"An organization of impersonal relations": The Internet and networked markets
First Monday

An organization of impersonal relations: The Internet and networked markets by Holly Kruse

As the Internet has enabled near–instantaneous transmission of buying and selling information by brokers through servers in financial markets, and from bettors to servers in parimutuel horse racing markets, transactions that once took place in face–to–face settings are now dependent on far–reaching global communication flows. This article explores the ways in which the Internet has influenced the dissemination of information in public, temporally driven, information–intensive markets. In looking specifically at the case of parimutuel horse racing, the article examines what implications the Internet has for networked markets and what it might indicate about mediated forms of presence.


Networks and markets
The Internet and parimutuel markets
Networked markets
Implications of network technologies: The case of parimutuel markets




In the first decades of the twentieth century, Australian George Julius perfected a kind of intranet. His totalizator was initially created as a voting machine, but when it didn’t well serve that purpose, Julius adapted it to total wagers and calculate odds at racetracks. The first crude totalizator was installed at a racetrack in Auckland, New Zealand in 1913, and by 1930 the machine had been improved and was widely adopted in Australia, the United States, and Europe. Prior to the totalizator’s invention, tickets for each race were printed in advance, and human clerks had to manually record bets (Conlon, 2002).

With the arrival of the totalizator, which was electrically connected to clerks’ selling machines, tickets issued for each horse in a race and for all horses in a race were automatically totaled. A bet placed with any clerk at any ticket–selling machine was instantly and automatically entered into the pool, which allowed quick calculation of pre–race odds and post–race payouts. Moreover, each “horse counter” on the totalizator machine had a trip switch that could be thrown if a horse was a late scratch in a race so that all betting on that horse would immediately stop at every selling machine. After betting closed on a race, the counters on the totalizator were quickly re–set to zero to enable it to be ready to take bets on the next race soon after the previous one had ended. Preparation for the next race was simple: merely change the paper roll on ticket–issuing machines because a different color was used for each race (Conlon, 2002).

This description of the new totalizator machine at Hialeah Park racetrack in a 1932 edition of the Miami Herald illustrates how problematic the previous system had been:

When the windows open there isn’t a printed ticket in the plant. Each [selling] machine is threaded with a roll of blank paper. When you make a bet and the operator presses his lever point, the machine pulls the strip of paper over the printing device, stamps it with the number of your horse, the number of the race, and the amount wagered.

This you see, eliminates printing thousands of extra tickets each day, enables the operators to throw open their windows immediately after the race, ready for the next race. And of course the use of the totalizator gives the bettors the chance to place their money right up to the minute the horses are ready to leave the post. [1]

Obviously, the introduction of the totalizator made processing bets on horse races a vastly more efficient process than it previously had been. One machine in Sydney in the early 1930s could record up to 250,000 bets each minute, a rate quite impossible to achieve with humans recording and totaling bets at a racetrack (Conlon, 2002).

We have become used to a world in which market transactions, like those handled by the totalizator, are mechanized and ultimately automated. Today, many transactions take place via the Internet. And indeed, as the example of the totalizator indicates, parimutuel horse race betting, which can now legally take place online in many states in the United States, as well as across Canada and Australia and many countries in Europe, has much to tell us about the development of communication technologies and how they influence the nature of information flows within countries and across continents.

In the United States and elsewhere, parimutuel wagering on horse racing traditionally took place in the public spaces of racetracks and urban and suburban off–track betting (OTB) facilities where prices have been and still are determined by the actions of others who are also participating in the market. Today, however, with communication and information networks allowing individuals to buy, sell, and bet in semi–public and private spaces, the sensitivity of participants to others in the marketplace — indeed to those who will affect their actions in a particular market at a particular time — has changed. What are we to make of the shift away from the embodied markets — those enacted largely in person, in face–to–face gathering places — that have existed on trading floors and at racetracks and OTBs, and the significance of the types of presence they help to create?

In the case of parimutuel wagering, for instance, network technology, especially the Internet, enables the near–instantaneous transmission of bets from bettors to wagering hubs and of odds information from tote systems to bettors, and the Internet allows bettors to quickly receive data on particular horses in upcoming races. These elements create an environment, characterized by its advantages and its risks, instantly available in virtual space across state, provincial, and national boundaries, expanding its presence to infiltrate new private and public sites, and thus creating new and different forms of connection among participants.



Networks and markets

Relatively recent communication technologies like the Internet were hardly the first applications to contribute to the disembodiment of presence in personal and group communication. As James Carey points out in his analysis of the electric telegraph, the device’s invention and adoption in the 1840s were particularly notable because for the first time in human history, communication was no longer dependent on transportation, and messages could be transmitted in near–real time over wires. One can argue that the invention of the telegraph dwarfs in importance the emergence of all of the significant communication technologies that followed: the telephone, the radio, the television, the Internet, the mobile phone, and the wireless Internet. Each one of these technologies merely improved upon on the telegraph’s ability to almost instantaneously send message content. Very early in the life of the telegraph, proponents saw its potential usefulness in creating and developing national, and later international, markets. The electric telegraph’s precursor, the optical telegraph — which relied on manually operated, line–of–sight signaling towers to relay messages — was seen by its primary developer, Claude Chappe, in part as being useful as a medium by which to relay commodity prices across western continental Europe, and ultimately from France to England and back by sending visually transmitted messages across the English Channel [2].

All markets, including financial markets, had been moving away from in–person contact for some time. The shift began in earnest in the mid–nineteenth century, as James Carey makes clear:

Before the telegraph, business relations were personal; that is, they were mediated through face–to–face relations, by personal correspondence, by contacts among people who, by and large, knew one another as actual persons ... . With the telegraph and, of course, the railroads and improvements in other techniques of transport and communication, the volume and speed of transactions demanded a new form of organization of essentially impersonal relations ... . [3]

As markets continue to move away from face–to–face systems and to electronic networks, non–human agents, and screen systems, some participants think important information is being lost. Deirdre Boden (2000) observes of remote futures markets that many brokers have resisted screen–based systems because they lack the rich information provided by interpersonal cues. She explains that brokers have been, and still largely are

...Acutely aware of, responsive to, and occasionally dependent upon more subtle, interpersonal cues such as voice, posture, gesture, pace, direction of gaze and general choreographed demeanor of other brokers in their vicinity and beyond. [4]

Indeed, one Eurodollar trader at the Merc says that he analyzes the body language of other traders in the pit and bases trading decisions on his observations, noting that, “Sometimes you see things in the pit that make you hit a bid on the screen” [5]. Donald MacKenzie and Yuval Millo (2003) make the same point in describing activity in the Chicago Board of Trade and Chicago Mercantile Exchange pits, where complicated hand signal systems have been developed among traders to augment the existing vocal systems used to make trades [6].

Still, it seems inevitable that markets will move to screen–only systems, with little face–to–face interaction among participants. Among stock market investors, home and work computers now allow them to directly participate in markets through the Internet, without ever dealing with a broker in person. The same shift is taking place in the world of horse race wagering, creating a situation that is largely analogous to the one currently found in financial markets and one in which the use of technology is helping to create new forms of presence and co–presence.



The Internet and parimutuel markets

In many ways, horse race wagering in North America resembles trading in financial markets. In the U.S., all legal horse race betting is parimutuel, which makes it significantly different from other forms of gambling, including lotteries, slot machines, and table games. It is a system of betting in which those who bet on the winners of a race share in the total amount wagered minus a percentage for the facility’s management and any government revenue. Odds in parimutuel betting are thus mutually determined among all people betting, based on the horses on which people are betting and how much money is being bet. As economist Martin Weitzman (1994) explains, the group of bettors collectively arrives “via a market type of mechanism at the return which each horse will pay” as each bettor “follows his or her personal motives or preferences” [7].

Today, the complicated process of managing the collective information flows through which odds are set, as with financial market flows, is handled by computers, though as direct descendents of the totalizator, they still tend to be referred to as tote machines. Indeed, the electric totalizator worked well through much of the twentieth century, when information flows were contained at individual racetracks at which patrons were betting only on races at that track and on one race at a time. In the United States and elsewhere during the 1970s and 1980s, however, OTBs, which allowed bettors to wager at locations other than a racetrack, and intertrack wagering sites (ITWs), which allowed patrons at one track to bet on races at another track, began to appear. The earliest OTBs in the U.S. did not have access to simulcast feeds of races that could be watched on television monitors, nor could wagerers at OTB or ITW locations bet on out–of–state races, but these limitations began to disappear during the last two decades of the twentieth century, so that:

Today, more than 85 percent of the money wagered on horse racing comes through simulcasts — the transmission of signals from racetracks to other racetracks, off–track wagering outlets, and account–betting services. Before 1982, almost all parimutuel handle was generated on track. [8]

By the 1990s, bettors in most U.S. states in which parimutuel wagering was legal could wager on multiple races at multiple tracks in multiple states within a matter of minutes or even seconds (Kruse, 2002). By necessity, totalizators at tracks were computerized and became hubs in large telecommunication networks through which incoming and outgoing information was transmitted across state, and in some cases national, boundaries.

Parimutuel networks in the United States greatly expanded in the late 1990s as more states decided to allow account wagering, enabling bettors to place wagers from home using the telephone, intranets, interactive television, and perhaps most importantly, the Internet. No longer did bettors need to visit a facility and place a bet with a human teller; just as participants in financial markets no longer needed to be on site. Even if bettors traveled to a track or OTB, they had the option of using self–service betting terminals — SAMs — that began to replace some parimutuel clerks, removing an element of interpersonal interaction from the venue. Now, in some states, information flows can legally enter and leave patrons’ homes through online services like Youbet.com, through interactive television and Internet services like the Television Games Network, and through telephone wagering services. Outside of the U.S., online services like the Ontario Jockey Club’s MyWEG in Canada and BSkyB’s SkyBet in Europe provide multi–channel account wagering.

In addition, inside the U.S., bettors can legally wager on races taking place in Australia, Europe, South Africa, and Asia through a variety of services. These sorts of global flows are even more complex when one takes into account the hundreds of illegal Internet betting services with hubs (essentially, servers) located in, among other sites, the Caribbean and Central and South America. National and international information flows converge in single pools at racetracks, creating a number of technological and human challenges related to the speed and ease of information access, and changing the nature of the parimutuel wagering experience.

The relocation of parimutuel wagering first to remote sites, and then into homes with Internet access, as with changes in financial market technologies, has a variety of social and economic implications, most obviously the diminishing number of face–to–face interactions among individuals that take place at betting locations like racetracks and OTBs. Also significant is the fact that the Internet can provide bettors with screens and screens of data that can be used in handicapping races and making bets, but the ability to access unparalleled amounts of information via the home (or office) computer means that certain social affordances may be lost: for instance, the spontaneous sharing of information among track patrons. Message boards enable discussion of races and handicapping angles, but as with financial and futures market traders, text and graphic information may not fully substitute for the presence of cues that may only be fully visible or audible in person, such as the appearance and behavior of a horse from the paddock to the starting gate, and the comments and other behaviors of patrons. On–site participants have been able to observe a variety of environmental cues and use them to make causal inferences about changes in betting pools and odds.

In fact, interpersonal exchanges among racetrack or OTB patrons have been trademarks of horse race wagering; visiting a racetrack or an OTB is a social activity. In her book The Racing Tribe, anthropologist Kate Fox (2005) observes that at racetracks strangers are comfortable making eye contact with each other, and often strike up conversations. She describes the behavior as not characteristic of a normal crowd, but rather of a small, friendly tribe. Indeed, when new parimutuel wagering facilities are built, facility designers take the social nature of horse race betting into account. One designer notes that horse race wagerers “interact among themselves, exchanging tips (and boosting their standings with peers when their tips pan out)” [9], while another observes that horse race bettors are “comparison shoppers” who like to explore various wagering options with other patrons [10]. Not unlike traders on an exchange floor, they are also mobile, moving at the track among television monitors, the paddock, betting windows, and seating areas. This mobility differentiates them from, for instance, slot machine players, who are not only unlike horse race bettors because they are not participating in a multi–player market, but because they engage for long periods of time in a repetitive, solitary, stationary activity. Yet as parimutuel wagering increasingly moves via communication networks into individualized and/or domestic spaces, these specific elements of the experience are changing and even disappearing.

Also disappearing when betting on horse racing on the Internet is the tradition of buying a Daily Racing Form and/or tip sheet to study before a day of parimutuel wagering. One of the advantages of the Internet over print is the easy availability of digitized information from numerous sources. Subscribers to online services have access to substantially more data to handicap races than could ever be contained in a print edition of, for instance, the Daily Racing Form. The print edition of the Form and the track’s program for the day do have certain advantages — or, to use the term Abigail Sellen and Richard Harper (2003) use in their book The Myth of the Paperless Office, affordances — over screen–based media. For instance, Sellen and Harper point out that one can carry, fold, and write on paper, as well as easily examine two print documents in quick succession; these are all useful characteristics both on site and at home when wagering on horse racing. The tabloid format of the Racing Form and the pamphlet format of racetrack programs make both especially easy to read, write on, and carry. At the track or OTB, bettors often use the program to write down the odds early in the betting cycle for a race, compare them to the morning line odds printed in the program, and keep running tallies of their gains and losses on each race. They also consult with others over the Racing Form, compare the Form’s handicappers’ picks to those of tip sheet handicappers, and carry the Form or program to the window to make bets. These are quite different tasks than clicking on links and through menus, and alternating between browser windows, activities crucial in online betting.

Additionally, one might note that online interfaces and telephone menu systems remove human interaction from placing bets from home. Internet wagering services provide tutorials on wagering that help make up for the absence of, for instance, the “beginner’s” betting window present at many tracks to assist new patrons. At–home wagerers, however — like amateur stock market players — may find themselves stymied when their questions are not easily answered by an online tutorial. On the other hand, betting from home means no standing in line to place a bet and no need to have cash on hand. Of course, one still waits in an online queue and faces delays in placing one’s bet — and one does not have the option of switching to another teller’s line to ensure getting one’s bet in before wagering closes for a race. This is just one example of how bettors and traders become aware of absent others who are in some way present in markets.



Networked markets

The expansion of communication and information flows in parimutuel and financial markets means that the decisions of participants are increasingly affected by a growing number of distant others. Technology stretches relations across space, allowing transnational information and investment flows into formerly local markets. The stretching of relations across space affects how participants experience time because, for instance, in parimutuel markets simulcast bettors are often wagering on multiple races at multiple tracks within a short time frame; and in financial markets, investors want to buy and sell at certain times based on time–sensitive prices. Race bettors must make decisions based on changing information for various races for which wagering is open, and their decisions, registered in the form of wagers that change the pools and thus the odds, influence others — both individually and collectively — across space and in time.

Being physical proximal to others creates a particularly immediate sense of presence, because in the absence of technological mediation, communication only with those in one’s physical proximity makes one quite aware of the ways in which communication is embodied in individuals. Carolyn Marvin and Peter Simonson (2004) argue of U.S. presidential campaigns and elections over time “that textualized messages displaced tangible bodies as the prevailing medium of ritual expression” [11], and the same is certainly true of parimutuel and financial markets as they have moved away from being live, on–site markets and toward being located in myriad remote locations linked by wired (and wireless) networks and enacted through screen systems. Some implications of divorcing message content from the human body (and of course, message from transportation) as manifested in early wired network technologies are explored by Jeffrey Sconce (2000) in his book Haunted Media: Electronic Presence from Telegraphy to Television. Sconce points to “the miraculous ‘disembodying’ presence” evoked by the telegraph, making it a technology particularly intriguing to mid–nineteenth century spiritualists who were interested in “the tantalizing possibility of a realm where intelligence and consciousness existed independent of the physical body” [12]. Later media technologies also evoked the idea that communication devices could separate being from body; television was lauded in its early years for allowing viewers to be present “in two places at once” [13] — presumably in one’s home and, virtually, in the setting of the television show one was watching — just as today wireless devices allow brokers in trading pits and bettors at OTBs to simultaneously inhabit the spaces of two (or more) markets at once.



Implications of network technologies: The case of parimutuel markets

As is apparent in both parimutuel and financial markets, networked communication technologies have not only helped move the action to remote locations, they have greatly expanded participation. In parimutuel markets, betting pools have grown as more bettors in far–flung locations are able to bet on races through the Internet. The introduction of these bettors into parimutuel wagering systems can greatly affect odds as pools become national, and in some cases, even international. Obviously, small bets have virtually no effect on odds except in the most minor markets, but communication networks allow high–stakes bettors to easily enter pools and have significant effects on odds at small, and of course even at large, racetracks.

It would be useful at this point to once again make a few observations about parimutuel betting. Because parimutuel odds are set by wagerers as they bet on horses in a race, in a manner similar to the way that prices are set in financial and futures markets, and not pre–set by the “house,” the odds reflect public expectation: i.e., the likelihood of a particular horse to win the race. The opinion of the public is then reflected in the race payoffs. Those who bet on a horse that seems unlikely to win but wins are rewarded for their risk. This has always been true of parimutuel wagering, but today, either on–site or at home, bettors, like the brokers described by Deirdre Boden, “watch screens and a continuous flow of information while simultaneously intervening in those flows to create new conditions of next actions by both themselves and others” [14]. Like futures markets, parimutuel horse race markets are terminal: in other words, time–bound. Bettors become particularly aware of the actions of others and the effects of those actions on odds, and thus on their own actions, as post time nears, odds change, and the market closes. Unlike futures trading, the close of the market is almost always imminent in horse race wagering because there are usually only thirty minutes between races at any track. In the U.S., the Kentucky Derby, the Kentucky Oaks, and the Breeders’ Cup races permit future book betting, which allows wagers to be placed during specified time periods months or weeks before the races, but these races are exceptions.

The importance of others’ actions in one’s own betting decisions in parimutuel markets means that technologies like the Internet that open markets to national and global actors and change the nature of information flows have significant effects on bettors’ decisions, and on their sense of the mediated but still human presences of other market participants. Before OTB and ITW wagering, bettors were quite aware of the presence of other bettors, because they were all physically located at the track. Almost everyone betting on races at that track — with the exception of those using illegal bookmakers — was on site. At the track one might glean information from other patrons, through conversation, eavesdropping, and other forms of observation; and one could also share legitimate information and opinions, or even mislead other bettors in an attempt to get a better price on a horse.

The co–presence of bettors might even allow them, unscientifically, to gauge factors affecting opinions about horses (vocally circulated bona fide information, as well as rumors about injuries, medication, pedigree, and ability, for instance) and thus affecting the odds. With the introduction of OTBs and ITW sites, although a bettor was present with a group of bettors who may (or may not) be wagering on the same race or races, there were also bettors located at remote sites across the state, nation, or world who were only present through their bets into pools, and whose presence was made manifest in the odds and pools displayed on monitors at a network of facilities. As communication and information networks allowed the number of bettors to significantly expand, eventually enabling solitary individuals to place bets over the Internet, the presence of fellow bettors became less immediate.

How might this change in the mode of presence affect parimutuel bettors and markets? We do know a few things about how horse racing markets tend to work. For instance, the horses selected by groups of bettors as those most likely to win — in other words, the favorites, the horses with the lowest odds — win most often. In fact, favorites win about one–third of the time. This statistic indicates to Thaler and Ziemba that “bettors in these markets have considerable expertise” [15]. Bettors are able to reliably and collectively choose the winners of races quite frequently from among the many horses in each race. This result is much better than that predicted by chance. We might wonder then whether the expansion of markets, as interactive services like YouBet.com in the U.S. and BSkyB in Europe appeal to non–hardcore racing fans, will dilute the expertise currently concentrated in betting markets because more uninformed bets will be placed as casual fans and users enter markets. If so, how will this dilution of expertise affect individual and collective behavior?

One thing that is clear is that no matter the size of the parimutuel pool, if a few players wager relatively large amounts on a race, they can and will influence the odds by lowering the odds on the horses on which they bet, increasing the odds on the other horses in the field, and thus influencing the actions of other bettors. Perhaps some players in the market will take the low odds on a particular horse as virtual insurance that the horse will win. Others might view a horse with low odds as a poor betting value and search for horses that promise better payoffs on which to bet. Further, we might assume that a few players investing large sums of money in a market will influence the behaviors of other players even in very large networked markets. Indeed, in n–player games (games in which several players may be involved, rather than just two), the addition of an extra player can shift the equilibrium and make almost all of the other players in the game change their actions.

Economists Rob Feeney and Stephen P. King (2001) stress that player actions are interdependent, and they are especially so when not all players have equal information [16]. Naturally, disparities in information held by players exist in parimutuel markets, perhaps especially in geographically dispersed, heterogeneous online markets. As players in a sequential game, parimutuel bettors make decisions about optimal actions based on what they know about the actions taken by preceding players, reflected in current odds. They may assume that these players have more or less information than they do, and they make predictions about what they think the players that follow them will do. For instance, if they think that the late money will continue to be placed on a horse that is already at a short price, they may well avoid betting on that horse, or de–emphasize it in exotic bets like exactas and trifectas, because they expect a low payoff if the horse wins.

In the terminal markets of horse race wagering, this chain of events leads bettors to wait to act until they have, they believe, the best possible information. A study by Charles Plott, Jorgen Wit, and Winston Wang (2003) found that most parimutuel betting action occurs within the last four minutes before betting is stopped [17]. They argue that it makes sense for players to wait until almost the last second, because this allows bettors the greatest amount of time to glean information from other bettors’ actions [18]. For instance, wagerers may want to see if there are “overlays” — horses that their information tells them have higher probabilities of winning than the current odds suggest — and then try to get the best price. Or bettors might wait so as not to tip their hands by betting early on unlikely horses about whom they have inside information, especially if they are making substantial bets (Asch, et al., 1994). They therefore are avoiding the risk that others will follow them and hurt the price they get on such horses. The fear of followers may well be warranted. Plott, et al. found that “the general pattern seems to be for individuals to ‘jump on’ the favorite with the crowd and ‘jump off’ of the least active markets in a manner typical of herding” [19].

These findings may be proof of the racetrack adage that the “smart money” comes in late. Those who bet early in parimutuel markets take risks by acting without adequate information. Feeney and King state:

Players who must make relatively early action choices may choose an action that ex ante appears to have a low expected payoff. Their decision is driven by the rational belief that later players will cluster on the outcome that ex ante appears more favorable. The interdependency of returns means that such clustering can result in an ex ante more favorable action having a lower expected return ex post. [20]

Given these player tendencies, the networking of odds information, through the Internet, has interesting implications for market decision–making.

For instance, as more data is transmitted through networks from distant locations into racetrack pools, any glitches in the transmission and reception of information can have serious effects for players in the market. Currently, for example, bettors can make wagers on a race until the starting gate opens. Yet in national and international computerized tote networks, there can be delays in information reaching the host server. This means that a player may decide to place a bet based on odds information available in the last moments before betting on a race closes, but she might not in fact get the odds on the horse that she expected. Final odds often do not appear until after the race goes off because of delayed transmission of information through the network. Wagers going through some servers might not be registered in the host track’s pool until the race has started, causing the odds to change while the race is in progress. Although such delays are to be expected in communication networks, changes in odds after betting has closed makes some players suspicious that people have found ways to place bets after races have started (Beyer, 2002).

Suspicions that irregularities exist in parimutuel communication and information networks that, especially in the age of the Internet, allow certain bettors to place wagers after a race starts are not entirely unfounded. For example, mutuel window clerks have long had the power to change tickets after races begin; specifically, they can cancel already purchased tickets from three to twelve seconds (depending on the track and jurisdiction) after a race has begun. This practice is allowed because it can protect tellers from disputes with bettors should tellers find that they have mis–entered patrons’ ticket information (Hegarty 2002). It is a practice, however, that also can cause odds to change after a race starts, and some players suspect that mutuel tellers might be abusing this power to cancel large bets after races begin and thus cause major shifts in odds and payouts.

The notion that mutuel clerks can work the system to change odds after races begin is not merely a product of paranoia. There are precedents for system glitches that have actually allowed clerks to take bets on races in progress, although the clerks may well not have known that the races had gone off. For instance, on 20 September 2003, a malfunction in tote company Autotote’s network let tellers at 26 remote sites in North America take bets on a race at Belmont Park after the horses had left the gate. In fact, at some sites, bets could be placed on the race for up to fifteen seconds after the race ended. Tote networks have safeguards that should automatically end betting when the starting gate opens, but in this case, a software patch meant to help Autotote’s network handle increased traffic caused a problem that kept this “stop betting” command from being sent to the affected sites. Thus, late bets — including bets placed by people who may have already known the race result — went into the parimutuel pool, and payoffs to winning bettors were smaller than they should have been (Drape, 2003; Hegarty, 2003).

Some critics of the current system have argued that significant shifts in odds as a race begins could in large part be avoided if the odds cycle was sped up. In other words, the odds that players see displayed on monitors in OTBs, on their computer screens, and/or on their home televisions would be updated more frequently, allowing players to get the latest and best information available within the last crucial minutes before wagering closes. In 2003, tote companies went from updating odds every minute to updating them every 45 seconds. Subsequently, some have gone to thirty–second updates, and officials at United Tote claim that the company can update win pool odds every ten seconds (MacDonald, 2003).

More frequent updating, however, would not eliminate shifts in odds after races start, and it would mean that already overloaded networks would need to handle even more information. Also, increasingly frequent updates may be of little use to bettors. Chris Scherf, executive vice president of the Thoroughbred Racing Association (TRA) echoes concerns of traders in financial markets over the glut of information to which they are exposed when he explains:

There is a human absorption factor with the timing of updates. If odds changed every five seconds, if that was technically possible, then people watching the odds would go blind in two minutes. You want to give people time to look at the odds and absorb the information. [21]

Limits on human information–processing abilities may mean that speeding up the odds cycle as much as is technologically possible is not the answer to problems in computerized networks. On the other hand, human abilities may prove irrelevant in the face of improving technologies. Currently, “program wagering,” in which computerized robotic agents, which constitute a non–human but in some ways quasi–human presence, aggregate, analyze, and act on information allows big bettors to use real–time odds unavailable to all participants in the market and to place substantial bets just before races go off [22]. Although the expansion of program wagering helps racetracks by increasing participation and thus the amount of money bet, a sticky issue it presents is how to make real–time odds information, and the tools with which to act on it, available to all who are virtually or physically present and participating in the market.





Taking advantage of weaknesses in communication networks to perpetrate market fraud is of course nothing new. In his book on the telegraph, The Victorian Internet, Tom Standage describes an instance in the 1840s in which two men colluded to bet on the winner of the English Derby after the race had been run, but before the results were known to people who were not at the track. The man at the track sent the message “YOUR LUGGAGE AND TARTAN WILL BE SAFE BY THE NEXT TRAIN.” The reference to “tartan” in fact told the message recipient the design on the winning horse’s silks, allowing the man to greatly profit by betting on the winner after the race had been run [23]. Similarly, in the previous decade two French bankers developed a complex system that used the electric telegraph’s predecessor, the optical telegraph, in a way that allowed them to get information about the stock market before anyone else, and then to act on that information [24].

The historical occurrence of such acts should make Internet market participants both aware and wary of the presence of others in major markets that are hybrid, largely technologically mediated, and still at least partly embodied. If we look at markets that are enacted both, and often simultaneously, in face–to–face contact and through network technology, we see that they allow their participants to display and process a variety of verbal and non–verbal information, both proximally and remotely, in communication–rich environments. As technologies allow market participants to engage primarily as Internet users rather than as embodied individuals, certainly their perceptions of others who are present change. In the case of parimutuel markets in the United States for the past two decades, and thanks to communication and information technologies, even as people are copresent with others at OTBs, they are also engaging with an event taking place at another location, present only in graphically displayed information flows and images on, and sounds from, television monitors. The same is and has been true of financial brokers and traders operating in offices across the nation and around the world.




As presence in markets is increasingly technologically mediated and, because of the Internet, removed from public space, the movement is to use technology to amplify the vectors through which immediacy and present–ness can be experienced. The emphasis on real–time odds and real–time stock quotes enabled by faster data processing allows users at remote locations, especially those participating from homes or offices, to feel confident that they are acting on the latest available information, even as they are deprived of verbal and non–verbal cues that once would have been used to make investment decisions.

Absent others are present in changing data: odds in parimutuel markets and prices in financial markets are determined by transnational networks of participants acting in near–simultaneity. Televised and streamed reports from the floors of financial exchanges or racetracks, accompanied by crawls containing real–time market data, emphasize to viewers and users the “liveness” and presentness of the events on which they, in concert with thousands of others, are presumably acting, or deciding to act: even if the illusion of liveness, and the ability of a huge number of remote participants to be part of the event, means increasingly that participants can be present in undetected, unreadable, and sometimes malicious ways. End of article


About the author

Holly Kruse is assistant professor of communication at the University of Tulsa. She has a Ph.D. in communication from the Institute of Communications Research at the University of Illinois.



1. Quoted in Conlon, 2002.

2. Standage, 1998, p. 15.

3. Carey, 1989, p. 205.

4. Boden, 2000, p. 186.

5. Quoted in Skertic, 2004, p. 1.

6. MacKenzie and Millo, 2003.

7. Weitzman, 1994, p. 49.

8. LaMarra, 2006, p. 62.

9. Di Ilio, 1993, p. 42.

10. Dissinger, 1995, p. 58.

11. Marvin and Simonson, 2004, p. 130.

12. Sconce, 2000, p. 44.

13. Sconce, 2000, p. 129.

14. Boden, 2000, p. 188.

15. Thaler and Ziemba, 1994, p. 256.

16. Feeney and King, 2001, p. 166.

17. Plott, et al., 2003, p. 341.

18. Plott, et al., 2003, p. 346.

19. Ibid.

20. Feeney and King, 2001, p. 166.

21. Quoted in DeRosa, 2002.

22. Davis, 2004, p. 5292.

23. Standage, 1998, p. 107.

24. Standage, 1998, p. 106.



Peter Asch, Burton G. Malkiel, and Richard E. Quandt, 1994. “Racetrack betting and informed behavior,” In: Donald B. Hausch, Victor S.Y. Lo, and William T. Ziemba (editors). Efficiency of racetrack betting markets. San Diego: Academic Press, pp. 299–306.

Andrew Beyer, 2002. “Integrity of parimutuel betting at risk,” Daily Racing Form (1 November), at http://www.drf.com/news.article/41929.html, referenced at “Tote system scandal update” (23 November 2002), at http://horseracing.about.com/library/weekly/aa112302a.htm, accessed 8 November 2007.

Deirdre Boden, 2000. “Worlds in action: Information, instantaneity and global futures trading,” In: Barbara Adam, Ulrich Beck, and Joost van Loon (editors). The risk society and beyond: Critical issues for social theory. London: Sage, pp. 183–197.

James W. Carey, 1989. Communication as culture: Essays on media and society. Boston: Unwin Hyman.

Brian Conlon, 2002. “Totalisator history,” at http://members.ozemail.com.au/~bconlon/, accessed 5 December 2002.

Scott Davis, 2004. “Shifting sands,” The Blood–Horse (3 October), p. 5292.

Ed DeRosa, 2002. “Tote companies begin testing 45–second update cycle,” Thoroughbred Times.com (9 July), at http://www.thoroughbredtimes.com/todaysnews/newsview_asp?recno23653, accessed 10 July 2002.

John B. Di Ilio, 1993. “How to plan an integrated facility,” Gaming & Wagering Business (15 May), pp. 42–44.

Donald H. Dissinger, 1995. “Track to the future,” International Gaming & Wagering Business (1 May), pp. 58–60.

Joe Drape, 2003. “Ways to keep schemers from beating the system,” New York Times (22 October), at http://www.nytimes.com/2003/10/22/sports/othersports/22RACI.html, accessed 22 October 2003.

Rob Feeney and Stephen P. King, 2001. “Sequential parimutuel games,” Economic Letters, volume 72, number 2, pp. 165–173. http://dx.doi.org/10.1016/S0165-1765(01)00416-5

Kate Fox, 2005. The racing tribe: Portrait of a British subculture. New Brunswick, N.J.: Transaction Publishers.

Matt Hegarty, 2003. “Accused teller no longer employed,” Daily Racing Form (30 October), at http://www.drf.com/news/article/51048.html, accessed 30 October 2003.

Matt Hegarty, 2002. “NTRA asks for tote upgrades,” Daily Racing Form (13 December), at http://www.drf.com/symposium/2002/news.html, accessed 13 December 2002.

Holly Kruse, 2002. “Narrowcast technology, interactivity, and the economic relations of space: The case Of horse race simulcasting,” New Media & Society, volume 4, number 3, pp. 385–404. http://dx.doi.org/10.1177/146144480200400304

Tom LaMarra, 2006. “The arrival Of intertrack wagering,” In: Staff of Blood–Horse Publications (editors). Horse racing’s top 100 moments. Lexington, Ky.: Blood–Horse Publications, pp. 62-65.

Michele MacDonald, 2003. “Tote companies working to improve posting of odds,” Thoroughbred Times.com (29 September), at http://www.thoroughbredtimes.com/todaysnews/newsview_asp?recno38303&subsec=1, accessed 29 September 2003.

Donald MacKenzie and Yuval Millo, 2003. “Negotiating a market, performing theory: The historical sociology of a financial derivatives exchange,” American Journal of Sociology, volume 109, pp. 107–145. http://dx.doi.org/10.1086/374404

Carolyn Marvin and Peter Simonson, 2004. “Voting alone: The decline of bodily mass communication and public sensationalism in presidential elections,” Communication and Critical/Cultural Studies, volume 1, number 2, pp. 127–150.

Charles Plott, Jorgen Wit, and Winston C. Yang, 2003. “Parimutuel betting markets as information aggregation devices: Experimental results,” Economic Theory, volume 22, pp. 311–351. http://dx.doi.org/10.1007/s00199-002-0306-7

Jeffrey Sconce, 2000. Haunted media: Electronic presence from telegraphy to television. Durham, N.C.: Duke University Press.

Abigail J. Sellen and Richard H.R. Harper, 2003. The myth of the paperless office. Cambridge, Mass.: MIT Press.

Mark Skertic, 2004. “Between the pit And PCs, traders find middle ground,” Chicago Tribune (10 April, business section), p. 1.

Tom Standage, 1998. The Victorian Internet: The remarkable story of the telegraph and the nineteenth century’s online pioneers. New York: Berkley.

Richard H. Thaler and William T. Ziemba, 1994. “Parimutuel betting markets: Racetracks and lotteries,” In: Donald B. Hausch, Victor S.Y. Lo, and William T. Ziemba (editors). Efficiency of racetrack betting markets. San Diego: Academic Press, pp. 255–268.

Martin Weitzman, 1994. “Utility analysis and group behavior: An empirical study,” In: Donald B. Hausch, Victor S.Y. Lo, and William T. Ziemba (editors). Efficiency of racetrack betting markets. San Diego: Academic Press, pp. 47–55.


Editorial history

Paper received 6 September 2007; accepted 21 October 2007.

Copyright ©2007, First Monday.

Copyright ©2007, Holly Kruse.

“An organization of impersonal relations”: The Internet and networked markets by Holly Kruse
First Monday, Volume 12 Number 11 - 5 November 2007

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.