this market. A monetary value could be placed on parking spaces, in that once someone forfeits their spot, they run the risk of not finding another one later and being forced to pay for a spot in a garage or paid lot. (The value would be whatever the local rates were). Some cities do employ meters (or more recently kiosks), where people do pay for street parking, but it is significantly cheaper than using a garage.

There are a few tradeoffs to be considered in this market. Street parking can be more convenient because of the parking space location, and the instant access to the road, (as opposed to driving up and down six levels of a parking garage looking for a spot). Those parking on the street have a higher risk of damage to their vehicle, and being ticketed for not paying the meter in time. There are several other factors that may sway people in the looking-for-a-spot pool to choose one over the other, and it make it an interesting topic to study.

A design choice in this market is the free-for-all nature of street parking. In most cases, people do not have reserved spots, and in a city full of people, there is almost always someone looking for a spot somewhere nearby, which makes the time that a spot is vacant very short. It makes for effective allocation of goods, as the utility of having a spot is not being wasted by leaving a spot vacant for an extended period of time. The change from free to paid parking on the street does not really affect the rate at which spots are filled. People in the city are desperate to park, and paying a small fee is usually not a problem. A change that would drastically affect the market would be removing street parking altogether. This would drive the value of the paid spaces in lots and garages way up. (So if you are on a city board, and own a garage, you might have some thinking to do…). This, however, would not benefit either pool. Those in a spot would not want to use their car for fear of not finding a spot in the now crowded garages. Those using their car would have to search for longer and drive farther to find an empty spot. Thus, the market is operating rather efficiently, despite the fact that it still may be a pain to find a spot sometimes.


Algorithms at 35,000. Matching Pilots with Routes

Everyday thousands of flights come and go throughout the United States. Tens of thousands of men and women work these flights, connecting people with what matters most in their lives.  How it is determined what pilot get assigned to particular flight lies like many things in the modern world, with a algorithm and a market. We will now explore the market of “bidding” a pilot’s routes. Unlike many jobs, the job of a pilot is somewhat unpredictable. Often pilots do not know if they will be jetting off to New York, or Tokyo or Milain until a week before the trip or sometimes with new pilots a few hours before the flight. Pilots for example do not typically work the same flight to the same city everyday. We will design our market as follows. Our agents will be pilots who will submit preferences for the set of legs that they want for a time period, typically 2 weeks. For example a Pilot A could have a preferences {(PHL->ITH->PHL),(PHL->SFO->PHL)} where in this small airline on a particular day they wish to work a flight from Philadelphia to Ithaca and a leg back rather than a leg to San Francisco and a flight back.  These routes or work duties are what is being exchanged. The order that we go through the pilot’s is based solely on seniority. Thus a older pilot can choose to go to Paris when they choose to, but a younger pilot may be making a lot of trips to Alaska in the winter for example.

The airline market is also (thankfully) a heavily regulated market so we will now define some constraints for our bidding economy. First, pilots can only work fly 100 hours in a 30 day period, to reduce fatigue. Thus while a pilot can get matched, or pick up, multiple routings he or she can only work a maximum hours of 100. This also applies per day, but than this would not be a valid market. For pilot’s only one person can get matched with a particular route. ie You can’t have 4 pilots work a flight you must have one first officer and one captain. For flight attendants this is different and you typically can have four or more flight attendants on the same routing. It is also interesting to look at references. Pilots with families usually value time at home, and routes that mirror a typical 9-5 job. Younger pilots typically value more interesting destinations  and ones that maximize their income. Whatever their choice is it is definitely not a typical 9-5 job.


The airline also has some vesting interest in the market. Their objective is to run a safe, efficient, and profitable airline.  Thus they have the following wants when they choose an assignment. They want to reduce turnover of new pilots as training costs are high. They also want to make sure all flights are matched. Thus a maximum matching must occur. ie every flight must have a crew. Additionally an airline would want to fill a schedule with the fewest amount of crew as possible to reduce costs. They would also like to reduce the amount of layovers in cities as this adds to hotel prices. Airlines would also like to reduce the amount of deadhead flights. Routes where the routing does not end in the origin yet crews are still paid to get back to the start. This is wasted money. Balancing this and other concerns determines how an airline will design its market.


Finally, it would be interesting to see how we could change this market slightly. For example, changing the market so new pilots get some input into what routes they would like to fly. This could potentially reduce turnover. Whatever the algorithm who flies your plane on your next route was simply a market matching.



Bidding Schedule Lines

« go backkeep looking »