That’s a good point. Unfortunately, it’s difficult for me to say anything particularly informative to address this, given that most of the criteria we have in mind are qualitative in nature and don’t really lend themselves to being operationalised in a way that would allow candidates to better assess their odds of success.

A couple of things I can say that may or may not still be somewhat helpful:

• There is neither a maximum nor a minimum number of applications we intend to fund. Instead, we intend to fund all of the applications that are above a certain bar in terms of our funding criteria. Given my current (very uncertain) assumptions about the program’s reach and the likely composition of the candidate pool, my best guess is that we will only end up awarding a handful of grants, but in principle, there is nothing preventing us from giving out a significantly greater number of scholarships if the candidate pool ends up being correspondingly larger and/or stronger than anticipated.
• If an applicant’s academic record is comparable to that of other folks who get admitted to these universities, they are unusually thoughtful about the topic of how to do good in the world, and all of this comes through in their application materials, their chances of success should be favourable. (I appreciate that the second criterion in particular is somewhat unhelpfully vague!)

One way to address this issue would be to open the program for applications earlier in the year and to evaluate applicants well ahead of the university application deadlines, so that candidates already have their scholarship offers in hand before they need to decide whether to apply to the relevant universities, which should make this decision easy for those who did end up receiving scholarship offers. (As BrianTan notes in another comment, we do plan to inform successful applicants for scholarships in the US before the relevant university application deadlines in order to allow them to include information about the scholarships they have been awarded in their university applications. However, as noted in my post, they will have to get started with the required advance preparation - in terms of taking the required standardised tests, collecting references, etc. - much earlier than this.)

Given the timing of when we started working on launching this program, the fact that we will need some time to process applications, and the fact that applications to the relevant universities require advance preparation, this sort of format wasn’t really an available option for us this year, but it is something we may or may not consider doing next time around (assuming we decide to run another iteration of the program).

Another thing I’ll note is that obviously, folks with sufficiently strong academic backgrounds who apply to these universities generally also have a non-zero chance of getting admitted to and receiving funding from the universities themselves (even if for one reason or another they don’t end up receiving a scholarship from us), and they should factor this into their decisions about whether to apply.

I strongly agree. My impression (not based on research, but merely on unfounded hypothesizing) is that either an non-US applicant must have a finances arranged for a US university prior to applying. I would recommend providing the approval of the scholarship in advance of being admitted to the school. This way the applicant could apply and honestly list his/her method of funding the degree. The funds would be released only after the applicant is admitted to the school.

Also, what about people who "ought" to apply but don't know it yet? Could there be an EA scholarship program for developing nations? 1 page application, a short test, then then anyone above a certain bar gets supported in a full funding application and if successful helped applying to a university.

If I had to guess I'd think there were many barriers in the way of students applying to foreign universities. 

The main reason is simply that it so happens that most of the very top universities are based either in the UK or the US. (The fact that ETH Zurich is the only non-UK/US university that is in the top-20 on both the QS and Times Higher Education rankings partly reflects this, although my sense is that these rankings have some pretty serious limitations and should be taken with a major pinch of salt.) I also think there are additional benefits associated with attending university in the UK/US, including in terms of opening up career opportunities in the English-speaking world.

I agree that ETH has some things going for it and including it might well have been a reasonable choice, although my impression is that its teaching language at the undergraduate level is German, which means that it’s not really a relevant option for the vast majority of potential applicants.

In general, the decisions about which universities to include involved a number of debatable judgement calls, so I think there is a decent amount of room for reasonable disagreement on this topic.

I think that ETH  Zurich is a particularly good example since it's one place below Imperial. Not sure I'd be convinced of this argument if it was #15, say. 

Do you want to see a write up by me, not the OP, that gives some structure/rationalization/justification about why this set of universities was chosen?

This is an awkward subject and I think it’s unlikely that you will get a verbose response. 

I am worried that there will be a lack of response, and this might be create a perception  about the objective value of candidates outside this set of universities, despite these beliefs not actually being held strongly by anyone.

Instead of this situation, I'd rather produce a balanced write up, that can still be be kicked around (and maybe get explicitly stomped on by the OP).

Agreed. There are top-ranked universities in other countries. Additionally, $X might allow 5 students to attend a good university, or 1 student to attend a great university. I'd suggest learning toward a more "diversified" model in which more students receive funding (a larger number of "bets" are taken).

Did you mean to say $12,000/year instead of $1,200/year tuition at ETH Zurich? 

I would encourage people to think about this. I think the Thiel fellowship is genius  - to motivated, self-driven people, university is almost solely for signalling, because they would learn a greater amount of more important knowledge more efficiently without it. A large prestigious scholarship,  especially  with the requirement to drop out of university, easily provides the same signalling. Since there are no fees involved, it is also much cheaper to offer (although a big dollar value is a good idea for signalling, so don't make it too small!), and can maybe be given to more people, who can work without being constrained by the university system. People should of course be selected based on demonstrated ability to work on self-driven projects.

You mean people drop out of university to do EA-aligned work or skill-building? I don't see how that is necessary or something we'd want to particularly encourage, though I'd like to hear OpenPhil's thoughts.

Another suggestion on how to make the application form slightly better is to state in the form's description that the form autosaves your answers even after you close the tab, allowing people to not have to complete the form in one go. (I presume Paperform does this for all forms).

We did consider asking for less detailed information in the financial information section for the exact reason you point out, but we ultimately felt that the current approach struck the best balance between a number of countervailing considerations. (For example, having to ask all of the most promising candidates to provide additional information at a second stage would have added to turnaround times, which would in turn have required us to set earlier application deadlines.)

Note that the financial section of our application form already requires applicants to provide meaningfully less granular information than the CSS/FAFSA forms and most of the university-specific financial aid forms I have seen, so I’m hoping that this won’t be too onerous on candidates.

Also, some people's parents would be quite hesitant to disclose what they actually make to their child. So that would just make it much harder for a student to get accurate figures on what their parents make. 

I'm not sure what the solution there is, but potentially you want to account for that scenario some way, such as giving the student the option to just give a rough ballpark of what their parents/family make, especially at just the 1st stage of the application. 

Thanks for the suggestion - I thought I had addressed this by asking for people’s “GPA/aggregate score”, but I’ve now added a few sentences to the question description to further clarify that we’re just looking for people to use whatever metric of aggregate performance is commonly used wherever it is they go to school (not necessarily a GPA).