I have shifted from a position of 'really? A third-person action D&D game?' to cautiously optimistic about the upcoming Warlock: Dungeons & Dragons. Partially because of a really solid showing at Gamescom, actually leveraging the prospect of playing a spellcaster with a bunch of interesting interactions and some swanky-looking combat.

My optimism continues to grow after an interview with IGN, in which general manager Dominic Guay tells the site that the game will be getting rid of traditional quest markers—there's a compass you can turn off, but that's about it: "We wanted them to engage with the world, to figure things out and feel that when they found things, it was based on their actions and not just blindly trusting a waypoint."

While I'm not always opposed to a waypoint—it really depends on the kind of game—I can say that a surplus of hand-holding is always irritating. Developer Invoke Studios has lucked into a topical time to talk about this stuff, too, since such obnoxious waypointery has made us feel a lot better about not getting the PS5 exclusive Wolverine anytime soon.

Guay continues that "the core team on this game have worked on other open world games before and we've done the, let's say, the heavy-handed guiding approach before. What ends up happening at some point—and it's normal—a lot of players will just [blindly follow] the waypoint, and then you lose that agency, you lose that feeling of curiosity because you're like, well, why even bother? And so we didn't want to go there. We wanted the player to explore."

Instead of typical quests, the game revolves around "Mysteries", which sound like a sort of light immersive sim way of going about things. Like asking for rumours in a D&D game, players can bumble around town chatting to NPCs to accrue clues and information—creating the illusion of stumbling across something organically.

"Not only are there multiple entry points, but depending on how you stumble upon them, it might change how you interact with them … You can chance upon that Mystery from a completely different place, and then it'll lead you toward that settlement, and you'll know before coming in there's something fishy going on here. So then you'll approach that settlement differently than in this case where you learn it from within."

Not all of these mysteries are so straightforward—but this all sounds pretty solid to me. What'll really make this system good is whether Invoke Studios can pull it off in a way that feels natural. Creating the illusion that I'm some super-smart warlock detective is neat on paper, but if every beggar I bump into starts talking about a sinister basement, I might feel just as coddled as I am with big glowing map markers and prompts. Or yellow paint. At least we aren't doing yellow paint again.

When I got to play a sliver of Company of Heroes 3's dynamic Italian campaign for the first time, it was hard not to be impressed. The team had taken all the best parts of Company of Heroes 2: Ardennes Assault, expanded it, and then added all sorts of novel new ideas. It felt like I was playing Total War: World War 2.

A year later, once I'd been given unfettered access to the campaign for my review, I was less impressed. Player feedback had inspired Relic to remove one of the new features I dug the most—controllable detatchments—but there was a larger, more fundamental problem: the dynamic campaign wasn't dynamic. The fascists largely sat back and let me liberate Italy.

The RTS battles remained incredible, and the second more traditional campaign was excellent, leading to me giving it a score of 82% in my Company of Heroes 3 review, calling it a "failed experiment but a great game". Broadly, though, the player reception was not great, and user reviews on Steam continue to sit at Mixed.

"We set out to build the biggest Company of Heroes game ever," Relic CEO Justin Dowdeswell tells us. "More factions at launch. Two full campaigns playable at launch. I think it was an ambitious goal we set, and that showed up in some of the issues of launch around the level of polish and some of the bugs that the players were encountering. We didn't quite land it."

The Italian campaign was conceived early in the game's development, but actually getting it to work was a "stumbling block" that never quite got resolved, Dowdeswell says. "Yeah, we bit off more than we could chew. The idea was great. The spirit of it was good. The team was energised. It was just more than we could get done in time. And it wasn't nearly as polished as it needed to be at launch."

Relic thought, or at least hoped, that everything would fall into place eventually. "For so many of us in the industry, things come together kind of late in the project," Dowdeswell says. "So when it's not there and you still have time, there's this belief: well, we'll get there, we always do. And this one, we didn't."

When that didn't happen, Relic had to deal with the fallout. "It's disappointing for sure," Dowdeswell says. "It's so much work. A big team spent years and years making that game, and to see it not land well with players, it hurts.

"Obviously we spent serious amounts of time in postmortem discussions: where did we go wrong? What can we do to fix it? And just taking the time to be really thoughtful about what the top issues were that we needed to address, and when we could get them addressed by. So that was really the focus. I think we got over the initial shock quickly, and then it was just heads down. Let's work on these problems and find ways to turn it around, basically, knowing we want to do that quickly."

But was there a way this could have been avoided? At the time, Relic thought it was doing things right: bringing in content creators, modders, competitive players and other members of the community to chat to the studio and, eventually, play the game during development.

"So I think that was good," Dowdeswell says. "But by the time we got to the place where we actually were ready to start sharing software, some of those decisions were so far back that it was too late to reverse them or change them, and it was also too late to really course correct by the time we knew the game had to come out."

This was all years ago, but to Relic's credit it hasn't given up on Company of Heroes 3. On its second anniversary, Company of Heroes 3 2.0 arrived with lots of free improvements and additions, on top of premium DLC. And earlier this year Relic launched Final Stand, a standalone wave-based roguelite. Unfortunately, like the base game, it's sitting at Mixed on Steam. Later this year, we're also getting the 3.0 update, which will include additions to both games.

Perhaps more exciting is the prospect of the original Company of Heroes returning in the form of a remaster, Company of Heroes: Definitive Edition, with enhanced textures and lighting, quality-of-life improvements and a 64-bit upgrade. It's due in 2027.

Following its split with Sega, Relic is now independent, but it remains dedicated to Company of Heroes. And while the latest iteration may have not quite panned out, I remain convinced that nobody does WW2 RTS battles quite like this lot.