Fansign Application Guide

Math behind draw methods, worked examples, round splitting, video and overseas entries, and how the numbers are computed.

Last updated

1. Three draw methods (as modelled here)

Most sellers do not publish their draw method, weights or caps. This simulator therefore treats the three methods fans commonly talk about as models (assumptions) and lets you compare how much the odds change when only the method changes.

  • Machine draw: 1 album = 1 ticket. All tickets go into one pool and winners are drawn without replacement.
  • Hand draw: people are drawn, weighted by album count, but weight stops growing at a cap (default 3, adjustable 1–5).
  • Mixed: part of the winner slots by machine, the rest by hand; you set the split (default 70% machine).

Same scenario (3,000 applicants · others average 25 albums · 100 winners): with 5 albums hand draw is better (3.3% vs machine 0.7%); with 300 albums machine draw is far better (32.9% vs hand 3.3%).

My albumsMachineHand (cap 3)Mixed (70% machine)
10.1%1.1%0.4%
50.7%3.3%1.5%
101.3%3.3%1.9%
303.9%3.3%3.7%
8010.1%3.3%8.1%
15018.1%3.3%13.9%
30032.9%3.3%25.2%
50048.6%3.3%37.8%

2. How odds change as you add albums (machine)

Machine-draw odds use sampling without replacement (hypergeometric): with M tickets in total, k of them yours and K winners, the chance none of yours is drawn is (M−k)/M × (M−k−1)/(M−1) × … (K terms); your odds are 1 minus that. Here the other applicants hold about 74,975 tickets.

  • 10 albums: ~1.3%
  • 30 albums: ~3.9%
  • 80 albums: ~10.1%
  • 150 albums: ~18.1%
  • 300 albums: ~32.9%
  • 500 albums: ~48.6%

Odds grow almost linearly while your share is small and flatten as it grows. Clearing 50% here takes 522 albums; the calculator’s "albums for 50%" finds this point for your inputs by the same binary search (1–10,000).

3. Is hand draw more equal?

In the model, yes. With a cap of 3, everyone with 3+ albums has the same weight, and the gap to a 1-album applicant is at most 3×. Here 1 album gives 1.1% and 3+ albums give 3.3% regardless of count. Sellers rarely confirm how they draw, so read this as "if it were a hand draw".

4. What the popularity tier changes

The tier multiplies your "average albums per applicant" input: niche ×0.7, mid ×1.0, major ×1.4, mega ×2.0. There are no official statistics on average entries or cutoffs by group, and community cutoff reports vary widely by event, seller and round. If you can estimate this event’s entry size, enter the average directly and keep the tier at mid (×1.0).

5. Video-call vs in-person fansigns

  • Applicant pool: video calls need no travel, so more regional and overseas fans may apply.
  • Winner slots and time per fan: these differ per event — enter the slot count from the announcement.
  • Duplicate wins: some events limit wins per person; check the notice before applying.

6. Notes for overseas fans

  • Seller conditions: some sellers ship and take payment internationally (e.g. Ktown4u), but eligibility and ID checks differ by seller and event — read the notice first.
  • Video events: no travel needed, but times are in KST.
  • Proxy applications: wins can be cancelled at the ID check; not recommended.
  • Transfers and resale: often banned by seller terms; detected wins can be cancelled.

7. Round splitting — the math

Spending the same 100 albums (scenario above, same competition every round, machine draw):

  • 100 albums × 1 round: 12.5%
  • 50 albums × 2 rounds: 1 − (1 − 6.5%)² = 12.5%
  • 20 albums × 5 rounds: 1 − (1 − 2.6%)⁵ = 12.5%

Under the machine model, the chance of winning at least once is practically the same either way. The benefit of splitting is sequencing: if you can see a round’s result before buying for the next, an early win lets you stop and lowers average spend. Real rounds have different applicant counts; the calculator’s "rounds" input treats each round as an independent draw under the same conditions.

8. Macros, scalping and gray areas

Application macros, transferring entries and reselling won slots often break seller terms, and detected cases can lead to cancelled wins or account action. This site does not endorse them and only simulates legitimate entries.

9. Leftover sealed albums

  • Resale: second-hand platforms (prices vary by album).
  • Donation: ask recipients in advance.
  • Recycling: sorting rules differ by local government.
  • Fandom sharing: many fans give the album body away in fan communities.

10. How this simulator computes — method and limits

  • Machine: single-round odds computed directly with the without-replacement product (exact up to 5,000 winners and 1M tickets; beyond that the (1 − k/M)^K approximation).
  • Hand: a per-person weighted draw approximated by assuming each winner is another applicant with average weight — an approximation, not an exact result.
  • Mixed: machine and hand stages combined as if independent; slightly optimistic versus real draws where a winner is removed from the next stage.
  • Multiple rounds: 10,000 Monte Carlo runs of independent rounds give "at least one win" and expected wins; the same value follows from 1 − (1 − p)^rounds, and the sampling error of 10,000 runs is at most about ±0.5 percentage points (one standard deviation).
  • Validation: not checked against real event outcomes. Inputs drive the result — treat it as a comparison aid.

11. Sharing results

"Share settings" copies a link carrying your inputs as URL parameters, so a friend can swap in their own album count. "Share to Threads" opens a Threads compose window with the result text; "Copy for KakaoTalk" copies the text and link for pasting into a chat.

12. More questions? See the FAQ

Seven common questions, including worked odds and the 50% reverse-solve, are on the FAQ page.

Sources

Checked 2026-11-06 · Draw methods and entry sizes have no official statistics, so the draw models and tier multipliers here are assumptions.

Author Jikwang Kim (maintainer)Last updated bal.pe.kr 마이크로 SaaS

Back to calculator