August 31st, 2014

This week inadvertently became almost entirely about rebalancing the stat system. It just plain took longer than expected to adjust all the monsters/equipment, and a lot of related stuff got tweaked in the process. The primary purpose of the overhaul was to make stats more readable. 1 strength directly responds to 1 more damage now, 1 defense is exactly 1 less damage, etc. Stat growth has become largely flat so numbers don’t really get out of control. I don’t know that I gained all that much from these changes, but it has been a needed clean up for awhile now so at least it’s out of the way.

One of the subtler changes to fall out of this is that I’ve made defense exclusive to enemies now. After running some calculations it seemed like the main purpose of defense, aside from keeping the size of numbers down, is that it has a much bigger impact on low level characters trying to fight high level characters rather than the other way around. So it’s an effective level gating mechanic. In terms of player vs player it’s actually ideal to make it so lower level characters stand a chance against higher level characters, but in terms of player vs monsters it’s more ideal to make stat growth meaningful to create a sense of progression. Hence, no defense in play when fighting players to keep PVP close while keeping defense in play for monsters as an effective way to keep players out of areas they shouldn’t be in and to make stats more relevant. The one major downside to this is that players won’t notice as much numerical growth in their damage numbers against enemies as they will in numerical growth against themselves. There are more convoluted ways to have a similar effect while letting damage numbers grow over time (ie, strictly in PVE have the gap in level determine a damage reduction), but I feel like making it based on a stat makes it easier for players to understand exactly what’s going on. It also allows me to have classes that make fighting monsters easier by reducing defense, but being totally worthless against other players which creates specialization (doubly so in co-op where a role could be dedicated to reducing defenses so the rest of the team can deal damage).

Another big change is the move to reduce all monster encounters to just 1 enemy. We pretty much all agreed that the multiple monster encounters are basically miserable to play. Which makes a lot of sense since they aren’t really interesting to solo. Sew proposed the idea of scaling enemy encounters against party size. So a 1 person gets 1 monster, 2 people get 2 monsters, etc. Multi-monster fights are more interesting with a party in play, and it helps reign back the extreme advantage that parties currently have in the game. So I’m heavily considering going with this route. The bigger problems are the extended issues: what happens when someone joins an existing battle, does a second monster spawn in? If so that’s a gigantic opportunity for griefing. If not then players are heavily rewarded to only join battles instead of being a party. It also opens up the opportunity for one player to get locked into a battle he can’t win when his party members die or leave the battle. A third problem is focus fire: with damage rates the way they are it would be totally trivial for monsters to focus fire a single party member to death in a single turn. The same goes for players. Something has to be done to make target selection more interesting than piling on to the same target. So there’s a lot of hard problems to be solved here first, but I’m hoping to move in this direction.

Meanwhile we’ve been working on designing the new battle UI. There’s some exciting stuff in there. In particular I’m looking forward to exposing passive abilities better, because it will allow me to have more gimmick-driven enemies without them being quite so confusing. Other cool stuff includes making the type system a whole lot less confusing to new players, faster to read ability selection, and really just less confusion for everyone involved across the board.

All throughout this my decision to stick to the current battle system continued to haunt me, since I had to test it quite a bit while rebalancing the stats. It’s probably fine? This is basically one of the biggest make or break decisions I’ll have made for this entire game. As it is, the game will be fairly easy for newcomers to pick up on. There will be a reasonable amount of RPG depth and trying out different combinations of builds. But it probably won’t be quite deep enough for an extremely dedicated competitive community or anything. Is this the better direction to go for in terms of being profitable? I have no idea. But it probably suits the game as a whole better. I want people to enjoy seeing the goofy story branch out in all kinds of surprising ways while casually backstabbing/cooperating their friends, and the lower barrier of entry is the best way to hit that mark. There’s still just enough depth here to keep the game interesting throughout.

August 24th, 2014

This week consisted of doing a bunch of minor tweaks, performing more tests, and making a decision on whether to do a major change to the battle system. The long and short of it is that as-is the battle system doesn’t allow for much in the way of creative combinations. Which is pretty much inevitable when players can only do a single action in a battle turn and whatever passives they have. There’s a lot of ways we could try adding multiple actions per turn to give players more opportunities for combination (and hence customization). Doing a card deck shuffle system has been pretty appealing to me (because it still minimizes the number of options players have per turn, thus minimizing turn time). On a less drastic note having players choose two actions in a turn was also interesting. But eventually I settled on leaving it where it’s at.

There’s quite a few factors behind this. The system is at a reasonably good place right now- PVP is viable against players with differing levels, battles never take too long, PVE has a decent (but not great) element of following patterns, people can pick up the system quickly, etc. The only major flaw is the lack of combo depth, and I think I can actually improve that through smart ability design and map system interaction rather than pouring another layer of complexity on it. There’s also the big factor that I only have 3 months left to overhaul the design of this thing and adding this stuff on to combat would take up a significant amount of time. So I’m leaving it alone for now, changing it is more likely to make a bigger mess than it is to elevate it.

So with that out of the way I’ve started taking a spread shot approach to improving the game from here. There’s basically 5 points that I want to improve on, 3 of those I had numerous ideas for (and mostly pertain to improving the newest additions) and just picked some of my favorite improvements and put ’em on the schedule for update 5. Let’s take a look at the currently focused elements:

1. Refine The Battle System

There’s no single major problem with the battle system at this point. Just lots and lots of minor things. Some of it is just usability, like stats being unreadable nonsense to players. Some of it is broken monsters. Classes and their abilities need a fair amount of rethinking to see if I can add more combo opportunities. The attack type system is overemphasized and needs toned down slightly.Equipment needs to be a more interesting choice. There are experiments I want to perform like having 4 types instead of 3. Just lots of little things to punch it up and make it more immediately readable.

2. Improve Racing

The new main quest basically turns the game into a race to a goal between the players. Which is actually a big step up from the directionless mess it was before. But the actual racing itself needs improvement since once a person takes a lead, there’s very few options to catch up to them. I Have a bunch of ideas for making this more interesting.

3. Improve Victory Resolution

Right now if two players make it to the game ending battle at the same time, it all comes down to turn luck to determine who wins. This isn’t satisfying at all. We had some convoluted ideas for how to fix this like letting players fight each other in a battle against a monster. But we eventually went with the much simpler idea of ending the game with a climatic PVP showdown between players. This will probably fix it just fine, and there’s plenty of ideas for variations in this method of resolution.

4. Improve Progression 

While requiring battles has made a world of difference for players in appreciating their progression as being useful, the actual acquisition of power is still pretty straight forward: kill monsters. I want to give players stronger reasons for going off the primary objective, which will make the race itself more interesting: try to take it on slightly underlevelled, or be slow and cautious about it? On top of that I want to give players more options in actually customizing the build of their character to adapt to the game situation. The current leading proposal for this is to have 5 or so resources that players have to turn in varying combinations of to unlock different parts of their progression. This is starting to become an optional point to work on (might slow down pacing too much and makes the learning curve harsher), but I still want to give it a shot before dropping it.

5. Add Chain Reactions

I’ve talked about this a billion times on this blog. I want systems to bounce off of each other in a natural way to create unexpected situations for players. Right now the leading idea for this is to allow the map tiles to have effects on them that affect the battles taking place in them in various ways (ie, double damage from a particular attack type, everyone takes damage each turn, etc). This is one of the optional points that I’m willing to drop if nothing works.

————-

Once I’m satisfied with these points, the main goal is going to be to implement prototype versions of each of the scenarios. From there it’ll be about balancing and coming up with the final classes, items, equipment, monsters, and so on. Ideally the entire final versions of the scenario quests will also be done by the end of the year, but I’d settle with the prototypes since the rest of the game will have been finalized. There’s plenty of other work to be done with polishing, finalizing the menus, etc as well which will happen in who knows what order (ideally we should start on the menus once I finalize the systems rather than waiting on content completion). I can almost see the light at the end of the tunnel, but it’s very far away.

 

Update 4 Test Report

Changes

  • Entirely new main quest prototype. Highlights include the ability for any step of the main quest to potentially win the game, timed objectives where all players lose the game if they expire (time limits often so low that players are forced to team up or lose), and escort missions where you go into new areas instead of back to the starting area.
  • New Monster Den system where every player has to fight a battle to unlock each tile on the map.
  • Switched to a hand made map instead of random (possibly temporary for testing), switched map to open world instead of locking areas based on the plot.
  • Revised most of the classes in the game to have more attack abilities, generally stronger attacks, and removed most class stat losses.

Test Results

I think one of the testers said it best: “It’s like playing a game now, instead of avoiding playing a game”. The new monster den system is basically the update we’ve desperately needed for awhile now. Battles actually feel like part of the game, and as a result it feels like there’s a complete game now instead of half of a game. It also dilutes the number of luck-based events that occur since the battles act as a mostly static constant of progression, instead of relying on the more heavily randomized events.

There’s still tons of problems to solve (pacing is screwed up because of so many battles, battle system flaws are way more obvious now, balance is a long ways off, etc) but this is basically the start of a base game that we can finally start to expand on instead of pulling out the carpet over and over again.

Next Up

One of the more prominent issues is trying to make the mid-game victory conditions feel a little more earned instead of anti-climatic. Even in cases where two players defeat the final boss, it’s difficult to determine a victor in a satisfying way. I’m currently considering stuff like having the two players duel for it. It’s a tricky problem.

Now that battles are actually part of the game all of the problems with them are much more obvious. I’m a little worried that the RPS system won’t stay entertaining in the long term with the AI. Having other players control the monsters might help, but I still don’t think there’s quite enough depth to make it last. The biggest thing is that since players make only one choice in battle, there isn’t much room for satisfying combos. There’s a lot of potential workarounds for this (and not just adding more battle commands, the brevity of choice is helpful for pacing), but I haven’t really decided which one is right. Something I can implement quickly is especially important at this point.

I’m kind of shocked that an idea actually worked for once. So much so that I don’t really know what exactly I’m going to work on next. I’ll probably start by trying to restore the pacing back to where it’s supposed to be (which probably just requires shrinking the map).

August 10th, 2014

At long last the game is basically how it used to be. Probably a few crashes still lurking in there that I need to sort out. But the conversion is done. There’s a looming technical issue that I’ll need to sort out at a later date, but for the time being I’m switching my gears entirely back to working on the game itself. If I get lucky, enough time will have passed for one of my libraries to fix the issue for me. If I’m not lucky, I’ll have to do something time consuming. But I’ll worry about that when there’s a game worth investing into.

I’m taking a more literal prototype approach to adjusting the game this go around. Instead of planning out and scripting a large chunk of the branching plot, I’m just building a simple prototype that only branches in a few specific ways that represent the different ways the plot can affect the gameplay. It’s much faster to build, and gives a solid idea of how the changes will work out before applying all the detail to it. Doesn’t give much sense of the flavor of the game, but I’m ignoring that for now.

Likewise I’m temporarily going to start making the map layouts by hand instead of with the random generator. For now this is a measure that lets me easily radically alter the structure of the map without rewiring the generator. But it also lets me get a sense of what matters in a map layout and what doesn’t, which will feed back into the generator once I know what to target. It may also turn out that there’s only a handful of set layouts worth having around, so map generation might get dropped entirely.

It’s all about saving time to try things out faster. Hopefully these changes will work out. But if not, it won’t have cost as much as it could have.

 

August 3rd, 2014

The conversion to the new system came close to completion this week, but was thwarted by one last issue in the new system itself which slowed it down by requiring minor design work. Otherwise, it’s pretty close to running again. Still a few extended issues in terms of adding font outlining and tooltips to the new system in order to fully restore previous functionality. But odds are I’ll get to jump back into working on the game proper this week.

The amount of time sunk into this overhaul matched my worst fear estimate, and the amount of remaining work will probably take it far past that. I suppose it had to be done, but right now the amount of system and game design work required feels like it’s threatening to swallow me whole.