Guild Clearing House (Milestone 4)

From crowdresearch
Jump to: navigation, search


Many in the cohort have expressed interest in Guilds, as organizational design, how they potentially connect with Skills matching tools, Pricing Engines, Learning Communities or as negotiating platforms, etc. Wether you want to collaborate, flesh out an idea, see who has a similar/shared idea, look for inspiration, this page is intended to be a clearing house for ideas/problems/solutions. One of the joys of this project has been the sharing of ideas and collaboration...So, use the page respectfully, just use it :) The idea of Guilds is not new, they were established many years ago, so no one in this century owns the guild idea :)...However, this cohort can certainly be the steward of the guild concept in the digital economy

Guilds have seen a renewed interest in the context of Massively Multi-Player Online Games (MMPOG). Players have been encouraged to regroup in so-called guilds in order to share experience and cooperate in game session. Participating in a guild has become a large part of the game experience.

The overall idea is to add a "social infrastructure" element into crowdworking.


Can Guilds address the following problems:

  • 1. Cold Start for workers
  • 2. Expertise and Competency expectation/standardization (Reputation/Trust). Can guilds solve the ironic relationship between pricing and skill exhibited in Mturk (
  • 3. Voice for workers: grievence and programatic. (Balance of Power)
  • 4. Universal pricing (See @Seko's Clustering solution...)
  • 5. Upward mobility/skills acquisition/professional development.
  • 6. Improve communication amongst workers
  • 7. Improve the depth of the crowd. Right now 10% of the workers in the crowd do 90% of the work. Can Guilds flatten out the crowd?
  • 8. Can Guilds be dropped into MTurk and fix problems or are they part of an integrated solution....Or, do Guilds only work in new holistic open systems.
  • 9. Creating a sense of belongingness/relatedness for the workers.

Guild Structure

Guilds have a long history in the pre-digital world as skilled craftsmen organized to protect and preserve a way of life, a craft/trade. With such a concentration of skills, Guilds held considerable negotiationing power in engaging with employers brought uniform quality to craftwork while protecting skilled tradesmen, improved or at least standardized wages and instituted robust training and apprentice programs that ensured quality and helped with cold start challenges. These conditions normalized relations between workers and employers by establishing and introducing standards, expectations and bureaucracy into economic transactions. Employers came to trust guilds as experts in their chosen field and workers found value in their annual dues.

As Trust and Power collide over issues of accountability, responsibility and competency, digital communities must address the basic premise of expertise and how that is defined. In a pre-digital world the acceptance of expertise was shaped by, but not limited to: educational pedigree, pre-ordained social status, cloistered/rural communities, etc. But, as universal education (and grade inflation J), post war social mobility and improved transportation systems radically changed our world, so did our idea of the expert. In a digital world in which we have access to more information than ever before, Experience and Experiences format a new social contract; especially when anonymity is factored into the assessment.

In a digital world, standards, the fundamental building blocks of a community can be captured, measured and used for the forces of collective good (Not the facebook kind J) to establish the definition of expertise, beyond reputation and perception. Once expertise/competency has been defined (as a quality quotient?) by converting actions, anecdotes and behaviors into quantitative metrics Balance, Symmetry, and Harmony, the natural state of efficient systems can be achieved. Just because an idea is old, doesn't mean it cannot be innovative....Hence, the digital Guild.

By introducing Guilds into Daemo, workers through the tenants of collective action, get a powerful voice in the transaction. Leveraging a cohesive membership group to define wages, standards and expectations in the digital world are no different than their antiquated brethren. For Guilds to do so, they establish criteria for entry, embody an ethic of expertise and create and reinforce the bonds of trust by delivering on their claims. Guilds must internally self regulate or risk losing prestige, standing, reputation. Yet, Requestors also benefit in that instead of having to trust (and go through the vetting process individually) they can place it with an organization, as inclusion in a guild requires entrance exams, apprenticeships or other processes that establish competency and expertise be it transactional and/or subject matter oriented. They can have a single point of contact which can lead to repetitive interaction which has comforting as well as transactional benefits (familiarity), be able to establish expectation as to time and quality.
Guild Organization/Framework
  • Leadership
  • Apprenticeship
  • rank and file
  • committees (Grievence/Pricing/Training/Vision&Development)

Needs Addressed by a Guild Structure

  • Workers need to feel they are being fairly compensated for their work.
  • Workers need to feel like they are treated fairly and respectfully, and have a voice in the platform
  • Workers need to be able to expose their skills so they can get work they are qualified for and advance their skills
  • Workers need to feel they belong to a group. The "relatedness" factor has been proven to be extremely important in the happiness and commitment of individuals whether it be workers in a company in which they enjoy belonging, or students in a school. Guild structure is answering to the issues of crowdworker loneliness seen in Week 3 testimonials.

  • Requesters need to be able to trust the results they get
  • Requesters need to have workers who have the appropriate skills and demographics do their tasks
  • Requesters need to find "qualitative workers" quickly and efficiently. A guild is akin to a "seal of quality" on top of potentially dozens of workers.

Solutions and Outcomes

  • 1. Establish a Culture of Skill. Organizations are moving from low cost to high skill labor
  • 2. Leverage ground truths to serve as a universal skills baseline...Consistency and predictability (Reliable)
  • 3. Guilds can scale and are reliable
  • 4. They embody Trust and the symmetrical distribution of Power
  • 5. Flat organizations work, because they are comprised of like minded folks, that share a similar mental model of 'reasonable"
  • 6. By establishing Norms...can create a matrix association of skills to wage.
  • 7. These norms can also provide the foundation for pricing engine.
  • 8. Economies of scale and relationship with MOOCs
  • 9. Create a certification process. (Digital and non Digital)
  • 10. Collaboration and Collective Intelligence

Guilds help manage workers at scale, which on a global level is important. It also helps focus activity and bring a singular voice and a universal standard to grievance management, skills development, wage negotiation, cold start of newbies, etc. Opportunity for entrepreneurs


  • Where does the evolution of Guilds go? Develop into Unions, Co-ops or evolve into companies?
  • Can Guilds go out into the market and secure business...yes?
  • At what point does a guild get too there a tipping point when the community fractures because (Divergence of interests, agendas, politics)?
  • How does the crowd absorb competing guilds
  • Can they structure in a way to provide benefits?
  • If a crowd has an effective pricing engine, does the system need guilds.
  • At the individual point of view, what type of behaviour is allowed ? Should multiple identities and participation in several groups at the same time be permitted ?

Tentative Reflections

Guilds should be free to organize as they want. They will be nonetheless influenced by the tools that are made available by the platform for them to function. A guild can certainly makes use of additional external systems but if the Daemo provides efficient mechanisms, these would shape the organizational structure of guilds. The offered functionalities will contribute to grant power to leaders in the group. For example, introducing pooling of jobs between members of the group (directly or indirectly) ultimately makes leaders in position to control revenues of these members.

Human nature being constant, groups of people will trend to the whole range of possible organizations and governance. While in the real world, localization has made people dependent of existing ruling groups, it is not necessarily the same in the virtual world where participation in a group can hardly been enforced. However we must take into account the possibilities of cyberbullying, especially in a context where reputation will have a direct impact on the capacity of someone to apply for a job. Mechanisms (and monitoring) must be in place to prevent deviant behaviour and make sure that fair competition remains prevalent. This is certainly of major importance for the adoption of the platform and its sustained success in the long run.

In terms of research foundations to help our understanding, we must certainly found interesting matters in social studies and game theory regarding group dynamics in a quite well defined context. Current state of affairs in MMPOG is another useful information source. Guilds in MMORPGs like World of Warcraft or even more massive (Dark Age of Camelot, etC.) are a demonstration of a virtual structure, a frame, making possible that a group of up to hundreds of diverse players from across the globe complete complex tasks (like killing boss in raids) together. Interactions with pricing and skill recognition are also quite certain as exemplified by the functioning of historical guilds in the Middle-Ages. These must be understood and utilized to skew globally the system towards a meritocratic community. The following paper gives an interesting and quantitative analysis of the social dynamics of the World of Warcraft platform:

  • Nicolas Ducheneaut, Nicholas Yee, Eric Nickell, and Robert J. Moore. 2006. "Alone together?": exploring the social dynamics of massively multiplayer online games. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '06), Rebecca Grinter, Thomas Rodden, Paul Aoki, Ed Cutrell, Robin Jeffries, and Gary Olson (Eds.). ACM, New York, NY, USA, 407-416.

Of particular interest, the article reveals that guilds bring inherently very loose tights between players with very low interactions between members. Additionally, sizes of guilds remain surprisingly limited: in the period of the study, the median size of guilds was 9 and most of them (90th percentile) count less than 35 people. Guilds appear then to be very unstable groups with an estimated 25% death rates by month and a large churn rate in striving ones. Of course, we are considering a game that could hardly be considered as the main activity of users and we can expect more involvement when dealing with people that make a leaving out a constant presence in a platform. The analysis stresses the importance of building an audience to one successes in game in games like WoW as well as the actual difficulty to manage large scale guilds with the game tooling set.

Possible / Necessary Features of a Guild-friendly Environment

  • Dedicated forum creation
  • Dedicated IRC channels
  • Personal page profile (multi-sided ?) -> see identity management in the platform
  • Task management environment
  • Ranking (ladder-board) and badges attribution capacity. Possible gamification (akin to guilds in MMORPGs).
  • Distributed administration capacity of dedicated tools
  • Anonymous voting mechanisms
  • Joining and leaving mechanisms
  • Mediation capacity in relations between guild representatives and requesters
  • creation of sub-guilds/sub-groups dedicated to certain fields. See the example of Wikistrat, where the thousands of analysts are a part of expertise sub-groups. Same with group-based MOOCs (see NovoEd, where learners of the same MOOCs are studying in groups).
  •  ???

The Skill Forest: a mechanism for matching task profiles to worker skills

From reputation to adequation


  • Problem: Reputation is a broken mechanism for building trust between requesters and workers.

Reputation systems pretend evaluate an individual through a single rating. This approach forbid a true assessment of any advanced skills, by definition not homogeneous. As a consequence, only the more basic micro-tasking can make use of reputation to select candidate workers, therefore putting an implicit limit to crowdsourcing platforms in addressing the more complex and value-added tasking. This limit severely reduces the available range of wages on the platform and helps in producing a downward spiral in quality of production.

  • Solution: abandon reputation for a real assessment of proficiency in specialized domains, and automatically assign tasks to workers while taking into account their particular skills.

If we can find a reliable process to evaluate more advanced and detailed rating system, we can provide requesters with critical information regarding the capacity of the crowdsourcing population to service more advanced tasks. This gradually must enable them to create job opportunities with more added value. Budget allocation will potentially raise in proportion in a more diverse market. This will enable various types of workers to find adequate deals according to their skills and expectations.

This solution must scale to accommodate large population of workers and requesters. We thus have to demonstrate the feasibility of automation of task allocation in such a diverse universe. It is also necessary to describe the way guilds can operate, that in particular in order to capture and validate member skill description.

  • Hypothesis : we consider that worker skills declaration is organized in the context of guild membership. The guild takes charge of the validation of these skills, so we do not have to include any particular check.
  • Description of a matching algorithm

Skill Forest definition: The system provides a skill ontology (big word to say a logical organization for skills from the more general to the particular).

E.g. Computer Science -> Software Development -> Python

This classification is not a tree, but a forest:

  • E.g. Computer Science -> Web Development -> Python exists
  • but also Computer Science -> Software Development -> Functional Programming -> Python

“Computer Science” is defined as a root skill (there is no more global skill that encompasses Computer Science in the ontology.

Skills are expressed as a proficiency level on a node. We expect than the deeper the node the easier it is to assess a proficiency level. It may be for example quite difficult to assess the level of someone in software development, but clearly it is simple to test someone on its knowledge in Python (task completions, tests, etc.).

So a skill profile can mention a level 3 Python capacity for example. When a level is assessed, it is also required to select a path in the ontology from a root skill associated with this assessment. Let consider this person has mainly used Python in the context of web development, the associated path with the level 3 capacity is then (Computer Science -> Web Development -> Python). We see that the level is not really rating a node, but a path in the ontology.

When a requester publishes a set of tasks, he defines a list of constraints. These constraints are again paths from root skills in the ontology with an associated minimal level for each path. What is a match ? There is a match if all the constraints are sub-paths of existing skills at the required level.

Example : a requester asks for a (Computer Science -> Web Development) @ Level >= 2

The worker is (Computer Science -> Web Development -> Python) @ Level = 3

We have a match as (Computer Science -> Web Development) is a subpath of (Computer Science -> Web Development -> Python) and Level is superior to 2 (3 in this case)

If the request had been (Computer Science -> Web Development -> Python -> Django) @ Level >= 2, this worker would not have been a match

Then, still considering the first constraint, if another worker is (Computer Science -> Software Development -> Python) @ Level = 3, again we do not have a match.

Wait a minute, the level for the worker results from a guild assessment while the constraint is defined by the requester. How can we be sure they share the same idea of a proficiency? We cannot. It is part of the competitive policy of the guild to tune its attribution of level to meet requester expectations. Of course, if we have only one guild, this will work the other way around, requesters will learn to adapt their expectation to the guild rating scale.

How do we introduce new skills? We must allow for evolution in order to follow the demand. The skill ontology can thus be extended at will by requesters. Of course, it will take some time for this evolution of demand to be reflected by an adaptation of the offer. As a footnote, I am more cautious regarding the capacity of workers to do the same, as it may open some unexpected capacity to game the system.

Why a guild? To work efficiently, the described system need to have a moderated system for skill assessment. With the explosion of possible paths, we consider not economical to centralize moderation. The guild structure provides a peer environment to control the correctness of worker declaration as it is in the interest of the community that requesters can select the adequate candidates. The satisfaction of requesters in term of work quality is a condition for their repeat submissions.

Historically, the guild structures have ensured customers of the proficiency level of their members. The apprenticeship concept played an important role in this respect. It is certainly something that can be adapted to help validation at some higher levels. At the same time, being a validating mentor can be a mandatory experience to reach the highest levels in a domain.

Guild Reputation: In a competing world of guilds, we may be back to the basic question of evaluating guilds themselves. One can consider this a lesser problem as the volume of guilds is one to two orders of magnitude less numerous than workers, but this can be a loophole due to their role in validating their members. This may be an opportunity to introduce Boomerang.

How to test the concept: A first social feedback must be organized to define the possible reactions of actual users (workers and requesters) in existing platforms. This feedback can take several forms: crowdsourced surveys, expert interviews,... We can then infer a model made of rules. The model can be simple enough to have analytical solutions, but more probably we would need to test its dynamics through the simulation of interactions.

Other considerations (on the guild structure, pricing, research protocol, etc.)

See discussion between @yoni.dayan & @trygveson about the subject here


  • (PierreF) Behavioral Mechanism Design: Optimal Crowdsourcing Contracts and Prospect Theory. David Easley, Arpita Ghosh. Proc. 16th ACM Conference on Economics and Computation (EC), 2015.

  • (PierreF) Galen Pickard, Wei Pan, Iyad Rahwan, Manuel Cebrian, Riley Crane, Anmol Madan, and Alex Pentland. Time-critical social mobilization. Science, 334:509–512, 2011.


  • (PierreF) Eleni Koutrouli and Aphrodite Tsalgatidou. Reputation Systems Evaluation Survey. ACM Comput. Surv. 48, 3, Article 35, December 2015.

DOI: - File:Reputation evaluation survey A35-koutrouli.pdf

  • (PierreF) S.R. Epstein. Craft Guilds, Apprenticeship, and Technological Change in Preindustrial Europe. The Journal of Economics History, Vol. 58, N°3, September 1998.

File:Craft guilds apprenticeship and technological change.pdf

Links to Experiments Page(s)


  • @trygveson
  • @arichmondfuller
  • @Amdp
  • @PierreF
  • @Ferlin87
  • @Anotherhuman
  • @yoni.dayan