Two ways to be turned down
There are two ways a selection process can tell you no. One is a bar. The other is a breakpoint. They feel identical in the inbox and they mean completely different things about you.
A bar is fixed in advance. Pass it and you are in. Fail it and you are out, and the failure is about the work in front of the reviewer.
A breakpoint moves. Every application is scored and ranked against the other applications in the same window, and a cut line is drawn wherever the money runs out. The applications above the line are funded. The ones below it are rejected for being below the line, which is a statement about the cohort, not about the person who sent it.
Most students read a rejection as a bar. That is why so many of them write the same email back asking what was wrong with their file.
What a breakpoint looks like in published numbers
There is a real programme that publishes its own breakpoints, which makes the mechanism visible. Sweden's climate investment support ranks every application on a single computed number: kilogrammes of CO2 equivalent reduced per Swedish krona invested. Then it sorts them and funds what fits the budget.
The published cut has moved a lot. The 2015 rounds cut at 1.00. The 2016 first round cut at 0.75. The 2016 rounds two and three cut at 1.00. The four 2017 rounds landed between 0.75 and 1.00. The current estimate sits near 0.5 to 0.75.
Same programme. Same rules. Same scoring method. A number that would have been rejected in 2015 got funded in 2016 and rejected again in 2017.
Your bar is set by the people in front of you
Now translate that to a selective programme or an internship.
When one cohort takes 25 people out of 25,000 applications, the ratio is one in a thousand. That number is real, and it is also the least useful thing in the room, because it says nothing about where the line fell on the day.
What matters is the shape of the competition. Twenty-five places means the line is drawn at the twenty-fifth ranked application, whatever that application scored. If the twenty-fifth application was strong that cycle, a good file gets rejected. If the cycle was quiet, a weaker one gets in.
This is why "what were you looking for" is the wrong question and "where did the line fall" is closer. It is also why applying once and concluding something permanent about your ability is a mistake.
Not every rejection is the ratio
Read the rejection data rather than the rejection letter. In that programme, 22 per cent of refusals are for insufficient reduction per krona invested. That is the breakpoint working as designed.
The rest are mostly administrative. Applications that were incomplete, where the programme does not accept supplements after the deadline. Applications withdrawn by the people who filed them. Work that had already started and disqualified the project. Documents that expired while waiting.
Those failures are not competitive. They are mechanical, and they are the ones you can actually prevent. A missing attachment is not a judgement about your ability. It is a closed file. Nobody is going to phone you and ask for the missing page, because the rules say they cannot.
So handle the mechanical part first. Complete, on time, all attachments, before the deadline rather than at it. Then spend your remaining effort on the part that moves the ratio.
What actually moves the ratio
In a ranking system, there are only two honest levers. Make the thing you are counting bigger, or make the denominator smaller.
For an application, the countable thing is proof. Not adjectives about your enthusiasm, but artefacts that exist outside the document: a repository with commits under your name, a page that loads, a dataset with a date on it, a written analysis of something real. Each one raises the score of the file without changing anything else about it.
The denominator is the cost of taking a chance on you. A reviewer who has to imagine your ability is paying a price. Every piece of evidence you supply, with a link and a date, lowers that price.
That is the entire game. Most people spend their effort on tone and very little on the two numbers.
The breakpoint will move again
The useful habit is to treat a cycle as a cycle. The line that refused you in one window is not the line that will refuse you in the next, and it is not a verdict on your competence. It is a comparison against a specific set of other people who happened to apply at the same time as you.
The probability that a genuinely strong file is refused by a moving cut line is not small. It is roughly the probability that the cohort around you was strong, and over a decade that programme funded 27,114 of the 42,326 applications it received, which means a large number of competent files were simply in the wrong window.
Two things follow. Apply in more than one window, because the cohort changes and so does the cut. And keep the file ready between windows, because the openings are short and the prepared are the ones who use them.
The people who get in are rarely the ones with the strongest single application. They are the ones still applying in the third window, with a better file each time.
I have built companies across twelve countries and hired into teams in most of them. The candidate who was refused once and rebuilt the file was, in almost every case, the one who eventually landed. The candidate who read one refusal as a permanent measurement of their ceiling was the one who stopped. Only one of those two readings is supported by the arithmetic.