Decisions + expected value: part 2
Expected value decisions can be time intensive. How do we move more quickly?
In my previous post, we learned what expected value was, and how to apply it to decision making. Generally, we are not looking for simply a result, but an expected outcome, and we are not only after a single outcome, but seek to apply the proper framework in the long run to achieve repeated success.
But getting the right data, processing that information, really thinking it through, and then calculating expected values can take a long time, especially when the problem is large and complex. And we want to avoid analysis paralysis at all costs, especially when we are working collaboratively with others. It is much better to make a decision quickly with 80% of the info, then waiting weeks for 100% while blocking your team. We’ll dig into this in depth later in the article. But first, let’s see how processing so much information works in poker.
In poker, it’s generally unacceptable to sit and think about a single decision for a few minutes or longer simply for courtesy and gamesmanship, but overall just for the health of the game. Amateurs can be put at an even worse disadvantage if pros are given a limitless amount of time to process a hand thoroughly, not to mention it’s an awful experience for everyone at the table.
In fact, in this year’s World Series of Poker main event final table, a player took an absurd 19 minutes to make a decision. Granted, there can be a *lot* to think about in some complex spots, especially one when you’re playing a very good player with $10 million dollars on the line, but taking this long is just not good for the game.
But these types of decisions demonstrate just how complex poker can be. First, poker is a game of incomplete information. And it’s the most critical pieces of information that are missing: your opponents’ hands, and the cards that are going to be dealt next from the deck. Having this info (which would be cheating), would make you practically unbeatable. So you must constantly make educated guesses based on the incomplete information you have at hand. And some of the best in the world can make some pretty sick guesses.
That might look like wizardry, but it’s not. These players spend an enormous amount of time studying and perfecting their craft. They go through many many repetitions of hand scenarios, analyze their mistakes, and keep improving. It’s the enormous amount of dedication they put into their games which make their decisions at the tables seem like magic. But it’s really just hard work and preparation.
Which brings us to the first principle of making fast decisions. Taking the time to master your craft. When I first started playing poker, it would take me a long time to run through the very basic analysis of a hand. But with time, that analysis became like instinct which I could do in seconds, which allowed me room to process more advanced analysis. Then that advanced analysis became like instinct over time, and so on and so forth. The more I studied and the more I played, the faster and more accurately I could process information, and the better and better I got at poker. The analysis I had to do every time eventually became intuition.
You may admire some business leaders who seem to similarly have a track record of making very good, intuitive decisions all the time, and without taking an exorbitant amount of time to do so. This is also not magic, but a lot of hard work and preparation behind the scenes, and often a lifetime of perfecting their craft. Similarly to poker players, at some point, basic decisions were struggles for these future business leaders. But over time, with more study, observation, and repetition, much of it became part of their intuition.
But all decisions can’t just be made with intuition. More often than not, deliberate analysis and statistical thinking are required to come to the right conclusions. So when do we choose which method? We can split these situations into two buckets: problems where risks that are well known, and problems where risks are not well known.
If risks are well known, you need intuition less. Data should be more readily available, thus statistical analysis makes more sense.
If risks are not well known, you need more intuition. Data is not readily available, so more qualitative analysis makes sense.
As examples, the risks of going skydiving are well known, because many people have gone skydiving. So a plethora of information is available for you to prepare yourself (or perhaps decide not to do it at all). On the other hand, eating a strange mushroom you found in the forest may have risks that are not well known. And maybe a Google search is not helping you identify what it is. So intuition will be your only option here. A prudent person probably wouldn’t eat it, which seems like a fine decision.
But if that prudent person doesn’t know anything about mushrooms, how would they have any intuition about it?
This is where heuristics come in. Often, Wikipedia will define things better than I can: any approach to problem solving or self-discovery that employs a practical method that is not guaranteed to be optimal, perfect, or rational, but is nevertheless sufficient for reaching an immediate, short-term goal or approximation. Some examples of heuristics are:
Rules of thumb
Availability heuristic
Educated guesses
Trial and error
Not knowing much about mushrooms but generally knowing strange mushrooms can make you sick is a good example of the availability heuristic. Baseball outfielders learn that if a fly ball is below the brim of your cap, you run forward. If it’s above the brim, you run backward. This is an example of a rule of thumb. So you can look at heuristics as a way to make a decision that should get you in the right ballpark of a correct decision, and the more knowledge and data you have in hand, the more accurate that decision will be.
So far, we’ve covered hard work and perfecting your craft to prepare you to build a strong base of experience and intuition to make tough decisions. Then we covered situations where risks are not well known and data is scarce, would be situations where more intuition will be needed. And then we saw how heuristics can help you make better approximations to get you close to a better decision.
But there is one more thing to consider when making fast vs slow decisions, and that is judging how much information is actually needed to make a decision. Generally, moving with less info can be faster but less accurate, while trying to gather more info will be slower but more accurate.
So when should you choose to take the time to gather more info to solve for accuracy? Well, not all decisions are the same. Some decisions that are reversible, or low risk, or will need less resources to complete, can and should be made with less info, and thus, more quickly.
So always think about trade-offs. When considering how much information to gather, you are trading off accuracy for speed. Fast decisions and a bias for action are always a good thing, but being reckless with decisions is certainly bad.
There is a very obvious feature you need to add to your product. Every competitor has it, every customer expects it, and it’s not that much effort to build. This is an obvious decision that doesn’t need a lot of time and analysis. But contrast that with a big bet on an innovation that no one is asking for, but you feel will fundamentally change the way that people do a task for the better. This is certainly a high risk, high reward decision that can use a little more time and research before pulling the trigger.
And for a team waiting for you to make a decision so they can get to work, speed is of the essence. Blocking a team when more analysis isn’t necessary is not only wasteful of resources, but can be demoralizing if people are looking for you to be decisive.
So consider this framework on the type of problem you have on hand, and follow with the appropriate decision making process the next time a big decision comes your way.
So in conclusion, never underestimate the impact of hard work and preparation on your success. And after you build that strong basis, use a combination of heuristics and data driven thinking to arrive at accurate decisions quickly. And know when to get in the lab and do a bunch of research and when you should just pull the trigger.
Your ability to make good decisions persistently and quickly will lead to many successes and accelerated career growth. And who knows, after getting these principles nailed down, maybe you can go and win the World Series of Poker yourself 😎