NiGHTS: Game Design Blog #1

After playing CIV V for awhile when it first came out, I found that games were starting to feel all too similar to one another. I was generally taking the same policies, the same tech paths, and building the same Happiness buildings. Trading with the AI was an option, but one I generally avoided, as Diplomacy itself seemed more of a nuisance than a necessity.

I'll say it, I preferred the local Happiness mechanic of CIV IV. The main reason I preferred local Happiness was that it never had an impact on more than one City - and at the same time, it made each City feel unique - effectively, it rewarded strong city placement by allowing a City to grow.

This leads into my largest gripe with global happiness - it doesn't feel rewarding - it feels punishing, (and boring) - with harsh penalties attached to negative thresholds. It didn't feel rewarding when I improved that source of Gold for additional Happiness, as at the same time it only gave me +1 gold - just like all other luxury improvements, (for the most part). Why bother even having 15 luxury resources when they all more or less grant the exact same bonuses for improvements? At this point I thought, "what would a game of CIV play/feel like if you were rewarded with Happiness instead of Unhappiness for growing your Cities?"

I tried out a game with this vanilla mechanic reversed, and I genuinely enjoyed it. All of a sudden there was this huge incentive to grow your population, and this brought a new sense of satisfaction to each improvement, even if it happened to offer the same bonus as every other improvement. But this system also lacked any balance/challenge whatsoever.

I then decided that I would increase the cost of Cities, but again - that would just seem like I was punishing the player all over again by restricting expansion. It wasn't until I looked through the XML that I realized it was even possible to ascribe Happiness/Unhappiness to military units. This was even fully supported - almost like Firaxis at one point tried this out themselves before reverting to the traditional format currently in CIV V.

In early versions of NiGHTS, I gave increasing Unhappiness values to units based on their strength - and this seemed to work. It countered the automatic Happiness that my citizens were producing - almost as if my citizens were providing the military with the support they needed to operate. A problem quickly arose in the form of unit upgrades. All of a sudden, that Warrior that had a 2 Unhappiness hit, was now a Swordsman, with a 5 Unhappiness hit. Math isn't fun, (for most people)... And I didn't want people to have to perform some basic equations each time they upgraded a unit to make sure they could afford the Unhappiness.

I decided a base number for all military units would work best, but this then presented me with the problem of what that exact number should be. Cities at the time produced 2 Unhappiness each. I couldn't have a military unit cost more Unhappiness than a City, I knew that much - but then how could I reign in the constant stream of Happiness produced by a City's citizens? After some initial experimentation, I finally settled on a system where Cities would cause the brunt of Unhappiness in each game in order to introduce a negative Unhappiness number from military units that would offset the Happiness produced by a City's citizens. Cities would produce +20 Unhappiness, military units would produce +5 Unhappiness, and citizens would generate +1 happiness each.

Now I had to figure out how much happiness luxury resources would give. I tried 5, (which wasn't enough to stay out of the red), and then 10, and honestly, 10 just seemed to work. I would need 1 luxury to fund 2 military units, or 2 luxuries to fund a new city. If I managed to settle in a prime spot with lots of food resources, then I could forgo focusing on luxuries and instead focus on building farms - but I would have to make sure that I always had a supply of happiness saved up in case of an early DOW or a pack of barbs.

Tying Unhappiness to military units, (and Happiness to citizens), all of a sudden brought back that feeling of one-more-turn that personally, I found to be lacking in CIV V. I now had legitimate choices to make that seemed to matter, and best of all - they felt unique to each new game based on my starting terrain, the location of barbs/AI, and my policy/tech paths.

A nice side effect of increasing the happiness generated by luxuries to +10, was that they immediately shot up in value in the eyes of the AI. All of a sudden, 500-600 gold per luxury was what was considered a fair trade between consenting parties. (This led to the entire restructuring of improvements/yields/buildings/production/culture/science - but I'll talk more about that in later design updates).

Last but not least, I turned my focus to the remaining sources of Happiness in the game: buildings, policies, and Natural Wonders. I knew that certain buildings would still need to provide Happiness, just in case a player didn't begin the game in a good starting location. The difference now, though - was that these buildings weren't the only/best option. They were an option - and they also could be avoided altogether in a successful game. As for policies, Happiness garnered from these would have be lowered by a fair amount - as the total Happiness generated in a game would predominately by produced by your citizens, (and other Happiness avenues like buildings/policies - would just fill in the gap when necessary).

Natural Wonders received one of the bigger Happiness changes, in that they no longer automatically gave Happiness upon being discovered. This in part helped reduce some of the luck element found in vanilla CIV V, where on top of their sometimes ridiculous bonuses, (the Fountain of Youth and El Dorado have been removed from NiGHTS), additional free Happiness just felt like overkill. Having said that, I knew I still wanted Natural Wonders to feel important and unique - so tying in with the eventual increase in production found in NiGHTS, Natural Wonders initially have much higher yields - (and they still grant happiness), but only if they're inside of your borders, (and only +5 at that point). This seemed like a fair balance in my test games, as they essentially provided the benefits of 2-3 starting improvements, (as opposed to an overload of initial happiness). With a Natural Wonder inside of your borders, you could still get out an extra 1-2 military units, but this would be more of a gradual process due to your increased production.

Diplomacy also changed quite a bit to correlate with the new Happiness values and to ensure that viable trades would occur at a higher frequency than in vanilla CIV V. I knew that overall aggressiveness of the AI's would have to be addressed, but in order to keep the same level of difficulty, (and increase it if possible) - the AI would have to focus on early expansion as opposed to early DOW's. I'll talk about this process in future design updates.

If you've gotten this far...

I'll start off by saying that the tech tree was the single most frustrating feature to create in NiGHTS. I'd played the Civ series a lot over the years - going back to Civ 2 - (although my introduction to Sid's games started with Pirates on an old Commodore 64) - but looking back, the Tech Tree was one part of those games that, until now, I always took for granted.

When I first started designing the tech tree in version 9x, I already had it sketched out on a pad of paper, ready to go - "I'll be done in an hour or two" I said to myself. Well, I was done in a little bit over an hour. After patting myself on the back for a job well done, I decided to fire it up and give it a run-through. And then I got tanks in 400 AD. What was originally a 1-2 hour hack-up in XML soon turned into a crash course in UI programming. The main problem with the tree was that units/buildings just weren't arriving at the proper times - and even when they were, it was only because I hadn't bulbed any free techs or signed any research agreements. When you factor these two mechanics into the overall research pace of the tree, it truly becomes a design nightmare.

I was using the CIV V tech tree as a template of sorts on how not to create the tree in NiGHTS. I knew that I wanted eras to come at more or less appropriate points in time, (and I wanted the Classical Era to be deeper than 1 row of techs). I also wanted to differentiate the individual tech costs, so that all techs in a specific Era wouldn't be the same.

Each time I re-designed the tree, I would try it out and attempt to game it, so to speak. I aimed for wonders and policies that gave free techs, and focused on getting Great Scientists so I could bulb as many techs as possible. I realized that most people didn't play the game this way - but for those that did, the overall design was going to have to be solid.

Eventually I had a system in place where I would play the game until I reached an era/unit/building that arrived too quickly compared to the scope of world history. At that point I would start over, add in a row of techs, add in a few prerequisites, push a few units back a tech or two, and start playing again from the beginning. After 3-4 days of this, I had completed a basic design I was happy with - but I then realized I had a new problem...

Units were coming at more or less the appropriate time - but some of them were being obsoleted and upgraded by newer units in a span of 5-10 turns. At this point I scrapped what I had and started from scratch. Now I wasn't even looking at the names for individual techs - I was throwing them around all over the place with no buildings or wonders attached to them, only units. I'm not going to lie - these test games were boring. But they allowed me to shape and design the tree without the added constraints of making sure this tech came before that one. It also allowed me to focus on the actual aesthetics of the tree itself. Without building/wonder/improvement icons cluttering everything up, I became aware of the individual tech connectors, and how in certain configurations, they just didn't look natural. While the overall shape of the tech tree was pretty much complete, I now focused on aligning techs - making sure that all the connectors had individual arcs, and that there was a sense of flow to the Tree, where paths didn't look stunted or out of place. I soon realized that going from row to row to row, while efficient - looked ugly. There needed to be longer connections weaving their way through the tree - and these needed to stand out and be obvious to the player. While the base tree in vanilla CIV V may be simple compared to earlier iterations, it succeeded at presenting obvious paths to take - (there just needed to be more of them at any given time). The tech tree at this point had roughly 110 techs in it. Strange as it may sound, it was at this point that I started to name each tech - and in the process, I carved the tree down to the 85ish currently in the mod.

Once that was finished, I discovered the either/or tags for techs. Naturally, I scrapped what I had and started from scratch for a third time. Another week or so of pruning led to the version that's currently in NiGHTS, (although the industrial/modern Era's did go through another brief overhaul).

Units were now properly placed, Era's were being entered around the appropriate time frames, and the AI was competently choosing it's Tech paths. During the third re-design, this was an area that I probably focused most of my attention on. I realized that the tech tree itself needed to present the player with a variety of choices that would differ from game to game - but more importantly, it needed to allow the AI to keep pace, even when it made choices that were less than efficient across all difficulty levels. At various points in the 3rd version of the tree, (usually when entering new Eras), tech paths converge, (to an extent). This meant that regardless of the path an AI chose to take, it would end up with a potential military build choice relevant to it's position on the tree at that specific point.

Even though I was now happy with the overall layout, something still seemed off. I played a few games of CIV II/III/IV at this point, trying to get a feel for what I was missing - and then it hit me: the UI for the Tech boxes was too large. I'll use the tech tree in CIV IV as an example here. While it was complex, (with multiple branching paths and dead ends), I never felt lost in any of my games. It just so happened that when looking at the Tree in IV, you could see almost 6 rows deep. Compare this to CIV V, and you can almost see 4 rows deep. Seeing more techs at any given time allows for some tactical planning that really just doesn't lend itself to a tree that constantly feels linear, (even though that's not always the case in CIV V).

So I now decided that I would trim the UI for the Tech boxes down. This meant that only 3 icons would be visible - but to compensate for this I made sure that each Tech was completely filled out with buildings/units/improvements. (Looking back on my time spent playing CIV V, I hadn't realized until now that there was a lot of space being wasted. Some Techs only yielded 1 building or unit, when there was room for 5 - (expansion)? After figuring out how to trim the boxes with my limited knowledge of LUA, I finally had a tech tree that I was sort of okay with. The only glaring issue that remained, was that players would have to hover their mouse over each tech and read its description when there was an either/or prerequisite.

That's when I remembered something from my earlier attempts at modding the tree. There was an XML tag that allowed you to disable certain techs - and when this happened, a locked icon would show up in front of the tech. Of course I didn't want to lock techs to signify they had an either/or prereq, but at the very least I now had something in the UI I could work with.

I ended up making copies of the current techs in the mod, copies that I disabled and placed underneath the active techs a player would be interacting with. I then replaced the lock icon with a science icon - and voila, I had UI indicators for either/or prerequisites, and all through basic XML. (The one drawback here is that these copies show up in the Civilopedia - which I'll admit can be confusing at times).

In the end, I guess the one thing to keep in mind when planning a Tech route in NIGHTS - don't ever bother checking the Civilopedia.