cover
ISBN: 0385721706

Cites sociology, social experiments and case studies to reinforce the theory that under the right conditions (4 of them) collective group
decisions can be better than those made by a few experts.

Impactful intro – cited the story of Francis Gallon, who sought to disprove the average intelligence of a crowd but proved instead what Surowiecki wrote: “under the right circumstances, groups are remarkably intelligent, and are often smarter than the smartest people in them” even if most are not especially well-informed or rational, it can reach a collectively wise decision. p.XIII

And paradoxically the best way to do this was if people acted and thought individually.

[If this holds true, then what makes the difference between the wisdom or folly of crowds is the circumstances. And how many in a crowd to make it substantial? And how independent were individual decisions made? The book addresses in subsequent chapters]

P7 the 1986 Challenger shuttle explosion, the US stock market was found to have accurately predicted the culprit was Morton Thiokol
(later found to be culpable for the defective O-rings)

P 10. Four conditions that characterised wise crowds:

  • diversity of opinions (each person has private info, even if it’s eccentric interpretation of known facts)
  • Independence (individual’s opinions are not determined by opinions of others)
  • Decentralization (people able to specialise and draw on local knowledge)
  • Aggregation (some means to turn private judgement to collective decision)

[Also adds there has to be some information available].

P15 google search results and how it taps on the principle of collective crowd preferences.

P29 (to prevent group-think) fostering diversity more impt in small groups and in formal org than in larger collectives like markets or
electorate (in the latter, diversity is almost guaranteed)

P 30 intelligence is relevant but “intelligence alone cannot guarantee you different perspectives on a problem”

P31 qualifies that diversity without being informed cannot bring about collective wisdom smarter than an expert’s. Diverse group of people
with varying degrees of insights and knowledge can make better decisions than one or two very smart people. Also, that new recruits to org bring value not because they possess more knowledge (they can’t, not as well as seasoned staff) but for the diversity they bring.

[I think it’s about letting the potential employer know about your diverse expertise. You have to bring some value at the very least! Indeed, if you read LKY’s memoirs, he didn’t run the govt entirely. It was a team of old guards who had diverse views, and made collective decisions].

P43 – “looking up the sky” social experiment, i.e. Social Proof.

P48 case study on statistical analysis of American football plays suggests that one has to deal with psychological inertia. Stats and
also tactical logic suggests football coaches are better off if they adopt a more aggressive play. But it’s against conventional wisdom and
practice. I think unless the team has supreme confidence in the decision of the coach, the coach has to convince the team why going
against convention makes sense. the case study is also an example where it’s not crowd wisdom as it lacked independent thought based on
information.

P53 information cascade. That it’s impt for decision to be made at the same time rather than in sequence and subsequent decisions being
“contaminated” by actions of others.

P57 “collective decisions are most likely to be good ones when they are made by people with diverse opinions reaching independent
conclusions, relying primarily on their on their private information.”

Mentions Sept 11 attack; how it wasn’t the failure of decentralization but the lack of aggregation of independent views.

P94 Stanley Milgram experiments with queue-jumping and asking people to give up their seats. Found that the hardest part was actually working up the courage to ask for the seat, rather than resistance from those being asked (one observation was that conventions, norms and unwritten rules can be so entrenched that it takes too much work to act on the contrary.)

P105 when it comes to stock market is more erratic.

P246 explains that buying a stock is a “dependent decision”. That most people buy a stock based on what they think others feel the stock is
worth (?)

P116 why people cooperate at all
[I can apply this to use of the public library – it’s a cooperative public that will make things work.]

P135 US TV industry – reverse example of how individually smart key TV networks make poor decisions collectively, when they don’t solve
‘cooperation problems’.

P139 experiment showed most people start out honestly but decide to become selfish when they see most others are. (negative anti-social
behaviours are results of the actions of a few?)

P147 Cites Singapore’s road pricing system mentioned as case study of resolving traffic coordination problems (letting the crowd make
decision on whether to drive, as opposed to telling who can drive – like Mexico’s odd/ even licence plate system; drivers but two cars
with odd/ even plates).

P157 diversity is essential to solving problems but poses a challenge for coordination problems.

P160 The SARS crisis as a case study for decentralized cooperative self-managing, coordination.

P165 allowing individual pursuits are better collectively (given a shared direction, I’d think) is better than ordering people what to
do.

P175. Space Shuttle mission Columbia disaster, the Mission Management Team (MMT) as case study of how not to run small groups.

P186/ 7. Hierarchies in small groups. Ideas that are heard first tend to be more influential. Harder for later dissenting views to break
what has been agreed (my experience says this is 50/ 50). Surowiecki suggests that strong advocates for ideas, even if idea is good, is
important.

P190. If an org sets up teams and merely expects them to be an advisory team than be able to make decisions, it means not realising
the potential for collective wisdom (if the end decision is still made by one individual).

p210 how employee participation is about giving decision-making powers also. Or else it might just be layers of red tape.

P212 decisions about local problems should be made by those close to the problem.

See P212. Studies suggests it’s enough that productivity increases when ppl know they can choose to make a decision/ choice, and they
usually do not exercise that choice.

P215 decentralization works only if everyone plays on the same team.

P222 the best CEOs recognise their limits of own knowledge and individual decision making. They build teams to help make those
decisions.

P228 suggests short sellers are reviled but they are necessary to keep process from escalating.

p241 bowling boom and bowling bubble of the post-WWII years.

P251 prob is when u you start following a crowd without doing your independent analysis.

See p255 the way information is phrased mattered. If it subtly hints either positively or negatively, it creates biases.

P257 mobs don’t suddenly turn violent. There is a buildup among the majority who are unsure, where the more people do so, more will act in
the same way.

P262 what voters think democracy is
[my view: “democracy is the process where free people all exercise their rights to allow a select few to dictate the lives of the many”]

One take-away: Collective outcomes may be good, but not necessarily so for the individual.

Advertisements