How I Learned to Stop Worrying and Love the swarm.

... Or the vote is a popularity contest

There are several ongoing discussions on the nature of the vote, rewards voting, etc. In this post I will present the ideas that the main purpose of voting is to be a popularity contest, which focus on votes, as a quality measure is misplaced, that the current mechanism is quite effective, and concern for the swarm vote are misleading.

Popularity voting measures, not quality

Downvotes are widely discouraged, except to report abuse, which means that each job not abuse with at least one upvote upvote should have a percentage of 100%. It is evident that no more upvote count can be used to compare messages to each other on the basis of the votes. (The effect of the weighting of the vote deliberately ignored here for simplicity.)

Since the vote occurs only as an exit, it is obvious that the primary effect of the vote is the popularity evaluate. Popularity in some cases can be correlated with quality, but difficult to characterize ways. For example, in some contexts, it is possible that less popularity correlates with higher quality. (Example: Consider the comments of expensive products, but of excellent quality compared with less less expensive, but still good products cheaper quality inferior products is likely to get more comments ..) Moreover, as I will cover later while its popularity is objective, quality is subjective.
Trying to replace the vote in popularity with quality rating is an error

A classification system that effectively measures the quality, if one could be created, but also need a second method to measure popularity. Popularity criterion should be used as hot lists, lists trends, etc.; If the content is "high quality" but is not popular is placed on the hot lists, most people see that you do not like or, perhaps worse, not interested at all.

Any effort to put therefore a measure of quality should be careful not to discourage participation in the essential function measure its popularity, either by distraction or require more thought to participate at all. notation functions should be separated from the functions of voting, and will probably be submitted under any outstanding user interface.

Another reason the focus should remain on the measurement of the popularity is that people are very concerned about popularity, even when quality measures are also available. The most widely followed statistics on Youtube is counting view. Twitter is as follows. Both platforms also have quality measures (up and down-votes in Youtube and tweets as favorites on Twitter), but they get much less attention. In traditional media, books are widely promoted based on their past or present in the list of best sellers appearance. Movies have great weekends opening or breaking box office records are a big problem.
Popular, but not the quality, the objective

A measure of popularity can always be calculated and compared. Some people are more popular than others (most people like them), some music is more popular than other types of music (more people hear it, or maybe I heard it more), etc. Quality, on the contrary, is highly subjective. People can agree that it is more popular, while disagreeing in every possible way about who should be more popular.

Of course, in a summary really like to measure quality in order to reward quality. However, attempts to create added quality measures suffer many perverse results and paradoxes that are well known in the classification systems of study and vote. For example, most would prefer to B, but the particular aggregate measure may classify more B than A. In an added reward system would mean more money going into the content that most identified as the lesser of the two. This is not a fatal flaw, of course, no one expects no system is perfect, but it raises questions.

In all likelihood, the best approach to reward the subjective quality is falling. Meanwhile suffers from the problem of micropayments cognitive load, it is very difficult to imagine simple aggregate measures that serve this function effectively. One approach that could be explored is the use of a filtering mechanism of collaboration (recommendation) to identify new probable high quality content to a particular user and then automatically "tip" with a portion of the funds linked to that user a fund earmarked. This raises several questions and the additional complications that are beyond the scope of this post.

The popularity is very popular

Its popularity is highly concentrated in any system. The most popular personalities have much, much more Twitter followers that even celebrities least as -more- . The most popular videos on YouTube remain much, much more views than 10% , etc. Once the content becomes popular that often then become even more popular ( often called " go viral " ) . It can be inferred , in fact, that the same popularity is very popular and that like people can be famous for being famous, the content can be popular because it is very popular. When this also serves as a focal point for comments and discussion , this can also make the content more valuable .

Conclusions
We must stop trying to make some sense of voting as a direct measure of quality, they are not, they are measures of popularity. Quality measures could also be used, but they are secondary to the main axis of the vote, which is to identify popular content in a system of social media.

The current voting system and rewards designed, apparently, to both encourage and reward participation, and reach a consensus on the popular content is very effective in performing this essential function. a subset of content is identified as popular and reward people who produce this content (even by accident). swarm of voting is not only inevitable, but necessary, because it ensures that some of the content will be very popular, in turn satisfy the human need for people to connect and communicate based on common characteristics, even the most fleeting.

Although nothing forces people collectively to choose the best content to be popular (and perhaps no mechanism could), small influences probably serve as a turning point for better content generally being chosen more often generally worse content. This will tend to reward content that is of superior quality. more direct mechanisms to reward quality content includes the fall, or possibly with further development of the idea, the automated warehouse.

In summary, the main changes in the arrangements and existing voting rewards are not necessary or actually serve any useful purpose, which is clearly identifiable at the time. Refinements and adjustments may in fact be necessary, but insufficient data from a small base of initial users that exists today is totally insufficient to identify changes if any would be useful. A rating system to evaluate and reward quality directly can be a useful addition, but designing one that works effectively is a major challenge.