The problem with the RTS genre

The problem with the RTS genre is that people will associate all the RTS genre with only games like age of empires,Starcraft,Command and Conquer,Warcraft and so on...these games are very similiar because they're are in a sub-genre of RTS that never had a name.Take a look at Wolfenstein 3D/DOOM,they are in a FPS sub-genre that is called old-school fps,now take a look at Arma/Battlefield,they are in the military fps sub-genre,they are very different from old-school fps but they are still called as a FPS.Now take a look at Command and Conquer/Starcraft(C&C and Stacraft are to the RTS's what Wolfenstein 3D and DOOM is to FPS's)they are RTS,they are similiar,but people never gave to these games a specific Sub-genre name so they are called just as a RTS,now...take a look at games like Myth,Commandos,Pharaoh,League of Legends,Tropico or Nexus: The Jupiter Incident,would you call these games as RTS's ? because technically they are !! just in a different sub-genre.

So...is not the RTS genre that is dying,is the Dune/C&C/Starcraft clones that are rare nowadays just like the Wolfenstein 3d/DOOM clones are rare nowadays.

Build up army and attack all at once?

Tropico is a village builder rather than a city builder.

RTS isn't dying, it's just no indie or budget games want to make those games. There are plenty of Wolfenstein clones I assure you. On the super cheap level. I'd prefer more Doom clones as we've discussed in other threads.

The "Souls Like" is a buzz word. I'd go with calling them HAGs. Hardcore Adventure Games.

26 minutes ago, geo said:

Build up army and attack all at once?

 

Tropico is a village builder rather than a city builder.

 

RTS isn't dying, it's just no indie or budget games want to make those games. There are plenty of Wolfenstein clones I assure you. On the super cheap level. I'd prefer more Doom clones as we've discussed in other threads.

 

The "Souls Like" is a buzz word. I'd go with calling them HAGs. Hardcore Adventure Games.

The fact is that i can name alot of successful real time strategy games made in modern days: Paradox Games(Grand strategy and RTS),Rimworld,Cities Skylines,Any moba,Battlefeet Gothic:Armada,Men of War,Steel Division,Total War Warhammer,They Are Billions,Wargame...people don't seem to undestand what RTS stand for,so they think Dune 2 clones are what the RTS genre is all about.

How is it a problem that specific styles/subgenres of RTS have no names?

I don't understand this thread.

8 hours ago, koke said:

 

The fact is that i can name alot of successful real time strategy games made in modern days: Paradox Games(Grand strategy and RTS),Rimworld,Cities Skylines,Any moba,Battlefeet Gothic:Armada,Men of War,Steel Division,Total War Warhammer,They Are Billions,Wargame...people don't seem to undestand what RTS stand for,so they think Dune 2 clones are what the RTS genre is all about.

I'd consider Cities: Skylines to be a city builder. Never played any of the others.

What about FTL? Real time strategy or not?

Tetris 99 has a lot of strategy and it's real time. You can't even pause it.

I would have thought the issues with RTS games are;

- More catered to a portion of the gaming market, resulting in less sales in comparison to other genre / trends

- Market further shortened by often requiring a mouse / keyboard to be a good experience

- Not easily accessible, requiring a bit of learning to feel ready to jump into a game

- Difficult to balance if you want to provide a unique variety of playable factions

- Often requires a bit of brain power to perform in the game, becoming less ideal for quick casual games

But not giving more classifications to the genre might contribute... somehow.

24 minutes ago, geo said:

I'd consider Cities: Skylines to be a city builder. Never played any of the others.

 

What about FTL? Real time strategy or not?

 

Tetris 99 has a lot of strategy and it's real time. You can't even pause it.

Cities: Skylines has thrown you off, but the other examples are definitely what would be commonly considered as Real Time Strategy.

The difference is that they do away with base building, scouting and resource extraction: one is about heavy micromanagement. Another has you thinking your units like you would in a 1700s ship battle, with directions, mutiny, etc. And the last one you coordinate massive armies. All are games where you give orders to multiple units from an overhead point of view in real time. As for moba they are basically a fork of Warcraft 3 systems with a huge emphasis on heroes.

Anyway,not related to the thread,but today i started playing Starcraft 2 trilogy campaigns for the first time(i'm a fan of the original starcraft campaigns),holy crap ! Starcraft 2 is so boring !!! everything from the cinematics,to the soundtrack,to the voice acting,everything feels like a generic blockbuster movie,SC2 lack of any soul or appeal, really ! feels like World of Warcraft.Now i understand why people hate Diablo 3 so much ! Blizzard nowadays makes every game looking like a generic boring Blockbuster movie to appeal to the masses,fuck them !!

I heard SC2 atleast has some good designed single player missions and challenging skirmish against AI,but i don't know if i can stand all the boring bullshit this game throw at my screen !

@koke Really? I actually found SC2's campaign quite good (atleast the Terran campaign is really good). The missions are really well designed with each mission having a unique gameplay mechanic. The plot is kinda cheesy but its okay for me and atleast the cutscenes can be easily skipped. I also liked stuff like choosing which mission to play first, whose side to support and being able to upgrade the units after completing a mission.

Whereas when I recently tried playing the original SC for the first time, I only completed the Terran campaign and gave up around the middle of Zerg campaign. The missions felt kind of tedious. Most missions were just about amassing a 200 supply army and then attacking the enemy. The annoying 12-unit selection also made matters worse.

13 minutes ago, ReaperAA said:

@koke Really? I actually found SC2's campaign quite good (atleast the Terran campaign is really good). The missions are really well designed with each mission having a unique gameplay mechanic. The plot is kinda cheesy but its okay for me and atleast the cutscenes can be easily skipped. I also liked stuff like choosing which mission to play first, whose side to support and being able to upgrade the units after completing a mission.

 

Whereas when I recently tried playing the original SC for the first time, I only completed the Terran campaign and gave up around the middle of Zerg campaign. The missions felt kind of tedious. Most missions were just about amassing a 200 supply army and then attacking the enemy. The annoying 12-unit selection also made matters worse.

The cutscenes and the story are the main reasons to play the campaigns,because if i want just gameplay,i would skip the campaign and play skirmish instead.Stacraft 1 missions are a hell of tedious ! i know ! but the cutscenes,soundtrack,atmosphere and story are much better than SC2,just compare both games intros and you will see that SC1 is alot more grim and dark than SC2,SC2 tries too much to be epic,but this shit doens't belong in Starcraft or Diablo universe,this shit belong to the Warcraft universe.

by the way that starcraft 1 brood war intro was inspired by this apocalypse now scene

The first problem with RTS games is that it's real-time, so that makes it hard to have a lot of things happening at several places at once. A human player simply can't keep up. So you have to eliminate the micromanagement as much as possible to make it interesting.

The second problem is that it's not actually strategy. With games like Dune II, Command & Conquer, Warcraft, Starcraft, that pioneered the genre, it's tactics. The strategy part is limited to the base development (the orders in which you go for buildings, units, and upgrades). Everything else, from where you send your resource harvesters to how you manage your combat units, is simple tactics.

I have never played any MOBA because I'm not interested in online multiplayer, so I may be mistaken, but my impression was that they're basically a sort of third-person Counterstrike. You manage one single unit in a team, it's just that there's a fantasy skin on it and you get a top-down view of the battlefield instead of seeing through your champion's eyes.

Actual strategy games are usually of the 4X variety: the Civilization, Master of Orion, etc. genre. In that you have the real-time variant mostly known from the various Paradox games like the Europa Universalis series, which is Real-Time With Pause. There you can pause the game to get the real-world time to read some message boxes, adjust some sliders and give some orders, and then unpause to see the effects, with the ability of pausing again if you need to intervene again.

On 3/27/2019 at 11:15 AM, koke said:

Take a look at Wolfenstein 3D/DOOM,they are in a FPS sub-genre that is called old-school fps

 

Well, when Wold3D and in particular Doom came out, they were The FPS par excellence. They couldn't be "old school" yet, and they pretty much defined the genre from then on. Heck, especially Doom caused every other FPS game released after it to be called a "Doom clone", at least during the 1990s.

There wasn't anything quite comparable at the time, unless you strive to find similarities between games like Midi Maze or Battlezone and Doom, or consider vehicle sims to be "FPS" as well. There were a lot of "virtual reality" games before Wolf3D and Doom that offered a first-person perspective as well, but they simply didn't leave a mark because several elements of what made Wolf3D and Doom so great were missing.

Any re-categorization as "old school" or whatnot, can only be done a-posteriori, by definition.

As for Dune II/Warcraft/Starcraft etc. clones.... I'd say that that particular RTS subgenre has been milked for all it was worth on the PC in the 1990s and 2000s, so of course fewer companies are going to take a new stab at it, at least on the PC. However, the genre is alive and well on mobile platforms, with titles that look suspiciously like their 1990s or 2000s counterparts ;-) At least, unlike FPS, that's actually a genre that makes sense on a touchscreen.

8 hours ago, koke said:

Now i understand why people hate Diablo 3 so much !

The only people who hate Diablo 3 are the stick-in-the-mud ultra fans of the first two. Everyone else correctly recognizes it as a fantastic game. It had a lot of issues when it was first released but they were all fixed years ago.

3 hours ago, Gez said:

The second problem is that it's not actually strategy. With games like Dune II, Command & Conquer, Warcraft, Starcraft, that pioneered the genre, it's tactics. The strategy part is limited to the base development (the orders in which you go for buildings, units, and upgrades). Everything else, from where you send your resource harvesters to how you manage your combat units, is simple tactics.

And this was precisely what led to imbalance/abuses in a lot of them: in the end, the gameplay in most of those games boiled down to finding the optimal race/building/tech progression, couple that with the quickest-to-build units that could pack a decent punch, and off you went, at least when playing against human opponents. This led the infamous "7-man/grunt rush" in Warcraft II, the infamous Zerg rushes in Startcraft etc.

Against the AI, well, things are both more boring and exciting at the same time. More boring, because the AI will simply cheat to outproduce and outupgrade you, and games against the computer are generally quite protracted compared to PvP. More interesting, because you get to fully develop bases and tech trees before ultimately winning, or manage to find alternative ways to break the AI. Pick your poison ;-)

Command and Conquer doesn't belong with the other games in that list. Well most of them do but the only one played competitively (C&C 3) doesn't fit at all. The majority of units are viable and many have hard counters + fog of war is active. If your opponent even manages to figure what you are up to and counter it you'll lose horribly even if you have "Better" tactics. The original base also usually have very low resources that a rush failing won't put you ahead like say warcraft 3 where the original base lasts the majority of the game. Some players even decide to rush expansions by moving their MCVs around, meanwhile in the other games the only rushes people really consider even viable are tank rush or stupid things like korean aircraft rushes in red alert 2.

Same with Age of Empire 2 and all the hard countering\hidden damage bonuses involved; Although it has a very cool projectile system that allows a player to come ahead a much more advanced players with more strategic engagements (And stupid things like height advantage in melee combat, those hills must be steep).

I have Command & Conquer: The First Decade; one day I'll have to play other games from it than the original C&C and original Red Alert.

Yea Yuri's Revenge, Tiberian Sun and Zero Hour are probably the best ones to try in that collection. Although I'd still hold C&C3 above all of them if you want more strategy involved. ZH does have some strategic sub-factions but RA2 has a huge amount of worthless units, especially infantry aside from some overpowered ones; However trying everything in it and defeating the AIs super weapons is where a lot of the fun is. Definitely focuses more on tactics than strategy over all, online games usually end with a tank rush killing a harvester\construction yard because everything is made of paper tissues in that game.

I love RTS. I play Kane's Wrath and BFME II almost everyday. Halo Wars was great but Halo Wars 2 really sucks. Age of Empires is fun but sooooo looong just to play one game. 

What the world really needs is a very high-quality Star Wars RTS. (Empire at War wasn't great IMO)

I find it hard to get into RTS games anymore nowadays, the genres of games I play have somewhat narrowed as I got older.

2 hours ago, PeterMoro said:

(Empire at War wasn't great IMO)

Well I wholeheartedly disagree. I remember being upset that EaW didn't have near the amount of content which Galactic Battlegrounds and Clone Campaigns had, however I grew to love it after time, it was like the opposite of GB since it focused more on space than on ground, Forces of Corruption expanded on it further and made it even more fun and allows for more modding possibilities.

I want to throw it out there. The Lord of the Rings RTS was awesome and should be recognised.

9 hours ago, Pegg said:

Yea Yuri's Revenge, Tiberian Sun and Zero Hour are probably the best ones to try in that collection. Although I'd still hold C&C3 above all of them if you want more strategy involved. ZH does have some strategic sub-factions but RA2 has a huge amount of worthless units, especially infantry aside from some overpowered ones; However trying everything in it and defeating the AIs super weapons is where a lot of the fun is. Definitely focuses more on tactics than strategy over all, online games usually end with a tank rush killing a harvester\construction yard because everything is made of paper tissues in that game.

RA2/Yuri's Revenge is my favourite C&C game. I really loved the unique units in that game especially the IFV, chrono troops, mirage tanks, terror drones, siege choppers and those UFO discs. But unfortunately that game did have some serious balance issues (Yuri's faction was OP) and the AI was also incredibly dumb. In my prime, I was easily able to beat 3 to 5 (depending on map) Brutal AIs.

Speaking of C&C games. Has anyone tried mods like "Twisted Insurrection" and "Mental Omega".

About Twisted Insurrection:

It is a standalone Tiberian Sun TC. It completely redesigns the game with new art, new units, new OST and a new plot following an alternate timeline where the Brotherhood won the First War (Mod's story follows from Tiberian Dawn's Nod ending). It fixes most of the issues people had with tiberian sun and amplifies the TS atmosphere 10-folds. In addition to all this, it comes with a new client/launcher that allows easy access to multiplayer using CNCNET and adds RA2-onwards style skirmish/multiplayer options like choosing teams, starting locations and hundreds of high quality maps. And a bonus: It even comes with a Tiberian Dawn mode (where u can play with TD's units).

About Mental Omega:

It is a Partial Conversion mod for Yuri's Revenge (It requires the orignal game's files to work). It is kind of an unofficial expansion pack for the game that redefines/remasters the plot and adds new units, new OST and a completely new 4th Faction. It improves the balance and makes the subfactions more unique and diverse. Just like Twisted Insurrection, it too comes with a new client/launcher for easy multiplayer access.

It comes with 50+ missions. However I should give a fair bit of warning. The mission are really hard, so it is recommended to play on easy difficulty (comparing difficulty of these missions with vanilla RA2 is like comparing modern doom slaughterfest wads with iwads)

16 hours ago, Maes said:

Against the AI, well, things are both more boring and exciting at the same time. More boring, because the AI will simply cheat to outproduce and outupgrade you, and games against the computer are generally quite protracted compared to PvP. More interesting, because you get to fully develop bases and tech trees before ultimately winning, or manage to find alternative ways to break the AI. Pick your poison ;-)

It is possible to make a fun RTS that doesn't let the AI cheat, but it's difficult to make the AI a challenge. I need to release my C&C clone - it uses a 99% fair/non-cheating AI (structure placement is pre-calculated, because I wanted to finish the game, and didn't want to figure out how to prevent the AI from building structures in a very stupid manner). But everything else is completely fair:

• The AI must scout to find your units and base
• The AI retaliates based on enemy units it can actually see
• The AI starts out with the same amount of cash, and builds at the same speed.

However, you can run in Expert mode, which gives the AI more money, to offset it not being up to human intelligence :)

I don't know if it's legally releasable: It uses original C&C's images and sounds, but the code is mine. I built it for 3 reasons:

1. Because I couldn't figure out how to cheat in the original
2. I wanted to play bigger maps!
3. I wanted to write some pathfinding algorithms and try my hand at optimizing them.

It has a built-in map editor, with some rudimentary scripting to try to duplicate some of C&C's slightly-scripted missions. Internally, animations are handled a lot like how Doom handles it: thingtypes are mapped to frames and states, which are mapped to action functions. Sound familiar? :)

By far, the coolest aspect of C&C is the pathfinding. Good pathfinding goes a long way towards making an AI seem really intelligent. The trickiest part was traffic control. Units negotiate with each other to reduce traffic, by avoiding mapping out single-file, absolute shortest paths. My final solution ended up a little buggy, but I left it because the bugs actually seemed to add some realism to multi-unit navigation.

To all programmers: I highly recommend trying your hand at pathfinding, if you've never given it a shot: It can be a hell of a lot of fun, and it presents some very complex challenges with real-time visual results.

Age of Empire 2 HD has an AI that doesn't cheat (isn't even allowed to omni move every unit individually). Needless to say people realized they don't want to play against a competent AI even if it doesn't cheat thanks to that AI.

@kb: Well, if anything, this -by no means little- programming endeavour of yours must have taught you that, well, computers are "dumb", so to speak, and won't do anything you didn't program them to (unless you use actual AI, e.g. neural networks, genetic algos and the such). Now, if a much more limited game such as chess is so difficult to play efficiently by means of a computer program, imagine what happens when your "chessboard" suddenly becomes a few thousands (or hundred thousands) "squares" large, your unit pool expands to several hundred, there are upgrades/tech trees to take into account etc. and you'll see that you cannot really "solve" a RTS game deterministically/with traditional programming. "NP complete problem" doesn't even begin to describe it.

Most RTS from the "golden era" of the genre really didn't have the luxury to use anything but deterministic/scripted AI due to limited development/runtime resources, and simply tried to simulate AI tactical superiority by giving them ridiculous strategic superiority, if that makes sense. In most cases, this works: even in a PvP scenario, the lesser player will likely think that his opponent must be cheating in order to be able to crush them so overwhelmingly, so rapidly ;-)

To translate this concept to something closer to home, Doom DM bots themselves can be made especially annoying/deadly if constantly put in situations where they can get a 1-shot kill with e.g. a railgun from beyond normal visual range with just a split-second window of LOS, like Quake 3's infamous Xaero. Yes, Xaero is eventually defeatable if you put your mind to it, but that's exactly how superior players feel to a "noob" in an actual Deathmatch, so in a sense, it's a good approximation.

Well, AI and neural nets are freakishly similar to pathfinders, when you get right down to it :)

Regarding Chess vs. RTS: In terms of decision tree size, you're right RTS quickly approaches infinity. But, realistically, the RTS has a few things going for it. First of all, it's real-time, so the number of moves that the human can make in, say, a few seconds, is rather limited. Also, if you think less in terms of actual units and actual locations, and lean more towards possible actions, it becomes much more managable.

My AI loop contains multiple tasks, which are executed round-robin, one task per tic. An entire cycle of tasks make take a half-second or so to run. Some of the tasks are as follows:

Building

• What infantry do I need and what can I afford this cycle?
• " vehicles "
• " structures "

Discovery

• Do I need scouts? If so, send some units scouting.

Defense

• Am I being attacked? If so, is it at my base, or at a remote location?
• If at my base, send a ton of ordinance quickly.
• If remote, send an appropriate response based on the size of the attacking army.

Offense

• Do I have an WMDs ready? If so, look for where I can cause the most damage, for a few tics.
• Have I found an enemy base? If so, can I afford to attack? Will the attack leave me defenseless at home?
• See any scouts? If so, pick them off at high priority.

This is a simplified description, but you get the point. It's basically rather simple decision tree, with a lot of tuning and balancing. The AI waits until it can take advantage of any of the above situations, without causing undesirable stats, like too little defense, or being unable to do repairs/rebuild troops. The AI is designed to "get pissed" when the base, or troops are attacked, and it send more troops than it thinks it needs, to ensure that it kicks ass. Otherwise, it tends to be conservative, until it feels especially stockpiled, when it tends to switch to kick ass mode!

Experienced humans can definitely goad it into recklessness, but, for being a non-cheating, single-depth thinker, it actually does pretty well. That all I really expected for single-depth thinking. The AI is synchronous, so it had to be very fast. I'd love to develop an actual long-term, multiple depth decision tree, which could be interleaved for fake threading, or it could actually be built to run in separate threads.

But, you know, I'm not sure if I want the computer to regularly kick my ass! As-is, you've got to be sharp to beat the AI, and if it's too easy, you can:

• Add more computer players (I currently support a max of 4 players, AI, or human, local or peer-to-peer LAN, enemy or coop)
• Let the computers cheat, by giving them more cash (Expert mode)
• Use computer-generated maps (I don't have this working yet, but I can't wait to try it! You have a huge advantage over the computer when you are familiar with the maps).

However, I do believe that deep thinking is not only possible, but very doable, by looking at your resources as groups of units vs. individuals, and by thinking of navigation in terms of "destinations" vs. "any point on the map." Instead of discrete commands like "move unit 112 to position (130,207)", a typical command might be "move enough close, fast units to the scene of previous battle #27 to be able to defeat all enemy units in the vicinity." Sounds like a more complicated command (and it is). But commands like that vastly reduce the search space required by an Alpha/Beta, MinMax-type engine. Because, there are only a few "important" destinations at any given time, and there are only a handful of "useful" unit groupings and quantities, which can be described with adjectives:

Unit types

• fast
• strong
• expendable (cheap) units
• nearby
• good scouts (fast and cheap)
• super-important (transports, harvesters, mobile construction units)

Quantities

• enough for a distraction
• enough to win a battle
• enough for a slaughter
• a few (scouts)

Locations

• Home base
• Enemy base
• Choke-points (bridges, and other high-traffic narrow pathways)
• Near-home resource areas (tiberium fields, etc)
• Near-enemy resource areas
• Good lookouts near home base, for defense
• Locations of previous battles

Thinking in these terms can drastically reduce the search space/depth required. When this is combined with decent single-depth rules like in my AI loop, faked or real multi-threading, and a possible, optional cash boost for computers, it should be relatively straight-forward to create an incredibly competent RTS AI.

Like I mentioned, it's not too bad as-is (I can beat it pretty regularly, but I have to be fast and precise.) Again, it's amazing how realistic good pathfinding makes a game, without even trying! At any rate, it was massively fun to build, and if I ever get some free time, I may pursue taking it to the next step.

Do you think it's legal to release it, using the original C&C resources, as a free game? What if I make it require the (freely-distributed) resources that come with the original? C&C units have 32 rotations, after all, and I'm no artist...

Can't it be released as a mod? Overwriting some files but not enough to run on its own. (No game assets to allow standalone runs)

Although you might have more luck asking on OpenRA or E-mailing someone at EA, although am not exactly sure who since westwood is gone lol.

On vendredi 29 mars 2019 at 10:23 AM, kb1 said:

I don't know if it's legally releasable: It uses original C&C's images and sounds, but the code is mine.

The original C&C is a free download from Electronic Arts nowadays, so as long as you make your engine able to read C&C's data directly, you can get away with distributing just your program, telling people to download the game. It's what things like OpenRA do.

On 3/29/2019 at 7:52 AM, Chezza said:

I want to throw it out there. The Lord of the Rings RTS was awesome and should be recognised.

Hopefully we get a digital release one day. I lost my disk long ago :-(