Phantom Blade Zero Launch Day Report: A 431,000 Concurrent Peak Tops the Bestseller Chart, and the ‘MC 88 vs. 77% Positive’ Technical Reputation Split#

Long Zhuri | October 29, 2026

Hello friends, I’m Long Zhuri. At 10:00 a.m. today, Phantom Blade Zero officially unlocked, launching on PC and PS5 at the same time, developed and published by Beijing Lingyoufang (S-Game). By the time this piece was written, the core launch-day figures had taken shape — although a few of them are still moving: a Steam concurrent peak of 431,541 players, the top of the global bestseller chart, a Metacritic PC critic average of 88, and a Steam user positive rate reported at 77%. That gap between a high critic score and player sentiment is what this article is here to take apart.

This is the closing piece in this site’s Phantom Blade Zero coverage line. The release date observation covered pricing, platforms and wishlist milestones; the final sprint rundown covered content scale and the Q4 release calendar; the pre-order data observation established the scope boundaries of third-party estimates; and the launch eve piece compiled the review embargo lift and the action checklist — for information those four already covered, this article only does the necessary recap and does not repeat it. Today I will do three things: lay out the day-one heat and sales-tier figures; dissect exactly where the “critics 88 vs. players 77%” reputation split splits; and finally give week-one purchase and onboarding advice.

Let me put the data scope up front: every figure below is labeled with the nature of its source — SteamDB statistics, third-party estimates, media reporting — each is its own pool, they cannot be mixed when cited, and no arithmetic may be performed across them. There is no official figure for precise day-one sales across all platforms, a point that will come up repeatedly, so let me put that on record now.

1. Day-One Heat Report: Past 100,000 in Under 20 Minutes, Peak Settles at 430,000#

Lay out this morning’s timeline and the rhythm is immediately clear:

Three figures are worth reading together.

First, the 100,000 mark took under 20 minutes. For a premium single-player game, the first surge in concurrent players comes from pre-order conversion — pre-order players could pre-download before unlock (the launch eve piece has the full pre-load and disk checklist), so unlock means an immediate flood, and this slope measures how efficiently the pre-order period built up a reservoir. Data from the first day of pre-orders (detailed in the next section) already forecast this, and today simply delivered it.

Second, the peak of 431,541 is still rising. The figure we can read now is SteamDB’s statistical scope for launch day, and Gamersky explicitly notes in its report that the peak is still climbing, with the final number to be confirmed by SteamDB and similar platforms after week one ends. So every report citing this figure today (including this one) is writing a “launch-day point-in-time value,” not a final value — like the positive rate, this is a number that moves.

Third, the bestseller chart is three charts at once. Beyond the global top spot, the China-region, Hong Kong and Taiwan bestseller charts also topped out simultaneously. Read alongside the regional pre-order structure in the next section, this “simultaneous first” is not surprising, but it does confirm one thing: the main buyers on launch day are the same crowd as the main buyers during the pre-order period, and the heat did not break off the instant the game unlocked.

Incidentally, on the streaming side: Twitch peak concurrent viewership exceeded 12,000. For a premium single-player game built around solo story action, streaming data is a leading indicator of long-tail heat; this level on day one is not phenomenal, but it is enough to watch.

Day-One Data at a Glance: Scope and Timeliness of Each Figure#

The easiest way for a report like this to go wrong is to mix figures of different scopes into one pot. The figures worth recording today differ in source and timeliness:

Keep this table and check it again after week one ends — three of these four figures are “moving,” and the only thing that can be settled today is that the heat delivered.

2. Sales Tiers: About 850,000 Pre-Orders Across All Platforms, Over 1 Million Including WeGame#

On sales, let me set up the scope table first, then talk observations:

Scope Progression: From Single Platform to All Platforms#

The pre-order data observation on this site dissected Alinea Analytics’ estimates during the pre-order period, when coverage was limited to the Steam single platform; that piece already wrote through the methodology and scope boundaries, and this article will not repeat them. The key change in today’s new set of figures is coverage: this version counts the PS5 side as well, giving an all-platform scope — the two versions of the estimate cover different platform ranges, so the figures cannot be directly compared, still less subtracted. As for the total passing 1 million copies once Tencent’s WeGame channel is included, that is SteamDB’s statistical scope, a different pool again from the firm’s estimate.

Rewind to the rhythm of the pre-order period and today’s report card is not without sources: the game opened all-platform pre-orders on August 12, topped the Steam global bestseller chart in under two hours, and on the first day of pre-orders alone sold over 650,000 copies on Steam by itself (media reporting scope). In other words, the standard edition at 268 RMB had already run a fairly steep curve during the pre-order period, and launch day simply continued that curve.

One more layer deserves separate mention: about 43 million USD of pre-order revenue arrived before unlock. For a premium single-player game, this means that an R&D investment of 300 to 400 million RMB (per Liang Qiwei, see section 5) began to be recouped before players ever got their hands on the game — pre-orders are not a bonus head start, they are the first pillar in a AAA project’s cash-flow structure. Whether that pillar stands firm says more about the commercial grade than the launch-day chart ranking does.

Benchmarking Against Black Myth: Wukong: What 73% Means#

Taking the 70 days before launch as the comparison point, this game’s pre-order sales at the same stage were about 73% of Black Myth: Wukong. This comparison needs to be read calmly: it shows Phantom Blade Zero did not reproduce the explosive slope of Black Myth: Wukong’s level, but that is not a negative judgment — a critic opening score of 88 is right there (expanded in the next section), and the base is already large enough to support a AAA commercial cycle. Put the two facts together and you get a position more honest than either/or narratives of “surpassed” or “fell short”: the explosion speed is slightly behind, the reputation starting point is quite respectable, and the commercial base is established.

On regional structure, players in Mainland China, Hong Kong, Macao and Taiwan contributed 62.5% of pre-order sales. That is the pre-order-period distribution, and the regional structure after launch could well change — day-one Twitch data reminds us that attention on overseas streaming exists, but the specific share awaits later figures.

Channel Structure Observation: PC Remains the Main Battlefield#

Steam’s share of about 73.8% of sales is the most industry-meaningful number in this set. It says that even with PS5 as a console option and WeGame as a domestic channel, the base of China-made premium AAA still sits firmly on Steam. Set against the background that Tencent holds about 25% of S-Game through an affiliated entity (see section 5), the WeGame channel pushes the total scope past the 1 million mark, but the bulk of the channel remains on Steam — a structure that is a direct reference for the platform strategy of China-made single-player games to come.

3. The Reputation Split: What Sits Between 88 and 77%#

Now to the core of this article. First the figures:

What an 88 from critics means was already sorted out in the launch eve piece, which went through the shape of the review embargo lift, so here I only take the score for comparison. The genuinely new information is the trend on the player side: the statistical scope at 12:00 noon on launch day was 79%, across 255 reviews in total, of which 162 were Simplified Chinese with a positive rate of 75%; as the review count grew, the overall positive rate fell back to 77%.

Taking this movement apart, four points are worth recording:

• Two points in time, two numbers: 79% at 12:00 noon, later falling back to 77%. 77% is not an “opening score,” it is a point-in-time value still moving on launch day, and any verdict based on it is premature;
• The subset scopes differ: in the 12:00 noon tally, the positive rate among Simplified Chinese reviews (75%) was below the overall scope at the time (79%). The two scopes cover different sample ranges, so they can only be read against each other, not converted;
• The sample base is still small: 255 reviews in total at 12:00 noon, and proportions on a base that small swing widely; 77% may still move up or down as review counts grow;
• “Mostly Positive” is a label: Steam assigns a review label by positive-rate band, and 77% corresponds to the current “Mostly Positive” tier. Whichever band the ratio moves into, the label follows, which makes this another metric worth watching through week one.

What Sits on Each Side of the Gap#

Positive-side voices concentrate on combat feel. What players repeatedly name are things like adjustable blade-clash sound, the killing-intent resource, and precise parrying — consistent with where critic reviews lean, and the action framework built around parrying and resource management, compiled in this site’s combat system feature, is the most solid part of this game’s design layer.

The negative side points almost entirely at launch-day technical issues, along three main lines: the lengthy shader compilation before entering the game, with running in the background lowering the compile process’s CPU priority and stretching compile time further; SteamAPI detection failure; and crashes. The official side has published an FAQ document alongside, compiling troubleshooting steps for crashes, patch downloads, and hardware settings.

Put the two sides side by side in a table and the shape of the gap becomes intuitive:

Put both ends together and the conclusion emerges: the main cause of this reputation split is technical adaptation, not game design. Between the environment critics review in and the wildly varied hardware and system environments of players on launch day sits one link called “engineering delivery.” The 88 proves the grade of the design layer; the 77% records the debt owed on launch-day technical infrastructure — both numbers are telling the truth, they are just not talking about the same thing.

This is another case study for the topic of “China-made AAA launch-day technical infrastructure”: when the industry concentrates its resources on content and visuals, engineering details like shader compilation strategy and launcher compatibility are becoming the actual variable in launch-day reputation. This is not a pitfall unique to Chinese teams, but every time it is stepped in, the price is paid out of the launch-day player positive rate.

4. Week-One Purchase and Onboarding Advice#

The conclusion first: reputation at the content level is solid; the sticking point is hardware adaptation. Whether to get in during week one depends on your hardware situation and your tolerance for launch-day technical problems. Three cases:

First-Launch Technical Checklist#

There are only three confirmed categories of launch-day technical issues, with three corresponding lines of action:

• Shader compilation: allow plenty of compile time before entering the game, and do not switch the process to the background while it compiles — running in the background lowers CPU priority and only stretches compilation further. Don’t assume the game has frozen halfway through compiling and kill the process to start over.
• SteamAPI detection failure: work through the troubleshooting steps in the official FAQ first; this one correlates strongly with the local environment, so blind reinstalling is not the priority option.
• Crashes: the official FAQ already compiles troubleshooting steps for crashes, patch downloads, and hardware settings, so go through the document item by item before considering other measures.

Editions and Prices#

On Steam in the China region, the digital standard edition is 268 RMB and the digital deluxe edition is 328 RMB, the same price band as Black Myth: Wukong. The specific differences between the two editions were compiled in full in this site’s pre-order and purchase guide and are not expanded on here. On timing alone: the price is an open card, there is no discount variable during the launch window, so waiting for patches and waiting for a discount cost the same thing — which means the only reason to “wait” is technical issues, not price.

One more note for PS5 players: per the compilation in this site’s launch eve piece, the physical disc version for PS5 will not arrive until 2027, so buying the PS5 version during the launch window means buying digital.

Week-One Observation Checklist#

Deciding “get in now or wait a bit longer” is essentially judging whether the technical issues converge within week one. This checklist can be worked through item by item against week one:

Note the scope of the last item: 86 and 88 are critic averages for two platforms; there is currently no public item-by-item comparison data on specific experience differences between platforms, and this article makes no inference.

5. Putting Day-One Results Back Into S-Game’s Ten-Year Frame#

Finally, let me place today’s report into a larger frame. S-Game was founded in 2011 and established itself early on with the Rainblood series and the Phantom Blade mobile games; after launching Phantom Blade 3 in 2021 it took a strategic investment of several hundred million RMB from Tencent, which holds about 25% through an affiliated entity; producer Liang Qiwei has disclosed that this game’s total R&D cost is between 300 and 400 million RMB.

The game is built on Unreal Engine 5 and is a dark wuxia third-person action adventure; the player takes the role of “Hun,” a lone swordsman with only 66 days left to live, fighting martial arts masters and monsters in the fictional “Shadow Realm.” From mobile-game methodology to AAA scale, the most expensive thing on that path is not the engine but the polishing cost of getting a detail like “adjustable blade-clash sound” named and praised by players — an R&D investment of 300 to 400 million RMB buys exactly that.

What does today’s report mean for S-Game? On heat, the 430,000 peak and the bestseller chart top prove that the reservoir built during the pre-order period was fully delivered; on reputation, the design foundation of 88 and the launch-day sentiment of 77% both hold true. For a team moving from mobile games to AAA, the real test on day one has never been the numbers — it is whether the next few weeks can patch that technical gap flat, and we will look at that again when week one ends.

Further Reading#

• Phantom Blade Zero feature section: story, combat, art and pre-order features are all in place, updated continuously with the actual release experience;
• Phantom Blade Zero dated for October 29 (September 28): release date observation covering the date, China-region pricing and wishlist milestones;
• Phantom Blade Zero sprinting toward the October 29 launch (October 19): final sprint rundown on content scale confirmation and the Q4 calendar melee;
• Phantom Blade Zero pre-orders at about 500,000 copies (October 27): market data observation on third-party estimate scope and market validation;
• Phantom Blade Zero unlocks tomorrow (October 28): review embargo roundup and launch-eve action checklist.

Editorial Chatter#

Having finished this report, the editorial room had its usual chat. Zixing stared at that negative review about shader compilation and shook his head: a design where the compile process drops in priority when backgrounded is the system actively telling players “just wait slowly,” and he plans to run a launch-performance test to see how much each CPU actually differs during the compile stage. Wubian was more interested in the channel structure: Steam at 73.8%, WeGame pushing the total past one million — this structure and the revenue-share logic of the mobile-game era are two completely different species, worth a piece of its own. Zhanlan Yingren only cares about the Shadow Realm itself — he says that counting from the Rainblood era, this team has been painting dark wuxia for over a decade, and today someone finally painted it with a AAA budget.

My own feeling is simpler: from pre-orders opening in August to the unlock today, we have written four pieces along this line, and each one answered the suspense left by the previous. Today’s answer is two-sided — the 88 says the game is worth playing, the 77% says its launch is not running steadily enough. A technical gap can be filled with patches, but every player only gets one first impression. So if you enter the Shadow Realm tonight, give shader compilation plenty of time, and don’t rush.

See you in the Shadow Realm.

— Long Zhuri, Editor-in-Chief, NTGame