User:OrenBochman/WGT/Utilities

Edits - The Stakes

Lets our imagine our four editor working so if L =2.5 and an avarage edit changes 30 D_F and 4 D_P

(Links between Article A and B can be viewed as allowing A to depend on B, essential importing it's memes without gaining length. But the contribution of linked memes is reduced by the number of links they are removed from each other.)

Creating a new article produces U(A) and is likely to introduce new completely new memes into the wiki.

By editing an exiting article an editor contributes U(S(A)-U(A) in each revision. Where S is the successor function. goal D_T and D_F may be altered. Since this is not a domain expert almost always D_T will become D_F.

To a domain expert D = D_T + D_F and he can increase U(A) consistently by either introducing more D_T or removing D_F. To a policy expert D_T and D_F look the same and so he can increase U(A) by increasing P_A towards Sum P. However in pursuing this

The reviewers dilemma is as follows: An editor makes a revision, the first reviewer must then decide to accept or reject it.

	Editor D (co-operates)	Editor P (defects)
Editor D (co-operates)	Each gets U(S(A) - U(A)	D get 0 P gets ()
Editor P (defects)	Editor A: goes free Editor B: 1 year	Each serves 3 months

Information assymetry

In the context of a wiki we diffrentiate between domain memes $D_{1}...D_{N}$ and $P_{1}..P_{N}$ policy memes.

the lack of of knowledge in a domain the degree of knowledge in a domain is the is definde by

• To a domain expert $D_{A}=D_{T}\bigoplus D_{F}$
• To a policy expert $D_{A}=D_{T}\cup D_{F}$
• This diffrence is the primary source of information assymetry in the game. Most domain requires years of study.

Let the $P$ policy be described by Policy memes and violations $P_{1}...P_{N}$.

• To a policy expert $P_{A}=P_{T}\bigoplus P_{F}$
• To a domain expert $P_{A}=P_{T}\cup P_{F}$

Notes:

• As ${\left\vert P\right\vert }$ increases there is a new role called a policy expert.
• This diffrence is a secondry source of information assymetry in the game. Learning to edit a wiki takes weeks of study.
• so ${\left\vert D\right\vert }>>{\left\vert P\right\vert }$
• overt a year change ${\left\vert P\right\vert }$ can change dramaticly, but ${\left\vert D\right\vert }$ is constant.

Utility of Article and Edit

Here are five stereotypical Utility functions:

• Wise - Objective utility function (information symmetry) or policy savvy domain expert

Failed to parse (unknown function "\begin{equation}"): {\displaystyle \begin{equation}\label{eq:someequation} U_W(A) = \frac{ \sum{D_T} - L \sum{D_F} }{\left\vert A \right\vert} . \frac{\sum{P_T}- L\sum{P_F}}{\left\vert P \right\vert} \end{equation} }

• DExper -the Domain Expert's utility of an article

$U_{D}(A)={\frac {\sum {D_{T}}-L\sum {D_{F}}}{\left\vert A\right\vert }}.{\frac {\sum {P_{A}}}{\left\vert P\right\vert }}\qquad \qquad \qquad (2)$

• the Policy expert's utility of an article

$U_{P}(A)={\frac {\sum {D}}{\left\vert A\right\vert }}.{\frac {\sum {P_{T}}-L\sum {P_{F}}}{\left\vert P\right\vert }}\qquad \qquad \qquad (3)$

• the Noobe's utility of an article

$U_{N}(A)={\frac {\sum {D}}{\left\vert A\right\vert }}.{\frac {\sum {P_{A}}}{\left\vert P\right\vert }}\qquad \qquad \qquad \qquad \qquad \qquad (4)$

• the freelader -

$U_{B}(A)={\frac {L\sum {D_{F}}}{\left\vert A\right\vert }}.{\frac {\sum {P_{T}}-L\sum {P_{F}}}{\left\vert P\right\vert }}\qquad (5)$

• the vandal -

$U_{W}(A)={\frac {\sum {D_{T}}-L\sum {D_{F}}}{\left\vert A\right\vert }}.{\frac {\sum {P_{A}}}{\left\vert P\right\vert }}\qquad \qquad \qquad (6)$

where L is the lemon coefficient and is greater than 1. also we have omitted the advantage conferred by links.

Sorting Gold Nuggets V.S. Lemmon

Lemmons are used in the sense of The Market for Lemons

There is an overall lack of ability to differentiate good information sources from bad ones. This includes good citations, from bad citations. Good external links from bad ones, reliable sources from unreliable ones. Using and appealing to domain experts to check and correct entries is one solution. However can these academics be asked to do so in good faith - if any ignoramus may revert their corrections. There is generally no way to assess the level of fact checking that has gone into an article.

The problem is that in every level a few bad lemons editors can drive out the really knowledgable ones from contributing.

One such situation involves say a high-school student who knows the rules of style but nothing more wishing to make content more encyclopedic. By reformatting the information to fit a more factual style, Good prose can become a mish mash of factoids that end up mentioning details in an inaccurate way. Inaccurate means wrong.