Making Moneyball Work

Screen Shot 2014-02-20 at 10.45.35 PMBreaking News: (July 16, 2036) The national Comprehensive and Reliable Assessment of Performance (CRAP) database reached its goal of 100 percent coverage with the last employer — Roto Rooter of Northern Idaho — getting connected to share employee performance data. Employers nationwide now have a central resource to evaluate candidates for jobs, using the concept of Moneyball that was developed in the late 20th century. The database, established by the Dream On Act, is administered by the BUFFOONS (Bureau of Unreliable and Freely Flexible Or Objectionable Numbers and Statistics) at the Department of Labor.

Maybe this will come to pass, but don’t hold your breath and be careful what you wish for. Let’s think about what it’ll take to make Moneyball work.

First, thanks to all that commented on my last article on the subject of Moneyball and suggested an ongoing conversation. The key is having “good, common, open data” (thanks Daryl Clements). Even if the data was available, we would need common standards for it to have any relevance. That is highly unlikely to occur, though there are precedents. In India, the software industry has established the National Skills Registry — a database of registered and verified knowledge workers in the industry.

Big Data and Moneyball

Several commenters pointed out that Moneyball is more about discovering undervalued assets (players/employees) than just using statistics to hire employees. This is where a big data infrastructure is likely to help in discovering nuggets of information buried in social networks and other sites, to suggest that a candidate has more (or less) to offer than is revealed in a resume or from interviews. For example, a chef whose recipes were posted on Pinterest; a customer service manager who promptly responded to complaints through tweets; or from reviews of work-related books or products posted by a candidate on Amazon. However, these pieces of information will never reach the level of objectivity or comprehensiveness achieved by baseball stats. It’s more likely that such data will help in sourcing, by revealing candidates who were not identified from common sources — the real undervalued assets.

For employers, statistics can also help reveal where they are most at risk for turnover. The SAS Institute has done pioneering work on this. Combine that with data like recent additions of references to LinkedIn profiles or comments on Facebook about other jobs and an employer can better plan for turnover and how to focus their recruiting strategies. Call it reverse Moneyball.

Article Continues Below

Where the concepts of Moneyball can really help are in changing the conversation around how to evaluate candidates. Billy Beane — the manager of the Oakland A’s — challenged the conventional wisdom about statistics typically used to gauge players, focused on speed and contact — such as stolen bases, RBIs, and batting average. He found that on-base percentage and slugging percentage were better predictors of success. That flew in the face of conventional wisdom and the beliefs of many baseball scouts and executives. Expect the same as a recruiter when you want to use a profile based on big data and confront a hiring manager who goes by gut feel or claims he can evaluate a candidate from their handshake.

Moneyball concepts are already being applied to college admissions by matching students with colleges. ConnectEDU is a company that uses data on students from seventh through twelfth grade to do so. They also are trying to do the same for employers hiring new grads. This is very feasible because the data is standardized, comprehensive, and accessible. Regrettably the data stops being all that once the new grad is hired into their first job.

We may have to wait for the buffoons to get their crap together before Moneyball becomes more common among employers.

Raghav Singh, director of analytics at Korn Ferry Futurestep, has developed and launched multiple software products and held leadership positions at several major recruiting technology vendors. His career has included work as a consultant on enterprise HR systems and as a recruiting and HRIT leader at several Fortune 500 companies. Opinions expressed here are his own.


7 Comments on “Making Moneyball Work

  1. Good God, Raghav, that’s the wrong way to open an article before I’ve had my coffee in the morning.

    There are reasons for optimism and pessimism on this front. Something like standard gauge in railroad history reveals that companies can cooperate when necessary and in their interest to develop standards for interchangeability. My guess would be that common standards for measuring the performance of employees are a long way off though, and complicated by the different practices and cultures of various workplaces. That employee A might be a good match for Company B and not Company C because of the methods both companies use is something that is implicitly acknowledged, but I think people would have trouble acknowledging it explicitly for various reasons. Especially when the end product/result both want is seen as identical for all intents and purposes.

  2. Raghav,

    Beane’s application of Moneyball was because he was looking for undervalued talent based on hard data. You cannot overlook that one incredibly evident and major shortfall of his concept’s application is that it failed to keep players over the long term and jettisoned talent as soon as it became budget-risky. Beane’s principles work splendidly when undervalued, talented players are “stuck” in their existing contract or situation. It almost guarantees those same players may skyrocket in value at the next opportunity.

    For players, it’s when they can negotiate a new contract. For business, it means any time someone undervalued becomes a known entity (and known can even mean to her-/himself). No company’s going to simplify the process of identifying and exchanging talent. The markets and investors would punish that company immediately.

    I’m not suggesting Moneyball’s useless. I’m suggesting that it not be taken where it cannot reasonably go.

    Other teams learned from Beane’s practices, but it took just a few years of application for them to literally start bidding on the undervalued again. Today’s baseball salaries and rosters are the highest at any point in MLB’s history. Moneyball wasn’t the reason – it was just an enlightening spark. The “winners” in baseball quickly returned to and have largely been the “spenders.” Even worse for the sport, mid-markets are now spending at levels that aren’t sustainable. So every few years there are smaller markets that basically gut their rosters to get expenses back down – only to go back on the spending spree when the conditions for winning look better.

  3. Thanks Raghav, Richard, and Daryl.
    I believe that an approach similar to this would be greatly beneficial to almost all stakeholders, EXCEPT FOR THOSE MAKING OUT VERY WELL THROUGH THE CURRENTLY INEFFICIENT AND DYSFUNCTIONAL STATUS-QUO. There are too many VPs of Talent Acquisition at too many large and/or prominent companies upholding their successful careers, too many well-healed sr. recruiting consultants making up what the previous category wants to hear as great and received wisdom, and too many 3PRs and technology vendors making lot$ off the Greed, Arrogance, Fear, Ignorance/Incompetence, and Stupidity (GAFIS) of those at the top. I’m afraid that a powerful jerk in charge beats solid, objective evidence 7 days out of 7. Until there is a strong incentive for change in this direction, I expect much more of the same…

    Happy Tuesday,

    Keith “Mr. Optimism” Halperin

  4. Doesn’t seem inappropriate to me. I give you a hypothetical two consultants, Consultant A and B. CEO C, asks A why he can’t get and retain talent, he replies:

    “Your talent brand is sub par, we need to beef it up with more social media presence, a talent community for PASSIVE candidates, and we can let everyone know what a wonderful and innovative place this is to work!”

    He poses the same question to B, and he replies:

    “I surveyed your salaries, you’re below market and everyone who has left for the last two years has gotten a minimum 25% increase by switching jobs. You smashed into three employee’s cars in the last week alone while parking, and they feel threatened in that if they go to their insurance to cover the damage, you’ll fire them. Also your company has an incredibly poor online reputation. Your Glassdoor score is 1.7 with 75 reviews, all of the reviews mention poor management and poor treatment, specifically constantly shifting goals and timelines, no planning, nepotism, favoritism, and screaming at employees in public and private. You need to start treating people better or you won’t get or retain the best.”

    Who do you think the CEO is going to believe and go with when it comes time to get some help with his recruiting? Incidentally, the above scenario B outlines is an actual one at a company I’m aware of.

  5. Thanks again, Richard.
    As we know from Rule 33 of the Consultant’s Code of Ethics:
    “It never hurts to suck up to the client.”


    Keith “Occasionally a Recruiting Consultant” Halperin

  6. Raghav,

    Applying Moneyball concepts and creating data linkages from education to workplace is an interesting concept. As the Chief Business Intelligence Officer at ConnectEDU, I can attest that we’ve experienced success in helping connect students to colleges and careers that best match their skills and aspirations, using both quantitative and qualitative data. And I agree that our success in the P20 (pre-K to higher ed) environment is due, in part, to the standardization of data in that context. As we work with postsecondary institutions to help connect students and entry-level candidates to best-fit career opportunities, we’ll need to explore ways of continuing that data linkage for the benefit of the student, institution, and employer. We’re already starting to do so with our career-focused solutions, which you can explore in more depth on our website:

Leave a Comment

Your email address will not be published. Required fields are marked *