vertices

Awesome additions guys! I've added a few and incorporated Pitangus' suggestion to divide these up for easier reading. I kept the personal statement and research experience together because there's some overlap in how you want to handle them.

INTELLECTUAL MERIT
- PROPOSAL
-- whether proposal has a clear hypothesis/question it is trying to address
-- whether proposal is novel/innovative
-- whether proposal is original (your own)
-- whether proposal is transformative (rare)
-- whether proposal is well-written
-- whether access to resources is addressed in proposal
-- whether expected results are addressed in proposal
-- whether validation of results are addressed in proposal
-- whether proposal is sufficiently detailed (this is a tough one... too detailed and it might not be well-written/clear and have no room for other objectives)
-- interest of proposal among scientists
-- relation of research to theories of founders/experts in the field
-- relation to previous work the applicant discusses in other essays/supporting material
-- whether the outcome measure is well-related or "distal" to the intervention (likely more of an issue in social sciences)
- PERSONAL STATEMENT/RESEARCH EXPERIENCE/SUPPORTING MATERIAL
-- strength of academic record
-- research capabilities, independent and in team
-- publications (includes submitted/in review articles, especially if you don't have other pubs yet)
-- presentations (includes talks and posters)
-- participating in any other funded proposals (e.g. help writing grants)
-- leadership potential of applicant, including personal qualities such as persistence and drive
-- previous research essay draws connections among past projects and internships, and explains their significance (both personal + BIs)
- ALL
* strength of recommendation letters and what they say about all of the the above (including proposal if applicable)

BROADER IMPACTS
- PROPOSAL
-- whether proposal integrates/supports science education with research
-- whether proposal integrates broadening diversity with research
-- whether proposal benefits society or some large group outside major field
-- potential of research to support "citizen science"
-- applicant's plan to disseminate results broadly, especially with non-science stakeholders if applicable, but further too
- PERSONAL STATEMENT/RESEARCH EXPERIENCE/SUPPORTING MATERIAL
-- benefits to society, education or diversity in all past research projects
-- applicant's activities to increase science education and recruit young scientists
-- applicant's activities to broaden diversity in science
-- applicant's leadership roles in broader impacts activities (including TAing)
-- quantity of applicant's BI activites (e.g. reviewer lists lots of examples or length of time)
-- applicant's plan to -continue- broader impacts activities
-- "unique perspective" due to unusual life circumstances which gives applicant insight into the needs of underserved communities
- ALL
* specifics for BI activities in all sources
* strength of recommendation letters and what they say about all of the the above

@Dynamom: Maybe next year's thread's kickoff should be a group mining expedition of all the previous threads. I read them all at some point (including all 170+ pages of the year-less one) and incorporated trends but didn't write the information down back in November when I probably should have.

I agree with the posts so far with regard to publications, collaborations, networking, and funding. One thing that hasn't been emphasized is communication skills. Write well and then write better. Present well and then present better. These are ways you get remembered, get papers accepted, and get funded. These signal to departments that you will continue to do those things and departments are looking for people who do those things.
 
Another thing you should so is start reading about academic job searches now. That will give you a better idea of what you're up against while hiring and really drive home what really matters. Your field, Computer Science, has a ton of advice on the web:
http://www.cs.virginia.edu/~weimer/grad-job-guide/guide/index.html
http://33bits.org/2012/10/01/five-surprises-from-the-computer-science-academic-job-search/
http://homes.cs.washington.edu/~mernst/advice/academic-job.html
http://www.pgbovine.net/guo-faculty-job-search.pdf
 
Note many talk about how much easier it is to get to the interview if you already know someone at the school (or if one of your recommenders do). Networking and collaborations.
 
I've heard from multiple people that the NSF GRFP is good if have if you want to be a professor, even more so than harder to get fellowship. Personally I don't think it would be a deal breaker. However, anything that differentiates your CV can help especially where you don't have connections.

On the other hand, there are more worthy applicants than there are awards, even with the numbers having been significantly bumped up. Those 2nd years that weren't awarded in previous years and are trying again aren't necessarily "worse." A lot of work goes into meeting that bar of 'award worthy' but in the end there's always an element that is outside of an applicant's control, like getting unlucky with a reviewer (someone whose other applications make your z-scores less competitive, someone who doesn't like your subfield and won't advocate for you, someone who is tired because your application was at the bottom of their stack, etc) or getting tanked by a letter writer like PhDplease! In a sense, these applicants are stiffer competition because they're starting with already strong applications from the previous year and making them even stronger. Several people here will tell you that they won on their second try and can articulate how they learned from the experience and addressed the reviewers' feedback.
 
I know my undergraduate application was a trainwreck. I think replacing the 'tossed together in the last week' with 'meticulously crafted using knowledge from 5,000+ GradCafe posts' probably helped. I've always believed my 1st year application was stronger than my 2nd year would have been because I didn't get any papers accepted my first year while I'm sure other applicants did. Further, would the growth in sophistication of my proposal 'kept up' with the growth in everyone else's? Who knows!?
 
Also, I had heard that while the awards are given to fields of study in proportion to applicants, they're given out by year roughly in thirds. I would guess that there are more 2nd year applicants than undergraduate applicants, making it a tougher pool. [ Does anyone have hard data? ] I know several people in my program who didn't apply until their 2nd year because they either didn't know about it or thought it impossible until they were encouraged by people who pointed out the acceptance rate lately has been too good not to try. Some people come from undergraduate institutions or labs that don't emphasize the award.
 
Finally, graduate school is a time of learning and growing. Someone who might not have been 'competitive' as an undergraduate may be quite competitive by the time they're a second year. We're not always going to be roughly the same level of "competitive" with respect to each other as we were in undergrad. It could be finding the right research project, coming in contact with great mentors, being in the right kind of environment, or having better resources to do either research or outreach that can help someone really shine. (At the same time, the reverse can be true making some "competitive" folks less so. It feels weird saying that because we're all graduate students which makes us all pretty strong to begin with.)
 
In summary, your competition isn't necessarily worse. Of course, I still believe your chances are too good not try. Good luck everyone!

On the other hand, there are more worthy applicants than there are awards, even with the numbers having been significantly bumped up. Those 2nd years that weren't awarded in previous years and are trying again aren't necessarily "worse." A lot of work goes into meeting that bar of 'award worthy' but in the end there's always an element that is outside of an applicant's control, like getting unlucky with a reviewer (someone whose other applications make your z-scores less competitive, someone who doesn't like your subfield and won't advocate for you, someone who is tired because your application was at the bottom of their stack, etc) or getting tanked by a letter writer like PhDplease! In a sense, these applicants are stiffer competition because they're starting with already strong applications from the previous year and making them even stronger. Several people here will tell you that they won on their second try and can articulate how they learned from the experience and addressed the reviewers' feedback.
 
I know my undergraduate application was a trainwreck. I think replacing the 'tossed together in the last week' with 'meticulously crafted using knowledge from 5,000+ GradCafe posts' probably helped. I've always believed my 1st year application was stronger than my 2nd year would have been because I didn't get any papers accepted my first year while I'm sure other applicants did. Further, would the growth in sophistication of my proposal 'kept up' with the growth in everyone else's? Who knows!?
 
Also, I had heard that while the awards are given to fields of study in proportion to applicants, they're given out by year roughly in thirds. I would guess that there are more 2nd year applicants than undergraduate applicants, making it a tougher pool. [ Does anyone have hard data? ] I know several people in my program who didn't apply until their 2nd year because they either didn't know about it or thought it impossible until they were encouraged by people who pointed out the acceptance rate lately has been too good not to try. Some people come from undergraduate institutions or labs that don't emphasize the award.
 
Finally, graduate school is a time of learning and growing. Someone who might not have been 'competitive' as an undergraduate may be quite competitive by the time they're a second year. We're not always going to be roughly the same level of "competitive" with respect to each other as we were in undergrad. It could be finding the right research project, coming in contact with great mentors, being in the right kind of environment, or having better resources to do either research or outreach that can help someone really shine. (At the same time, the reverse can be true making some "competitive" folks less so. It feels weird saying that because we're all graduate students which makes us all pretty strong to begin with.)
 
In summary, your competition isn't necessarily worse. Of course, I still believe your chances are too good not try. Good luck everyone!

About Broader Impacts:

1. Whatever you do for your broader impacts criteria should be something you would do regardless of whether or not you were applying for the NSF.  Don't go tutor women in science or poor rural math students just because you are applying for a grant.  Do it because you genuinely want to increase the number of poor rural students in pure mathematics or the number of women in science, because you think that those are important and valuable things.  What's important for the NSF application is not so much that you've done these things but that you're able to talk about why you've done them and why you think they are important.  "I think they're important because NSF wants me to do them to get this grant" is not going to cut it. 

That goes for anything that you do to satisfy your broader impacts criterion - whether it be promoting teaching and learning in science, enhancing the infrastructure for research and education, broadening scientific idea dissemination, or the impact of your research on society.

2. You can't start doing a Broader Impacts activity a few weeks before the NSF deadline and expect it to win you cookies. That comes across as quite obviously doing something for the purpose of winnng the grant, and not because you value the contribution that thing makes to the science.  They want people who are going to continue to think about the ways in which science (both their science and science in general) contributes to the wider world, and to take action to involve the wider world in science.  Again, they are giving you taxpayer money, so why should they is the question.

Here's what I think.  Think about what made you passionate about your science in the first place.  Why do you love what you do?  Now think about all the geeky stuff you've done in service to that passion.  Chances are at least some of that stuff fits the Broader Impacts criterion.  I judged the NYCSEF not (only) because I thought "I can do some great service and encourage high school students to enter and be passionate about science," but also "I get to see some really cool high school science fair projects, awesome!!"  Every serious science student I know has done something seriously geeky "for fun" that they didn't have to do but wanted to because SCIENCE.  Think about those.

*
panabtl - a little of both.  Your proposal should have a background and framework that is grounded in things you have already done/have some experience and knowledge in, but the proposal should also show that you want to learn new techniques and expand both your own knowledge and science in some way.

Broader impacts is simply promoting science outside of academia.  Broadening the participation of underrepresented groups (whether they be ethnic/racial minorities, poor people, rural people, immigrants, etc.) is only ONE of the broader impacts examples.
 
From: http://www.nsf.gov/pubs/2007/nsf07046/nsf07046.jsp
 
Caveat lector - the following list is neither prescriptive nor exhaustive and should not be read in ways that constrain the creativity of researchers in proposing activities with broader impact. However, in all instances a proposal must be specific in how it addresses the Broader Impacts criterion.
 
Advance discovery and understanding while promoting teaching, training, and learning, for example, by training graduate students, mentoring postdoctoral researchers and junior faculty,  involving undergraduates in research experiences, and participating in the recruitment, training, and professional development of K-12 mathematics and science teachers.
Broaden participation of under-represented groups, for example,by establishing collaborations with students and faculty from institutions and organizations serving women, minorities, and other groups under-represented in the mathematical sciences.
Enhance infrastructure for research and education, for example, by establishing collaborations with researchers in industry and government laboratories, developing partnerships with international academic institutions and organizations, and building networks of U.S. colleges and universities.
Broaden dissemination to enhance scientific and technological understanding, for example, by presenting results of research and education projects in formats useful to students, scientists and engineers, members of Congress, teachers, and the general public.
Benefits to society may occur, for example, when results of research and education projects are applied to other fields of science and technology to create startup companies, to improve commercial technology, to inform public policy, and to enhance national security.
 
You don't have to tutor or mentor minorities - or anyone - to fulfill broader impacts.  One of my BI statements was judging the NYC Science & Engineering Fair.  Another was how I hoped my scientific research would impact health policy, specifically AIDS outreach and prevention work.  I talked about talking about PhD work and science in classrooms with children when I volunteered in elementary schools.  Developing applied projects with non-academic labs could be broader impact.  Teaching an "understanding science" seminar to business people could be BI; writing science articles for newspapers or magazines could be BI; creating a startup or using your research to design a smartphone app could be BI.
 
People always think "broader impacts" is just a cute buzz word for "diversity" and it's not.  It's "Why should anyone care about your research?  Why should we take taxpayer money to fund you?  What are you ever going to do to give back to society and repay this debt you've incurred?"
 
In fact, I would say it's probably beneficial to your application to talk about several different kinds of ways to impact the community and world more broadly, not just the one about underrepresented minorities.

Summarizing briefly, XSEDE is a program/system that gives access to computing, storage and software resources through a unified, trying-to-be-friendly interface with existing support staff. If you have a project that could benefit from having a computer program run on dedicated or parallel machines, you could use the XSEDE resources. If you need storage space for your data, you may be able to obtain space. That's a very short version. If it sounds right for you, check out the site under 'Resources'.
 
My understanding is that being awarded or offered HM gives you the ability to write proposals for XSEDE resources on your own, whereas other students would have to ask their advisor or a post-doc at their organization.

Offered Award! I thought I'd offer a couple of tips, especially for anyone who may be perusing this board hoping to apply in the future:
 
If you are in a field where it is extremely rare to have publications early in your grad career (for example, almost everything is single authored in anthropology, and based on extensive fieldwork we do later in our grad career), be sure to present research at conferences, submit papers for awards, try to get things published on blogs, etc. If you're prepping to apply for the GRFP, dust off any research papers you've written  in the past and get them presented/read/submitted somewhere! There are lots of undergraduate paper competitions through AAA, ASA, etc. that no one really knows about. Smaller regional conferences are a supportive, non-intimidating way to get your work out there. And that stuff counts for the GRFP!
 
A hypothesis is totally necessary, even for fields like Anthro where that's not usually done. A contingency plan if your hypothesis is totally wrong = better. I expect to go to the field and find _______. If that is not the case, I plan to ___________. Either way, my project will answer this question:___________. From what I understand, they like this because they know even if your plan is an utter failure, the NSF will still not be wasting its money because you're going to re-route the project and it's still going to be great.
 
Broader impacts is so, so important. Multiple reviewers mentioned my past experience/future work with marginalized populations. In the future, I argued that my research will benefit these marginalized groups, and also that I will be working directly with those populations during the research process. Involve non-scientists IN your research if possible! Make a connection to education.
 
My advisor is an NSF reviewer. In the project proposal, she recommends one section for Broader Impacts and one section for Intellectual Merit. Don't think that either of these things will be implied by the rest of your proposal! You need to dedicate specific paragraphs to these things. In the Personal STatement, don't mention things you've done in your past without tying each one to BI and/or IM. Example: I volunteered at _______ place, which had the broader impact of reaching ____ group with scientific awesomeness.
 
Argue that your planned/current graduate program is the correct place for you to be-- this can be done by your recommenders, and/or in your personal statement. You don't have to re-hash your entire application or grad school or mention specific profs, just argue there are programs/resources/centers/labs that you need. One reviewer said he was confident about the success of my research based upon the program's fit for me.
 
Tell your letter writers they should be addressing BI and IM in their letters, and provide them with resources to know what exactly that means.
 
NUMBERS. Bullet points. Sub-titles and sections (Introduction/Context, Hypothesis, Methods, etc). Formatting is crucial to make your statement easy to read and remember. Again, this is something Anthro does not do a whole lot of, but the NSF loooooves it. Example: Personal statement: (this could be in your intro paragraph) I am prepared to do amazing research for three reasons: 1) I have a strong academic background, 2) I have a commitment to working with X population, as demonstrated through past research and volunteer experience, 3) my current grad program provides me with the resources to be successful. Research statement: This project has several broader impacts that will affect X population or Y body of knowledge: (a) __. Specifically, I will_____. (____ ©____.
 
The cookie-cutterness may be a turn-off for the social and interperative sciences, but you must must speak NSF's language to have any chance at an award.

Hey... are you guys seeing the maintenance message I am now... https://www.fastlane.nsf.gov/
 
03/28/13  -  GRFP/FastLane will be unavailable from 11:00 PM ET Thursday March 28th - 5:00 AM ET Friday, March 29th for scheduled maintenance. We apologize for any inconvenience

Hey... are you guys seeing the maintenance message I am now... https://www.fastlane.nsf.gov/
 
03/28/13  -  GRFP/FastLane will be unavailable from 11:00 PM ET Thursday March 28th - 5:00 AM ET Friday, March 29th for scheduled maintenance. We apologize for any inconvenience

Hey... are you guys seeing the maintenance message I am now... https://www.fastlane.nsf.gov/
 
03/28/13  -  GRFP/FastLane will be unavailable from 11:00 PM ET Thursday March 28th - 5:00 AM ET Friday, March 29th for scheduled maintenance. We apologize for any inconvenience

I think it sometimes gets thrown around at live presentations or when asking experts. Here's some of our members discussing it in last year's thread:  

The NSF GRFP is is tough on the emotions because competition is strong but the chances (and the payoff) are too good not to apply. I guess that's preparing us for future proposal writing.
 
Just remember, even if you aren't awarded one this year:
1) You have helped your field just by applying since a higher number of applicants from your field means more money for your field.
2) Your experience this year will definitely help you next year.
3) There is an element of luck to this ("Will I get a reviewer who just doesn't get it at all? Will I get a reviewer whose imbalanced scoring ends up hurting my total score?") so by starting now you're maximizing your chances.

The presentation they usually put up tends to give a ballpark about the number of applicants in the previous year.

This presentation quotes the 12,000 for 2011 and 10,000 in 2008 on the last page: http://www.eng.umd.edu/sites/default/files/research/2012-NSF-GRFP-Presentation-OFFICIAL.pdf
This one puts 2010 at 12,000 too: http://www.nsfgrfp.org/assets/File/2011%20NSF%20GRFP%20Presentation.pdf
and 2012 at 12,000: http://www.nsfgrfp.org/about_the_program/promotional_materials

This presentation claims a more exact 9,347 for 2008 (announced in 2009): http://www.math.lsa.umich.edu/~millerpd/docs/501_Fall09/grfp_talk.pptx This is more in line with the 10% figure for that year. This webpage also quotes that number: http://www.ohsu.edu/xd/education/schools/school-of-medicine/departments/basic-science-departments/environmental-biomolecular-systems/news/nsf_fellowship_2009.cfm

Meanwhile, gradcafe quoting a now defunct page says 8,146 eligible applications (out of 8,582 total) for awarded 2008 (application started 2007):

The other way to look at it is 2,000 awards out of 12,000 applicants. One in six.

The awards are given out proportionally per field of study so just by applying you're helping your field get a larger portion of the the total. This also means you can still go by the roughly one in six measure for your field.

I think it also works out that awards tend to split evenly by year (undergraduate, 1st year, 2nd year), but I seem to recall this isn't by policy but instead just how it usually ends up.

For the past three years, the hours have been different on the GRFP night. It's been 11pm to 5am. The midnight to 8am has been for other stuff.
"GRFP will be unavailable from 11:00 PM ET Monday April 5th - 5:00 AM ET Tuesday, April 6th for scheduled maintenance. We apologize for any inconvenience." ( )

and

"04/04/11 - GRFP/FastLane will be unavailable from 11:00PM ET Monday, April 4th - 5:00AM ET Tuesday, April 5th for scheduled maintenance. We apologize for any inconvenience." ( )

and

"FastLane will be unavailable from 11:00 PM ET Thursday March 29th - 5:00 AM ET Friday, March 30th for scheduled maintenance." ( )

...but to support guttata's enjoyment, you never know...

For the past three years, the hours have been different on the GRFP night. It's been 11pm to 5am. The midnight to 8am has been for other stuff.
"GRFP will be unavailable from 11:00 PM ET Monday April 5th - 5:00 AM ET Tuesday, April 6th for scheduled maintenance. We apologize for any inconvenience." ( )

and

"04/04/11 - GRFP/FastLane will be unavailable from 11:00PM ET Monday, April 4th - 5:00AM ET Tuesday, April 5th for scheduled maintenance. We apologize for any inconvenience." ( )

and

"FastLane will be unavailable from 11:00 PM ET Thursday March 29th - 5:00 AM ET Friday, March 30th for scheduled maintenance." ( )

...but to support guttata's enjoyment, you never know...

I see in the NSF's "Notice No. 133" released on February 27, 2103 regarding the "Impact on FY 2013 Sequestration Order on NSF Awards":
 
http://www.nsf.gov/pubs/2013/in133/in133.pdf?WT.mc_id=USNSF_80
-----------------------------------------------------------------------------
We intend to make the necessary FY 2013 reductions with as little disruption as possible to established commitments, and are using the following set of core principles to guide out sequestration planning activities:
 
* Protect commitments to NSF's core mission and maintain existing awards;
* Protect the NSF workforce; and
* Protect STEM human capital development programs
 
-----------------------------------------------------------------------------
  I hope that NSF GRFP falls under the 3rd core principle....

I emailed the GRFP general info address about whether the sequester will impact the number of awards, and I got this response -
 
 Thank you for your e-mail. To date we have not received any information about the impact of the sequester on the GRFP.

I agree with the other posters that it is not a big deal to ask for the extra night, especially since the school mentioned it specifically. I can't imagine no one else will need to stay. Even if it was just your friend, it's not unreasonable because everyone has different travel restrictions. Furthermore, it allows your friend more flexibility if things run late or there are unplanned after-visit opportunities (e.g. some current students invite him out for that last night).
 
If he wants to get back sooner (maybe he wants to take less vacation days) and he has time to get back to the hotel before hist train, he could check his luggage there for the last day. If it's a motel or a small hotel, he might want to call and verify they have that service.

anybody read about the "sham phd"? http://mygraduateschool.wordpress.com/2012/08/09/the-sham-ph-d/

people who use these services can't fake their way through their entire career. Sooner or later, it's going to catch up to them. My research group encountered one of these. Came into our group as a post doc. Didn't do sht. And then literally disappeared, and never came back. I'm not losing sleep over this sort of thing....

Unless you can have ghost writers write your references. That would be scary...

I agree with comments here, especially fuzzy logician. Very bad idea. Grad school means a ton of study, reading, research and writing (especially in PoliSci, the OP's field). If you can't write a 2-3 page statement, why go to grad school? 
 
But people actually doing it? This article was highlighted on LinkedIn. http://chronicle.com/article/The-Shadow-Scholar/125329/
I don't know if he's mostly talking about international students whose main struggle is writing academic English, but it was depressing. 

Still neither. You're hardworking and smart, but made a mistake. Everyone has a bad test or a bad assignment or an off day.

Awesome additions guys! I've added a few and incorporated Pitangus' suggestion to divide these up for easier reading. I kept the personal statement and research experience together because there's some overlap in how you want to handle them.

INTELLECTUAL MERIT
- PROPOSAL
-- whether proposal has a clear hypothesis/question it is trying to address
-- whether proposal is novel/innovative
-- whether proposal is original (your own)
-- whether proposal is transformative (rare)
-- whether proposal is well-written
-- whether access to resources is addressed in proposal
-- whether expected results are addressed in proposal
-- whether validation of results are addressed in proposal
-- whether proposal is sufficiently detailed (this is a tough one... too detailed and it might not be well-written/clear and have no room for other objectives)
-- interest of proposal among scientists
-- relation of research to theories of founders/experts in the field
-- relation to previous work the applicant discusses in other essays/supporting material
-- whether the outcome measure is well-related or "distal" to the intervention (likely more of an issue in social sciences)
- PERSONAL STATEMENT/RESEARCH EXPERIENCE/SUPPORTING MATERIAL
-- strength of academic record
-- research capabilities, independent and in team
-- publications (includes submitted/in review articles, especially if you don't have other pubs yet)
-- presentations (includes talks and posters)
-- participating in any other funded proposals (e.g. help writing grants)
-- leadership potential of applicant, including personal qualities such as persistence and drive
-- previous research essay draws connections among past projects and internships, and explains their significance (both personal + BIs)
- ALL
* strength of recommendation letters and what they say about all of the the above (including proposal if applicable)

BROADER IMPACTS
- PROPOSAL
-- whether proposal integrates/supports science education with research
-- whether proposal integrates broadening diversity with research
-- whether proposal benefits society or some large group outside major field
-- potential of research to support "citizen science"
-- applicant's plan to disseminate results broadly, especially with non-science stakeholders if applicable, but further too
- PERSONAL STATEMENT/RESEARCH EXPERIENCE/SUPPORTING MATERIAL
-- benefits to society, education or diversity in all past research projects
-- applicant's activities to increase science education and recruit young scientists
-- applicant's activities to broaden diversity in science
-- applicant's leadership roles in broader impacts activities (including TAing)
-- quantity of applicant's BI activites (e.g. reviewer lists lots of examples or length of time)
-- applicant's plan to -continue- broader impacts activities
-- "unique perspective" due to unusual life circumstances which gives applicant insight into the needs of underserved communities
- ALL
* specifics for BI activities in all sources
* strength of recommendation letters and what they say about all of the the above

@Dynamom: Maybe next year's thread's kickoff should be a group mining expedition of all the previous threads. I read them all at some point (including all 170+ pages of the year-less one) and incorporated trends but didn't write the information down back in November when I probably should have.

Graduate school applications in the US are a lot less of a numbers game than undergraduate in the US. GRE scores, GPA, and the like are all about not raising red flags and being filtered out than they are about rising to the top. These rankings are probably not much different. You use the numbers to validate your application. You use the narrative of your recommendation letters, your personal statement, your writing samples, your research experience and publication record to actually get in.

I doubt the kind of inflation that makes a difference goes on. The professors doing the rankings and reading the rankings are part of a small community. If someone seriously inflates their rankings, it will get noticed. It's not worth their reputation with their peers, especially for a student who needs inflating. Whether it's top 1% or top 10%, what's being said (and not said) in the letter is going to be telling the reviewing professors the real story.