Jeez, over three months since my last post. Project Nexus came out; Played up 'til the MAG Agent in the final stage then got my ass handed to me. Great stuff! Diablo 3 also came out. Abstaining from that 'til I'm done with MR2. On that note, back to topic.

Last time I posted I was doing some closed-group testing. The fine folks that participated caught a ton of bugs, most of which have been fixed. They also gave me a lot of feedback on a topic that I wrestle with constantly: Game Difficulty.

Anybody who's reading this has probably played MR1. A lot of the complaints I got for that game was from people getting stuck. "Exploding clowns keep killing me" or "Can't kill the monocle guys (Saucy!)" or "Can't find Keycard 2" and so on. People were getting stuck, not because the game was particularly hard, but because they were missing a clue. To survive the clowns you have to figure out their blast range. To beat the Saucys you have to notice that they say "OW" when hit by the same type of weapon they're holding. And my favorite, the area with Keycard 2 can be found if you look at the background and realize that there's a breakable wall.

Those things were deliberate. They were put there to challenge the player. Challenge makes the game interesting, whether its combat or exploration. As a designer, I want the player to struggle, yes. But I also want the player to succeed. I want the player to feel the satisfaction that comes from beating an enemy that seems over powered. That is the point of Game Difficulty.

But there are limits to what the average player can handle. My testers tell me that I am making the game too hard. From my perspective, every situation I present them with is winnable. Do I stick to my guns and leave it, knowing that it may turn a player off to the game? Do I compromise my design and make it easier?

It's a struggle for sure, but a struggle worth having.

Right now I'm building a version for public testing. As many of you know, I can only work on this on weekends so I expect that the test version won't be out for about 2 or 3 weeks. But for those who are still waiting, thank you for your patience and support.

-RevolverRoach

IMAGE: A combat scenario that most of my testers failed to beat and my notes on how to make it easier. And yes, I do most of my planning on paper. I've got math notebooks full of these things.

As usual, I'm posting here to show that I am, indeed, still alive. Also, I am posting because at the moment I am incredibly frustrated with ActionScript.

WARNING: Programmer rant ahead.

MR2 is in the testing and bug fixing part of production. The small group of people I have testing the game have been doing an excellent job breaking it and exposing all the bugs. 80% of the bugs are related to saving and loading data and I'm finding that these bugs are the hardest to catch.

Here's an example of an easy fix. Let's say that out of 10 testers, 5 find the same bug. I can compare their 5 game states, see what conditions they share, and from that I can deduce what the bug is. This is about a 2-3 hour process of coding and retesting. In the grand scheme of things, 3 hours is not a lot of time.

Now an example of a hard fix. Let's say that out of the same 10 testers, only 1 guy experienced a bug. With just one error log to work with, I have no point of comparison. Worst case scenario: he can't re-create the bug. I have to go over each of his save states until I find the one where the error appeared. This can take hours or days, depending on how obvious the problem is.

And it's not like I can write it off as a fluke. 1 out of 10 is still 10%. In the first game I wrote off the zombie hands bug. When people started complaining about it, I heard loud and clear. But try as I might, I could not figure out what was causing it. If I can't recreate the bug then I can't fix it. Simple as that. To this day that bug is still in there and people still email me about it. I don't know whether that's my failing as a programmer or just one of those annoying quirks of Flash. Either way, debugging sucks.

If you have read any articles on programming with SharedObjects please point me to them. I have spent hundreds of hours reading through forums and tutorials and found very few to be helpful.

It seems that the only thing people can agree upon is that SharedObjects are a clunky, ugly, unreliable way to save data. It's fine for simple games. A game with level select only needs to save the number of levels you've unlocked and maybe a rank/high score for each level. Examples of this are Super Meat Boy and most tower defense games. A game with progression (a.k.a. levelling) needs to save which skills you've unlocked/bought and how much experience/money you've earned. Examples of this are Toss the Turtle and Sonny.

Those examples I gave may over-simplify, but the point is that as long as you keep your data to a minimum, SharedObject works fine. If you all your tracking is a player's progress and unlocks, SharedObject works fine. Once you start allowing a player to customize, that's when the system starts breaking down.

To illustrate, MR1 saves the following info.

1. Equipment of each character
- Headgear
- Weapon 1
- Weapon 2
- Armor
2. Map Exploration
- 1 Boolean (true/false) for each of the 200+ rooms
- 1 Number of the room of last save location
3. Uncollected Items
- 1 Array (up to 3 items) for each room with a locker
4. Collected Items
- 1 Array for Headgear
- 1 Array for Weapons
- 1 Array for Armor
- Number of Flashbangs
- Number of Smoke Grenades
- Number of Medkits
- Saved PID Energy
5. Special events. 1 Boolean (true/false) for each:
- Savior Boss
- MAG Zombie Boss
- Tricky Boss
- Final Boss

For a flash game, that's a pretty long list of save data. The upside: I was able to make a game with exploration and character customization. According to emails and reviews, those were the two best aspects of the game. The downside: I tended to lose track of data. Even with a ton of error checks built into the program, sometimes I'd just miss things.

For example, the zombie hands bug happens when the game can't read your Weapon1 save data. It defaults to '0' which in the game engine is the number assigned to zombie hands. That's as far as I got on that bug. I don't know why it screwed up only Weapon1 and no other equipment. Suffice to say, the more complex you make a game, the more likely things will go wrong.

There are some very good, complex flash games out there with character stats, skill trees, inventory, randomly generated dungeons, etc. Of course most of them bypass the whole SharedObject issue by using a SQL database. I'd love to do the same but that requires you to use some site's API (ex:Mochi, ArmorGames) which IMO are poorly documented. I looked into the NG API but there's almost no info on that.

Bah. Grumble. Blargh. It's just frustrating all around.

So yeah, the final word is that MR2 is still alive, mostly done, but incredibly broken. Also, I want to announce that once the game is in a more stable state, I'll do some sort of closed beta test. I'll slowly filter people in from NG forums because honestly, I've been working on this game for so long I've lost perspective. I have no idea if the public will see it the same way I do.

Peace out,
- Revolverroach

P.S. Have you guys played Abobo's Big Adventure? That game took me on a trip down memory lane. Too bad the non-NES generation won't understand why that game is so awesome. Nostalgia is one of those things that you can't pass on.

Q: When is MR2 is coming out?
A: When it's done.

Made the mistake before of promising a deadline and didn't make it. Not gonna do that again. On a lighter note, here is a sneak peek into the making of the game's art assets.

STEP 1] Hand sketch of a background, character or doodads
STEP 2] Scan sketch, choose color palettes
STEP 3] Re-create in flash using vector shapes

The whole process takes around 5 hours as long as I don't make too many mistakes. I'm doing this to over 100 sketches which contain the BGs and doodads that will be used to build the 300+ "rooms" in MR2. Do the math.

Sketches I'll be using to make the backgrounds in MR2

A few of the sketches I use to plan out keyframes for animation.

MR2 > Status: Not dead
revolverroach > Status: Working

Making a new post because the comments thread in the previous one is getting a bit long. So first off.

= Madness Retaliation 2 is DELAYED to MAY 1 =

My deepest apologies to the people who were upset at not seeing MR2 on Friday. My heartfelt thanks to the people who defended me on the comments thread.

Reasons for delay:
1.) My graphic artist quit at the last moment. I wish I could source this out to other people, but the graphics need to be quite specific. Basically, I'll need to do 'em all myself again.
2.) My IRL work has been very demanding. And unfortunately, IRL money is necessary to stay alive (and make games)

And as for why I'm only saying this now instead of Friday: I haven't been able to afford to pay my ISP bill. No internet means I can't post or reply. And this will hold true for the rest of MR2's dev period.

So flamers and haters feel free to torch this post's comment thread. I won't be able to read 'em 'til the end of the month anyway.

And for the fans and supporters I say:

"I see in your eyes the same fear that would take the heart of me. A day may come when the courage of men fails, when we forsake our friends and break all bonds of fellowship, but it is not this day."

Hold fast friends. I will get this damn game done in time.

EDIT: April 17, 2011
Madness Retaliation 2 Release Date move to May 1. Explanations in next post CLICK TO READ.

==========

Since so many people are asking, I wanted to do a quick shout out to let you know when the next Madness Retaliation game is coming out and what new features are in store for you.

- 18 Character classes with unique skill sets!
- 4 full-sized areas to crush your enemies in!
- More Weapons!
- More Armor! (can I has gloves?)
- Explosives!
- Sniper Rifles (effin campers rejoice)
- Really big guns!
- More Exploding Clowns! (nooooo...)

and the biggest request of all

- Password System! In addition to the normal save system, you can use passwords to jump to specific points of the game. That way, even if you switch computers or your save gets screwed up, you can recover most of your progress. I know it's not the "save to NG profile" people have been asking for, but its the best I can do ATM. eff off.

When's the release date?

APRIL 15, 2011

... and now I return to my hermit cave to code for your entertainment.

Ah. So it's 1/11/2011 and the end of my vacation. I'd be sad except that I really am looking forward to immersing myself in coding for the next month.

I actually didn't plan to post on this date that everyone is making a fuss over (especially G4tv. Hope Candace Bailey turns out to be a good host. Olivia, we miss you) I end my vacation today because the work has piled up and I need to go trim it down a bit. But first...

FINISHING UP WITH TORCHLIGHT

Torchlight took the best game mechanics of Diablo and created a smooth and satisfying single player experience. I spent maybe 30% of my game arranging my inventory, transmuting gems, browsing stores and gambling for good items. Looting was a breeze, especially with the help of your pet-turned-salesman.

The skill system, is admittedly not as deep as Diablo's. But that's not necessarily a bad thing. Since the only pre-requisites for unlocking skills is character level, you don't get "locked" into a specific skill tree. This allowed me to invest singular points into skills just to see how they function. I didn't have to create multiple characters just to see what a classes ultimate spells look like.

The interface functioned extremely well, especially once my hands memorized the hotkeys. It was so smooth that I would be pulling out equip or skill menus in the middle of battle. Mechanics-wise, my only complaint is that some of the enemy hit-boxes were off. I found myself relying on shift-clicking (hold position attack), even for my Destroyer who is a melee fighter.

I did a complete run with a Vanquisher and played through half as Destroyer and Alchemist. However, the flat storyline gave me no compelling reason to keep playing. To be specific, the lack of a worthy enemy kept me from feeling that i was on a grand quest. I felt more like an exterminator than an adventurer.

In Torchlight, the primary quest is to go deep into a dungeon to find the source of corruption. Fair enough. Eventually you learn that its being caused by an ancient and powerful evil. Good, good. And the evil is manifesting in an old man who, at the end of each area, throws a boss at you... Wait, what? That's my big bad guy? Oh wait, there's a giant demon at the end. Ok, phew. That's more like it. But still, I spent 75% of my game thinking that my main antagonist was a weak looking old man. I know that a greater evil had been hinted at in the narratives, but because you never see him, he makes no impression.

Let's compare it to Diablo II. You follow a cloaked man who leaves a taint of evil in his wake. You see clips of him in a tavern massacre. You see him fight a sword-wielding angel with nothing but his fists. And win. You see his transformation, his ascension as Diablo, Lord of Flames.

There was something special in the way that the storyline of Diablo II unfolded that made you eager to see more. Between the cinematics and the dialog between you and the townspeople, a picture of a world on the brink of destruction was painted in my mind, and I felt the need to see it through.

Another gripe: Torchlight had a voice-over narrative as you entered each area, but it was never enough to really get you into the game. It didn't help that the background during the narrative was a picture of a book, the supposed journal that the text was being read from. There were no cinematics, not even a "puppet show" to help illustrate the situation. I would've settled for roving-cam footage of the area.

Overall, Torchlight is a solid game that is worth a try. But apart from mods, I see no reason to play through it more than once.

I hit a mental wall while developing Madness Retaliation 2. Usually when this happens, I go play some videogames. StarCraft 2 would've been my first choice, but my install broke while patching. It frustrated me so much that I uninstalled the damned thing. It'll be a while before I plug back into it seeing as my laptop's DVD drive is broken.

So then I decided to play something a friend recommended: Torchlight. I'd heard a little about this game. I know I'm ages late playing it, but at least it's a fairly recent title. Otherwise I'd be playing "Zelda: A Link to the Past" again.

Feature-wise, Torchlight is more stream-lined than Diablo 2.
-The dungeons are more linear, saving you exploration time.
-The pet allows for selling items without leaving the dungeon so that you sustain your immersion.
-Item costs at shops are affordable (at least for the first 50 levels)
-The inventory system is greatly simplified, which I like. I hated inventory management in D2. I spent about 30% of my play time organizing my stash and pack. Confounded OCD.
-The Shift+click mechanic makes moving items much faster. Right-clicking to equip an item was also a nice touch.
-The minimalist interface was easy to pick up despite the fact that I had not played D2 in years.
-Picking up items was a breeze. Alt+clickclickclickclick. It's not quite as easy as Dungeon Seige's automated item pick-up, but I've no real gripes with that. Also, I hated Dungeon Seige.

One of the major features of Torchlight is that Skills have no prerequisites (apart from Level) which allows you to experiment with builds. This is a stark contrast to Diablo 2's Skill Tree system which restricted you to certain paths. While every Diablo, WoW, and DotA player knows that investing heavily into a small number of skills is the best way to play, having the chance to play with the other spells is a great thing to have. Otherwise, you'd never get to see the fancy animations and visual effects that the game has to offer.

And yet in my eyes, Torchlight still pales in comparison to Diablo 2. And it is all because Torchlight falters in a single category: Story.

...more on that in the next post. I've got to finish my Vanquisher's Normal run.