Hiring Strategy: Researchers proposed a model to maximize success in professional recruitment

When it comes to hiring new employees, big companies often have to choose from hundreds of candidates, a process that requires time and resources. Can Mathematics help to streamlin these procedures? At least in the widespread sense, perhaps yes.

A paper published in Statistical Mechanical Journal: Principle and Experiment By Pavel Krarapivski, a statistical physicist at the University of Boston, proposes an algorithm that identifies three hiring strategies, corresponding to different purposes of each company.

Krapivsky took inspiration from the famous “Secretary problem” or “optimal marriage problem”. In one of its several versions, a princess will have to choose her future husband from a pool of 100 candidates at a grand reception. However, strict rules are applied: she can meet only one suit at a time and has limited time to know it.

At the end of each encounter, he must immediately decide whether to accept or reject the autism. She cannot see the previous candidates again, nor can she ask any of them to wait while she consider others. How can the princess make the best option?

The secret lies in a number: 37, to be accurate (if you think about 42, extend your hand). “If we divide 100 by 2.718, which is the number of eullars – one of the most famous in mathematical history – we get around 37,” Krepivski explains.

Practically, this means that the princess should evaluate and reject the first 37 candidates, keeping an eye on her quality. Starting with the candidate No. 38, he should select the first one who is better than all he has met earlier. According to krapivsky, this strategy guarantees the best possible results under obstacles given.

The method is so reliable that Johannes Capler is also rumored – although there is no concrete evidence – it has been used to select his second wife. “He studied the problem in detail, spent a year to do so instead of his own great research, and then made an alternative,” recalls Krarapivski.

Krapivsky improved the problem in a more modern context, applied to hiring practices in large companies. The original idea is the same: The company has a single parameter to assess the quality of a candidate and should decide whether they have to hire immediately or reject them without reconsideration. In addition, in this model, new work employees cannot be dismissed.

“I don’t like to firing people,” cribski jokes. Contrary to the secretary problem, the section of candidates here is constantly and potentially infinite, making the model more realistic to modern workplaces, where decisions are made on the basis of immediate business requirements.

The study examines three separate recruitment strategies:

• The maximum improvement strategy (MIS) decides that a candidate is hired only when his score is higher than any rented employee.
• The average improvement strategy (AIS) allows a candidate to be hired if their score is more than the average score of all current employees. On the other hand, the local reform strategy (LIS), each candidate is being evaluated by a randomly selected employee or a small hiring committee and is hired only when his score is crossed by interviewer or all committee members.

Unlike the problem of optimal marriage, no one is the best strategy – in addition, the option depends on the company’s purpose. If the goal is to maximize long -term quality, MIS is the best approach, but it is the result of hiring slow. If priority is to balance quality and speed of hiring, then AIS is a proper agreement. If rapid work is more important than quality, then LIS is the most effective strategy.

“Of course, these are simplifications,” Crappive Note, “but they can still be useful.” For example, a model presented in paper, can serve as the foundation of algorithms used in social networks and digital platforms.

These include not only platforms designed for job discoveries, such as dating apps such as LinkedIn, or Tinder, which match future suggestions based on previous “swipes”, but also those who control material selection, resource management and artificial intelligence.

“Many of these are actually based on very simple algorithms, such as we suggest what we see on YouTube,” Krapivsky’s conclusion.