In a world that runs on ceremonial expectations of optimal performances, but where it is rarely in your best interests to actually deliver optimal performances, practicing mediocrity necessarily involves capability masking: the act of hiding the true extent of your capabilities.
Capability masking is the opposite of “fake it till you make it” behavior, and comes in two varieties, illustrated below, both of which are involved in the behavior commonly referred to as sandbagging.
Capability masking has to be done in a subtle way. You can’t just pick a suboptimal performance level that’s in your own best interests and then nail it precisely without breaking a sweat. Sandbagging is an artistic performance, not a throttle setting, and it’s worth learning to do well.
Controlled precision in hitting a performance level is a clear tell that you’re not going all out to the edge of your performance envelope. That would belie the prevailing power narrative built around the counterparty’s right to expect optimal performance, and your obligation to deliver it. It might set up a challenge to power and authority that neither party wants to navigate at that moment.
So you have to signal an optimal performance capability that’s lower than your actual capability, and also slightly different from your desired performance level, and then miss the promised level and hit the desired level, by either undershooting or overshooting. The nature of the miss supplies the fodder for rationalizing performance that’s different from what has been ceremonially promised.
In Type 1 capability masking, you undershoot the promised performance level (often with deliberate, visible mistakes) and offer appropriate apologies and excuses.
In Type 2, you overshoot the promised performance level, while still falling short of your actual capability, explaining the excess performance away as non-repeatable luck.
Which you pick depends on 3 variables: the actual performance level that is in your best interests to deliver, your estimate of the actual limits on the counterparty’s power to coerce more performance out of you, and the level of expectations you want to set for the future.
In both the cases illustrated above, the actual depth of your reserve capability is being hidden from the counterparty, and the actual limits on the counterparty’s power to extract performance are being hidden from you.
The ceremonial setting of mutual expectations, consistent with nominal rights and obligations, serves to prop up a performed narrative of relative nominal power. The actual delivered performance serves the immediate functional need, and updates an uncertain, evolving mutual assessment of actual relative power.
Note the information deficits on both sides. They may be aware you’re masking your capability but will be uncertain about the extent of the masking. You on the other hand may be aware they are overstating the extent of their power to coerce more performance out of you, but will be uncertain about the extent of the overstatement.
In both cases, the knowledge state of the counterparty could range from completely clueless to completely complicit. When they are completely clueless, your excuses or “I just got lucky” stories have to be very persuasive. When they are complicit, they can be phoned in just well enough convince casual onlookers. The point is to hit a performance level that’s acceptable to both parties now, while preserving a fiction around the performance level that ought to be on display.
We often recognize capability masking as part of a principal-agent information asymmetry problem, but power masking is also usually present. If you as an agent have an incentive to mask the true extent of your capabilities, the principal usually has an incentive to signal more power than they have, while masking the actual limit on power, so you’re never quite sure how much sandbagging you can get away with.
The result is an equilibrium of Mutually Assured Mediocrity (MAM), as both principal and agent operate below actual capabilities to coerce and deliver performance respectively, via power masking and capability masking. MAM is an energy efficient way to fulfill needs.
The effective psychological transaction is: “I will pretend to believe your explanations for your performance gap if you pretend to accept my posture of power.”
Non-volitional optimal performances requires the powerful party to have extraordinary amounts of power that can be nakedly exercised, with no masking. Think Pharoahs with an army of slavedrivers armed with whips. You’re going to work yourself nearly to death because sandbagging could mean actual death.
On the other extreme, when you suspect that the nominally powerful party has very little actual power, you can be increasingly insolent and open in your sandbagging, all the way to nakedly suboptimal performance or non-performance. The fiction can get really thin, just enough to preserve a very thin veneer of nominal power being respected. Such a situation is often on the edge of open rebellion and a power shift. Either party might break. The pressure of increasingly unsustainable face-saving may trigger a costly power display on the part of the principal, or cosmetic pretense may give way to open insubordination on the part of the agent.
Capability masking is a big part of what I call optimization theater. Our world runs on transactions that are ceremonially based on one party (usually the stronger one) demanding optimal performance from another and the other promising that performance. Job candidates promise to “do their best”. Sales people tell you they’re giving you the best deal possible. Politicians promise to work all out to fulfill campaign promises. Parents pretend in playing with children that they’re trying hard. Children promise to clean their rooms.
The ceremony of demanding and promising optimal performance is about validating the nominal relative power of the two powers, while allowing the actual relative power to govern the working relationship. Optimization theater has 4 effects generally.
- An immediate functional need is acceptably fulfilled by a mediocre performance
- A narrative of stable power relations with concomitant right to expect/obligation to deliver optimal effort is perpetuated
- Actual depth of capability is masked by a weaker party.
- Actual limits on power are masked by the stronger party.
This is a good thing. Optimization theater serves as a check and balance on fake-it-till-you-make-it false consciousness. Instead of pretending we can do more than we actually can, we pretend we can do less. Instead of a fragile pattern of constantly overreaching and failing, driven by clueless true believers in false consciousness myths, we have a pattern of constantly underpromising, and sometimes overdelivering, by more self-aware actors. Optimization theater is how we learn about actual distributions of power and capabilities, pick battles at the right times to reshape them, and keeping the world running with minimal expenditure of social energy at other times.
The evolution of a society is an arms race between the fragile evolution of false consciousness myths and the robust evolution of mutually assured mediocrity equilibria. That’s how you slouch towards utopia.
About Venkatesh Rao
Rao must be suffering from chronic depression. Beyond his dystopian worldview, he fails to offer a reasonable rationale for why either service providers or their customers would behave this way. What are the gains either psychological or material for both parties engaging in a duplicitous fraud?
One builds a reputation that moves you up the career ladder by always exceeding customer expectations – you do not negotiate a suboptimal goal – they have a problem and your role is to help them get it fixed. If your fix is suboptimal, you are unlikely to be invited back.
If I had an employee who behaved like this article describes they would find themselves very unemployed. My only experience that comes close to what Rao describes is when you hire somebody who lacks talent and tries to make it look as if they are exceeding expectations by whining about barriers to success rather than describing strategies to overcome them – not so tricky to detect.
Tom, I read it quite differently. I read the situation as negotiation around how much past “good enough to solve the problem” you go.
So yes, there is a problem and it needs fixing. But we could maybe get away with just fixing that problem, and I can avoid the effort of fixing all the future problems in the same solution space at the same time. It’s mutually beneficial because it’s cheaper up front (good for customers quarter), and repeat business down the line (good for your long term viability).
You are correct that a customer wants their expectations exceeded, everyone does, they want you to exceed their expectations in performance, in discounts, and in timeliness.
But what if their expectations are accurately anchored to your peak performance? Well then exceeding expectations by going “beyond 100%” is just a recipe for burnout, or in the case of pricing, bankruptcy (but both are basically linked to resource efficiency, just with personal limits not being directly substitutable and so characterised by easily priceable external flows), and so you do not price at the minimum cost you can achieve, giving them maximum value for money, you price at what you think the market can bear, ie. some deviation from the local signalled price given by some proximity weighted averaging function.
We can see in the libor rate an example of people intentionally, over multiple years, anchoring a lower expectation which they could then exceed. That was financial fraud, but because of it’s large scale collaborative character, in contrast, at an individual level giving public prices that do not fully reflect an efficiency gain you just made, while giving the occasional discount to loyal customers that is well within your profit margin, is a form of implicitly maintaining a similar difference between the public market rate and the theoretical minimum price it could achieve. You are doing it to increase your own profit, but you are also causing the price to stick at a higher stable value which others can also profit from.
– This is particularly common in remote work involving computing, where you can suddenly optimise away a task that your client still believes you are doing, and will often accept your prices on the basis of something being in some sense a big job. It shouldn’t really matter in terms of economic theory; if you can do something in half the time for half the price as the next competitor, and you do it in 75% of the time for the same price, then they should really just suck it up and get used to the idea of you making a nice profit on the back of your technical competitive advantage. But in practice, once people know you can do it cheaply, they will often try to apply pressure in indirect ways to get you to compress your cost structure, particularly if they are under pressure themselves. And so people don’t necessarily deceive their clients, but they don’t give the full information about their cost structure, labour utilisation and efficiencies, these are rightly considered commercially sensitive pieces of information, even if in a theoretical market they would be perfectly accurate and available public information.
Indeed, it is only in this more realistic kind of framework that it is possible to consistently exceed expectations, if those expectations are truly updating based on the nature of performance, then you can look at trends of performance, and automatically adjust your performance expectation to include expected productivity gains.
If your staff do approximately 5% better than expected every year, simply adjust the expectations to match, and remember that this is model agnostic, it is simply a matter of projections based on a time series. Any consistent pattern of exceeding expectations should be tautological.
And so if your expectations were true performance predictions, then exceeding them should be down to external noise, not due to the consistent performance improvements of your agent.
If they are on the other hand ersatz motivational or normative “expectations”, based not on what you think they can physiologically or psychologically achieve but what you assess is reasonable or possible to expect of people, given the market conditions, your level of power, and so on, then you return to the same game, but from the other side.
*compress down to your cost structure
*and substitute oxymoronic for tautological, I really should write these slower.
I am very glad that I donot work for an employer like Tom.
Problem is known. In most case fix is known. What is “suboptimal” like whats the price of fix,how many days to fix is mostly unknown. That’s where sandbagging comes in. Without sandbagging everyones life would be like
that of oarsmen on pharoah ship
Could you define false consciousness myths? Or have you already, somewhere?
It just seems a significant concept if it’s the other party in the arms race.
An archetypal example would be the current President of the United States, where his appearance of performance is applied to two audiences, himself, and his targeted base, discarding all appearance of success from any other angle. You can think of it as like a carefully angled instagram photo, if you also try to simultaneously forget how you made it, taking the subject position of your audience, while also making sure you know how to do it again.
Many other potential lessons from the hustler in chief.
When is performance optimal? Peak performance usually isn’t sustainable. Do you need an employee that works two hours at peak performance and spends six filling the day with menial work or one that spends six hours at 80% of peak performance before filling the day? Is your business about quality or quantity? And even 80% is of course ‘80% such that you still have energy to live your private live’. Otherwise it’s again unsustainable and performance will eventually suffer, gradually or abruptly.
I’d like to share my first hand observation of this:
By the first grade my now 10-year old son figured well out how to give the least effort and still get passing marks. He decided on his own that school offered little value to him and preferred to spend his time day dreaming and scheming in a Feynman like thought experiment format. Badminton with game-theory.
Despite his sweet demeanor there’s a ruthless lion who cares not for the opinions of sheep. (Not a wishful projection, just play a strategy game with him. Winning nearly every time) He has yet to show any real leadership skills, so while not a front man, he may become the fool who whispers in the figure head’s ear – or we’ll see if he’s backbone to become the silent shot caller.
Cloaking ones’s capability seems a willful deceit, so appears to me not a strategy for the herd but for those who wound play with, take interest and rise to power.