アズプロは3,000万件の事前登録を達成できそう？【海外の反応】

https://reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2Fasnpaag37sfh1.png%3Fwidth%3D1920%26format%3Dpng%26auto%3Dwebp%26s%3De104ad4a3cb1c0eeb6f71e28b789d7b7a1c58a23

The pre-reg campaign tops out at a 30M reward, and I’ve been logging the global counter every 30 minutes for about a year (~18,000 readings). Here’s whether it actually gets there.

Short version: 20M is close to certain and 25M is likely (that’s the free Anais, the reward most people want), but 30M is roughly a coin flip that depends on how big the launch is. A 2026 launch looks likely (Nexon’s Q1 investor deck lists AP as a 2026 release), and pre-reg spikes hardest right before a game ships, so 30M comes down to whether that launch spike is big enough before the campaign closes.

# Where we are

16.5M as of July 27. The 15M reward is already claimed; 30M is roughly double the current count.

|Milestone|Reward|Left to go|
|:-|:-|:-|
|20 million|Pure Memory Trace ×5|3.5M|
|25 million|Character “Anais”|8.5M|
|30 million|Pure Memory Trace ×12|13.5M|
|5M social followers|Kibo “Starhue Floating Cloud”|not reached|

The tier that matters is 25M: a free Anais, and it’s what most people are pre-registering for. 20M and 30M are Pure Memory Trace (pull materials), and 5M followers is a Kibo skin.

https://reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2Fo1xmcatzfsfh1.png%3Fwidth%3D1650%26format%3Dpng%26auto%3Dwebp%26s%3De73db2ac5d8d3af5822d6cb00fd1ee4732ba7e93

Nothing much happened for most of 2025 (5M to ~7.5M, then flat for months). Spring 2026 is when every region started testing and it climbed sharply. 30M needs +13.5M, close to everything the game has gained across the whole tracked period.

# Growth only happens during tests

On a quiet day the count adds ~2,700. During a test or marketing beat it adds tens of thousands: CBT2 sign-ups peaked around 130k/day, the Japan test hit ~114k/day, and the single biggest day was 185,000 during the first CN beta.

https://reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2Fn4i6wrv5gsfh1.png%3Fwidth%3D1650%26format%3Dpng%26auto%3Dwebp%26s%3D2ee8dac652833305ad739af3bfa33d82b87563b5

Same pattern by month, each orange bar a test:

https://reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2F73vtcnc6gsfh1.png%3Fwidth%3D1650%26format%3Dpng%26auto%3Dwebp%26s%3D5f1b452f753552a6e7c7111ba2db9f205522f76d

Every test so far, sign-up window vs the test itself:

|Test|Region / publisher|Sign-up window|Test dates|
|:-|:-|:-|:-|
|CBT1 (Star Arrival Test)|China, Manjuu|closed Jul 28, 2025|Aug 15–18, 2025|
|CBT2 (Constant Sequence Test)|China, Manjuu|Feb 24 – Mar 17, 2026|Apr 3–11, 2026|
|Korea CBT|Korea, Nexon|Mar 14 – Apr 24, 2026|May 15–18, 2026|
|Taiwan CBT|Taiwan, NIJIGEN|May 15 – May 29, 2026|Jun 12–15, 2026|
|Japan CBT|Japan, Yostar|Jun 1 – Jun 25, 2026|Jul 2–10, 2026|
|CBT3 (Journey Trace Test)|China, Manjuu|Jul 10 – Aug 20, 2026|not yet announced|

The sign-up window pulls far more than the test week (Japan: ~1.8M during sign-ups, ~0.7M during the test). The run of regional tests plus the milestone campaign took the count from ~11.7M to 16.5M since May, then it flattened. CBT3 sign-ups are open till Aug 20 but there’s no live test right now, so it’s back to a few hundred a day. Next real jump is CBT3’s test, then launch.

# Can it reach 30M?

Not on daily traffic. At 2,700/day, +13.5M would take about 14 years. To hit 30M by Dec 31 the count would need to average ~86,000/day for the rest of the year, which is the busiest week on record held for five straight months. It’s cleared 86k on only 26 days all year, never more than a week in a row. And every one of those 26 days fell inside a test or its sign-up window, none of them out of nowhere:

|Window|Days over 86k|Peak day|
|:-|:-|:-|
|CBT1 test (Aug 2025)|5|185k|
|CBT2 sign-ups (late Feb – early Mar)|8|150k|
|CBT2 test (Apr)|6|87k|
|Taiwan test (Jun)|1|86k|
|Japan test (Jul)|6|114k|

https://reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2F0f4yb017gsfh1.png%3Fwidth%3D1650%26format%3Dpng%26auto%3Dwebp%26s%3D75865b23ab4667d83eb296642fd041f66db7f11a

So it’s all about the launch. There’s no AP launch data yet, so the reference is comparable launches:

|Game|Pre-registrations at launch|
|:-|:-|
|Genshin Impact|~10M|
|Honkai: Star Rail|~20M|
|Neverness to Everness|~30M|
|Wuthering Waves|~30M|
|Zenless Zone Zero|~40M|

AP is already at 16.5M months out, and 30M sits squarely in the WuWa/NTE range (both topped out there; Zenless hit ~40M). Working from the current count, a modest launch lands around 24M and a big one reaches 30M. I’m not charting that, since the launch size is a guess and it would just be a made-up curve. Most of those peer numbers are advertised milestones, so real totals ran higher, which works in AP’s favor.

* 20M: very likely. Only needs +3.5M; the pre-launch ramp covers it.
* 25M (Anais): likely. A normal launch run-up reaches it.
* 30M: uncertain. Needs the launch to nearly double the count.

If the launch lands in 2026 with time to spare, 30M is reachable. A small launch, or one that slips to December, probably pushes it into early 2027.

# AP vs Ananta

Same tracking, for comparison. A year ago AP was 4.4M behind Ananta; it’s now within 0.9M (AP added 11.4M this year to Ananta’s 7.9M). Different games, not a direct comparison, but the growth since regional testing is steep.

https://reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2Fwifhf7l7gsfh1.png%3Fwidth%3D1650%26format%3Dpng%26auto%3Dwebp%26s%3D23d395cb5eb61aafbd0af1ea4a1dd6b160e3549a

# Which regions drive it

One shared counter, so it can’t be split cleanly, but you can see how much each region’s test moved it.

https://reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2F5b8iqk08gsfh1.png%3Fwidth%3D1500%26format%3Dpng%26auto%3Dwebp%26s%3Dead4e84f6d5e3d2371b46627549c04c0a682b1f8

China by far: the biggest spikes, and the daily rhythm runs on Chinese clock time (drops to ~21% overnight in CN). Japan is a clear second, Taiwan moderate, Korea barely registered. That Korea number is low enough that I suspect Nexon tracks Korean sign-ups separately, so I wouldn’t read it as low interest.

# Could the numbers be faked?

Big gacha games almost always “reach” their top milestone, because announcing a reward and then missing it is bad PR. The rare exception is Limbus Company (300k goal, hit ~180k, just didn’t award the top reward). So AP will probably announce hitting 30M regardless, which makes the real question whether they hit it honestly.

I ran the usual fake-data checks over the full year (Benford’s Law, last-digit, round-number clustering, runs test, day/night timing, event correlation) and everything comes back clean: the count sleeps overnight in CN like a real audience, round numbers show up slightly less than random, ~90% of big jumps sit inside a real event. But that’s weak evidence. These checks are public math, so a careful dev could produce numbers that pass all of them; what they’d catch is lazy manipulation, like a rushed bump before a deadline. The clearer warning sign would be a fast climb to 25M/30M right before Dec 31 or a Nexon earnings date with no event behind it. Nothing like that so far.

# TL;DR

* ~16.5M now. 30M means +13.5M, roughly double.
* The count only grows during tests and launches (~2,700/day quiet, up to 114k during an event), so 30M needs a big launch, not grind.
* A 2026 launch looks likely, so 20M is close to certain, 25M (Anais) likely, and 30M a coin flip on launch size and timing.
* The fake-data checks are clean, but that’s weak evidence since a careful dev could pass them. Watch for a jump to 25M/30M right before a deadline with no test behind it.

Pre-reg is free and takes a few seconds, and the count only jumps when a lot of people sign up at once, so if you want the free Anais or the 30M rewards, CBT3’s test and launch are the windows that matter. Mostly curious what people think about launch timing, since that’s the whole ballgame for 30M.