Hawk

I basically dropped the blog this year for various reasons. That was probably a negative overall, so let’s bring it back up to speed.

The primary focus for Update 9- which here in November still hasn’t been finished- was squarely on making the game more intelligible to normal human beings. The scope of this ranges from improving the way the game presents information, a complete redesign of the user interface, and even shaving down mechanics to be more easily understood. To this end I started out by writing a series of proposals for things to improve.

The first one was to finally implement the new map art style we had been planning for months prior. The game has undergone dramatic shifts in overworld movement over the years that finally required changing how we present maps

The first style allowed “free roam” over the map similar to most tactics games. It was basically just a random first guess to see how it felt. How far players could move was mostly just determined by a movement stat and terrain bonuses.

The second style evolved into a more limited “node movement” where players could only move between key locations, a response to the empty spaces being a waste of time. The exact movement mechanics varied, but by the end players had “stamina” which dictated how many nodes they could move in a turn.

The main problem with the second style was that it didn’t give players very strong goals, they mostly roamed around powering up and ignoring the main quest goals and battles- making for drawn out, noncompetitive games where a major system felt irrelevant. This style eventually evolved to create the concept of “gate” nodes that required players to complete battles or other challenges before they could freely walk onto them, solving the detached battles problem.

 

Making quest goals more obviously pressing was the primary objective of the third map style, which turned the map into a linear board game where players could choose which lane to go down- making it far more obvious that the goal of the game was to reach the end of the board before other players. In general I like to think of the format as a dungeon where players try to go as deep as possible before retreating to town. Movement turned into dice rolls to create a gambling tension, and also functioned as compressing the older random events into a much faster presentation by being part of the movement (sadly at the loss of world flavor).

At this point the original isometric graphics hardly made any sense with the new linear map style. Scrapping the existing art was a hard choice, but when combined with a new art direction aimed at greater productivity we knew it was the right choice. The end result to date:

You might think rebuilding our entire map engine and adjusting the game to use it would have been the biggest time investment. The reality is that it was entirely finished by the end of January. Updating the user interface would go on to consume the rest of the year, and probably beyond. Next time I’ll talk about how that process went down.

In short programming is well under way, but there’s very little interesting to talk about with it. The tasks aren’t interesting enough to be worth discussing outside of a tutorial setting, and design worked has ceased for the time being. Next week I might have something to talk about with respect to some trial runs of smaller map sizes, but for now just assume nothing worth talking about happened when I don’t post.

I spent most of this week working on yet another draft of all the classes. I feel pretty good about the progress I’ve made in that I’ve come up with a solid template for building PVP class’ attacks- now that I look at the previous draft it’s quite clear I was very scattershot with them last time. Coming up with a good template for map abilities is a little trickier, and I haven’t quite nailed it down yet.

At this point I finally realized that I haven’t really been applying an iterative approach to these parts of the game. I’ve been building lengthy class lists and then just inserting them into the game wholesale. This worked pretty well for the Co-Op classes since party dynamics are pretty key, but it keeps flopping when it comes to the PVP classes.

So I think I’m going to take a different approach this time. I’m going to start with only two classes, and I’m going to make them the most basic all-around entry level classes possible. Once I’m comfortable with what makes those classes tick, I’ll start to expand the roster. I won’t just be applying this to the classes, though. New equipment and card designs will be built in the same way.

After I finish writing this entry I’m going to sit down and design a very basic set of starter classes, equipment, items, and cards. After that it’ll be off to the races to implement the new changes. Once implemented, I’ll start testing them out and getting a feel for how everything works at the most basic level and make adjustments needed. I’ll start to understand the feel of these systems, and be able to add new things with that feel in mind. New stuff will be added in small bursts, and tested/adjusted after each addition.

At this point I’m sitting here with a completed design doc that outlines the changes for this next update. Not really happy about how long the whole process took, though mostly satisfied with the theoretical results. Most the changes are light, and I’ll be able to produce the update rather quickly as a result. They seem to hit all the right notes. Next up I need to update the design of the classes as well as design equipment/cards that fit the new systems. I might start doing some programming in between since I’m uneasy with how long this design phase has gotten.

There are a few things that leave me a little uneasy. In the old build positive spaces were able to either give players gold or experience. In the new system these spaces just give a random reward from a deck. While this allows for more flavor (and the deck in question can be manipulated, which adds its own layer of strategy), it also means that is one less consideration (“do I need gold or experience?”) for players when deciding where to move. Other things have been added to give this consideration, but I fear whether it’s enough. Time will tell.

Other than that, I mostly just can’t wait to see this new plan in action.

This week started with me confronting what had been giving me such unease about the new dungeon system: pacing. Right now we can fit roughly 8 battles into a 4 player game while staying within the hour mark. That would allow for, say, 4 dungeons that consist of two battles. Not a very satisfying dungeon crawl. General pacing would have to be improved to make room for more dungeons.

I then had a realization that we could dramatically improve multiplayer pacing by resolving everyone’s combat simultaneously during the end of the round. Which then led to the realization that we could improve it even further than that by making sure _everyone_ is in combat on the same schedule by placing battles on a timer. Naturally the problem is that doing this requires redesigning pretty much the rest of the game and even theme around it.

I wrestled with the concept for most of the week but eventually realized it was too dramatic a shift for this late in the game. I do think it would result in a better game, though. Something for the sequel.

With that realization I went back to the dungeons and realized that even with the relatively minor pacing tweak there just wasn’t enough headroom for the dungeon system. And in the process of considering throwing away most of the existent game I began to realize something. Our current branch system isn’t that bad. It’s just lacking content to give the branches relevance. We already have a hint of it with classes that allow pvp interactions on the map- better not get too close to that class, or they’ll bite you! Better follow that class if you want to steal their heals! The real problem is that the lack of player abilities that allow for map placement was missing which would add a longer lasting layer of branch strategy. And a little more fundamentally, the branches are lacking a certain reward for traversing them in order to create situations where you want something a branch has but don’t want to deal with another player’s traps.

I got so used to never adding much content to each version of the game that I have begun to actively misdiagnose the problems with each version as systems rather than just missing content. Honestly, I suspect previous versions were similarly mistreated.

So at this point it’s more a matter of adding a little bit of extra seasoning to the current version before going on to adding the missing content to it. I’ll also need to trim the length of the branch system a little bit and start to speed up some animations that slow the pacing down to improve it a little less invasively. I still feel a little uneasy about some of these additions but at least the problem of game length that has haunted me isn’t too hard to fix with the branch system.

 

This week was mostly a week of second guessing last week. The main conflict was deciding between the old structure (single linear path with many branches) and the new structure (many linear paths with no branches). I had many ideas to make the branching system more interesting. I was worried about removing movement choice so aggressively would make the game feel a little too devoid of player choice. I also felt having many linear paths would remove the simplicity of making progressing to the end being the main objective at all times. But for all the ideas I had for improving it, none of them really solved the core problem of making advancing on the map have a more satisfying sense of progression. Many of them actively encouraged backtracking, which goes against the central push your luck format of the game. Eventually I came to terms with this and feel pretty confident about the new dungeon format.

So at this point I have all the pieces that make up the game strewn about the floor and I am carefully putting them back together. Throwing out the pieces that no longer make sense, fleshing out the new ideas to support the central pillars of the game. It’s going pretty slowly for whatever reason, but I’m finally starting to put the game back together. In the process of doing so I actually reformatted the event tile system in such a way that really excites me (in short, players will be able to manipulate the luck of events to a degree- either to their benefit or the detriment of others). Combine that with the new dungeon structure and I’m starting to really feel good about the game. Hopefully this design phase will come to a close this week.

At some level my design work comes down to stripping down a perceived problem to its simplest form before any real progress can be made on fixing it. In some cases this means untangling connected problems to find the origin. In this case the tangled mess was thinking of lacking progression and lacking maps being separate problems, when in reality the lacking maps were much of the source of the lacking progression. Now that the connection is clearer it’s easier to solve: the maps lack any form of “conquering” progression to them, which in turn makes any progression system, no matter how sophisticated, feel lacking in the accomplishment that makes progression feel good. Stupid simple on paper, but took awhile to boil it down to this. Now that I have it’s mostly a matter of reformulating the map around it. The leading contender right now is the “critical path” system:

It’s pretty obviously designed around solving the problem. By segmenting it into chunks with defined end goals, it’s easy to instill a feeling of “conquering”/”creating order” in players. It also allows for creating longer chunks within the dungeons without ballooning the critical path’s backtracking. This allows for a greater feeling of endurance without creating as many dead space turns getting back to place (the backtracking itself is still important to create that feeling of endurance because if there is no cost for retreating then players will always retreat at the slightest sign of danger). Another factor is that building around more linear pathways will allow me to create more well-defined areas, as opposed to the current system of creating an indistinct mush world since it’s trivial for players to sidestep danger rather than having to deal with it. In that regard it also creates more choices for players in terms of having to decide what challenge to undertake.

There are downsides to it. Having several longer tracks without connections makes it easier for players to distance themselves from each other by each going into different dungeons, diminishing player interactions. It’s also more difficult to explain the structure to players: how many dungeons before it’s safe to take on the next boss? On paper it should be “at least one dungeon makes it possible but risky, all dungeons makes it trivial”. But communicating that to players is tricky, especially when class advantages/disadvantages, luck, and player parties can make the extreme ends invalid.

So I’m not totally satisfied with this variant quite yet. Still exploring other options. It’s closer, though.

———————–

Probably the best thing to come out of this design phase so far was a suggestion to make the battle group on a tile static instead of random each time. It’s an idea I’ve been dancing around with in terms of less random ways of picking battles. Flat out making them static was something my brain was avoiding for some reason, but actually makes way more sense: areas become more distinct (that’s the tile with karate chimp!), class advantages/disadvantages play out consistently for each player (that tile was rough with my build, but easy for that other guy!) instead of players lucking out by getting an enemy they’re good at, and player engagement is increased because you can figure out the next enemy’s pattern by observing the guy ahead of you. It’s something so simple that it’s easy to miss, but probably going to be one of the biggest improvements of this design phase.