Thanks. Some of these seem to be good ideas indeed.
Isn't it an issue though that in a lot of these cases you/your colleagues have large direct impact on the outcome, and knowing the prediction itself can change your/their behaviour? E.g., if you have 90% public prediction that Adam will complete task X and 10% that Bob will complete Y, their knowledge of these can impact how they approach their work. Maybe Bob will .
That being said, it can still be an effective productivity tool, but it no longer directly measures the outcome of the stated prediction. Instead, it will become a reflective statement ("The probability of Bob completing Y provided he knows this prediction is 10%").
1I think in many cases this is a feature rather than a bug, I'd be curious as to whether you had an example of a time when having made the prediction would be deleterious to realising the desired outcome? Assuming as Adam says that you're more interested in the outcome than the prediction being accurate.
1Assigning a low probability that I will do a task in time is a self-fulfilling prophecy. Because the expected utility (probability times utility) is low, the motivation to do the task decreases. Ideally I would never assign probabilities to acts when choosing what to do, and only compare their utilities.
I think it's still very useful to be able to predict your own behaviour (including in the case where you know you've made a prediction about it).
Things can get weird if you care more about the outcome of the prediction than the outcome of the event in itself, but this should rarely be the case - and is worth avoiding, I think.
Your mind is a prediction machine, constantly trying to anticipate the world around you and altering its forecasts based on new information. It’s always doing this as a background process. But what would happen if you deliberately trained this skill? Could you get better at predicting your projects, your life, and the future?
Sadly, I don’t have a crystal ball, but I have found trying to see the future valuable. I can't be certain about what will happen, but by deliberately practising the skill of prediction, I can create feedback loops that allow me to better anticipate what the future will bring. This has indeed helped me become happier, better to interact with, and more impactful.
Concretely, making probabilistic predictions has helped me:
I’m building Fatebook to make it as quick and easy as possible for you to harness the power of prediction and reap some of the benefits above, while remaining completely free. The first step to getting value from predictions is to figure out what to make those predictions about, which I’ve had plenty of time to tinker with.
Which questions are worth asking?
Practice and creativity have allowed me to get more value out of the questions I ask, and the predictions I make. For example, it has helped me communicate with my team better around project timelines, hiring decisions and strategy prioritization. We share predictions with one another using Fatebook for Slack on questions like:
This question allows me to set expectations very accurately with my team. If writing the blog post is taking longer than expected, I can decrease my prediction that it’ll be done by Friday, which is a low-friction way of letting others know about my updated timelines.
Sharing predictions makes it much easier to reach consensus about what work should be top priority. If we’re aiming to grow and my team all predict that implementing the “sign up with email” feature is more likely to get us 500 new users, then it’s obvious we should do it first. Or, if the probability of either feature attracting new users is low, that’s a sign we should consider other options. When there’s disagreement between team members, it is easier to have a constructive discussion using concrete numbers - and, crucially, when we come back and resolve the questions, there’s a chance to reflect as a team on which factors were important and which were overlooked.
We use questions like “will we hire Alice?” to share our current expectations about working with candidates at a glance - if my prediction that we hire someone is meaningfully higher than my teammates, this is a good way to initiate a discussion about why I’d be so excited to work with them. It helps us be aligned throughout the hiring process, so we don’t have mismatched expectations at the point when we need to decide whether or not to make an offer.
When I submit code changes to a project (known as a pull request, or PR), I want the reviewer to accept it without significant alterations; otherwise the ensuing back-and-forth will be an inefficient use of both of our time. Similarly, I don’t want to later have to undo any of those changes, or to find unexpected behaviour when those changes go live. To maximise the chance of my pull requests going smoothly, I wrote a script that asks me to make predictions about these outcomes before submitting my PR.
If my prediction that the PR gets approved first-time is low, say below 75%, then this is a sign that I ought to spend more time polishing the changes before submitting them for review. I can spend some more time going over my own code and checking that it meets the required standards, before again making a prediction - this process can be repeated until I’d be shocked if the reviewer did not accept all of my changes.
I often feel anxious about sending important emails, particularly to people I don’t know well. In the case where I need to email one of our funders, Bob, I can make predictions to help:
The first question helps keep me accountable - it would feel painful to acknowledge that something I thought was 95% likely to happen did not come to pass, which might be the nudge I need to send the email in the first place. The second helps me think rationally. If there is 1 out of 100 worlds where Bob is gravely offended by my email, and replies saying that he thinks I’m a terrible person and that he’s never going to do business with me again, I have a natural tendency to only think about that singular world while sending the email. However, if I make a prediction and am forced to admit that in the other 99 worlds, Bob either doesn’t respond, or responds warmly, it is easier to appease the part of my brain that considers emailing Bob as a kind of threat.
Another set of questions I found helpful recently revolved around finding a new job. The first kind of question was to try to honestly estimate my chance of success for any given position, eg:
These are useful to understand how to allocate my effort between different stages of the funnel - if one of the three companies in the last question is 90% likely to give me an offer, it’s probably not worth it for me to be sending out fresh applications. Similarly, if I notice that I am overconfident on questions like the first one, and get fewer interviews then I would’ve expected, this is a useful signal that I need to make some changes to my CV and try to get some more referrals. This proved extremely useful in my recent job search for knowing when to switch between mostly sending out new applications and mostly preparing for interviews.
There are other questions I used to help me to decide whether or not I ought to take a job, for example:
Often in these situations I have a strong gut-feeling about which company to join, but it’s also a valuable exercise to see whether the numbers match up. Sometimes these intuition pump-style questions have been able to shift my actual intuitions. You could apply a similar kind of question to e.g. finding a new flat to rent, or a new romantic partner.
I’ve found these long-term “hill climb” predictions valuable too. Consider questions like:
If I string together some solid running weeks and nail my track sessions then I might be able to update the chance of completing the ultramarathon to 85%. This isn’t as satisfactory as crossing the finish line, but it does feel like a small reward for my hard work so far, and it might help motivate me to continue training as hard. Conversely, if I skip my Sunday long run in favour of a boozy brunch three weeks in a row, I’ll likely decrease the probability to 75%. This doesn’t feel great, but it’s not as bad as missing the goal altogether - in fact it might be just the harsh reminder I need that if I want to achieve my goals then I’m going to have to make sacrifices for them, and now is the time to course-correct.
How to track your predictions
In order to get the maximum benefit from creating predictions, deliberate practice is required. I suggest the following steps:
If you’re still unsure about how making predictions could be useful to you, then I’d like to help. Comment below or email me at jonny@sage-future.org with a little bit about a problem/goal/thing you care about, and I’ll help you brainstorm predictions you could make that would be potentially valuable to you.