[mod] Step research

[karamel]

Yet another research mod…
---------------------------------------------------------------
Important
Step research is now outdated. Check Battleplan (viewtopic.php?f=49&t=11294) for its evolution
---------------------------------------------------------------
Edit: see latest posts to get the latest version

The basis of step research is to get a few clear mecanics for playing with them better (or at least enjoy faster). See the mod manual to get the paradigms and build strategies with them.

Below is the outdated initial post
---------------------------------------------------------------

The main principles of step research is to make linear branches and cap extends with the research branch. It is a full research tree rework but does not add any new feature.

For example if you follow the machinegun branch you will have this path in v0.1: twin mg -> damage -> damage -> heavy mg -> dmg -> rof -> dmg -> assault gun -> dmg -> dmg -> rof -> dmg -> twin AG -> dmg -> rof
To get along this path you will need to research synaptic link data analysis and so, otherwise you will be capped to the first damage.

Check the attached spreadsheet to see all branches (branches path vertically, synaptic link requirement horizontally with the first column).

Also things available at start are rather different. You don't have to research for VTOL, Cyborg, machinegun and tank trap (and command relay but it's useless…). Hover is thrown further in the tree like bombs and bodies and mortars are more accessible.

I know that for now it will break a lot of things, balance and so (and AI). From my first tests I can tell that the research is strongly slowed but it's clear and offers more choices right from start (mortar + trap, machinegun VTOLs, flamer bug…). Still there is a lot of work on what is available when and modifying research cost/time.


All of this is GPLed and for 3.1.

[karamel]

New version.

• Research cost/time are uniformized. Time is incremented the further you go in a branch and it takes a bit more than one hour to go to the end. It takes the same time to research all things in a step in one branch than researching the synaptic links to go to next step.
• Only 4 steps: start, Synaptic Link Data Analysis, Dedicated Synaptic Link Data Analysis and Neural Synapse Research Brain (mk2 and 3 don't make step anymore but are still required)

I tested it with some AI and the result is that Nexus and Semperfi are lost, mainly being obstinated in spaming machinegun cyborgs. Semperfi does a bit better. NullBot adapts not so bad to this research tree with various quality depending on personality.

I suggest then playing with NullBots (see viewtopic.php?f=10&t=8826) with Step Research. I tried to adapt Nexus but it's too hard to make it do something sensible. I don't think there is much to do for NullBot to get new opportunities, maybe just giving more templates will do the trick.

So I think the next step will be tuning NullBot to have a good AI with this mod activated (need to see if it can detect the mod or ship Step Research with a modified version )

[karamel]

Version 0.3

• Defense, propulsion and body researches are free except fortresses and Dragon (but still have the same requirement), Bug is given at startup
• MRA, Ripple, Seraph and Archangel are anti-tank, Plasmite bomb is artillery
• Hurricane and Whirlwind are machinegun and share the same branch. All AA-Gun upgrades became useless and were removed.
• A slightly modified version of Nullbot is embeded
• Nexus resistance circuits are reachable

The first point is just to remove "side effects" researches. Bodies are attached to metals and heat armors (cyborg upgrades are more effective to compensate one body), propulsions to engines and defenses to turrets.

The second point is just for consistency. Otherwise the rocket branch provided weapons for everything and all bombs are artillery except Plasmite.

The third one is to give the machine gun branch an AA weapon and it looks more logical than having a dedicated AA machinegun branch.

Balance is still not fixed at all but the basic scheme seems to get in shape. No need to spend a lot of time ajusting stats and requirement if the next changes breaks all the tuning. At the begining I wanted to do just a research tree mod, to be compatible with other balance mods or so but I think now that it is pointless. So I began to edit stats but still willing to stay close to the regular Warzone to switch from one to the other easily.

[iap]

Hi, You're putting a lot of effort to this, and no one answers.
Sadly I can't check it now, but as a noob, at least in this sense, can you explain please what is the purpose of a "step" research? Is this like other RTS where you have an "era" (upgrade a village to town and a town to a city)?

[karamel]

Hello

You have it. The steps means you get along the tech tree step by step, that is there is only one research item available per branch and there are 4 "eras". Researching an item in a branch unlocks the next one.

If you don't research any synaptic link upgrades you will be stuck with low tech, when researching Synaptic Link Data Analysis (the first upgrade) you access techs from the 2nd era, of course only if you have researched the begining of the branch, because there is still only one item available at a time for each branch. Once you research Dedicated Synaptic Link Data Analysis you get the 3rd era, and Neural Synapse Research Brain the last one.

Check the spreadsheet on the first post to get a clearer view of branches and eras. Each column is a branch and you need to research synaptic upgrades (from the first column) to have the techs listed in further rows.

[Goth Zagog-Thou]

Hiya Karamel. Sorry that I haven't given your mod a go yet. I will try it out and post my thoughts about it. Looks interesting.

And I apologize for not giving it more attention. Thank you for your work, and I'll give it a spin momentarily.

[Iluvalar]

Hey karamel ! Good job ! I like that overall. Don't worry you will come to know me, I'm giving more complaints then flowers , but I repeat that mod is not bad at all !

1) You have a design fail. If you glue the components with their upgrades that much. The players will always have both the upgrades and the components. You are killing the last few situation where the players could use an old technology. I know the situation like that are already rare in the standard game, but the mods should improve the diversity of weapon not reducing it IMHO.

2) Your design also push the players into "squaring" as the progression rate of the research is slowing down during the game, players will tend to research a bit of everything which will lead to every playing having identical researches (with the exception of the weapon line maybe)

3) You are in my turf now ! Get out ! lol . Did you looked at my NRS mod for 2.3.9 ? Your ideas are good, and you go in the same direction as me. I have automatic tools and some pretty advanced mods. I personally dont mind competition at all. But you might fall in love with NRS .

4) You have a lot of balance to do now... It might be the hardest part for you... I've been decimated by 2 lancer cyborgs.

5) Things are going slow, if you want to have some public in the multiplayer games, I suggest you make the research a bit faster. If you need help to make some global changes to make one SP and one MP mod, I have some interesting tools for that.

[karamel]

Hello

I have read a lot from you and I appreciate the flowers and the critics. To be honest I didn't know what to do next, you gave me some advices.

1) I was a principle at start to mix weapons and upgrades. The idea behind this is to know if you see heavy cannon for example it means that they are powerful. You can't show a fully upgraded light cannon stronger than a not upgraded heavy one. But maybe this situation is just stupid or is a clever bluff strategy (don't know).

2) I agree with it. I noticed lately that a team of 2 could research a lot of branches at once (I didn't played much team games). A team of three may get all the techs With the few online experiences I had I feel it good to be able to switch quickly (but not too much) on an old branch to have a chance against a branch rush and not finish the game at first encounter at "bet what I'm rushing". But this is pure theory, maybe it's wrong…

Maybe a solution is to split branches in two (devices/upgrades) to lessen squaring, make team choices larger, enable bluff researches and speed up rushes (you still have to complete all techs in an era to continue so switching will still occupy labs/time)

3) I tested it once but I was affraid of all the new things I don't have a lot of time to play but the introduction inspired me for Step Research

4) I know, hardest for me as I don't played much online to understand the flaws and AI is not a good balancer. It will be a very long task.

5) A big slowdown is the burry of power and research modules in the tree. Maybe I can make them available sooner and highen the build prices to make them a strategy choice instead of a must have (and make them an early VTOLs target?). And just say that a step is 10 rushed minutes long for all instead of lowering them (meaning you can have seraph viper wheels at 30min with 3 labs instead of after one hour).

Thanks for your feedback, I read you have a lot of tools for balancing and some other things but it looked complicated at the first glance and was more on basic principles of Step Research. I'm interesded in what you have but again don't have a lot of time. So feel free to guide me to what you think is appropriate or even contribute (I think I can setup a git public repo)

[Iluvalar]

1) I believe it's part of the fun to have to spy on the enemy advancement. After a little time, it become obvious that your units survive better if the enemy had no research at all in a weapon. If you pay a little attention. It is safe to guess that if a opponent use a weapon, he is also researching it. So you dont need much visual clue.

Also, you will notice in NRS that new components lines are a bit more rewarding. To encourage players that have to replace their old design. So if the players want to focus on a weapon line, unless they try a bluff, they will spend at least a little for new components.

2) Even if there is more lines that you can fill with your labs, if the cost of the research increase faster than the gain, players will tend to switch from one line to another (30% dmg + 30% HP + 30% power instead of 120% dmg). Since you are trying to straighten the standard research tree, I'm not sure how you could fix that...

3) You should try it again and not be scared, all new structures except the spectral generator are mandatory. In fact, the AI dont even use anything else new. The new stuff are designed to be fair but not better than the normal stuff.

4)...
5) In NRS you can have a seraph viper wheels at 15 min with 1 labs and I still have complaints about how it's slow in MP... hehe

[karamel]

Version 0.4, with the feedback from Illuvalar and deep changes

• All modules (factory, power and research) are available at start
• Weapon and upgrades are split in 2 branches, almost all branches were split
• No longer synaptic links requirement (that's the end of the "eras")
• Bodies and propulsions are no longer free and have their own branch
• All research costs were changed

Power and research module should speed up a bit the game, but they have a cost: a power module is paid back after around 1 minute when supplied by 4 derricks and a research module costs the same as a new lab (+ build time) so building them right from start may not be a good idea. And they no longer give extra HP so beware of VTOLs (while I write this I notice I haven't touched the factory module at all)

The tree was reworked to split upgrades and devices, there is almost twice as much branches as before and there's no link between them. You can have a genuine plasma cannon or a fully upgraded light cannon. This should also make team game a bit more versatile. Some weapons moved, bombs and bunker buster are now bound to the mortar tree (need bombard to start researching them), Nexus turret and EMP are bound to wide spectrum sensor (like in the original tree) and hardcrete wall was set to be researchable at start.

Then, with the fall of eras, you don't need to lock a lab with research upgrades, it's just a bonus like the others that you can research or not.

And the last big change: research costs were completely changed. It's now a flat cost, the first upgrade costs the same as the last one, but you have spend a lot on it in the end. 2 exception for synaptic link upgrades, which costs a bit more each time but not more than the bonus it gives (to reduce the exponential bonus) and weapons that may cost more if the gap is important (like between machinegun and twin machinegun and heavy machinegun and assault gun).

Of course, none of this is balanced

[Iluvalar]

2 exception for synaptic link upgrades, which costs a bit more each time but not more than the bonus it gives (to reduce the exponential bonus)

It's the wrong way around. The real chokepoint in the game is the cost, not the time. As a result, you can double the research speed, and the player will not have double research. He will only have his researches half a step before the opponents.

As a result : Poduction upgrades (lab, engineering and factory) should follow their own progression which is highly exponential.


Warning : The more you listen to me, the more all this will become a clone of NRS

[karamel]

Version 0.5, back to basics.

• Cyborg engineer is available without researching engineering
• AI was slightly modified
• Defense researches are no longer free
• A bug on Cannon fortress requirements was fixed
• All modules, VTOLs and Cyborg factory need to be researched, as well as Tank trap
• Engineering and manufacturing are now split in 2 branches

Well, all of this makes Step Research closer to the regular game.

[Goth Zagog-Thou]

Good stuff, karamel. I'll try this new version at my earliest convenience.

[karamel]

Hello, I'm back with a new PC (thus Warzone and some other games run like a charm)

So I took some of the dust off step research and here is the 0.6 version. From the turn it took it is no longer a research tree mod but a less complex game. The idea (for now) is to get a more "noob friendly" game with less subtilities but still a lot of things (there is a clear difference between viper and bug for example). The idea is to have clear concepts all the way to reduce the learning curve. Step research is though not a good name anymore.

So some concrete changes to illustrate :

• Weights and engine power has changed
• Assault cannon and Hyper velocity cannon were removed (for now)
• Old designs are always available
• Hover is researchable at start and is a requirement for VTOL

The major change is unit speed changes. Max speed is doubled but almost never reached, weight were changed to have gradual speed with designs. For now light weapons are a lot faster than before, but upgraded weapons are always heavier than the previous (thus slower).

So New Paradigm bodies are lighter than Project (thus faster), Collective heavier and stronger (slower) and Nexus looks like luxurious high grade bodies (nothing new). The bigger the better, to summarize. Larger bodies are faster and stronger, especially with heavy weapons, but are less numerous. Also the bigger the gun is the heavier it is (most of the time), so weaker weapons are always faster. Engine improvements have some effects now.

Before getting flamed for the Assault an HPV cannon I justify their removal for leaving the cannon line at a similar size than other branches and to have just the idea "cannons are all-rounder slow weapons with high HP" (thus not much VTOLing with them). An assault cannon branch will probably come with the concept "light all-rounder weapons".

Hover is downed in the tree from behind tracks to have 2 branches: strong slow propulsions and fast light ones (with VTOL).

I started to write a few lines to explain the basics concepts. The idea is to know what things can do with a 10 pages booklet and get it in 5 plays but still have a lot of way to use it.

Still AI is not a way to see if it is playable well (walls should contain flamers, not everything and high speed is nothing when not hit and run is done) and I lack a lot of online experience. Even if I have a lot of fun tweaking Warzone it is better when it is used for anything . Feedbacks are highly appreciable.

[karamel]

New version 0.7, with it's manual/manifest. The booklet is not 10 pages long in the end but 21 but with plenty of images and decoration things. It can be considered of what Warzone with Step Research should be.

The question about non dedicated anti-air weapons is still open. Should there be or not? If yes should the modifier agains VTOLs be flattened (like it is for dedicated weapons)?

Changes from 0.6 :

• Degressive research cost, the more you advance in a branch, the quicker it becomes
• Heat and regular armor improvements are merged
• Edited engine and weight a bit to fit the "the bigger the better" paradigm
• Edited VTOL weights and engine

I tried to port it to Warzone 2.3 but it is more time consuming that I expected. I may retry it if it has some interests.