Stanford or Harvard? Forum

[Machine Spirit]

its 110%....

...Did you even read my post? We all know that it adds up to 110%. The question was over whether or not those two figures are even in the same statistical pool though, and thus whether they should be added in the first place. It wasn't that big a deal. Moving right along now...

I agree with dbt's commentary on Stanford's comparative significance as a transfer school.

[Haribo]

43% of people who get accepted to Stanford go there.
67% of people who get accepted to Harvard go there.

Empirically, most people who have to make the call probably go with Harvard.
Edit: source: http://www.lawschoolnumbers.com/

This is a terrible statistic, given the difference in class sizes between Harvard and Stanford (and the elephant in the room, which is Yale, which has much higher yield rates and probably gets an equal number of people from Stanford and Harvard - skewing the numbers badly.)

I wish disco_barred was still handing out "Dumbest Comment of the Day" awards. Math fail.

Well, this doesn't really apply to this post, as the OP is transferring, for all the reasons already stated, but I feel obligated every time the issue comes up to disprove the idea that yield rates mean anything (and every time some idiot who doesn't understand percentages disagrees with me.)

Let's do a little back of the envelope calculation.

Stanford accepts 400 students, Harvard accepts 800, Yale accepts 200. Let's assume all 200 of those Yale acceptees decide to attend Yale. Stanford will end up with a 50% yield rate, and Harvard with a 75% yield rate. Does the difference in yield rate mean anything? Obviously not.

Now those numbers weren't perfect - Yale isn't 100%, and Stanford and Harvard are both slightly lower (43 and 67) but they should show the limitations of using yield rate when class sizes are wildly different.

Sorry OP, /end dethread, this really doesn't have anything to do with your question, but it's a bit of a pet peeve of mine.

[Na_Swatch]

Actually Haribo, the gap between 43% and 68% is significant enough that, factoring in the difference in class sizes, you still display a higher preference for H versus Y.

Basically, for simplification purposes, assume like you did that 200 Yale admits also got into Harvard and Stanford. (A generalization based on the 75% Yale yield. If you use the actual 75%, the math is more complicated but you get the same results.)

With class sizes being (slight rounding here):
200 Yale
200 Stanford
550 Harvard

This means that
Harvard, at 68% yield, admits = 810
Stanford, at 43% yield, admits = 465

810 - 200 Yale Admits - 550 Harvard Admits = 60 turn down Harvard (Excluding Yale Admits)
465 - 200 Yale Admits - 200 Stanford Admits = 65 turn down Stanford (Excluding Yale Admits)

Thus you have more people turning down Stanford than Harvard, even though the Harvard class is more than double that of Stanford. Again the difference is not as exaggerated as it appears, due to the fact that Stanford's smaller class size means that there are more people who are only accepted to Harvard out of HYS than people only accepted to Stanford out of HYS.

However, this definitely illustrates basic yield choice between a person facing Harvard or Stanford is not equal. If it was, then the ratio to fill a portion of the class would be same for the first 200 students at each school. This means that to fill the first 200 of Stanford (entire class) you would need to accept 265 students (excluding the Yale 200). Similarly to fill the first 200 Harvard students you would need to accept 265 students (excluding the Yale 200). At this point, that would mean the remaining yield on the last 300 Harvard students accepted would be over 100%, illustrating that a 1:1 choice selection ratio given both Harvard and Stanford as options is not the case.

Anyways, some confounding factors such as Yale is actually 75% and the effect of say schools like Columbia with scholarships will skew these results slightly and obviously the gap between H and Y is much, much narrower than the outward appearance of 68% vs 43% yield. Just wanted to show that mathematically, these numbers still indicate at least a slight preference for Harvard over Stanford.

Also, I am utterly bored waiting for my megavideo 50 minute wait time to be over so I can continue watching my movie which is why my analysis got so long.

[Haribo]

Actually Haribo, the gap between 43% and 68% is significant enough that, factoring in the difference in class sizes, you still display a higher preference for H versus Y.

Basically, for simplification purposes, assume like you did that 200 Yale admits also got into Harvard and Stanford. (A generalization based on the 75% Yale yield. If you use the actual 75%, the math is more complicated but you get the same results.)

With class sizes being (slight rounding here):
200 Yale
200 Stanford
550 Harvard

This means that
Harvard, at 68% yield, admits = 810
Stanford, at 43% yield, admits = 465

810 - 200 Yale Admits - 550 Harvard Admits = 60 turn down Harvard (Excluding Yale Admits)
465 - 200 Yale Admits - 200 Stanford Admits = 65 turn down Stanford (Excluding Yale Admits)

Thus you have more people turning down Stanford than Harvard, even though the Harvard class is more than double that of Stanford. Again the difference is not as exaggerated as it appears, due to the fact that Stanford's smaller class size means that there are more people who are only accepted to Harvard out of HYS than people only accepted to Stanford out of HYS.

However, this definitely illustrates basic yield choice between a person facing Harvard or Stanford is not equal. If it was, then the ratio to fill a portion of the class would be same for the first 200 students at each school. This means that to fill the first 200 of Stanford (entire class) you would need to accept 265 students (excluding the Yale 200). Similarly to fill the first 200 Harvard students you would need to accept 265 students (excluding the Yale 200). At this point, that would mean the remaining yield on the last 300 Harvard students accepted would be over 100%, illustrating that a 1:1 choice selection ratio given both Harvard and Stanford as options is not the case.

Anyways, some confounding factors such as Yale is actually 75% and the effect of say schools like Columbia with scholarships will skew these results slightly and obviously the gap between H and Y is much, much narrower than the outward appearance of 68% vs 43% yield. Just wanted to show that mathematically, these numbers still indicate at least a slight preference for Harvard over Stanford.

Also, I am utterly bored waiting for my Megavideo 50 minute to be over so I can continue watching my movie which is why my analysis got so long.

Well, first of all I think both of our analysis are flawed by rounding. Since we'll never know exactly how many Yale students were dual-enrolled at Harvard and Stanford, neither can be perfect. However, your analysis shows some pretty heavy-duty rounding errors.

Anyway, I was bored as well so I decided to run the actual numbers based on LSN.

Stanford: 171 attending/398 admitted
Harvard: 558 attending/834 admitted
Yale: 189 attending/249 admitted

Calculations:
Stanford: 398-189-171 = 38 students chose to enroll elsewhere
Harvard: 834-189-558 = 87 students chose to enroll elsewhere

These numbers are quite different from your own

Of course, we should correct them for class size. Stanford now has a 78% yield rate, and Harvard is sitting at 84%. Harvard still has a slight edge, but I'd argue at 6% it's not worth writing home about.

Obviously, this is kind of a stupid, nit-picky little pet peeve of mine, but I don't get to play with numbers nearly enough anymore. Thanks for your reply, it was fun

[Na_Swatch]

Haha yeah I know what you mean about playing with numbers, its pretty fun (I didn't use actual numbers before because I didn't know the exact admit/ attending for HYS)... However for your calculations there you're omitting the extra 25% who decided not to attend Yale and likely chose one of Stanford or Harvard instead. These would be added back into the applicant pool that would be competitive for Stanford and Harvard increasing your yield estimate difference b/w the two by a little more.

And yeah I did a rough estimate with real numbers and came out with the slight edge for Harvard... I'd say 6% ~ 9% isn't insignificant when its repeated year in and year out... About what you'd expect imo. Still definitely not a big one at all, which I made clear in that straight up comparing 68% to 43% is deceptive

[Salted5]

Wow love the math analysis. In my defense, I did say that the comparative yield numbers probably mean that most people who have the choice choose harvard. No way to really know for sure without really datamining LSN.

[MarkTwain]

Yield for 1Ls is not really relevant. You're a transfer. You probably can't find yield information for transfers, but you should be looking at the opportunities to excel as a transfer at either school. From what I can tell, for the reasons I mentioned (small class size, LR shot), Stanford is better for transfers.

Agreed.

There are a lot of unique reasons to choose one over the other in the transfer context. Ex: Very good chance at SLR (anywhere from 40-60%), no chance at HLR.

[sparky1pixie]

Tuesday HLS
Friday SLS

[mallard]

I would pick Stanford since Haribo is there. I love that yield analysis.

[snowboarder]

One thing that hasn't been mentioned is course selection and relative strengths in different areas. If I were deciding, HLS would get a big bump for the sheer number of courses and opportunities. If you know what you want to do, I would see what opportunities each school offers in that field. Obv you can't really go wrong either way

[Na_Swatch]

I would pick Stanford since Haribo is there. I love that yield analysis.

Malllard you never gave me that PM about your experience with grades/ studying at HLS!