The article describes the decision process of bees in determining where to locate a new hive. More detail on this strategy can be found in an article about Cornell biologist Thomas Seeley’s research, Strategy: Groups vote on hive locations: honeybees:

“In one test they put out five nest boxes, four that weren’t quite big enough and one that was just about perfect. Scout bees soon appeared at all five. When they returned to the swarm, each performed a waggle dance urging other scouts to go have a look. (These dances include a code giving directions to a box’s location.) The strength of each dance reflected the scout’s enthusiasm for the site. After a while, dozens of scouts were dancing their little feet off, some for one site, some for another, and a small cloud of bees was buzzing around each box.

“The decisive moment didn’t take place in the main cluster of bees, but out at the boxes, where scouts were building up. As soon as the number of scouts visible near the entrance to a box reached about 15—a threshold confirmed by other experiments—the bees at that box sensed that a quorum had been reached, and they returned to the swarm with the news.

“‘It was a race,’ Seeley says. ‘Which site was going to build up 15 bees first?’

“Scouts from the chosen box then spread through the swarm, signaling that it was time to move. Once all the bees had warmed up, they lifted off for their new home, which, to no one’s surprise, turned out to be the best of the five boxes.

Pentland’s article focuses on the two processes at play in the bees’ decision making: the centralized process of sending out scouts to gather information and having them return to report back on what they have found and the richly connected network in which various scouts are dancing in the hive and other scouts are looking at the dancers and deciding which one to follow. He writes that creative teams may exhibit the same oscillation between these two processes.

Our experience with globally dispersed innovation teams over the years has shown this pattern. In their scouting phase, teams make good use of collaborative tools to aggregate such information as opportunities they have uncovered and key senior supporters. And idea markets enable team members to dance in front of their favorite idea. But invariably teams request more face-to-face time because it is more efficient.

This is why we consider it critical that teams begin their work with a substantial amount of time together. The experience forces everyone to listen and contribute  – demands which weaken with distance and mediating technology – and enables the team to rapidly share large amounts of information on multiple levels. The momentum the teams build up in their time together can carry them through the much more difficult experience of working virtually.

We invited all 52 participants to invest in the “stocks” of any of the ideas under development. Participation was completely optional and had no bearing on their performance in the program. In the seven weeks the market ran, over 2/3 of them actively traded, which, for a first-time market is an extremely high participation rate. Of course it helped that many of the participants are professional traders.

A handful of program alumni from the previous year also traded. Though it wasn’t a large group, they were paying attention to the innovative ideas coming out of the program and they were offering their insights – two objectives that are difficult to motivate in people who don’t have a stake in the outcome.

Because the ideas on which the participants were trading might take months or even years to come to market, we could not make the final valuation of the stocks dependent on, say, generated returns. Instead we asked the senior executives to “grade” each proposal on a specific set of criteria including strategic alignment, revenue or cost savings potential and thoroughness of the presentation. The aggregate grade became the final price. So in effect, traders were betting on how the senior executives would respond to the projects.

The market predicted which two ideas the senior executives liked the best. It also predicted three of the top five projects and four of the bottom five projects. So, this limited sample indicates that a group of young high-potential employees representing a broad cross section of roles, locations and functions, can think like their leaders.

For future iterations we are considering breaking out the various project criteria as separately traded stocks. This makes the market more complex but should provide better feedback on whether the market is indicating, for example, that people believe the idea is a good one but the team didn’t pitch it very well. We also plan to encourage earlier and more active participation by a wider group of stakeholders and alumni.

Our clients told us that this first experience with a prediction market opened their eyes to the power of collaborative tools for innovation. There is no question about this becoming a regular part of our programs and we are already discussing wider applications.

I am pleased to announce that this week we’ll be launching a prediction market as part of the innovation program Mohr Collaborative is running for one of our large financial services clients. My colleague Art Hutchinson (who blogs regularly about prediction markets, most recently here) and I have worked with David Perry of Consensus Point to provide the market technology. Participants in the current year’s program as well as program alumni will trade on the ideas in development. As many of our participants are professional traders, there is virtually no learning curve. I will write more as the market develops.

Accuracy
Wolfers and Zitzewitz recently published Interpreting Prediction Market Prices as Probabilities that claims that “prediction market prices are usually close to the mean beliefs of traders” and concludes…

Limitations

Thin Markets

Possibilities for Arbitrage

Gaming the Market

Small Probablility Events

Criteria for Success
For a prediction market to work well
1. Contracts must be clear, easily understood, and easily adjudicated.
2. A motivation to trade must exist. Perhaps simply through the thrill of pitting one’s judgment against others
3. There must be some disagreement about likely outcomes. “Disagreement is unlikely among fully rational traders with common priors. It is more likely to occur when traders are overconfident in the quality of their private information or in their ability to process public information or when they have priors that are sufficiently different to allow them to agree to disagree.”
4. There must be useful intelligence to aggregate. Public information cannot be selective, inaccurate, or misleading.

Types of contracts
All contracts assume risk neutrality – that risk doesn’t affect investors’ decisions becuase the amounts being wagered are small.

Winner-takes-all: contract pays off if and only if a specific event occurs. The price on a winner-take-all market represents the market’s expectation of the probability that an event will occur.

Index: contract pays off an amount that varies based on a numeric outcome, say, the percentage of  the popular vote or the number of printers sold. The contract price represents the mean value that the market assigns to the outcome.

“Spread” betting: traders bid on the cutoff that determines whether an event occurs, like point-spread betting in football, where the bet is either that one team will win by at least a certain number of points, or will not. The price of the bet is fixed, but the size of the spread can adjust. When spread betting is combined with an even-money bet (that is, winners double their money while losers receive zero), the outcome can yield the market’s expectation of the median outcome because this is only a fair bet if a payoff is as likely to occur as not.

Families of winner-takes-all contracts can reveal the probability distribution of the market’s expectations.

Why users innovate for themselves

Users do it themselves rather than hiring a customizer because of agency costs (i.e., cost of monitoring the agent), because their needs are unique and they want to get precisely what they want, and also becuase they enjoy innovating.

Because innovation by users is widely distributed, they need a way to leverage and combine their efforts and avoid more than one user developing the same thing independently (market failure). This problem can be avoided through “innovation communities.” Innovation communities are successful (even though you might assume that free riding would kill them) because innovators get some private rewards that are not shared equally by free riders – the innovations work best for the innovators so they don’t lose competitive advantage by revealing them. “Open source software projects are object lessons that teach us that users can create, produce, diffuse, provide user field support for, update, and use complex products by and for themselves in the context of user innovation communities.”

Why (lead) user innovation can be better than manufacturer-driven innovation

To innovate you need two types of information: 1) what the need is and the context in which it will be applied and 2) how the need has been previously solved. Users have more of 1) and manufacturers have more of 2) and each tends to rely more on what they have. So in solving the same problem a user might come up with a new way of doing something while a manufacturer may come up with a modification to an existing solution.

“Manufacturers have an incentive to develop innovations that utilize their existing capabilities—that are “sustaining” for them. Customers know this, and a customer that is considering switching to new technology is unlikely to request it from a supplier that would consider it to be disruptive” assuming the supplier won’t or can’t do it and fearing the supplier will decrease service in expectation of losing the customerLead user innovations are “significantly more novel than those generated by non-LU methods. They were also found to address more original or newer customer needs, to have significantly higher market share, to have greater potential to develop into an entire product line, and to be more strategically important.”
At 3M “lead user projects were found to generate ideas for new product lines, while traditional market-research methods were found to produce ideas for incremental improvements to existing product lines.”

How firms can make profitable use of user innovation

“Firms can proactively affect the rate and direction of user innovation.”

(1) Produce user-developed innovations for general commercial sale and/or offer custom manufacturing to specific users.

(2) Sell kits of product design tools and/or “product platforms” to ease users’ innovation-related tasks—“toolkits for user innovation design” as is done in the semiconductor industry….Need-intensive subtasks are then assigned to users along with a kit of tools.”

Example:
“StataCorp of College Station, Texas. StataCorp produces and sells Stata, a proprietary software program designed for statistics. It sells the basic system bundled with a number of families of statistical tests and with design tools that enable users to develop new tests for operation on the Stata platform. Many advanced customers freely reveal tests they have developed, other users then visit these sites to download, use, test, comment on, and improve tests, much as users do in open source software communities. StataCorp personnel monitor the activity at user sites, and note the new tests that are of interest to many users. They then bring the most popular tests into their product portfolio as Stata modules. To do this, they rewrite the user’s software code while adhering to the principles pioneered by the user-innovator. They then subject the module to extensive validation testing…. The net result is a symbiotic relationship. User-innovators are publicly credited by Stata for their ideas, and benefit by having their modules professionally tested. StataCorp gains a new commercial test module, rewritten and sold under its own copyright. Add-ons developed by users that are freely revealed will increase StataCorp’s profits more than will equivalent add-ons developed and sold by manufacturers (Jokisch 2001). Similar strategies are pursued by manufacturers of simulator software (Henkel and Thies 2003).

Note, however, that StataCorp, in order to protect its proprietary position, does not reveal the core of its software program to users, and does not allow any user to modify it. This creates problems for those users who need to make modifications to the core in order to solve particular problems they encounter. Users with problems of this nature and users especially concerned about price have the option of turning to non-proprietary free statistical software packages available on the web, such as the “R” project (www.r-project.org). These alternatives are developed and supported by user communities and are available as open source software. The eventual effect of open source software alternatives on the viability of the business models of commercial vendors such as StataCorp and its competitors remains to be seen.”

(3) Sell products or services that are complementary to user-developed innovations. “Opportunities to provide profitable complements are not necessarily obvious at first glance, and providers often reap benefits without being aware of the user innovation for which they are providing a complement.”

Manufacturers should “systematically search for and further develop innovations by lead users.”

The traditional focus on target-market users means that lead users are considered outliers. Market research also usually identifies the need but not user-developed solutions.“Listening to your customers” is not the same thing as searching for lead users (Danneels 2004). Many lead users have no incentive to lead, mislead, or even contact suppliers that might eventually benefit from or be disrupted by their innovations. They are simply solving their own needs via in-house innovation.

Lead users are a much broader category than customers of a specific firm (e.g., in advanced analog markets), and many have incentives that differ from those of customers. Lead users don’t care about whether their innovation is distruptive to the manufacturer.
“user-developed innovations that are most radical (and profitable) relative to conventional thinking often come from lead users in “advanced analog” fields. (users who have related but more extreme needs than any users in the target market).” Example: ABS came from aircraft braking which had more extreme needs than automobile braking.

How to find advanced analog lead users: “Pyramiding” = ask target market lead users to nominate. People with rare interests tend to know others like them.

How to find lead users in target markets: “at specialized sites or events”

One might think that an alternative approach would be to identify lead users before they have innovated but user innovation is likely to be a widely distributed phenomenon, and it would be difficult to predict in advance which users are most likely to develop very valuable innovations.

A video tutorial on identifying lead users: link
Application:
In the 3M experiment 3 or 4–member “lead user teams” from the marketing and technical depts. were coached through a process: “Teams began by identifying important market trends. Then, they engaged in pyramiding to identify lead users with respect to each trend both within the target market and in advanced analog markets. Information from a number of innovating lead users was then combined by the team to create a new product concept and business plan…” (von Hippel, Thomke, and Sonnack 1999).” link to paper

Result:
“Annual sales of LU product ideas generated by the average LU project at 3M are conservatively projected to be $146 million after five years–more than eight times higher than forecast sales for the average contemporaneously conducted “traditional” project. Each funded LU project is projected to create a new major product line for a 3M division. As a direct result, divisions funding LU project ideas are projecting their highest rate of major product line generation in the past 50 years.”

Reflections on market efficiency and the knowledge economy

• Markets become more efficient as the information accessible to transaction participants improves. New means of information aggregation are radically reducing the cost and disrupting the business models of firms that specialize in information collection
• (Foray, D. 2004. Economics of Knowledge. MIT Press.) positions users at the heart of knowledge production. He says that
  one major challenge for management is to capture the knowledge being generated by users “on line” during the process of doing and producing….He discusses implications of the distributed nature of knowledge production among users and others, and notes that the increased capabilities of information and communication technologies tend to reduce innovators’ ability to control the knowledge they create. He proposes that the most effective knowledge management policies and practices will be biased toward knowledge sharing.
• (Weber, S. 2004. The Success of Open Source. Harvard University Press.) The notion of open-sourcing as a strategic organizational decision can be seen as an efficiency choice around distributed innovation, just as outsourcing was an efficiency choice around transactions costs.

·        In 6 out of 8 events for which official forecasts were available the market predictions were closer to the actual outcome than the official forecast.

·        The probability distributions calculated from market prices were consistent with actual outcomes.

·        The market made accurate qualitative predictions about the direction that the actual outcome will occur (above or below) relative to the official forecast.

I’m separating my notes into two posts. First, the nuts and bolts about how it was done and second, some of the scientific issues.

How it was Done

The Prediction

Typically, the prediction was for monthly sales for a month three months in the future.

Business Constraints

·        Hesitation to engage employees in an exercise in which they might lose money. Solution: provide a small amount of cash to each participant before the market sessions – this constrains the amount of stakes a participant can have in the market and affects incentives to trade.

·        Market has to offer useful information. E.g., if forecasts are not valuable if they are made with horizons less than 3 months then market sessions need to be conducted 3 months before the event to be predicted.

Who participates

·        Relatively small number of participants chosen. Selected specifically from different parts of the business operation because they were thought to have different patterns of information about the targeted event. These patterns of information, including market intelligence, specific information about big clients, and pricing strategies, were in need of  aggregation. No public summaries of information available to the participants during the operation of the IAM. The official forecasts were not known until after the IAM closed.

·        Participants need to be selected carefully – don’t want to “miss” a person with much information but it might not be efficient to include many people without any relevant information. Little is known theoretically about the information size relative to the market that might be required for effective information aggregation to take place.

·        Laboratory experiments have suggested that a small number of uninformed participants provide both market liquidity and a function of adding “consistency” to the market through a process of “reading” and “interpreting” the actions of others. So, around five subjects was recruited from HP Labs (with little or no information) in each experiment.

Preparing the participants

15-20 minute instruction session from: explained the structure of incentives, the market mechanism and the web interface. In addition, the participants were told the goals of the experiment and were told that their participation was important for HP business. Contact information provided and participants were encouraged to call if they encountered difficulties.

Defining the Contracts to be Traded

Most similar to the IEM “winner take all” markets (state contingent securities).Traded a complete set of state contingent contracts (Arrow-Debreu securities). The space of possible outcomes was partitioned into about 10 intervals. Each interval was given a name and with each interval there was an associated security with the same name that traded in a market with that name. Thus the interval 0-100 would be associated with a security named 0-100 that traded in a market named 0-100. The interval 101-200 would be associated with a security named 101-200, etc.

Payoff

If the final outcome fell in an interval, the corresponding security would pay, say, one dollar per share at the end of the experiment. All other securities would pay nothing. A higher payoff per share would place more value on the share but the payoff per share interacts with the total cost of the exercise and the potential volume of trades and related market liquidity.

How to Start

Each participant is given a portfolio of shares in markets and cash to start. Could start with equal shares in all securities or could start with shares in every other security, alternating which security was first across participants. The unequal distribution of endowments was used to encourage trading by attempting to make sure that the initial endowments of securities did not approximate the ultimate equilibrium.

Market Mechanism

Web based, double auction markets. Marketscape software (Laboratory of Economics and Political Science at Caltech.)

·        All the markets for an event were organized on a single web page for easy access.

·        Links to a complete time series of trades available.

·        Links available to HP data bases, which allowed participants to review data held by HP.

·        A participant could enter a buy offer, a sell offer or acceptance of an offer through the web form on the page. Orders were compared to the other side immediately. If a trade was possible, it was executed and if not the order was placed in an order book. The best offers were listed on the main market web page. The whole book of offers was available for each market at the click of a button.

·        Participation was anonymous. However, each participant was assigned a subject ID number for each experiment. During the experiment, the subject ID number of the person who made offers and transactions were public knowledge. Participants had the ability to track behavior of other subjects with   in the same experiment if they wished to.

When should market be open?

·        In all cases the information was gathered for a week with the markets being open during lunch and in the evening every day. Management did not want participants being preoccupied with the task during the working day when pressing issues needed attention.

·        It is desirable to have a schedule (for example, 24 hours for a week) to minimize conflict with other activities.

·        It is not desirable to leave market open for long periods because participants will often find a lack of activity in the market and thus lose interest.

Advantages of Prediction Market Over Other Forecasting Methods

·        The methodology is flexible. It can be used to aggregate any type of information possessed by different people. It involves a natural methodology for quantifying subjective, qualitative, information and giving weights to the opinion of different people for the purpose of information aggregation. The task is performed giving not only a point forecast but also a complete probability over the range for which the value of some unknown variable is to be predicted.

·        The methodology is scalable by number of participants, timing of participants and location of participants. There are no practical limits to the number of people that can participate. With markets conducted over the Internet, hundreds and even thousands of people can participate either at the same time or at different times. Traditionally, businesses collect and aggregate information through a process of meetings, which not only limits the number of participants but also the time frame for information collection.

·        The methodology tends to be incentive compatible. Incentives to hide information, misrepresent information or simply ignore requests for information are either eliminated or limited. Furthermore the markets are designed to give incentives to the participants’ to acquire information about future events and use this information wisely in the market.

Observations

·        Theoretical arbitrage profits existed. In all the experiments, prices summed to be greater than the winning payoff. However, to take advantage of the arbitrage conditions, individuals needed to execute multiple trades when fluctuations of prices were substantial. So it is likely that there were actually no practical arbitrage opportunities. Why in all 12 experiments was the sum of the prices always above the winning payoff?

·        No significant trends in the sequences of predictions are observed. So it doesn’t appear there was any response to changing market information during the trading. Maybe all the information aggregated quickly at the beginning.

Scientific issues

·        How is the performance of the system related to the psychology and decision biases of individuals?

·        How can one deal with incentive problems in which individuals might large incentives to conceal or misrepresent what they know?

·        What rules and mechanisms might be needed for different underlying information structures?

·        If markets are thin or the number of participants few, how will the performance of the system be affected?

·        How can we find the people with the relevant information and how do we know that they knew something of relevance anyway? If the participants know nothing, the mechanism will produce nothing.

·        Can a prediction market not only produce a prediction but also simultaneously help management ascertain which participants have information. That is, can it be designed to attract those with good information and discourage those with bad information?

Scientific Literature

The experimental demonstration is first found in Plott and Sunder (1982, 1988). This early paper demonstrated that the ability of markets to aggregate information is sensitive to the market architecture. In particular, this early work demonstrated that compound securities are not as reliable as indicators as a complete set of state dependent instruments. The conditions under which a single compound security is reliable are isolated in Forsythe and Lundholm (1990) The need for selecting proper instruments is underlined by demonstrations of markets that can equilibrate at patterns that are not fully revealing of information such as cascades (Anderson and Holt, 1997; Hung and Plott, 2001) or misleading such as mirages (Camerer and Wiegelt, 1991) or bubbles ( Smith et al, 1988; King et al. 1993; Porter and Smith, 994; Lei et al, 2001). In fact, some types of market organization facilitate no information aggregation at all as is the case of the winners curse in sealed bid auction markets (Kagel and Levin, 1986; Lind and Plott, 1991). See Sunder (1995) for a summary, or aspects of search (Sunder, 1992).

“They need not be perfect in order to be compelling. Compared to any pragmatic forecasting alternative, prediction markets remain remarkably resilient to manipulation, and uniquely (if not perfectly) efficient at assessing the impact and importance of vast amounts distributed information.”

Examples:

• The Iowa Electronic Markets (IEM). IEM traders with real money at stake called the presidential race for Bush in September.
  “This is the fifth straight presidential election that the Iowa Electronic Markets has called correctly. Other markets did as well or better – including forecasting the tightness of the race.What I find even more fascinating is that each candidate took every single one of the states in which he was running over 50% likelihood according to a smoothing formula I applied to pricing data gleaned from Tradesports in the final days. Every one. As I watched it play out through the night, it felt like I’d been let in on prophecy.“
• The Des Moines Register carried this story two weeks ago, describing how the Iowa Electronic Markets are involving physicians in markets to predict the location, timing, and severity of flu outbreaks this Winter. …It was 90 percent accurate during a short pilot project last year

Following are from an article in the Australian paper, The Age

• The Hollywood Stock Exchange, an affiliate of online trading firm Cantor Index Ltd, allows people to buy and sell virtual shares in movies, celebrities and music. To pay for pseudo-shares, they use pseudo-money in the form of “Hollywood Dollars”. This allows people to bet on such questions as total box office returns and Oscar winners. Because the data outperforms industry forecasts, it is also syndicated as market research.
• Three years ago, Goldman Sachs and Deutsche Bank launched a market for economic statistics futures including employment, industrial output, retail sales and inflation. The Chicago Mercantile Exchange now trades in inflation futures contracts.
• Some companies have also experimented with prediction markets. Hewlett-Packard, for example, set one up that reportedly generated more accurate forecasts of sales than its own internal processes. Siemens had one that predicted the German conglomerate would fail to deliver on a software project in time, in defiance of its established management systems that insisted the deadline would be met. Management was wrong.

Prediction markets for business

• “By accepting the superiority of managed, credential-based, hierarchical information flows across the board, organizations are handicapping themselves in evaluating early signals on some of the most important open ended strategic questions they need to confront. And while such mechanisms are powerful and necessary for many kinds of tasks, they’re poorly suited to the task of harvesting, assimilating, and assessing distributed intelligence in the near real-time. I.e., on questions with high uncertainty and little or no past precedent, it’s only smart business to acknowledge the possibility that the best answers may sometimes arrive from truly unexpected sources.“
• “…where I think prediction markets will find traction in a corporate context is in highlighting where management teams ought to pay more attention (i.e., do more research), and in making groupthink denials of emerging external trends and inflection points far more difficult to sustain. This will be particularly true of prediction markets that are populated across corporate boundaries -e.g., including customers.“

Comparing a prediction market to the stock market: the importance of making contracts specific
“I suspect that other hypothetical contracts, written so as to avoid clear moral hazard issues could have (and still might) shed strategically important light on how the market for over-the-counter pain medications will evolve. For example, a winner-takes-all contract for 2006 stating that: “Two companies hold 90% share in Cox-2 therapies”, might have signaled much earlier that this was a market bound to consolidate. The reasons could have had nothing to do with the safety of any particular drug (e.g., difficulty in establishing a compelling third brand, patient-perceived efficacy, mergers and acquisitions, etc.) What keeps the idea of trading in such specific contracts interesting vs. simply watching the share price of major pharmaceutical companies on Wall Street is that…such markets can highlight the potential for specific change events with much greater precision.”

Critical success factors

• Select the right kinds of (non-random) questions
• recruit/attract a solid pool of informed/active tradesr
• Only rely upon the predictions of markets that show robust and honest trading activity
• Limit the stakes so as not to create moral hazard. There’s plenty of evidence to suggest that web games and carefully constrained real money markets are as or more effective than high stakes wagers at predicting the future outcome of uncertain but non-random events
• Make it easy to figure out what’s going on
• Be clear about trading parameters, e.g., market open and close
• Make it obvious how trading values translate into probabilities or vote shares
• Make it easy to compare the prices of the underlying commodities

Problems with Prediction Markets

• Skewed participation
• Poor question articulation
• Inadequate information input
• Thin floats and big spreads.
• Can prediction markets be deliberately influenced?
  “…prediction markets are proving as resilient to deliberate influence as theorists have long said they would be.” Example: The Bush relection futures at Tradesports survived a speculative attack: “‘There is now no question whatsoever that the Bush re-election futures contract at Tradesports.com is being manipulated. Yesterday the price of the futures were sold down from about 55 (indicating the market’s estimate of a 55% probability of Bush’s re-election) to 10 (indicating on a 10% probability) with a single 10,000-lot order entered by a single trader. An order that size represents twice the normal volume of an entire typical day’s trading’….The ‘price’ of a Bush future quickly rebounded to the mid 50’s.“

Robin Hanson on “Manipulators increase information market accuracy“.“Why does… manipulation seem to be less of a problem than many fear it should be? One possible explanation is the view that a manipulative trader is in essence a ‘noise’ trader in the sense that his trades are based on considerations other than his best estimate of asset value… when potentially informed traders have deep pockets relative to the volume of noise trading, increases in trading noise do not directly effect price accuracy… by inducing more traders to become better informed, an increase in noise trading indirectly improves the accuracy of market prices.” (emphasis added)

How prediction markets can separate the wheat from the chaff
“…let prediction markets directly incent information gathering and sharing, highlighting those individuals with the best grip on reality in particular areas, i.e., those with knowledge that’s objectively valuable to the enterprise. Mary Murphy-Hoye of Intel (a pioneer in using prediction markets) made this point directly in a Time Magazine feature article last summer entitled “The End of Management”:

‘I can now tell if planners are any good, because they’re making money or they’re not making money.’

She highlights a little talked-about, but important flip side to the whole debate: shouldn’t good knowledge management discipline marginalize those who obfuscate, dilute or detract from institutional knowledge-building to the same degree that it elevates (and enriches) those who add to it?“

Here are my notes on The Wisdom of Crowds by James Surowiecki

Overview

Under the right circumstances, groups are smarter than the smartest people in them, even if the group doesn’t contain expert members.

Experts are as likely to disagree as agree. Experts’ individual consistency is also 0.5. Experts overestimate the likelihood that they are correct – little correlation between self-assessment and performance. Therefore, however well informed and sophisticated and expert is, his advice should be pooled with that of others and the larger the group the better. So we should stop “chasing the expert” and ask the crowd. We continue to chase the expert because we assume average means least common denominator (Larrick and Soll) and we are fooled by randomness (Taleb).

Criteria for good collective decision making

The group must be big enough and diverse enough, the members must be forming opinions independently, and the group must be decentralized.

Diversity

To solve cognition problems you must 1) uncover alternatives and 2) decide among them. Diversity is needed for both 1 and 2. Diversity adds perspectives and weakens destructive characteristics of group decision making. A successful system recognizes losers and kills them quickly. Homogenous groups spend too much time exploiting and not enough time exploring (James March). Homogenous groups are susceptible to groupthink, willing to rationalize away counterarguments and often convinced that dissent is bad. Diversity is more important than individual intelligence but members must be somewhat informed. There has to be some information – can’t have a completely ignorant group.

Independence