Highguard: how 61 Apex veterans and Tencent's secret money built 2026's fastest flop

Some failures take years to brew. Highguard took 45 days. On January 26, 2026, it launched boasting a legendary studio, plenty of money, and the most coveted showcase slot in the industry. On March 12, its servers went dark for good. In between fits one of the most instructive stories of the year: a game that, by a lot of accounts, was dead before it existed.

What's interesting about Highguard isn't that it failed. Plenty of games fail. What's interesting is how every star aligned to make it fail this fast and this publicly.

The "indie studio" that wasn't so indie

Wildlight Entertainment introduced itself to the world as an independent studio small, with the money going straight into the game. On paper, a lovely story. The reality had some asterisks.

For starters, these weren't rookies. The studio brought together around a hundred employees, most of them from Respawn, including much of the team that built Apex Legends and Titanfall. At the helm, two heavyweights: Dusty Welch (CEO, former Respawn COO and former GM of Apex) and Chad Grenier (studio head and game director, twelve years at Respawn directing Apex). Even the soundtrack carried a prestige name: Cris Velasco, composer for God of War and Mass Effect.

And then there was the detail that blew up the "indie" narrative the moment it surfaced: Wildlight's primary financial backer was Tencent, through its TiMi Studio Group subsidiary, as Bloomberg and Game File reported. Nobody had mentioned it before launch. When players found out afterward, it felt like a con. A studio that sold closeness and modest pockets turned out to be funded by one of the largest giants on the planet. That sense of having been sold a dressed-up version stuck to the game.

A game with two souls fighting over one body

Highguard started life as a survival shooter in the vein of Rust. That was its first pivot. What reached stores was something else: a 3-versus-3 PvP "raid shooter" with medieval fantasy trappings. You played a "Warden," an arcane gunslinger, and every match began by picking and fortifying a castle literally. Four phases: a minute reinforcing walls, two riding mounts across a huge map hunting for better gear, a brawl over a sword (the "Shieldbreaker"), and a final push to blow up the enemy's generators.

It sounds ambitious, and it was. The trouble is you could see the seams. According to Jason Schreier at Bloomberg, some of the weakest systems were leftovers from the game's earlier survival concept. Four years of development and two different games trying to share one body. Players spotted it immediately: the base raiding clicked, but the crafting and survival bits felt undercooked.

The specific complaints were brutal and very precise:

• Maps too big for 3v3. Enormous and empty, designed for more players than each match actually held. The community christened the result a "horse-riding simulator": long stretches looking for action, very short fights, repeat. Wildlight responded by adding a 5v5 mode weeks later, but by then the queues were empty.
• Boring looting. Mining, gathering, opening chests, with no enemies or environmental threats to create tension.
• Disastrous technical performance. Players on extremely high-end rigs reported blurry visuals, forced motion blur, and erratic framerates. When someone with a top PC says "you have to try hard to make it run this badly," the word of mouth is already lost.
• Zero visual identity. Wardens distinguished only by a couple of cooldown abilities and forgettable designs.

The Game Awards "kiss of death"

Here enters the most-discussed character in the whole story: Geoff Keighley and the Game Awards stage.

Wildlight's plan was an Apex Legends-style surprise drop: announce, go quiet, and the next thing players see is the game already available. For the announcement they landed the jackpot — the "one last thing," the show's closing reveal, the slot historically reserved for legendary sequels and established studios. Welch told it without flinching: Keighley has been a friend of the studio since the Titanfall days, played Highguard, loved it, and offered them that closing spot.

And that's where the trap was, even if nobody saw it coming. That slot works as an implicit promise of greatness. When the "one last thing" turned out to be yet another colorful hero shooter from a studio almost nobody knew, the audience felt let down before they'd even touched the game. The memes started that same night. Keighley, who'd done nothing but play enthusiastic hype man, ended up cast as the villain who "doomed" Highguard. To be fair to him: he denied having any financial stake in the game and, per reports, gave away that pricey slot for free. He simply misjudged how it would land with his enormous audience.

The numbers of a collapse

The start wasn't even bad in absolute terms:

• Day 1: a peak of 97,249 concurrent players on Steam.
• Reception: roughly 16,500 reviews, around 71% negative. "Mostly Negative" on Steam from day one.
• February 11 (day 16): mass layoffs.
• A few days later: Wildlight went from about 100 employees to fewer than 20.
• By around day 38: a 24-hour peak of 460 players. A drop of more than 99% from launch.
• March 12 (day 45): permanent server shutdown.

Over its lifetime, more than two million players passed through Highguard. Most left almost as fast as they arrived.

The blow that sealed it: Tencent cut the funding

This explains why the ending was so brutally quick. Tencent's funding was tied to specific metrics, above all player retention. When those numbers didn't show up and they didn't, from the first week continuing to invest stopped making sense for the Chinese giant. The money evaporated around February 11, and with it any chance of "giving the game time." This wasn't death by impatience; it was contingent funding switched off like a light.

Welch had boasted to Variety weeks earlier about having "a year's worth of content nearly complete" and a team that "wasn't going anywhere." Days later, the tap closed. Players got refunds and, although neither company confirmed it with a formal statement, everything points to Wildlight Entertainment shutting down. Grenier said his goodbye on LinkedIn in a resigned tone: not the outcome they'd hoped for, but he was walking away with a pile of lessons for the rest of his career.

Did they kill it, or did it kill itself?

This is the question that really matters, and the honest answer is uncomfortable: both, feeding off each other.

There was a pile-on, yes. Josh Sobel, the studio's former lead tech artist, described it bluntly: within minutes it was decided the game was "dead on arrival," and content creators had free ragebait for a month. Every studio video sank under downvotes, and the comment sections filled with "Concord 2" and "Titanfall 3 died for this." More than 14,000 review bombs from users with under an hour played. Sobel had to make his account private over the harassment. That part was unfair, and it's worth saying.

But it's also true that the game left openings everywhere: two poorly fused designs, empty maps, weak performance, a fuzzy identity, and a marketing campaign that the CEO himself admitted failed to explain what the hell Highguard even was. The surprise drop only works if the product sells itself in the first five minutes. Highguard couldn't.

In the end, Highguard is both stories at once: a game that didn't live up to its showcase, and an industry that forgives less and less. It died in 45 days. And, lessons aside, it left almost nothing behind.