DNF Duel OBT2 Postmortem
What’s working and what’s not
Full disclosure: Late last year, when the first DNFD beta (and the second KOFXV beta) launched, I didn’t get to play it. I had just started working at a school and one of the kids passed along maybe-COVID to me (not all the symptoms quite lined up, and I’m double-vaccinated.) I ended up mostly KOed until the final night of the beta, and when I finally got to play, my ISP pulled the plug for three hours. Most of my perspective on OBT2 is completely fresh, although I was familiar with a few issues with OBT1 (the inability to play the game when it launched, the weird CPU battles, no training mode, some gameplay conceits, etc.) This was also why I never reviewed the second KOFXV beta, although I did get to play a bit more of that than I did of DNFD OBT1. (I also never got to play the first KOFXV beta because I got locked out for not having PS+, a bug which they acknowledged but never corrected during the beta period.)
This beta was not very feature rich when compared to the GGST or KOFXV OBTs, so my discussion will be focused on what it did have and how the gameplay felt. There will be a separate section with opinions about what the beta lacked: specifically, training and tutorial modes, and the queue-based matchmaking system that other ASW titles use for ranked games. We also already know that rollback in general and ASW’s version of it is super good, having experienced it in GGST and several retrofit titles, so I’m not going to spend more time on that than I have to.
What’s in the box?
When you booted up DNFD OBT2, there were only a few settings accessible from the main menu. There was no customization for audio or graphics whatsoever, although you could rebind your controls. The only two things you could select were Customization (a largely ignored feature given how many people I played against left their default name and profile as Adventurer with the Berzerker avatar) and the Online play. From the Online menu, you could either create your own room (with a plethora of parameters and rotation settings) or search for an existing room (with only two parameters to filter from: the rotation setting and the size of the room.) The avatar lobbies took more after Guilty Gear Xrd, and Granblue Fantasy Versus than the pseudo-2D voxel lobbies that Strive has, but have the very robust feature set all three lobby systems are known for- notably, players can engage on rotations on a single cabinet while others can be simultaneously spectating or running sets on the other available cabinets.
There were a few obvious things that were irritating about the lobby search filter. For one thing, the beta provided no invite or search parameters that let you find specific rooms. From what I played of the KOFXV beta, and from what I remember of the GGST betas, this was still true then, but at least in GGST you could seek out a designated Park or Tower floor and have a better time faster. While I assume proper invite and search functions will exist in the final version of the game, it was annoying in the moment since you would have to constantly refresh the lobby list to find your friends. Secondly, the connection bars on the side of the room seemed to be filtered very oddly. Rooms that I knew for a fact were in my region, or were coast-to-coast, were listed as one or two bar connections, but rooms located in Japan or Korea were three or four bar connections. I assume this is an egregious bug, and I left feedback about that in my survey response. Of course, with rollback, most of the long-distance connections I played were stable anyway, but it was still strange that East Coast or Midwest lobbies were ‘listed’ as potentially unplayable for me, while Transpacific connections (more than double the real-world distance) were totally fine.
DNFD as a Fighting Game
During the beta, I mostly stuck to playing Striker, but I also tried out Dragon Knight in some other sets so I could get a perspective on how other characters felt. I had heard comments informed by OBT1 about how the gatling tables in the game felt very limiting for most characters and I didn’t understand what they were talking about until I switched from Striker to Dragon Knight. For some context, I’ll try to explain the game controls and mechanics as succinctly as I can.
DNFD is a six-button game and five of those buttons are attack buttons. Offense is performed with light attack A, heavy attack B, the Skill button (S,) the MP Skill button (MS,) and the Awakening Skill button (AS.) Skills are traditional special moves, located on a single button, and are performed with single-direction inputs. These are not necessarily command normals, however- some Skills can be tools like fireballs or Rekka series or command grabs. MP Skills are tied to meter, carving out a niche as pseudo-supers, kind of like Melty Blood’s EX moves, and can also be performed with single-direction inputs. However, you spend less MP and refill the gauge faster if you perform the corresponding motion input (except for 5MS and j.MS, which every character has, with no equivalent motion input.) Every character has a 236, 623, and 214 input that can also be performed with 6, 2, and 4 respectively, and most characters’ 623MS/2MS are reversals of some kind. If you’ve depleted your MP, your character enters Exhaustion, where MP takes a bit longer to come back. Throws are performed by pressing A+B or the throw macro.
Awakening Skills are one-button supers that usually need to be comboed into and can only be performed when your character reaches Awakening- a Guts-like state when you’ve hit the last 30% or so of your red health bar, which also fully unlocks the 200-unit MP gauge. Awakening also provides each character a unique buff- for Striker, it was Power Fist, which buffed her damage and decreased her proration, and for Dragon Knight it was Dragon Force, which increased the amount of MP she got back. Burning your AS, whether on hit, whiff, or block, also burns this buff, meaning you can’t use AS again for the rest of the round. AS are also not invincible supers- I interrupted many of them, and had my own interrupted, with simple jabs. The Guard button, the game’s sixth button, can be used to block in lieu of holding 3 or 4 (although you cannot airblock at all in this game and you must still manually change your guard to block high or low,) and you can also use it with directional inputs to roll. Roll is a strike- and projectile-invuln movement option that crosses up, useful for Striker and Grappler especially in a game where everyone but them has half- or full-screen attacks. Notably, G will automatically side-switch, easily beating crossups in a similar way as SOCD, only with a single button press. There is also a guard meter that depletes as you block, in addition to the chip damage you already accrue. Guard Crush will open you up to offense when the meter is emptied.
The final major mechanic of DNFD is Conversion, the main cancel mechanic of the game. When blocking or taking damage, your character’s red health converts to white health, which then recovers over time a la Vampire Savior. Conversion is performed with B+S or the Conversion macro, and spends your white health on a cancel, analogous to Roman Cancels from Guilty Gear or Baroque Cancels from Tatsunoko vs. Capcom. This cancel also recovers MP, letting you start offense again if you’ve burned all your meter. You need white health to access Conversion, so you cannot always do it at will, but you can cancel out of most attacks, and even cancel in neutral like Xrd YRC (although it doesn’t slow the screen enough for you to steal turns.) Your white health can also be depleted forcibly by certain moves that trigger “Critical” or “Brutal” hits, which are usually MS moves designed as combo enders. The Conversion command is also used for Guard Cancels (your traditional Alpha Counter or Dead Angle,) but it costs MP rather than white health.
Get Nathan Dhami’s stories in your inbox
Join Medium for free to get updates from this writer.
The gatling system in this game is simple. A chains into B into S, which you can then cancel into MS. However, once you’ve exhausted your gatling chain, your turn is over, and not all normals can chain into themselves across all characters, meaning you may have to Convert or spend meter on an MS move to create space or remain plus. This was part of why Striker felt fun to me, whereas Dragon Knight felt very limiting to control. Striker had a unique Reverse Beat system where she could go from As and Bs into Skills or backwards and confirm a combo off anything or remain in her blockstring pressure without Converting. Her most infamous tool in the beta was her j.S divekick, Air Walk, which she could chain into j.A or j.B to remain plus on an already plus divekick and get more overheads. The rest of her Skills were also re-beatable and jump cancellable, which gave her a combo system that was way more freeform (to quote Maximillian, “loose”) than Dragon Knight, who was using the game’s typical gatling table.
If you’ve been following along, you can probably see how offense in this game plays out. Not every character has grounded high-low, and while every character can roll for an easy crossup, every character can also just hold G to block it automatically. There are a handful of ways that offense starts:
- Opening up your opponent on a high-low, which is hard for most characters to do due to either lacking the tools or the game speed. I can’t think of characters besides Kunoichi, Dragon Knight, and Striker that had good overhead options baked into their kits. (The movement of the game is mostly like GBVS, but there’s no universal overheads or sweeps like in that game.) Getting quick overheads off jump-ins usually requires Conversion.
- Cranking the guard bar, which is, again, hard for most characters to do. Kunoichi, Inquisitor, Striker, and Crusader may have an easier time with this since they have tools specifically designed to force their opponent to block and trigger Guard Crush, but otherwise it’s difficult to actually keep your opponent blocking for that long.
- Whiff punishing or counter hits, which is actually easy to do for even stubby-button characters like Striker and Grappler due to their shoulder-charge Skills and MP Skills. The rest of the cast makes do with the aforementioned big buttons found across every toolkit. Once you get a hit, you can convert (lowercase or capital C) into damage and a knockdown, and then start your loop.
- Good ol’ strike-throw. Again, some characters are better at this than others. Grappler is an obvious example, but other characters like Inquisitor or Striker have comically plus tools like Burning Wheel or Air Walk where they can just make you hold the move and then throw you while you’re panicking. Of course, there isn’t a lot of damage to (lowercase C) convert into after throwing somebody, and if you don’t have them in the corner, you essentially reset neutral.
Of course, this could all change with future patches and further metagame optimization. This was what offense looked like on day minus 90 or so, with no way to lab the game unless you had a second PS4 or a patient friend. I also want to point out that I had a lot of fun with the game- it felt really good pushing my way in with Striker and learning how to trigger Guard Crushes, or just opening them up with my silly divekick. However, I’m also not completely blind to how the game looked and felt, and I was aware that a lot of my offense, ‘loose’ as it was, was neutered by very strong defense. In a reductive sense, my offense boiled down to baiting my opponent into pressing a button, or relying heavily on strike-throw.
There are many ways this could be addressed in a future iteration of DNFD:
- I think the easiest way out is to tweak how meters work or to switch up properties of existing moves. Guard bar damage can be increased, and its recovery time can be decreased, and more characters can be given highs or lows rather than remaining as mid dispensers.
- Removing the SOCD-like properties of the guard button can allow it to retain its utility as a roll button and easy block while also making crossups possible. Make it so that players must change the side they’re blocking on even if they’re holding G. I’ve heard (but haven’t confirmed) that blocking with G cranks the guard bar further, but I don’t know if this is just hearsay, and even if it is true, I don’t know that this is disincentive enough to not use it.
- The game is in dire need of movement buffs. OBT2 increased walk speed, but both walk and dash speed probably could be tuned up higher to close the gap on characters with big buttons. Not every character needs additional air movement, since it’s clear that they want Kunoichi and Dragon Knight to be unique, but adding things like shorthops for instant overheads without spending meter could be really useful. (If a character with a hoverdash gets added into the game later, God help you all.)
Like I said before, I think the game is fun despite its current state of balance. I’m aware of how the game looks and feels in the version we played this past weekend, but I still had a good time. This may paradoxically sound like a condemnation, but it was immensely satisfying to just ignorantly press buttons and see what kind of stupid stuff I could get away with. The power levels of the cast are obviously high, and everyone has really messed up options on paper. I just want that power to be unlocked so that offense can become more varied than simply ‘make your opponent block forever until they panic, or you throw them.’ I’m also optimistic about the final version of the game, considering the pedigree of Arc System Works and Eighting, so I’m looking forward to trying out the final version and seeing how it gets changed and improved.
What the beta lacked
The discussion about whether the DNFD beta should have included a training mode or other tutorial features was played out more than most other topics about the game. I’m sort of on the middle of the road, with a slight lean towards “it would have been nice if it was in the beta.” Obviously, the beta was mostly designed to test the online and make sure that matchmaking and the rollback netcode worked… but there were balance changes between OBT1 and OBT2, in addition to a new character, and it would have been nice to have an environment where those changes could be tested. Also, the fact that balance changes were included at all obviously meant that the game feel was just as important to the beta as the online play. Having said that, it may still not be important for a beta to have a training mode since everything may be totally different on release, but it might also still be useful to have something besides spectating that players can do while waiting for a match, and it may also help people get a basic familiarity with the game’s systems. This is compounded by the fact that GGST and KOFXV both had betas with tutorials and training modes in them. Since I’m not someone that spends a lot of time in the lab, it was no skin off my back, but I also agree that it was odd that the feature was totally absent in OBT2 after it was a highly requested feature from OBT1.
Then there’s the lobby system. This conversation has likewise been beaten to death, and I’ve already written three different articles discussing the merits of both queue-based and lobby-based matchmaking systems. I want to clarify what I mentioned earlier, where I believe that there will eventually be additional search parameters, invite features, and a ranked queue system for matchmaking in the final release much like there has been in every other ASW game (Strive’s Quick Start setting works like a queue even if it places your avatar in the ‘physical lobby.’) However, I acknowledge that in general the consensus seems to be that people would prefer to wait in a queue for a match, play someone in a short set, and then get sent back to wait again for another opponent, rather than look for an open lobby and sit at a cabinet with someone for a longer session. Since I’m in the minority, rather than trying to convince everybody that ‘avatar lobbies are good, actually,’ I will explain my position with anecdotal experience, which I will concede is heavily opinionated.
When I tried to play KOFXV’s beta, the matchmaking system was basically the same as it was in the DNFD OBT, where I could only go into lobbies and could not reliably find my friends. The lobbies had no all-play, meaning that players who all piled into a single room could do nothing but spectate and wait for their turn. As a result, lobbies end up being limited to two-man rooms where you just find an opponent and then play them. In both the beta and the live version of the game, if I go into these rooms, and my opponent doesn’t want to play with me for whatever reason, I just get kicked with no explanation. If I want to play with my friends (and I have a lot of friends who want to play KOFXV,) we’re all forced to figure out who’s going to play who, and in which lobby, so that we can all be playing at once instead of just waiting and spectating- and there’s still no training mode in lobbies, only in the ranked and casual matchmaking queues. If I want to find a match in the queues, it takes forever even if I open the search parameters, and when I do that the chances of me finding a low-quality match (either due to skill disparity or connection) increases. I don’t want to spend my time waiting for a match, I want to play a match. One of my favorite games currently, Melty Blood Type Lumina, has lobbies that function similarly to KOFXV’s, but at least there I can wait in training mode, and I can also make it so that the two players on the cabinet have to rotate off to let others on after X amount of sets.
ASW’s avatar lobbies have all-play and rotations simultaneously once you get into the lobby. Whether you’re playing in a big casual lobby or an 8–9-man room that is privately hosted, you can see right away who is waiting for a match and just go right to them, and then you can play with them until one of you hops off the cabinet. My friends and I can all pile into a single room and rotate off each other and play at the same time without waiting or making multiple lobbies- we can engage cleanly in parallel play. While DNFD doesn’t have training mode in lobbies (I hope they add this later- to my knowledge GBVS still doesn’t have this) it still drastically reduces the amount of wait time before actually getting into a game, at least once people join the lobby. In previous games, I’ve found on- and offline FGC communities by meeting people through the lobby systems, something I would never have done in a regular queue-based system where no one wants to interact with each other. I also don’t personally think that the lobby system is any more or less obfuscating than simply having a ‘Play’ button, since in games that do have that feature you still have to toggle parameters and game mode filters. I should also point out that both avatar lobbies and text-based lobbies and matchmaking queues are equally error-prone, and that the game we just played was in beta. Of course, I do want the surrounding features to be cleaned up so that it’s faster for me to find where my friends are playing, or so I can filter out certain connections before a match, but the avatar lobby system itself is not inherently the issue.
Conclusion
Overall, I have very mixed opinions on DNF Duel’s beta as a full product, as a representative of the final product, and as a fighting game at its core. I was able to enjoy the beta without it having the tutorial features that many people were demanding, although I would have probably had more fun if I had those features. I would have liked the matchmaking system to be improved overall, but I’m glad that the avatar lobbies let me experience a bunch of matches in parallel to other users rather than everyone just waiting forever. I think the game itself is a blast to play, but there’s some glaring pain points in its balance that I hope are addressed by launch or some subsequent version. My overall personal impressions are probably more optimistic than the average player, but I’m not ignorant to the idea that the game probably needs a lot more polish before- and even after- its release. I’m also having a really hard time picking out which Striker color I want to use, because they all look really nice. (I’m probably sticking to 6.)
Also, don’t forget to take the surveys! ASW has put out two surveys asking about the state of the game, with the first one regarding the general feel of the beta and the second one asking specific questions about balance. If you want the game to be improved, letting them know through official channels is the best way to do it! We’ve seen this at work with GGST and P4AU’s re-release. Let your voice be heard!
