The Surprising Problem of Too Much Talent | Hacker News Reader

The Surprising Problem of Too Much Talent

scientificamerican.com

38 pointsjonbaer43 comments

I always roll my eyes internally when I hear people talking about how they only hire A-players.

The way I see it, you want A-players in certain positions where their performance and influence drives the company forward - "moves the needle", to use a clichė. You don't want A-players in mundane/support roles because they'll end up getting frustrated and likely become disruptive (the exception is when you put an A-player in a support role so she can learn the ropes/basics before you move them to the role you actually want them to fulfil).

Doesn't that perfectly describe Google? I mean, plenty of A players, not enough A player work to go around, at least what I have read from Michael Church.

Sounds exactly like Google. Here is this really cool new thing we built, but the work to finish it and provide support is boring so it doesn't get done. Fast forward a couple of years and they kill it saying no one used it and it provided no revenue.

I would like to see corroborating evidence from a more reliable source.

Some nuance should be applied here: this is usually said in the context of startups.

In the first few years of a startup, there's no such thing as a mundane support role. The first person you hire for support is your future "VP Customer Support". The first few people will form the support culture and processes of a company for the future.

In this context, all hires should be A-players, because they are supposed to contribute way more than just fulfil the role in their job description.

There are people who are passionate about doing X. They know how to do X very well, might even be a world expert in X. Now, you might even have one of these people at your startup, or maybe not because it can aponly accept jack of all trades at the beginning? Also, keep in mind that advancement in technology typically occurs through specialization, not generalization.

So do you really want your machine learning star working on customer support? If they spent 8 years a PhD to become an expert in that, how would they feel?

If someone becomes disruptive when they get frustrated, then they are not an A player.

Part of being an "A player" is to be professional, including in their communications and behavior. IF they are having problems with their role, speaking to someone about it and getting the problems resolved is appropriate, not disrupting the business.

This sounds like a fantasy. Everyone has different likes, dislikes, and motivation levels depending on the environment they're in. Also, "professionalism" is mainly a stick management uses to beat up on employees when they neglect to align their incentives with the company (because it's expensive and possibly tricky to give points on the package or real bonuses).

They usually get frustrated and leave the company. Also, not all A players make great people managers, which is to say the company went for A-level engineering talent but not A-level people management talent.

That's not the definition of an A player, that's the definition of a unicorn. There are very few talented people who remain productive if they don't feel they're valued / doing valuable work.

I said nothing about remaining productive. Of course productivity suffers along with morale. That is not the same thing as actively disrupting a business. Quitting and/or slacking off are common reactions. But if someone gets passive-aggressive and/or otherwise actively disrupts a team? That makes them toxic. And no matter how talented they are, a toxic individual is not an A player.

Most businesses that pride themselves on hiring only the "best" are of course not doing so, but merely using such phrases to boost the morale of the mediocre body of wage slaves they have on staff. If they truly hired the best, their salaries would be the best. The harsh reality is that most people are close to average. And that's OK. You don't need to be a genius to churn out the next incremental update of a SAP IDOC.

I really don't agree with this. I think a lot of people can be exceptional alone, if they direct their efforts towards that, more so if they have coaches, support structures, etc. But it similarly requires a kind of intelligence and the same levels of a different kind of effort, to work effectively with a group.

It's a kind of collective intelligence, almost like trusting your arm to always throw the ball in the way you have practiced, you have to trust that your team mates will be able to compute or think or calculate in ways that you simply don't have time for, or don't know how to know how to do, because there aren't 10 yous, and even if there were 10 yous, you'd probably be blinding yourself half the time you try to lead yourself.

The difference between a good director / lead / manager and a bad one is stark.

I don't think it's hard to turn a single person into a star. I think it's really hard to keep people motivated on a unified goal, keep all those people thinking about what needs to be built, while none of that thinking contributes to egocentricity, passive aggressiveness, defeatist attitudes, etc. There is so much conflict that creates tiny little cracks through an otherwise solid foundation of skill, but all those cracks can add up. Do you spend that five minutes solving a problem with your SAP coworker or do you spend that five minutes grabbing coffee and grumbling to yourself about what an idiot that guy is? For people who are aware of the cracks and go out of their way to fix them, instead of stampeding over others with their supposed skill and intelligence. I mean I would be an idiot if I ignored all the times my ego got in the way of my success, and I would likewise be an idiot if I ignored the fact that I must be aware of all of my social inadequacies, and I must be willing to continue to learn from them, if I actually want to succeed.

Working in teams is completely different from working alone, and working in teams is more than being able to do your part. Working in teams while programming is being psychologically extremely close to all your teammates, whether you realize it or not. There's some stuff that you just have to be able to sync together with and then discard, without even really realizing what happened. Half of the things people wind up arguing about, those things wind up being hilarious when looked at in retrospect, if they are overcome. It's hard to even understand how one could both simultaneously suppress and ignore their social emotional connectivity while acting as though it does not exist, when it clearly does. We like to think we are these efficient, effective, calculating machines, but we aren't. This is a big reason as to why work experience is more valuable than intellectual or educational experience.

I really think the mediocrity of development occurs with otherwise extremely intelligent people, people capable of solving huge variation of technical and conceptual problems, it comes from those people continuously thinking that it is their intellectual and technical skill that needs to be constantly improved and updated, rather than confronting how much time is literally wasted on poor communication, difficulty resolving social conflicts, and rolling your eyes at every motivational, team building activity. You either like the people you work with or you don't, and some people are unfortunately very dependent on jobs they chose out of necessity, rather than jobs they chose because those jobs and those people make that person happy. Whether you cut people some slack when they need it, you apologize when you make mistakes, or you ignore those things and act like a mini tyrant, these things make big differences. It's often not when we are at are best that determines the success of a team, but much of the time when we are at our worst.

I think it's easy to be talented when you get to decide what to be talented in. But most of the time you don't get to do that. If the project is to make a secure desktop application communicating over the internet in a certain time frame and you aren't effective at working in teams, network security and/or user experience etc. you can claim to be as talented as you want, you probably won't deliver good results anyway. You probably can't even properly judge what a good result is.

This reminds me of the argument from Chasing Stars by Groysberg. It is very difficult to separate an individual's performance from that of their environment. A lot of factors come together, including individual skill, to influence someone's performance. If you take that same individual and dump them into another environment it is very difficult (or impossible) to reproduce comparable levels of performance. Applying this to teams, there are a lot of factors necessary for them to generate a high level of performance that goes way beyond the individual talent/knowledge of each developer. That's why its necessary to hiring needs to be about more than just knowledge but personality and cultural fit as well. Then you need to look at all the support structures that you mentioned.

It's like the term "competitive salary" in a programming advertisement. Sure. If you had a competitive salary I'd be hearing about you from my friends, not reading your job advertisement.

I can see how this can be true. I've seen a team of "star" programmers fail massively on what should be a simple project. They over designed, they used all the new and latest, framework, language, plus spent all their time reinventing the wheel. They forgot about time/budget constraint.

Meanwhile, a much less talented scrappy team, just figures, "Hey!, we ain't so smart so we are going to get all the help we can" No new technology, just small established libraries and simple design, KISS, and what do you know? they shipped!

You can load up your sales team with talented sales agent and see massive output, they don't need much team interaction. But a dev team? If they are super smart, the problem better be tough or they will invent new problems to solve.

...and Agile is mostly about restraining enthusiasm to small, testable efforts. Which causes much moaning and misery among the most talented developers.

What we need is, some kind of process that sorts the ambitious parts to the talented/experienced devs, and the rote jobs to the new folks. Let the horses run!

I'm not sure how this applies to football (soccer). As far as I can see, Barcelona have the most superstar team in world football, and are winning everything. The next two teams with the most superstars (Madrid and Munich) are winning everything else.

Also, isn't this impossible, too? A team that wins everything over a long period will by default have their players considered superstars.

Well, yesterday Barcelona lost 4-0 against Athletic de Bilbao who won't/can't sign players from outside the Basque Country.

Yes, the 2015 La Ligua winners, Copa de Rey winners and current European champions have indeed just lost a single match.

I'm not saying they're invincible, just very good.

Barcelona has an undeniable king above all superstars, Messi. This give perspective to all others, Neymar and Suarez just know Messi is better. And the last batch of BCN superstars were formed by stars whose play style was inherently teamworking: Iniesta and Xavi were midfielders whose biggest strenght was prepare the game for others.

And Barcelona have always had its share of average, but impressively dedicated team players, like Puyol, Piquet and others.

Real Madrid isnt the same and do not have the same performance as well.

Puyol was by no means an average player and neither is Pique. These are players who were/are considered to be among the top five center backs in the world at their peak. Isn't that by definition a superstar?

Yeah there's limit to the logic, but it's also known that summing the best parts (in isolation) mean you will have the best product. The opposite is also true, Messi doesn't shine as much with the national selection, he needs his other team mates to be at its best.

> A team that wins everything over a long period will by default have their players considered superstars.

I was about to conclude like you.

I find the conclusions made by this paper to be dubious. Perhaps the reporting is weak and did not properly delve into the methodology used by the authors of the paper to draw their conclusions. But based on the article, couldn't it be possible that teams that have the higher percentage of superstars opted to load up on star players and therefore had less resources to flesh out the rest of the team. What would be the point of having let's say five or six world class players in soccer if you can only recruit below average players for the rest of the positions.

It would have been much more 'surprising' if the research had shown the opposite to be true.

Why? Is it generally a known fact that you don't want a team of superstars?

I thought it was.

From sports teams to software engineers, people with exceptional talent also seem to have exceptionally large egos to match. A team of normal people can usually tolerate & use one of them to their advantage, but a whole team of super-egos? What would happen if _everyone_ working on the linux kernel had the skill & personality of Linus Torvalds? There's no way that team would function. :)

You are confusing talent and personality. You can be untalented with a difficult personality and vice versa.

He's not confusing talent and personality. He clearly wrote that there seems to be a correlation between the two: "people with exceptional talent also seem to have exceptionally large egos to match"

Yes, but I don't agree with that hypothesis. From what I've seen, ego tends to peak around the "above average" range where they are trying to convince everyone (including themselves?) that they are extraordinary.

The truly amazing people demonstrate it in action not words.

A team of superstars is fine, but you can't have a team of all leaders. This is where personality comes into play. The team has to let someone lead. If you can find a small group of really talented people who also have their egos in check to follow a lead, amazing things can get done.

One of the benefits of microservices and highly modularized software is that you can have a team of superstars and still give the individuals the challenges they crave.

Interestingly, both the study and the article in question don't mention one of the most famous examples of an "all-star team," the 1992 U.S. Men's Olympic Basketball Team, who won gold and led at an average of 44 points per game.[1]

The study measures success over a period of many years, but perhaps there really is no substitute for an "all-star team" when you need to get shit done in a short period of time (as in startup culture).

[1]https://en.wikipedia.org/wiki/1992_United_States_men%27s_Oly...

I would argue that this is deeply correlated with the reward structure. If you only have rockstars but they are rewarded must strongly due to individual accomplishments they will tend to lower the team play, if instead the rewarding is more calibrated twards team achievements, then there will be much less worriying on who is getting credit for what. Thus increasing collaboration. (I am assuming that the entire team is mainly A players and Bs are filtered out, thus there isn't too much variance in individual productivity)

You want to gather the most talented and hardest working individuals with the combined skills that you need that can also work well together as a group. Sometimes skill trumps group dynamic and sometimes group dynamic trumps skill. Your job when assembling a team is to make those judgement calls. It's not difficult to understand.

One of the interesting corollaries to this is that the right person for a role at one point in a company / product is not necessarily the right person at a future time.

For example, the rogue / maverick type can be very important at early stages. They drive innovation, take risks, and get absurd amounts of things done. But, later that same risk taking can become absolutely destructive.

There is a good reason why millions of years of evolution has created a minority of "talented" people. Societies need super-talented people and leaders (these are often not the same people, btw), but it also needs legions of pedestrian followers. Troops. Individually, each troop is not very impressive, but as a collective, they are what makes a vision happen.

Here's a recent example of an overtalented hornet's nest: https://www.trumid.com/Team.aspx

I give them 6 months.

Evolution doesn't optimize for the species, though.

I'm not talented enough to understand what that means... please elaborate.

Yeah, sorry, that was confusing. I guess I was just trying to point out that for behavior to arise that is beneficial to a species (like the eusocial behavior in bees), it has to increase fitness for everybody, otherwise that kind of behavior would never evolve. If you want to draw the parallel to creatives vs. workers in the modern workplace, I don't really see that analog or metaphor (or literal interpretation for that matter) working very well – one half gets all the money and recognition, the other half gets zilch.

It just seems very unlikely to me to posit that human talent is variable because of genetic specialization within the species. It's much more likely that talent is influenced by thousands of genes, and lots of different combinations lead to lots of different outcomes. Not a biologist though.

I agree with this. However if you accept that the circumstances of a society's existence are highly variable too, that each of the many different genetic outcomes you speak of is relevant only under very specific circumstances, then we could have a situation where a "talent" is viable at some times, and not viable at others. In that case, we have an evolutionary interest in maintaining many different talents, many of which will remain latent/unused, in our multi-individual society, until circumstances wish?

Furthermore, and unrelated, it is also my personal opinion that highly "talented" people in the definition that society places on the word, often have deficiencies in terms of other prerequisites for success. It's a bit of a zero-sum game in some ways. I happen to be an excellent mathematician, but I massively envy the (far less inspired) discipline that one of my colleagues has allowing him to be much more effective in the day to day grind of business. Basically, I'm highly talented by most definitions, more than him (again by the societal definition) but I'm not very effective at making money in a world where that seems to be the scoreboard. Try as I might, I just do not have his work ethic. I just cannot do it. So who has the talent? Him or me?

This reminds me a comment about the last World Cup: Argentina had Messi, Portugal had Cristiano Ronaldo, Brazil had Neymar, Germany had a team.

"If I have seen further it is by standing on the shoulders of giants." --Newton

Reply on news.ycombinator.com