What Sports Can Teach Us About Evaluating Software Development Talent

Various professional sports take different approaches to evaluate talent, and yet the contributions of athletes in team sports is difficult to predict.  The NFL conducts a combine, yet they make gaffes such as drafting Tom Brady as a low draft pick.

Athletes perform in public and are watched by experienced scouts, yet the scouts did not accurately predict how well Duncan Keith would perform.

Billy Bean famously upended baseball scouting by switching to metrics that predicted success better than the subjective evaluations of scouts.

Is it easier to predict performance in individual sports, such as track and field, golf, or tennis?

Did Dan Marino ever win a superbowl? No. To what extent does this reflect his ability? Is it the quarterback's fault if a team does not have a good running game?

In business, we mostly perform in private and performance is much less amenable to evaluation with metrics. Also, business is largely a team sport.

The larger the team, the more likely that individual performance can be expected to vary from predictions.

Given the extensive advantages that professional sports teams have over traditional businesses in evaluating talent, and the errors that they make in those evaluations, what would lead us to believe that businesses have any hope of properly evaluating software developers?

Is there a way to structure work that avoids the requirement for accurate evaluation?  Is there a way to be robust or even anti-fragile to evaluations? There are articles that discuss how 10x engineers are only 10x in a certain context.

Agile is an attempt to be robust to estimation errors. Estimation should only be used to get in the ballpark. Maybe the same is true of evaluation? Should the goals of software developer evaluation vary depending on factors of the hiring team (size, maturity, experience with the problem domain, maturity of the problem domain pioneers/settlers/town planners).

It is funny to hear the scouting reports for athletes like Tom Brady and Aaron Rogers.  The scouts said many negative things about the people who were to become some of the best quarterbacks ever.

Cubicles are a High Tech Favela

My AVP (Area Vice President) told me today that my cube is a shared resource and needed to be centrally coordinated with the office manager.  This supposedly explained why it took 3 weeks for me to get permission to move to another cubicle when I was transferred to a different team.

This made me realize that I have no ownership, no equity, in the place where I spend the majority of my waking life.  It reminded me of the work that won Hernando de Soto the Nobel Prize in Economics.  He demonstrated that the reason for generational poverty in Peru is because the squatters in impoverished areas have no property rights and therefore cannot pass their property to their heirs.  They can build businesses and enjoy the fruits of the business while they live, but they cannot legally pass the title.  Thus, the equity dies with the founder.

This is also true when you work for someone else.  You can make a living so long as you are employed, but you take nothing with you when you leave and cannot pass any title to your heirs.

Of course, the AVP made the mistake of breaking the fourth wall by reminding me that I am not an owner.  All of the agents of a company should want the employees to behave like owners, which is easier if they feel some level of ownership.

How good should Adele's guitarist be?

If an act is a vocal act, the guitar and all of the other instruments are there to complement the singer, not distract from her.

Same with the focus for a product.  The main feature needs to differentiate the product.  The rest of the features are supporting acts that must not overshadow the main feature.

But if the guitarist was better, maybe you could get guitar fans to come to Adele concerts. But you would start to lose the vocal fans who don't want to sit through long guitar solos.  The guitar fans can go see Kenny Wayne Sheppard.

Adele has a good guitarist, but he knows how to support and complement the vocals. KWS has a good lead singer, but he does not overshadow the guitar.

If All of Your Developers Were Wearing Parkas ...

wouldn't you find a way to give them a warmer work space?

So when all of your developers are wearing headphones, why don't you find a way to give them a quieter work space?

Management is All About "And"

The essence of management is and.

• Short term and long term.
• Employees and customers and shareholders.
• New product development and maintenance.

Some people can focus on one part of the and (short term or long term), but the organization won't be successful unless the manager can manage all parts of the and.

Managing the and is subjective and messy and difficult and hard to measure.  That's why humans are required to manage organizations--perhaps with the assistance of a spreadsheet or a system or other tools. There are no simple answers or formulas.  

When is the Cavalry Showing Up?

Spoiler alert: Never.

If you want something to happen, you are going to have to make it happen.  Any time spent looking for the cavalry, looking for that perfect job, looking for someone to save you, is wasted.

As the generous folks at Rally Software say, "Create Your Own Reality".

Or as Seth Godin says, "Pick Yourself". (PDF he wrote here)

Wall Street Tries to Get What it Wants

Seth Godin wrote a post today titled Wall Street gets what it wants about what happens when a company goes public.  This brings several points to mind:

1. Founders never seem to understand that when they sell their company, they don't own it anymore.
1. After the sale, the founders have the money and the buyer has the company.  That's why they gave you that money.  Does the car dealership still own a car after they sell it?
2. The buyer will often say that they won't change anything after the sale.  They might even mean it.  There is usually a honeymoon period of stability, but things always change.
2. Public companies require growth to satisfy shareholders. But infinite growth is not sustainable. Therefore, investments in public companies are not sustainable and are suitable only for speculation.
3. In the search for growth, companies will often do stupid things that hasten their demise.
1. This is often because the best course of action is to "stay the course", but shareholders insist that you try something different.
4. Wall Street does not want reliable incremental growth.  Reliable incremental growth can be discounted to a present value and doesn't change.  You would be better off owning a bond.  People buy stocks because they are hoping for a surprise on the up side. Infinite up side surprises are not possible, let alone sustainable.
5. Over time, private or hybrid companies will replace public companies as investment vehicles.
1. We are already seeing the move to private equity.
2. Venture capitalists are extracting more the of the value from new companies.  This is demonstrated by the trend of large IPOs.
3. Hybrid companies such as Google have two classes of stock.  One class is for the founders who will continue to control the company, and one class is for investors who are allowed to come along for the ride.  These companies will outperform pure public companies because they can avoid much of the pressure from shareholders.
4. Further evidence is provided by Michael Dell taking his company private again.  He actually cares about the company.  That correlates well with success. He is now publicly commenting on the dumb things companies are forced to do by "activist investors".
6. Another variation of private companies is employee owned companies.  I really like this method of ownership because it aligns ownership with the people who have the most skin in the game (employees).
1. With liquid markets, shareholders can jump in and out of a company many times a minute.
2. Employees incur massive transaction costs in switching jobs.  Therefore, employees have much more skin in the game than the putative owners of the firm, which flips many of the theories of capitalism on their head.
3. Ownership of private companies is fairly illiquid, so owners of private firms do have skin in the game.  In fact, they may have higher switching costs than some percentage of their employees, which led to many of the theories of managing the firm, agency issues, etc.

Don't beat your goldfish because it won't fetch your slippers

Don't beat your goldfish because it won't fetch your slippers.

If someone is clearly not the right person for a certain task, don't keep punishing them for not performing.

Someone was asking me about a business situation where someone was neglecting something that I thought they should do.  She was surprised that I wasn't fighting this.  I responded that there was no point in fighting it, because the person didn't believe in it, wasn't equipped to do it, and wasn't interested in becoming able to do it.

Sometimes Bells and Whistles Aren't

People in product development will sometimes glibly refer to features as "Bells and Whistles", implying that they are not core features and can be removed from the product.

However, if you are designing a warning signal for a train crossing, the bells and whistles are fundamental to the proper functioning of the signal.  It's not always obvious which features are the bells and whistles.

Don't Do Anything That Would Make a Good Interview Question

Interview questions are designed to be tricky to separate the wheat from the chaff.  Doing things like this in your daily work probably indicates that you are being too clever.

If something would make a good interview question, it means that many people will not be able to maintain your code.