2025-2026 Wake Forest

Started by wysdoc
This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
Advertisement - Members don't see this ad
That was carnage. The following is most definitely tone deaf. Using chat GPT I made this write up on the waitlist. I hope you like it, I was hungover and it was kind of fun to make:


Estimating Waitlist Admission Probability at Wake Forest School of Medicine (WFUSM) Using Public Aggregate Data and Self-Reported Applicant Outcomes (2025–2026 Cycle)

Introduction
Wake Forest does not publicly report the size of its waitlist in a way applicants can directly use to compute individual probabilities. However, the admissions system still has structural constraints that allow a reasonable, transparent estimate of the average probability of receiving a waitlist offer.

The core challenge is that “waitlist movement” is governed by a small set of quantities that interact mechanically:
  1. How many seats must ultimately be filled by waitlist candidates (waitlist matriculants).
  2. How many waitlist offers are required to produce that many matriculants (depends on yield).
  3. How many people are on the waitlist (the denominator for any average probability).
This write-up documents, in explicit detail, how those quantities were obtained or bounded, and how they were combined into an estimate for the 2025–2026 cycle.

Methods
Data sources
We used two categories of information:

A. High-reliability aggregate inputs (structural constraints)
These are the variables that typically change slowly year-to-year and are usually reported in official aggregate form:
  1. Class size (C)
    We used a class size of approximately C = 200 seats (as discussed previously from MSAR-style aggregate reporting). This is treated as a structural constant because schools rarely change class size dramatically from one year to the next without announcing it.
  2. Fraction of the entering class that ultimately comes from the waitlist (f_WL)
    We used f_WL ≈ 0.30 (about 30% of the class). This implies that the number of waitlist matriculants is approximately:
    M_WL = C * f_WL
    With C ≈ 200 and f_WL ≈ 0.30, this yields:
    M_WL ≈ 200 * 0.30 = 60 waitlist matriculants
Important distinction: matriculants are not offers. “60 matriculants from the waitlist” means ~60 people who were originally waitlisted ultimately enrolled, not that only 60 people were offered admission off the waitlist.

B. Lower-reliability but useful self-reported data (behavioral proportions and cross-checks)

To understand decision patterns within the applicant pool (and to sanity-check plausible waitlist sizes), we used the self-reported dataset from admit.org for the Wake Forest 2025–2026 cycle. This dataset was treated as a convenience sample, not a census, meaning it can inform plausible proportions but cannot be assumed to match the school’s true totals.

Operationally, the dataset we analyzed (exported to spreadsheet form) contained:
  • N_total = 964 Wake Forest applicant entries
  • N_int = 114 entries with an interview date recorded
Among the 114 interviewed entries:
  • Accepted: 53
  • Waitlisted: 49
  • Rejected: 3
  • No decision reported (unresolved): 9
So, within the interviewed subset, the observed decision breakdown was:
  • Accept rate among interviewed (counting unresolved as not-accepted):
    A0_obs = 53 / 114 = 46.5%
  • Waitlist rate among interviewed (counting unresolved as not-waitlisted):
    WL_obs = 49 / 114 = 43.0%
  • Rejection rate among interviewed:
    R_obs = 3 / 114 = 2.6%
  • Unresolved rate:
    U_obs = 9 / 114 = 7.9%
Because those 9 unresolved interviewed cases could later become accepted, waitlisted, or rejected, we explicitly stress-tested three variants:
  1. Proportional allocation (neutral): distribute unresolved in the same proportions as resolved cases
    Among resolved interviewees (114 − 9 = 105), the shares were:
    Accepted = 53/105 = 50.5%
    Waitlisted = 49/105 = 46.7%
    Rejected = 3/105 = 2.9%
    Allocating 9 unresolved proportionally yields roughly: +4 to 5 accepted, +4 waitlisted, ~0 rejected.
  2. Worst-case for an applicant’s waitlist crowding: all 9 unresolved become waitlisted
    This increases the implied waitlist fraction among interviewed to:
    (49 + 9) / 114 = 58/114 = 50.9%
  3. Best-case for waitlist crowding: all 9 unresolved become rejections
    This decreases the implied waitlist fraction among interviewed to:
    49 / 114 = 43.0%
These stress tests matter because they constrain plausible waitlist size if you also assume a plausible number of total interviews.

Core model
The objective was to estimate the average probability that a randomly positioned waitlisted person receives an offer off the waitlist.

We defined:

C = class size (seats), approximately 200
f_WL = fraction of the class that ultimately matriculates from the waitlist, approximately 0.30
M_WL = number of waitlist matriculants = C * f_WL
y_WL = yield on waitlist offers (the fraction of waitlist offers that become matriculants)
O_WL = number of waitlist offers required = M_WL / y_WL
W = total number of waitlisted individuals (waitlist size)
p_WL = probability an average waitlisted person receives an offer = O_WL / W

Thus the key relationship is:

p_WL = (C * f_WL) / (y_WL * W)

Assumptions for waitlist yield (y_WL)
Wake does not publish “waitlist yield.” We therefore modeled y_WL as a plausible range rather than a point estimate. The range used was:

y_WL ≈ 0.50 to 0.65

Interpretation: if Wake makes 100 waitlist offers and 60 of them enroll, the yield is 60%. If yield is lower, Wake must issue more offers to get the same number of matriculants.

Assumptions for waitlist size (W)
Waitlist size is the single largest unknown. We bounded it using two approaches:

  1. “Typical” institutional scale argument (baseline range)
    For a school of ~200 seats interviewing on the order of ~500 applicants, a waitlist size in the ballpark of 250–350 is commonly plausible. This is not a claim about Wake’s true number.

  2. Self-reported cross-check using the admit.org interviewed subset

    Within the admit.org sample, about 43% to 51% of interviewed applicants were waitlisted (depending on how the 9 unresolved are treated). If one assumes total interviews of roughly ~500, then an implied waitlist size would be approximately:

W ≈ 500 * (0.43 to 0.51) = 215 to 255

Because admit.org is not a representative sample (and because not all interviewees update outcomes), we did not force W to equal this implied value. Instead, we treated it as a reason that the lower end of the baseline range (near ~250–300) is plausible, while still retaining a wider range (250–350) for robustness.


Results
Step 1: Estimate waitlist matriculants
Using C ≈ 200 and f_WL ≈ 0.30:
M_WL ≈ 60

Step 2: Convert waitlist matriculants to waitlist offers
With y_WL in [0.50, 0.65]:

If y_WL = 0.50:
O_WL = 60 / 0.50 = 120 waitlist offers

If y_WL = 0.65:
O_WL = 60 / 0.65 = 92 waitlist offers (rounded from 92.3)

So the implied number of waitlist offers is approximately:
O_WL ≈ 92 to 120

Step 3: Convert waitlist offers to an average waitlist probability
We then divide by waitlist size W.

Illustrative mid-point scenario (often a useful anchor):
C = 200, f_WL = 0.30, y_WL = 0.60, W = 300
M_WL = 60
O_WL = 60 / 0.60 = 100
p_WL = 100 / 300 = 33.3%

Range using the baseline waitlist-size bracket (W = 250 to 350)
  • Best end of this bracket (smaller waitlist, lower yield):
    O_WL ≈ 120 and W = 250 gives p_WL ≈ 48%
  • Worst end of this bracket (larger waitlist, higher yield):
    O_WL ≈ 92 and W = 350 gives p_WL ≈ 26%
This produces a broad but mechanically consistent range:
p_WL ≈ 26% to 48%

Most-plausible central band
When we focus on “middle” assumptions (W around ~300; yields ~0.55–0.65; f_WL near ~0.30), the implied probability concentrates around the low-to-mid 30s. That is where the earlier “~30–37%” estimate came from: it is essentially the interquartile region of plausible inputs, not a hard bound.

A practical summary consistent with the above mechanics is:

  • Typical/central estimate: about 30% to 37%
  • Broader sensitivity envelope: about 26% to 45% (and up to the high 40s under particularly favorable combinations)
Admit.org cross-check: interviewee outcome proportions (context, not a direct probability)

Within the admit.org interviewed subset (N_int = 114), the observed acceptance fraction was 46.5% (53/114) at the time of export, and the waitlist fraction was 43.0% (49/114). This supports two qualitative points:

  1. A large portion of interviewed applicants end up in the “not immediately rejected” pool.
  2. The waitlist appears to be a substantial component of the post-interview funnel, which is consistent with the need for significant waitlist matriculation (around 60 seats).

Discussion
This model estimates an average probability for someone on the waitlist, not a personalized probability conditional on rank.

In real admissions, waitlists are not random. Schools often rank or tier the list, and movement is concentrated near the top. That means:
  • If you are near the top third, your probability is higher than the average computed here.
  • If you are near the bottom third, your probability is lower than the average computed here.
  • Without rank information, the best defensible estimate is an average over positions.
Why the waitlist size dominates the uncertainty
The equation p_WL = (C * f_WL) / (y_WL * W) makes the sensitivity obvious:
  • If W is 250 vs 350, probability changes by a factor of 350/250 = 1.4 (a 40% swing).
  • If yield is 0.50 vs 0.65, probability changes by a factor of 0.65/0.50 = 1.3 (a 30% swing).
  • If f_WL shifts from 0.25 to 0.35, probability changes by a factor of 0.35/0.25 = 1.4 (again a 40% swing).
So even small uncertainty in these inputs produces meaningful changes in predicted probability, which is why we reported ranges rather than a single number.

How the admit.org dataset helps, despite being biased

Self-reported datasets have well-known biases:

  • People who receive “interesting” outcomes (acceptance/waitlist) may be more likely to update.
  • People who are rejected may stop updating.
  • Interview dates and decision timestamps may be missing or inconsistent.
  • The sample is not random and may not reflect the school’s true denominators.
Despite that, the dataset is still valuable for bounding plausible proportions (for example, “is immediate post-interview rejection common?” and “is waitlisting a large chunk of the interview cohort?”). It also helped us explicitly handle missingness via stress tests (the 9 unresolved interviewed cases).

Key limitation: “odds this year” still inherit last-year structure
Even if the admit.org sample is from the 2025–2026 cycle, the most important quantitative driver of the waitlist model is still the class-fill requirement (how many seats must be filled after initial offers decline). That is fundamentally a yield problem. Without official, cycle-specific yield data from Wake (overall yield and waitlist yield), any “this year” estimate necessarily assumes the school behaves broadly like recent years.

Conclusions
  1. Using class size (~200) and the reported fraction of the class that comes from the waitlist (~30%), Wake needs roughly 60 waitlist matriculants each year.
  2. 60 matriculants implies substantially more waitlist offers, on the order of about 92 to 120.
  3. The missing piece is the waitlist size. Using a plausible waitlist-size range (roughly 250–350, with admit.org suggesting the lower end is not crazy), the implied average probability of receiving a waitlist offer is most defensibly centered in the low-to-mid 30% range.
  4. A reasonable headline summary for a typical waitlisted applicant is approximately 30% to 37%, with a wider sensitivity range of roughly the high 20s into the low-to-mid 40s depending primarily on waitlist size and yield.
  5. Individual rank/tiering can move a person substantially above or below this average, but rank is usually unobservable to applicants; therefore, ranges and sensitivity analysis are the most honest way to report uncertainty.
 
Hurricanado said:
I feel that applicants should indeed be very aware, going in, that there are no merit scholarships given at this medical school. Wake Forest Med seems deeply entrenched in DEI. They don't hide that and why not believe what they communicate. DEI would logically/quite possibly apply in the admissions decisions and there is also a very high ratio of female to male students in its accepted classes. It's a decent medical school and gets tons of applications, maybe in part because it looks more mid and lower stat in the stat breakdowns on MSAR and so on, but that doesn't paint the whole picture. If you don't have the means to toss away money or desire to be making a donation and spend your time with the application stuff here too, think hard. Also, if you're not who/what they're apparently looking for here &/or are going in thinking you might score some scholarship money, think hard about keeping it on your application list.
Click to expand...
current student here with a merit scholarship (merit is literally in the name of the scholarship) so just wanted to share for the knowledge of prospective students as decisions are being made that you can’t say they never give scholarships. I also have no financial need/DEI factors (when I applied or currently), so the above statement is not entirely absolute 🙂
 
If they gave merit money they certainly didn't post it on their website or talk about it when I was applying at least. I even asked once and was told no. Cool for anyone who actually got any though.
 
Advertisement - Members don't see this ad
thatssotrue said:
current student here with a merit scholarship (merit is literally in the name of the scholarship) so just wanted to share for the knowledge of prospective students as decisions are being made that you can’t say they never give scholarships. I also have no financial need/DEI factors (when I applied or currently), so the above statement is not entirely absolute 🙂
Click to expand...
Did you have to fill out the CSS Profile for this scholarship?
 
Do we think we have to wait until April 15th to find out if we got any scholarships? Or do you think they’ll release them sooner?
 
Also, Ik its early to consider whether ppl with A's are going to WFUSM but just as a early sign, are there ppl with As to WFUSM looking elsewhere?? WL warrior speaking
 
Does anyone have any context or info about how Wake Forest tends to pick from their waitlist? A fellow waitlist warrior asking
 
dogman12 said:
Does anyone have any context or info about how Wake Forest tends to pick from their waitlist? A fellow waitlist warrior asking
Click to expand...
I'd also like to know the answer to this. However, I would assume that it's not released. That being said, IF someone has information I am all ears.
 
Are there any accepted/waitlisted students who have been accepted at another school they like more and are deciding to matriculate elsewhere? Just a fellow WL applicant wondering..
 
Advertisement - Members don't see this ad
Advertisement - Members don't see this ad
spark1206 said:
No news for me today, hoping for tomorrow 🙏🏻
Click to expand...
Nothing here either. FYI for anyone that came here because admit closed, you can still access the 2025-2026 thread if you click on School Forums. There is a drop down to change the application year from 2026-2027 back to 2025-2026. Took me a while to realize, thought I would share lol.
 
You must log in or register to reply here.

Similar threads

L
2025-2026 UCLA (PRIME)
2 3
Replies
124
Views
27K
stethosquirrel
S
wysdoc
2025-2026 University of Georgia School of Medicine
2 3 4 5 6
Replies
281
Views
96K
cheeseplease
C
wysdoc
2025-2026 Stanford
8 9 10 11 12
Replies
596
Views
187K
holajola
H
wysdoc
2025-2026 Temple
10 11 12 13 14
Replies
691
Views
204K
hopefulgiraffe
hopefulgiraffe
wysdoc
2025-2026 Tufts
5 6 7 8 9
Replies
421
Views
195K
lebronjames1243
lebronjames1243