Game Curmudgeons

Oh, zombies. Will you ever stop bringing me joy?

Probably, when you rise up and kill us all.

But until then, I’ll enjoy mowing you down in video games. And even when it becomes reality I’ll enjoy shooting you from long range, and hopefully I’ll be aided by plants then too, because it turns out they can be better than shotguns.

Plants Vs. Zombies is the newest entry to the PopCap catalog of casual-friendly hardcore-loved games. It’s a fresh take on the very very crowded tower defense genre, where instead of the standard placing defenses alongside a path travelled single file by enemies, defenses are put directly in the way of the invading zombies who generally stick to one of up to six paths. Plants generally need to be in the same line to be effective, but later plants can reach multiple rows, and move zombies into other lanes.

Instead of generating resources between waves, as is fairly standard, players occasionally get a free boost of randomly generated sun, but more often have to plant flowers or mushrooms which create the resource used for just about everything. Sunflowers and mushrooms create solar energy at a regular rate, which is used to buy more plants to fend the zombies off. Naturally the energy generating plants can’t do it themselves, and the stronger plants will require more solar power (and often they’ve a long recharge time before another can be planted). There’s a tradeoff between wide coverage with more plants to chew up before zombies reach the end or powerful but (mostly) vulnerable plants.  Can you afford to put the big plant in one row without losing another? Should you sacrifice one row to fortify another?

For a casual defense game, Plants Vs. Zombies pushes players into some very hard decisions, particularly at night (when sun is rare but special mushrooms are powerful allies) or with special terrain. It also changes things up with a mini-game break or a new area to defend just when it needs to in the story mode. Areas get progressively more challenging, from terrain, zombies (some of which who’ve got abilities that will force rapid changes to your plan), and environmental circumstances, but they’re always fun, even with a hard learning curve occasionally as new zombies show up, with new abilities that will absolutely wreck your shit if you aren’t prepared for them.

And even when they do, or don’t, damn is it fun and intense to get through. At marked points on a timer (so you always know how far you are from the end), zombies will come in a big wave. Usually you have time to prepare if you do it smartly and have plenty of energy ready, but even then certain zombies can make it a crapshoot and a desperate race to remove and replace plants with an appropriate counter-measure.

Plants Vs. Zombies has a crisp visual style, with that tendency toward simple shaded graphics PopCap loves to use. Cute dancing flowers, cute but clearly less than savory zombies. Everything has a very smooth motion to it, but occasionally it seems to become disjointed when pieces of zombies fall off. Minor complaint, though, sometimes things just seem too independant, like ther’s no real contact going on. Perhaps it’s just things falling slightly out of sync.

Of course, this is the kind of game where missing something can get a player’s brain eaten, and it’s essential to mention how clearly the presentation informs players of the situation. Different sounds announce different zombies, they announce incoming waves, they announce certain impacts and special effects even. Normally it would be incredibly easy for visual clutter to overcome, but zombies have unique enough shapes between types and are generally never very overlapped, so you can see what’s going on. The game’s angle is somewhere around a 3/4 overhead, and the lanes are each clearly visible and separated by alternating grass colors, which also marks in a subtle checkerboard pattern where plants can be placed as well as helping players get a better handle on how fast the danger is approaching.

The sense of humor and fun shines through extraordinarily well through the visuals and audio as well, of course. Dancing sunflowers, Michael Jackson zombies, screen door shields, and corn plants that occasionally throw butter can’t be described in any way as serious, nor can the way zombie arms and heads pop off after a good hit, and give an easy idea of how much health one has left. Great, GREAT visual feedback. And knowing the producer sacrificed his head by being pelted repeatedly with butter to get the sound right just makes it better. True story. There was no sound good enough, he went into foley, and got pelted with butter. Apparently, it’s VERY hard to wash from one’s hair.

Great background music sets it off, with an oddly cheery theme in its ominousness, but it wouldn’t be fitting any other way. And of course, while there’s not much, the writing that IS present is spectacular, generally in the form of Crazy Dave’s rantings, or notes from the zombies. There’s a reliance on visual humor more than the normal PopCap outing, but that’s far from a complaint when it’s so well done.

As usual, bonuses abount. Minigames abound for players who’ve finished the main story mode, there’s a zen garden, survival modes, and all sorts of unique challenges to work through. Plus, a significant amount of plant upgrades, special items, and general shenanigans to purchase in Crazy Dave’s shop with coins collected from levels and the zen garden. And rakes! Everyone loves when a zombie dies by stepping on a rake, afterall.

PvZ is definitely more of a challenge and much more active than the usual casual game. Instead of the turn based nature of Bejeweled or Peggle, zombies are just sent at you constantly in most modes. It’s ideal for a quick action fix, but lends itself to accidental marathons all too well. That “Just one round of survival” mentality NEVER stays there. The perpetual “one more turn” of a great game is present here, but is that really a surprise?

Plants Vs Zombies, completely unsurprisingly, gets 5 stars. It starts off a little slow, but picks up rapidly and the value for the price is excellent. It’s easy to get into, hard to put down, superbly polished. And buttering zombies never gets old.

This one sure took a while, didn’t it? Well, apparently, so did patching out the issues with Demigod’s connectivity, a major factor for a game based on multiplayer.

For what it’s worth, Brad Wardell put up an informative post about it, available here.

Now, on with the show.

Demigod is somewhere between Diablo, Dynasty Warriors, and Warcraft. Action RPG, musou, and real time strategy are all smashed together in a team-based competitive game expected to last between 30 and 60 minutes a round. The goal is 30, and it does happen sometimes, but only when one gets painfully steamrolled (perhaps I could say when you steamroll someone else, but alas, that’s yet to happen to me).

Players are given control of a hero unit, one of eight currently, amidst a constantly moving battlefield littered with grunts of all sorts, defensive emplacements, and the occasional other demigod. Essentially you’re dropped with another demigod or two into a prebuilt battlefield, taking on the opposing demigods, generally with the goal of crushing their citadel. In some ways, it’s not unlike jumping into the middle of an RTS game. The bases are established, units are spawning, and defenses are up. It’s up to players to break the stalemate, through their own power or by adding to the team’s strengths. Upgrades are bought for demigods and minions (which half of them get, the generals, while the assasins are one man armies) in the form of armor and items from the shop. Team upgrades, such as reduced death penalties, stronger and different types of reinforcement waves (which come on a constant basis), stronger buildings, etc., are bought from the citadel as the war rank goes up (basically the team level, determined by…all sorts of things!).

Gold is the primary resource of the game, critical for upgrades and items, which lend those oh so important advantages in combat between demigods and grunts, and even against fortifications and citadels, though there are plenty of resource controlling flags around the levels, and control to them is key to winning. Some provide experience bonuses, some regeneration, some more gold, some faster cooldowns on abilities, but all are extremely useful, and tend to be the centers of direct demigod on demigod conflicts.

Otherwise, players spend their time crushing grunts for experience, trying to move their own forward so that they can push through enemy defenses, which means more experience, more flags, and another step closer to destroying the citadel (the most common game mode, and really, the most fun). The game has a very subtle ebb and flow at first, but once someone breaks through the wall, there’s often a real snowball effect and the team on the losing side has to rally hard to end the push fast, lest the momentum become too great (which it most certainly does, and big pushes tend to be the game winners rather than small movements).

It’s a unique experience, and with a decent variety of maps and characters to control, as well quite diverse skill trees and upgrade options, there’s a lot to experience. Dynamic is definitely the word, especially as more demigods are on the way, and likely more maps. I’ve yet to play a game that went like the last one, and while the balance isn’t quite perfect yet, daily patches are making improvements constantly, and, excepting for when you get a clueless partner and an experienced enemy team, it’s a lot of fun.

And it’s pretty! Really, really pretty. The combination of Stardock’s technical trickery (they have a way of loading massive textures and assets into a small space and running it brilliantly) plus GPG’s artistic style works splendidly. The normally sci-fi oriented teams have taken a more fantasy oriented approach, creating fantastic levels, backgrounds, and models. The smallest grunts are all fantastically rendered, and the level of detail added to the individual demigods can be absolutely amazing, particularly The Rook. A living tower that stands far, far above everything else and can be upgraded to have smaller units on him working independantly. Archers, for example, can be seen in the turret on his shoulder when players zoom in, once the upgrade it purchased at least.

Excellent effects and animation bring the battles to life, with clear, recognizable sounds helping players sort out a bit of the chaos in the battle thanks to the unique sounds most abilities have, and several buildings. Throw in some beautiful musical scoring, and the presentation hits AAA levels on a game that’s close to budget priced (considering current owners are getting half-off coupons, it really is budget priced for some). And yet, it manages to not tax the system for the most part. A few frame stutters here and there, but it’s mostly a very smooth experience. I should note, however, some people ARE having issues with audio reverb and horrible framerates. The reason isn’t known yet, but it’s currently being worked on after more network patching (right this second, there’s testing going on for proxy servers to resolve lingering connection issues).

Occasionally, though, the chaos of crowded battlefields can make it tricky to get an ability fired off on the right target, and there really should be more documentation. Basic things like attack-move go unmentioned, for example. The first few games are a trial by fire and best played with the single-player AI (which, it should be noted, is a lot of fun and hard provides a good challenge without cheating like AI tends to in an RTS), but after some warm-up, it’s easy to jump in. Demigod could very likely pick up casual players. There’s depth, but it’s reasonably easy to jump in and won’t take hours of your time to finish a game. Frustration is generally low, though there’s always some initial confusion and challenge in learning a very unique game. Still, those complaints move aside quickly in favor of a lot of fun.

Demigod gets 4.5 out of 5 stars. The base game is excellent but there’s still some lingering technical issues making it hard to connect to other players and causing the occasional crash or stutter for a small amount of players. If the game continues to get the polish and up to twice daily patches, it’s going to be a full five stars. For now, though, it’s falling a little short of that.

While I’m very happy there’s a new series entry, which I hope tends more toward 2 than 3, I have to ask…

What the fuck is a Thiaf, and what marketer or meddling administrator chose the name? Did they not know 3 is used for E and 4 is used for A? Not that you can just insert a number at random?

Actually, that might not be true. But I can’t tell. See, there’s this thing you’ve done with the Street Fighter 2 HD Remix demo and the Marvel vs Capcom 2 demo, which is to make them ONLY playable via local multiplayer. Some of us own one controller. Some of us play alone mostly. Some of us just might not have anyone who gives a shit to play it with us.

In each of those cases, you’ve potentially lost a sale (hell, I know you lost me buying SF2HD since I couldn’t try it out). If you’re going to try and convince us to buy a game, what good does making it so some people don’t get to play do you? How are you going to convince everyone who owns MVC2 already to buy it again without letting them all play?

BROKEN Steel. Get it? Because the installer for PC and live achievements on Xbox are broken! Damn I’m clever. Come on Bethesda, I love the stuff you’re making, but it never seems to work right.

Okay, that bad joke aside, I’ve talked a bit to Brad Wardell, AKA, the CEO of Stardock, and he’s been active on plenty of forums as of late trying to take care of Demigod’s MP issues. He’s been spectacularly honest about how it’s all going and it’s great to see how much effort is being put into fixing this immediately, because when it works, Demigod is a ton of fun (thankfully, there are third party services like GameRanger that work well for playing right now). But for too many people, the P2P networking just isn’t cutting it, which Stardock is actively working to fix. In fact, I just got this from Wardell:

“We’re actually here at the office right now (11pm) with a dev team of 11 people working on this issue.”

Since he’s gone into quite a bit of detail on things, I’m going to put up some quotes from recent forum posts of his. It offers a better insight than anything I could paraphrase anyway!

“The Demigod launch is definitely really bringing up to roost the sad state of networking affairs on the PC.

In Supreme Commander, GPGNet was used and what it did was NAT traversal between players. Players who couldn’t be connected would show up with yellow or red pings in the SupCom lobby and the host would have to decide to kick them out.

It wasn’t a huge deal in SupCom because most people didn’t play MP and those who did, were technical enough to open up their ports.

With Demigod, the difference is that the player has to connect to everyone already in the lobby, not just the host in order to get into the lobby. That means, if there is 1 user in the lobby who can’t connect to more than a couple of people (for a variety of reasons) they can keep everyone else out.

The false assumption in Demigod was that a person who could connect to 1 person could connect to 8 other people and that’s just not the case. That edge case wasn’t identified.

The next false assumption is that because there were few complaints about Company of Heroes or SupCom, that the NAT system in Demigod would work fine. But CoH and SupCom are mostly played single player by the more casual part of the base. By contrast, Demigod is mostly played multiplayer.

As a result, a lot of people who normally wouldn’t even be trying an RTS MP are now in that mix which aggravated the first false assumption.

It particularly aggravates it because for Demigod, a new type of “Super NAT traversal” was made. For those who have opened ports and such, it works like every other game. But for those who don’t have normal network configurations, it apparently does a lot more but the result is that it uses a LOT of server resources to do it. As a result, what worked fine for a random sample of say 500 people fell apart at 5000 people and you still end up with edge cases that can thwart everyone.

Here’s an example edge case: Some ISPs we believe are limiting the number of peer connections they allow a user to make somehow. That person connects to host of a game fine,  gets in. But then, nobody else can connect to them and he blocks everyone from getting into the game.

This was figured out quite quickly. But the solution isn’t something that can be engineered overnight.

Here’s basically how the solution works:

1. You have to let people get into the lobby even if they’re not connected to everyone else.

2. Once in the lobby you try to direct connect or NAT connect to each other. If that fails, we try to use proxy servers to connect them. if that fails, they will appear with a yellow ping and the host will have to decide whether to kick them.

This is non-trivial to implement because, as some people know, Stardock didn’t develop Demigod. GPG did, including the networking. The networking in Demigod works fine as a unit test just like the connection servers work fine as a unit test. But integrating the two without requiring a lot of changes to Demigod is a significant task.

It is my hope that it will be solved tomorrow.

Now, that said, and this is depressing, in the last 2 weeks I’ve talked to enough people in the industry off the record to realize that in terms of RTS’s, what Demigod is experiencing isn’t atypical. It’s just that Demigod’s *multiplayer* userbase is a different demographic than the multiplayer user base of say a Supreme Commander or a Company of Heroes.

But regardless, it has to be made bullet proof and THEN once it is made bullet proof, Stardock will have to put in extra effort to make sure the MP community stays large through contests with prizes, coupons, you name it.”

The first connectivity patch is scheduled for tomorrow at this point, though that may change.

There’s some real, brutal honesty too about the success or failure of the game, and the understanding that from a player side, it doesn’t matter why it’s happening, only that it is.

“Well my feeling is, if Stardock/GPG can’t solve the MP issue this week then Demigod deserves to go down in flames.

And before someone reading this thinks I’m trolling, I’m the CEO of Stardock (I have gotten flamed on our forums for “attacking” Stardock and Demigod).

I don’t CARE what the technical issues are. I know what they are — against my will I know them. But I don’t care and neither do gamers. The damn MP just needs to freaking work for the vast majority of users. Period. End of story.

If I wasn’t on the inside, I’d be screaming “how the hell did the MP in this game get released?!” Except in beta, it worked really well. We had 10 player MP games online routinely.

Long story short: ANY MP centric game must have a very large free PUBLIC beta test. Period. As in, 10s of thousands of players.”

I don’t think any of us who haven’t programmed a multiplayer game realized just how far back things are with network. Windows has integrated plenty of standards for video and audio, but for networking, nothing.

“With the information we know NOW, we know how things should have been done. Unlike with 3D where we have DirectX to make use of, MP on the PC is still basically like the DOS days where each game has to pretty much do this stuff from scratch.

Even with Demigod, we licensed a recommended third party network library (Raknet) which, as [the prior poster] pointed out, has problems with multiple IPs.”

Hopefully you’ve learned something interesting from this, I know I have. And it’s nice to know that Stardock is going to keep working until this is fixed, even if it takes some time. There’s not a lot of companies I’d believe when they say they’re putting customer satisfaction on top, but not many companies let people get a refund on the game when it had major launch problems like Stardock did.

“I also feel that I have a moral obligation to tell customers what is happening and what we are doing about it. These people paid $40 for the game and they don’t care whose problem it is. They just want to play their game.

It’s not that I’m emotionally involved with the game. But I feel personally responsible for any problems the game has and I think that’s a good thing.

Everyone who starts a business has a motivation to do so. My motivation is because I wanted to have a company that does things a certain way. And one of those things is responsibility to the customer.”