XCOM Inspired Fantasy Game

[Nikita_Sadkov]

Stumbled upon a Japanese clone of Gollop's Magic & Mayhem, but then again, one can call Gollop's game a clone of Spellcraft

[wcho035]

If you want to have a look of the ship from the screenshot I have posted. Just download my Hybrid mod from the Mod release page. In the New Battle, I have the boat there for anyone whom is interested.

[Nikita_Sadkov]

Finally found some time to design proper distance combat. Before that ranged units had all the same mechanics as melee, which is a bit unrealistic. Ranged units also had fairly low range of only a few cells, like in Starcraft. Guess I'm one of a few persons who really hates Starcraft/Age of Empires ranged units mechanics, which don't really take into account distance, visibility or height, and Starcraft players don't really feel threatened by ranged attackers. I also believe Warcraft/Starcraft is one of the worst RTS games, learning nothing from competitors, like Warhammer, which it originally cloned.

In tabletop Warhammer ranged units were able to shot over whole battlefield if not impeded by walls, and same was the case with its computer implementations, like Shadow of the Horned Rats, predating Starcraft, but feeling much more massive, especially with proper 3d engine and huge armies. But Blizzard were a bit conservative and kept cloning to death the same Dune 2 concept.

Anyway, I had to devise a way to calculate range penalties without using random numbers. I used the already existing attack stat (which basically says how many move points that unit can put into attack per turn or how accurate it is), and used its max value as a cutoff, after which more attack points spent wont would result in less defense points penetrated. Same way, if enemy gets too close, ranged unit will lose some of its attack points.

[Yankes]

Finally found some time to design proper distance combat. Before that ranged units had all the same mechanics as melee, which is a bit unrealistic. Ranged units also had fairly low range of only a few cells, like in Starcraft. Guess I'm one of a few persons who really hates Starcraft/Age of Empires ranged units mechanics, which don't really take into account distance, visibility or height, and Starcraft players don't really feel threatened by ranged attackers. I also believe Warcraft/Starcraft is one of the worst RTS games, learning nothing from competitors, like Warhammer, which it originally cloned.

In tabletop Warhammer ranged units were able to shot over whole battlefield if not impeded by walls, and same was the case with its computer implementations, like Shadow of the Horned Rats, predating Starcraft, but feeling much more massive, especially with proper 3d engine and huge armies. But Blizzard were a bit conservative and kept cloning to death the same Dune 2 concept.

I think you should more think about starcraft as chess than simulation of battle field. In Supreme Commander you have weapons that can do exactly as you like, shoot for very long distances (and only if you have LoF, because bullets can hit ground).

"Starcraft players don't really feel threatened by ranged attackers"
this is probably not true, I see many times players avoiding combat because enemy have lot of range units or thanks to range they dominate game. Overall this is more question of balance. Many builds have same win chance, this mean range attacks cannot have real life advantage other wise would dominate game.

[Nikita_Sadkov]

In Supreme Commander you have weapons that can do exactly as you like, shoot for very long distances (and only if you have LoF, because bullets can hit ground).

Unfortunately I've never played Total Annihilation/Supreme Commander, but I heard these are really good RTS games, compared to most of the Dune 2 clones. I was always puzzled why people were so prone to clone Dune 2, when countless original RTS concepts are possible. Just look at numerous Japanese RTS games!

[Nikita_Sadkov]

I remember they ported the original Baldur's Gate for the original PlayStation, but found there is no market, and it was never released, outside of leaked version. Now they finally properly ported it for consoles. While the game is not new, the port is still interesting how they solved the human input device issues, because consoles don't have mouse and keyboard. I doubt they will ever do a proper remake, because it took them so many years to consider remaking Final Fantasy VII, which is an immensely more popular game and with more relevant story, while Baldur's Gate setting is just too dated and boring - no huge robots and flying ships.

It is strange that 1998 was the year when they discovered the pausable real-time: both Might & Magic 6 and Baldur's Gate allowed pausing timeflow, but Might & Magic 6 had much smarter design, allowing for completely turn-based combat inside real time engine. Vert similar to XCOM Apocalypse, which was released in 1997. I'm sure they  all were inspired by Gollop, because it looks too big for a coincidence. Japanese game design for example evolved completely differently from strictly turn-based gameplay: they introduced these filling with time cooldown bars, which give player some time conceive some tactics, but don't block the enemy from moving, compared to pausable realtime or purely turn-based approach. So I'm surprised Japanese haven't made multiplayer RPGs with it. And then just dropped it in favor of Baldur's Gate pausable realtime in Final Fantasy 12, which looks like a Baldur's Gate clone with bigger everything, including even the programmable AI of party members.

Also, there were many purely turn-based Japanese games, most never released outside of Japan (like Guarding Recall). They mostly use interchanging turns scheme, were each player moves all his units during given turn, but a few, like Final Fantasy Tactics, used that speed based scheduling, where each unit gains turns independently, according to its speed. I'm still unsure which is better, but I guess speed based scheduling could be better for multiplayer, so players don't have to white for too long. Other approach is Chess, where player can move only single game piece per turn (i.e. instead of units having speed, player has it).

One unusual game is Age of Wonders, a Civilization/Master of Magic clone, which had simultaneous turns - whoever clicks faster moves his units first. It was really confusing and orders of magnitude worse than say Japanese ATB system. I do consider implementing limited realtime for my game, but based on this extended Japanese approach, using cooldown bars botch for each unit and for the player (yup, copying Chess, with its limit on moves per turn). But yeah, there are countless ways one can implement time in a game, but the main rule one has to design his engine initially realtime, and then build a turn-based structure on top of it, otherwise nasty conflict can be possible, similar to multithreaded programming problems (i.e. if two units are ordered to move into the same cell, what would happen?).

[Nikita_Sadkov]

Thinking of it, there were a few original RTS games, like Creeper World and Netstorm. But Creeper World isn't really a proper RTS, because there is no other players, just player vs environment. It is closer to a puzzle game. I doubt games like Majesty or Kohan are original because they are basically just dumbed down usual RTS games. I.e. in Majesty player is limited to base building, which picks determines economy and army composition, but units fight for themselves. In Kohan there is a dumbed down base building - instead of a base, player has cities, like in Civilization, but still a neat proof that all turn-based concepts can be use in RTS games too. Say Warcraft 3 is basically a Heroes of Might & Magic game in real time, because it introduces large number of map structures and mobs guarding them. Mobs can also prevent players from attack each other in early game, if placed at bottlenecks, serving as a kind of grace period timer.

[Nikita_Sadkov]

Ok. Since I'm continuing the development of my game, I need money to buy some itch.io assets, so I want to convert a few BTC into PayPal USD. CoinBase requires a stolen European or American ID card to connect with PayPal, and they could block transaction if they notice a forged identity. So I decided to try Russian solution, which promises to send anything anywhere. Most of Russian sites exchange only larger sums at once, but one, Magnatus, appears to exchange any sums. For now I've sent them 10USD worth of BTC, and waiting for the result. Hope they don't steal such small sums. Guess it is the one of the profits of knowing Russian language - you can find and use shady barely legal services without google translate

So what BTC exchanges you people use here?

Anyway, I need PayPal USD to buy these assets: https://galefirerpg.itch.io/isometric-dungeon-designer

Which I plan to used as base and retouch into pixelart, since license allows it and it is easier to work with existing art than doing research to designing your own, especially since I'm doing everything myself. But then again, working alone means you have no responsibilities before other team members

[Nikita_Sadkov]

Ok. I thought Russians defrauded me: they claimed transfer cannot be made, due to some nonsensical PayPal error (imagine a high profile American service used by millions having an error!). I expected nothing less from Russians. But surprisingly once in a life-time Russians told truth:
https://www.paypal.com/en/cgi-bin/webscr?cmd=_display-country-functionality-outside

PayPal indeed doesn't work with Ukraine, although it works even with Albania, Bahrain, Morocco and Uruguay! Only Ukraine and some worst African countries, like Niger and Nigeria, are not allowed to receive money through PayPal, which appears to have no plans to ever support them. Guess it will be ages until racism finally gets defeated and everyone will get access to clean water and PayPal for that matter.

These Russians exchanger refunded half of the money I tried to transfer, but there were other problems discovered with my Ukrainian issued USD Mastercard, which makes it totally useless for the purposes of video game development, so I cant buy these assets Such issues, like me being a persecuted illegal immigrant and having no access to proper banking, are actually the hardest for this project. But, well, that is how gamedev works in Ukraine for you. Otherwise I would be working on some higher profile project with actual 3d graphics

Anyway, I keep working on the world map, implementing various Master of Magic style parts, since the game has to be a seamless hybrid between the two classic Microprose games. For now each city has associated castle with garrison and the actual city or village part, where tax subjects live. When player A attacks player B's city, player B has a choice to either allow player A pillaging it or using the garrison for the city defense. To actually conquer the city, player A has to attack the city's castle, which can have some fortifications, requiring powerful spells or siege equipment to break through (or maybe digging though with the imp). Think the first Age of Wonders, where siege equipment was mandatory for the city siege and destroyed battering ram meant that attacker had to retreat.

[Nikita_Sadkov]

In a fantasy game instead of laser guns charges you have arrows, greek fire and spell charges. There are several ways to implement sharing them between multiple parties: one way would be a single warehouse available to all players squads, other is a per-squad warehouse. The former is easier to implement and for player to use, but I decided on a per-party storage, because it makes it too gamey, and in future it would allow for supply chains and implementing specific scenarios, where some side has limited ammunition. I.e. if you implement a LOTR inspired map, where fellowship has to split, and it would look silly if two separate entities, many miles apart, share food rations. Still it would be a nice game-design exercise to reduce original XCOM down to a table top version.

The gif is from some crazy Japanese indie game, where player controls squads, instead of separate units.

[Nikita_Sadkov]

I remember back during the reign of DOS there were several exploration games, among them is Merchant Prince and Hammer of the Gods. These games had very unique fog of war implementation. Usually in games like Civilization or Dune 2 the undiscovered territory is just covered in black, but in Merchant prince it still had some vague form, which got more incorrect the further player moved from the start location.

These games were actually based on a 1991 Japanese game, called Atlas, which was about exploring the world and making trade routes. Same way, Tycoon series also appears to be based on Japanese game from the same company Artdink, which is like Japanese Microprose, but still alive today. In particular they recently remastered their Atlas game, although making it a bit more casual. Still love the flat world resting on a turtle idea Other curious Artdink's  game is Lunatic Dawn, which deviates greatly from your typical linear story driven JRPGs and has more freedom than most western RPGs. Hope it will get a proper translation one day, because I'm a bit too old to start learning Japanese

For now I think about implementing that Merchant Prince fog of war system for world map, because just black area is a bit uninspiring. Unfortunately I use purely software rendering so doing properly such effects can be a bit tricky.

[Nikita_Sadkov]

Another way to implement time comes from roguelikes, where monsters move after player moves. It is nice, because player doesn't have to press end-turn button, but unfortunately a bit limited. Although there are a few rather complex games using such scheme. One of them is Valkyrie Profile 2 (not to be confused with Valkyrie Chronicles). While the first VP game had the usual turn scheme, the second one even allows creating several squads, which move rogue-like style. The scheme can be adapted for multiplayer, if say the more one player moves, the more other player can move. If one player moves too far, he/she gets some handicap, and or opposing player gains bonus.

[Nikita_Sadkov]

Been working on a city view prototype. Since the game is inspired by Master of Magic, it should have similar city view, where player can quickly asses what buildings are completed. Maybe even drawing available for recruiting units near inn. Since compared to heroes of might and magic, the pool is limited, there should not be a problem displaying all recruitable units. As a side note, I'm considering to introduce that reputation/fame system, similar to Lord of Magic. Lords of Magic was another of these turn-based strategies with "pause", while that game was unfinished and had very messy design, it was very innovative. I've already borrowed several features from it. Guess game design is an evolutionary process, some early games must fail, until people find how to properly implement their features.

In particular, I've mixed Lords of Magic with Hammer of the Gods, hoping to ameliorate shortcomings of both games. But the quest system is now bound to the buildings the city has. I.e. the more higher tier the building combination, the harder the quest it gives and the better is reward. For example, if dwarven stronghold is present, dwarves could give the quest to kill the dragon, raiding their gold mines. Magic guild could also require some unlocking spells.

Anyway, there are a lot of buildings already and they all got assigned a role, with some buildings excluding each other (i.e. there could be either dwarven stronghold or orc fortress, because both settle in the mountains). For now I'm considering also implementing terrain dependencies for specific buildings. I.e. hydra and snakemen would require swamp. Such decisions are nice, due to how chaos theory works - they create natural chaos on a static map, without using random number generator. But Sid Meyer has already noticed that, even if he is no mathematician.

[Nikita_Sadkov]

While the site exploration itself uses discrete square lattice for movement, the world map movement is continuous due to several reasons:
1. Homage to the original XCOM. Although I've dropped the realtime component.
2. To emphasize scale, which player can't micromanage by steps, and have to think strategically (i.e. should I send these units over mountain?).
3. To give player a break from the grid world.
4. Continuous maps look a bit better, than tile based (i.e. square shaped shore is always annoying).
5. To allow myself testing different algorithms, so I can asses better what needs to be change in my programming language.

Anyway, there is a bit similar to XCOM fantasy game, called Ogre Battle, which also had continuous realtime world map movement. It had no aliens, but a revolution, so cities were liberated, instead of being defended. Surprisingly it was released a year prior to XCOM, but still not enough prior time for accusing XCOM of borrowing game design decisions. Although the following game of that designer Japanese (Yasumi Matsuno, who is like Japanese Julian Gollop), Tactics Ogre (the precursor to Final Fantasy Tactics), probably had some inspirations from XCOM, although its world map was far more limited and discrete.

Anyway, Ogre Battle squads moved over continuous 2d space, accounting for terrain features (compared to XCOM, which had just aircrafts completely ignoring terrain), so movement over mountains took really long amount of time - you could made a coffee and then return back just in time. Other than that, the game had unusual approach to recruiting new units. Beside humans, recruited by liberating cities, monsters like wolves and griffins were recruited by traveling over their specific habitat locations. Anyway, the squads had leaders, and killed leader meant that squad had to retreat for regrouping. World map also had time of day, and specific units, like vampires and werewolves were vulnerable during daytime. Ogre Battle also had "reputation" resource, which controlled what units player can recruit, how much income player gets, how efficient are different spells and what quests will be available.

So how it all comes to my game? Well, my world map is continuous, yet turn based, so I can't use Ogre Battle's approach of wasting player's time, annoying him/her in process. I can't also count pixels or some other infinitesimals, because that would be too hard for player to grasp and too tricky to compute, wasting a lot of AI cycles. So I decided to implement simpler method: just entering mountains will take 3/4 of the movement. Now party is marked as mountain-ready for the remaining of current turn and can move spend the remaining 1/4 of its movement for moving over the mountains. Same with river, which cost 1/2 movement to cross. It also makes programming enemy AI easier, since it can now cross any terrain, without being blocked.

Same with water, with the exception that only small parties can enter water, and player must have free ships for them to use. Obviously naval fleet costs money to upkeep. I've simplified it to that table-top level, knowing how hard it is to program ship interaction even on lattice-based maps, especially for AI (before that I made fully functioning Warcraft 2 clone with AI: https://www.youtube.com/watch?v=-k8jkeFfnl0 ). Games like Civilization, Heroes of Might & Magic and Warcraft 2 had non-working ship AI, where AI had really hard time crossing the water. And when I fixed the AI in my Warcraft 2 clone, I found that usual campaign maps became unplayable, because designers relied on the AI being broken, in fact the game had special handcrafted AI for each campaign map, and all these AIs was non-functioning in a special way.

Now for that to look non-ridiculous, I need a single border of water with all the ship and somewhat abstract looking world map. Anyway, aforementioned Hammer of the Gods, had similar solution, although ships were carried by squads!!! That was the most laughable thing about it: carrying a viking drakkar boat over long distances.

In addition there are roads. These are just a few pixels in thin, so tracking them properly for player and AI will be very hard. Yet all parties have guard range, which specifies nearby locations such party has access to. If say a city is in guard range, then party can enter. Same with roads: if a road is in guard range, unit takes less time to move there. Yeah, continuous space is measured with ranges, so even infinitesimally small items can be isolated.

[Nikita_Sadkov]

Finally implemented some world map pathfinding AI, instead of units going straight in line. Due to continuous space I can't use the usual diffuse approach (i.e. the Dijkstra's Algorithm), but since the number of obstacles is usually limited, and the world map is no a maze, I don't really need the diffuse approach, and can just try random paths while on of them works, which happens in most cases. Random paths are also nice since they make AI totally unpredictable. But yeah, random is still very resource consuming, so for an RTS game one would still use Dijkstra on top of some space partitioning structure, like quadtree, but space partitioning is tricky. Although one can also try further optimize random paths, by using something like neural nets, but then again, neural nets are structurally equivalent to space partitioning trees, so one will get the same result, just looking a bit different.

Anyway, now I have an easy way to control AI difficulty - just reduce the number of paths, and the AI will become even more dumber.