How to judge your chances of success

80,000 Hours is a non-profit that gives you the information you need to find a fulfilling, high-impact career. Our advice is all free, tailored for talented graduates, and based on five years of research alongside academics at Oxford. Start with our career guide.


You want to make a difference. This means being as successful as possible in whatever high-impact path you choose to pursue. In some recent posts I raised a worry that we might overestimate our chances of success, due to a tendency to think of ourselves as better than average, and to pay too little attention to information about the average person. This could lead you to pick careers that turn out not to suit.

However, some people have made the point that we also don’t want to go too far the other way and underestimate ourselves. Sometimes we do have reason to think we’re better than average at something, and we shouldn’t ignore this. Underestimating our chances of success might stop us from pursuing potentially high impact paths.1 What if you think you really are better than average? After all, whilst no-one is above average at everything, almost everyone will be above average at something.

In this post I’ll talk through some of the best ways we’ve found for working out your chances of success in a career.

In summary, the process is:

  1. Work out which factors (personality traits, skills and abilities) are most relevant to success in the field you’re considering
  2. Find ways to objectively measure yourself on these factors
  3. Given this information, narrow down your reference class to those similar to you
  4. Get your “base rate” from this class

Think about what’s relevant

You might have really solid evidence that you’re slimmer than the average person, but this isn’t going to alter your chances of success in most careers – unless you want to be a supermodel. It might sound obvious, but the best place to start is by thinking about which factors (such as personality traits, skills, abilities) are most relevant to success in the domain you’re considering, and then see how you measure up on those.

Some measures are going to be relevant to pretty much any career: intelligence is arguably one, although it may well be more important in some domains than in others. A meta-analysis looking at the relation between the “Big Five” personality dimensions (extraversion, neuroticism, agreeableness, openness and conscientiousness) and job success in a number of domains suggests that high conscientiousness is a consistent predictor of success across domains. The relationship between the other traits and success varies more across job areas.2

In some jobs the most relevant skills are going to be fairly obvious: being highly numerate is going to increase your chances of success in accounting, for example; being exceptionally good at public speaking will help you with advocacy. Sometimes it might be less clear what the most important considerations are. It might help to speak to someone with experience in the area, or look at typical job descriptions.

Make sure you’re justified

Say you’re pretty sure you’re above average intelligence. What is your justification for this? Going with a “hunch” isn’t good enough: you need to know what that hunch is based on.

There are some pretty good objective measures of intelligence. IQ tests and your past academic records are two pretty good ways. Try to get some data on how this places you: if you’ve got 2:1 degree from one of the top 10 universities in the country, find out what percentage of people in your age group have the same achievement. Intelligence tests will generally tell you what percentile your result puts you in.3

Similarly, there are a number of personality tests that measure the Big Five personality traits.4 It’s harder to see how objective a measure these tests are. They generally involve stating how much you agree or disagree with a statement about yourself such as “I am good at a lot of things”, and you’re clearly biased. If you do your best to answer honestly, though, these tests should give you a rough idea of where you fall on the relevant personality traits. It might help to try taking a number of different ones.

If you think you’re above average in any respect, try to find some objective measure to justify this. If you can use multiple different ways of doing this and put a number to it, even better.

Find the appropriate average

I argued previously that the best place to start when judging your chances of success is by looking at the average person. This isn’t strictly true. If you’re a PhD student at Yale trying to work out your chances of becoming a tenured professor, you’re obviously not going to start out comparing yourself to the “average” U.S. citizen, who probably doesn’t have a college degree.5 You’re going to compare yourself to the average PhD student at an Ivy league college.

Looking at the relevant “average” is crucial to accurately judging your chances: you need to work out what the appropriate reference class is.

Choosing an appropriate reference class

Sometimes it’s obvious what the appropriate reference class is, other times it’s harder. A general strategy would be to ask yourself the following question: what specific information do I have about myself that is both reliable and relevant to the outcome I’m trying to predict?

Suppose Daniel is our doctoral student trying to predict his chances of receiving tenure. What’s the relevant information here? The fact that he is doing a PhD, and that he is doing it at an Ivy League university, are both pretty clearly relevant in the sense that they clearly affect a person’s chances of becoming a tenured professor. (There might be other pieces of relevant information, such as his position within his class, or level of conscientiousness, but let’s keep it simple for now.) Irrelevant information, on the other hand, would be the fact that Daniel has dark hair. There’s no evidence that people with dark hair are more likely to become tenured professors.

What about the reliability of this information? I think Daniel can be pretty sure that the information that he is doing a PhD at Yale is reliable. An unsupported judgement about his position in the class, on the other hand, might not be entirely reliable information, so shouldn’t be taken into account until he has evidence to back up this judgement.

Taking the outside view

This method of basing our predictions on a class of roughly similar cases is what Daniel Kahneman refers to as “the outside view.” In contrast, to take the “inside view” is to focus on the specifics of your own case when making predictions. Due to the fact that we tend to be over-optimistic when considering the specifics of our own case, taking the inside view has the tendency to lead to the planning fallacy (a tendency to underestimate the amount of time it will take to complete an activity). In particular, people seem to find it hard to fully envision the possible circumstances that could lead to failure in their own case. Taking the outside view is also commonly referred to as reference class forecasting: using information about other cases in the relevant reference class to make forecasts.

Taking the outside view is the best known way to take account of any solid evidence that your chances are better than average, without falling prey to the base rate fallacy. If you want to know your chances of success, best to start with the average person’s chances of success: but sometimes you might have good reason to make certain assumptions about what this “average person” looks like.

However, sometimes you might have information suggesting you differ from average that’s strong enough to warrant consideration, but not strong enough or of the right nature to justify changing the entire reference class. Suppose Daniel has good reason to believe he is more conscientious than average (he did a number of personality tests, checked his responses with a number of people who knew him well to make sure he wasn’t overly biased in his answers, etc…). If he tried to change the reference class here, it would have to be something like “PhD students at Ivy League universities with the same level of conscientiousness as Daniel” which seems difficult to identify in practice, and is also likely to be too small to be useful. But it doesn’t seem like he should ignore this information either. In these cases, we need a way of updating on baseline predictions given evidence, dependent on the strength and reliability of that evidence. This isn’t easy: I’ll talk about how to do this in an upcoming post.

For more on how to combine the inside and outside view, we recommend this excellent summary by Luke Muehlhauser.

You might also enjoy:

Want to be successful? Know your odds

Do you think you’re better than average?

References and notes

  1. It might even be argued that underconfidence is more problematic than overconfidence: as people who are overconfident are likely to be more optimistic, work harder, which thus increases their chances of success. There certainly may be some truth in this: the optimism bias certainly has its benefits in some contexts. I’m hopefully going to write more about the optimism bias in general soon. 
  2. The Big Five Personality Dimensions and Job Performance: A Meta-Analysis, Barrick, Murray R.; Mount, Michael K. (1991), Personnel Psychology 
  3. The most accurate IQ tests require payment, but this free version seems to be pretty good and recommended by a number of different sources. 
  4. A few decent free online tests for the Big Five personality traits: (1)[], (2)[]
    (3)[], (4)[]
    Note that the accuracy of these tests will depend on the honesty of your answers, which are highly subjective – so I would recommend taking a number of different ones, thinking thoroughly about your answers, and taking the results with a pinch of salt. This website has some tips for getting the most out of these tests. 
  5. ~30% of U.S. citizens aged 25 and over surveyed in 2011 had a minimum of a bachelor’s degree.