Hacker News new | past | comments | ask | show | jobs | submit login
Ask PG: Application Rating System?
13 points by bharath on Oct 5, 2007 | hide | past | favorite | 16 comments
<p>From your latest essay:<p>"Sometimes it literally is software, like Hacker News and our application rating system."<p>This is the 1st time I've heard about you guys using some kind of automated application rating system. Would you be willing to discuss how it works?



To get to the interview stage:

Good college +2

Done with schooling +3

Single Founder -5

No Hackers on Team -4

Mention Lisp +2

Mention Smalltalk +2

Mention Python +1

Zero Hackers Willing to Relocate -5

Past or Recent TechCrunch Coverage +2

Past or Recent Mainstream Coverage +2

Google search finds many top blogs or major news sources linking to current or past project +3

Team of hackers previously founded a web startup +1

Team of non-hackers previously founded a web startup +2

Not drinking age -3

Average age 23 to 26 +4

Average age 30 to 35 -1

Average age 36 to 45 -2

Average age older than 45 -10

Knows idea should be a web app +3

Short and sweet application +3

Estimated scores for some companies at the time of application: Xobni 12 (MIT hackers), Auctomatic 13 (No hackers at time of application), Justin.tv 25 (Lots of pluses, no negations)

I'd say at 10 you have a good chance at an interview, at 13-15 an excellent chance. This is derived from Paul's writings, speeches, videos, and comments.

The actual idea doesn't matter because Paul will either tell you he likes it or recommend changing it during the interview. If you're not done with college, now is the time to say you will focus on the project instead.

Again, this is just a rough guide and I have zero qualifications to even write this. After the interview, only your execution matters.


I predict that tomorrow someone will take these scores and turn them into a webapp that rates applications. It looks like ycombinatorrater.com is available...


"Good college +2" seems counter to some of his latest writings. I guess "No Hackers on Team" would be a way greater penalty. I'm curious: why is the bonus for "founded a web startup" higher for a team of non-hackers?


If non hackers created a successful web startup, then they must have some other kind of talent going for them, which is a plus since most others in the YC community have the hacker thing going for them.


Yeah, but looking at the Y Combinator funding trends, the majority of their companies that succeed are by founders from "good colleges."


Correlation does not imply causation. http://en.wikipedia.org/wiki/Correlation_does_not_imply_caus...

Being a motivated high achiever probably correlates well with going to a good college, and with getting into YC.


Right, thus going to a good college = +2.


Well whatever the case, your college didn't teach you statistics.


I completely understand the difference between correlation and causation. But in the grand scheme of things, the human brain is designed to recognize correlation when developing internal rating systems. You would have to be stupid to not automatically give more clout to a graduate of Harvard than to a drop out of some state university. The state university student may or may not be less qualified than the Harvard graduate, but in the vast majority of situations he will be.

Even if YC doesn't give "real" points to graduates/students of good universities, just the mere emotional response to reading "harvard" or "yale" or "berkeley" sparks an automatically positive regard for the applicant. Isn't that enough cause?

Thus, good college = +2.


From the gospel, according to Paul... (September, 2007)

"Between the volume of people we judge and the rapid, unequivocal test that's applied to our choices, Y Combinator has been an unprecedented opportunity for learning how to pick winners. One of the most surprising things we've learned is how little it matters where people went to college.

I thought I'd already been cured of caring about that. There's nothing like going to grad school at Harvard to cure you of any illusions you might have about the average Harvard undergrad. And yet Y Combinator showed us we were still overestimating people who'd been to elite colleges. We'd interview people from MIT or Harvard or Stanford and sometimes find ourselves thinking: they must be smarter than they seem. It took us a few iterations to learn to trust our senses.

Practically everyone thinks that someone who went to MIT or Harvard or Stanford must be smart. Even people who hate you for it believe it.

But when you think about what it means to have gone to an elite college, how could this be true? We're talking about a decision made by admissions officers--basically, HR people--based on a cursory examination of a huge pile of depressingly similar applications submitted by seventeen year olds. And what do they have to go on? An easily gamed standardized test; a short essay telling you what the kid thinks you want to hear; an interview with a random alum; a high school record that's largely an index of obedience. Who would rely on such a test?"

college = 0


You don't lose points for college in my system. Remember, Paul says he accepts everybody who is good enough, not the top 20 teams. Whichever way you reach 13-15 on my scale, you're likely to get an interview. We'll have to see.

"And yet Y Combinator showed us we were still overestimating people who'd been to elite colleges."

Paul "still overestimates" such students, so they're more likely to get an interview. And since he used the word "still", it means he feels he's still biased at this present time. How do you think these test cases came about, anyway?

"college = 0"

I doubt that this can ever be true. Paul still asks for the students' colleges. In other words, he's still figuring this out. What he means is, nobody needs to have a degree or a good college to create a successful startup. That is true. But I do think there are good signs about a person that uniquely pre-qualify them if the student is from a good college:

1) The fact that the school accepted them, matters least (like Paul says in the quote above.)

2) However, I do think Paul considers it a sign of dedication if a student from MIT wants to leave or delay entry in favor of YC, or if they've recently graduated from Harvard or Stanford and instead of proceeding to graduate school or possibly a six-figure job, to spend a couple of years living on ramen noodles.

3) Regardless of what Paul thinks, VC's probably love that stuff, if the startup doesn't have lots of traction yet.

The big VC investments seems to go to those from elite schools: MIT (no product when selected by YC, 4m VC before product was finalized), Cambridge (no product when selected by YC, 400k VC), Yale (no product when selected by YC, unknown VC money), Stanford (likely no product when selected by YC, Sequoia Capital VC).

4) I suggested good colleges would get +2, not just elite colleges.


>Past or Recent TechCrunch Coverage +2

Does Mashable count? :p


It isn't the rating part that's automated. It's just a convenient tool for us to read and score applications quickly.


What makes you think that the rating software is "automated"? I bet it's just a tool used by the YC partners. It could be as simple as a human-assisted radix sort, so the 1000 applications first get binned into 10 bins, then each of the apps in each bin are further subdivided on a 1 to 10 scale, etc.


The best applicants get more points ;)

In order to maximise this you need to

a) have a good idea b) have a good team c) be able to execute

I believe this should be sufficient to 'game' the system ;)


I think the application would pull out all news.yc submissions and comments for the applicants and probably even show the top 10 google search results for each founder's name. Additionally, I think all the 4 partners go through all applications and they all probably rate it. The rating system probably binds the end result for easy consumption.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: