Data Scientist: Potential Superstars in Prediction Markets

Big Data Evangelist, IBM

Prediction markets are where data scientists will attain superstar status. It’s no coincidence that the current age of the “superstar” in professional sports began in the 1970s, when the legal constraints that had prevented the most accomplished athletes from seeking top dollar on the open market began to fade. Nowadays, every self-respecting sports star counts the days till they can truly cash in as a “free agent” while they’re still in the pink of their career.

I tend to giggle when I hear that data scientists are “superstars” or “rockstars” or something to that effect, which implies that they’re well-known public figures. In the real-world, most data scientists are on the payroll of one or another private company or government agency, and are little known outside those organizations or the professional groups to which they belong. And there’s fundamentally no problem with that. They’re not slaves, and, to varying degrees (subject to noncompetes and all that) they may seek employment elsewhere.

The prospect of superstardom for data scientists depends on the development of a highly visible, prestigious open-market for their services. This market may in fact be emerging with the development of prediction markets, which deliver on-demand access to industry- or domain-specific, analytics-driven predictions.

Prediction markets have taken root in electronics, telecommunications, manufacturing, automotive, retailing, pharmaceutical, banking and other sectors. Most such initiatives rely heavily on crowdsourcing of predictions and sentiments expressed by disparate individuals and organizations in the industries of interest. They also rely on analytic infrastructure, both to aggregate, filter and distill the crowdsourced “wisdom,” and also to auto-generate fresh predictions from this and other relevant data. And, of course, they depend on staffs of domain-knowledgeable data scientists to develop and maintain the filters, as well as train and guide the underlying segmentation, predictive and other statistical and linguistic models. Essentially, the data scientists are the core assets of these marketplaces, which, potentially, can become the “big leagues” of business-oriented data science. These marketplaces depend critically on data scientists’ predictive slugging percentages to knock business results clear out of the park.

In addition, open-source data-science expertise marketplaces such as have created something resembling free agency for ambitious statistical and predictive modelers. To the extent that data scientists (individually and in teams) can build powerful models that win the monetary prizes in these contests (which are typically sponsored by a particular client that wishes to outsource some tough data science challenge), they can advertise their stardom while enriching their bank accounts.

Superstardom means little if you can’t monetize your status. The development of a truly competitive open marketplace for data scientists will boost the best of them to a new, more prestigious, potentially lucrative plateau in their careers.

What is a data scientist?

See some of James' other posts on data scientists