Reverts d2k balancing to the original and another loading message.

[Unit158]

This is more of something that I want to discuss atm, it is clearly not suitable for merge :P.

If someone could find a good price for Fremen and the saboteur, I would be grateful.

[Mailaender]

Pinging @psydev as he did the d2k rebalancing for the last release.

It is currently impossible to merge because you started with an outdated revision of bleed. To update your branch use the following commands:

git remote add openra https://github.com/OpenRA/OpenRA.git
git fetch openra
git rebase openra/bleed

[Mailaender]

Not sure why you came up with this. It is not necessary. We have the Invulnerable: https://github.com/OpenRA/OpenRA/wiki/Traits for that.

[Unit158]

I didn't know about Invulnerable: at the time.

[Unit158]

Something I would like to add, I think it would be best to go to the original's values, then change from there. The original settings did not even resemble the original; it felt like a completely different game.

[Mailaender]

I started with the values from the RA mod, then @psydev rearranged them for a more fun game. I don't know where Dune 2000 stored the original game values and I never bothered to search more intensively for them as http://www.gamestar.de/spiele/dune-2000/36070.html awarded it being the worst balanced RTS of all time although the factions are almost identical. It felt like stone, bigger stone, even greater stone than stone, paper, scissor to the reviewers at that time which leads to either trike rush or teching up for one type tank rushes because the units don't complement, but obsolete each other.

[Unit158]

Eh, I am not sure I'd agree with that. I think they just got too pissed off to try, ha. In any case, a trike rush without troopers = dead. I would hardly say that anything obsoletes another unit, as if you simply send a rush of 1 type unit, you will lose. Sardaukar are probably the only unit which obsoletes another unit - the light infantry/trike. Even at that, they cost a ton.

[Mailaender]

I am actually okay with every change that makes this less like Red Alert on Arrakis so keep going. You get free hands on balancing. However I will decline anything that adds regressions or unused junk values.

[Mailaender]

This is too slow. We need it faster because #2246 is not yet implemented.

[pchote]

My brief experience playing with the original D2K is that carryalls are late game and high-tech units, and that harvesters are slow. I think we should try this slower value if it is indeed consistent with the original.

[Mailaender]

In fact you are right. It could also make the game slightly less spammy.

[pchote]

IMO we should merge this once the technical issues have been fixed, and let any balance questions be identified / resolved from playtest feedback. This could always be reverted if it is found to be worse than the current balance.

[Unit158]

hack, tries to give the LMG instant-hit capabilities, as most machine guns in the original were instant-hit. If someone comes up with a decent speed, please replace this.

[pchote]

1c256 is about 30 cells per second. 2097152c0 is excessively large, and may cause overflow issues. A value like 1000c0 is sufficient for instant-hits, as we are unlikely to ever support maps >1000x1000.

[psydev]

I know I haven't been around lately as my life became focused on school, but I am still interested in this game.
When I did the balance for D2K (and CNC), I tried a holistic approach, where each unit had a purpose in the game, and had strengths and weaknesses, and was supposed to fit in a wider, purpose-designed tapestry. I am fully supportive of modifications to make gameplay better, but it should be noted that tinkering here and there without a grander vision in mind can undo some of the purpose of what I did in the past.

I am happy to talk about what I did and why. I could write about it in a document if you want.

I made a strategy guide to Dune 2000 here: https://github.com/OpenRA/OpenRA/wiki/Strategy-Guide-for-Dune-2000-Mod

[psydev]

If you want to look at the original values from Dune 2000, you can download a program called TibEd which can open (and modify) the data files from a Dune 2000 installation.
They are quite different from what we have.

I made an excel spreadsheet with the values for each unit (damage per second, range, % vs. armor, etc.), for the original game, then the first OpenRA version, and then my modifications.

[Unit158]

Sorry for being blunt, but your balances had broken the balancing from the original. I admire the effort, but as it stood, the mod was nearly unplayable for my father, who had lots of experience with the original. The previous balancing had ruined a great deal of his strategies. In my experience playing against the AI, your trikes were god-like.

I almost completely changed the balancing, giving the game a feel like the original (in my opinion anyway, I haven't had enough testers to properly tweak things for this engine). The values, if not being the same as the original's, are similar as the original's. Only time-based issues were not accurately ported.

On 2014-02-12, at 11:12 PM, "psydev" notifications@github.com wrote:

If you want to look at the original values from Dune 2000, you can download a program called TibEd which can open (and modify) the data files from a Dune 2000 installation.
They are quite different from what we have.

�
Reply to this email directly or view it on GitHub.

[psydev]

I appreciate your candid feedback. But I was not trying to stick to any "balance" from the original; I doubt the original had much balance to speak of. The factions were quite similar--painfully so. Naturally, with all mods in OpenRA, your strategies in the original games will not work at all in the new game. This is true for every mod.

I don't think there is much gameplay value in the original. As mentioned by Mailaender earlier, it has practically the worst multiplayer in RTS history. The game was known for trike rushes and maybe a quick tech-up to a desired unit (e.g. Devastator). It's not really something worth emulating.

I think the unit balance is pretty decent as it stands now. I think if we want to make the game more interesting, a better option would be to look at the buildings and tech tree, and maybe unit costs. We could, for instance, implement upgradable barracks, factories, etc. Another possibliity is to enhance the starport to make it more interesting. I'm not saying that changes to unit balances should not be considered, but I think trying to become more like the original is a terrible idea, and the wrong direction of development.

[Unit158]

As for Mailaender, he did not play it. He simply read a review, or so he told me. Your balance did not fix trike rushes either. As I pointed out with a previous comment, it is better to use their balancing (which was done by professionals, not amateurs!), then move out from there. OpenRA still has the feel of RA, I didn't think the same was true for D2K. Also, in my tests, many units would simply make each other obsolete.

EDIT: unless you have some sort of plan to prove how units were supposed to interact with each other, I am going to be fairly close minded. Sorry.

[Mailaender]

Please remove the ending : copy pasta here and everywhere else.

[Mailaender]

Needs a rebase.

[Unit158]

Reopened, but if anyone merges this prematurely, you get to finish the job. I still have a lot to do

[Mailaender]

Needs a rebase.

[psydev]

I am busy with school right now as I'm in the last couple weeks, but I would be happy to balance-test the game with anyone whenever I'm on IRC. (will probably be weekend).

[Unit158]

Work on this was destroyed with my old internal branch.

[Unit158]

Reopened because I feel like working on it again :3

[psydev]

I'm not sure what you're intending, but here's some advice if you're interested in starting with small tweaks.
I think the trike is OP, as most people have mentioned. One thing that could be done there is reducing its power vs. quads, and/or reducing its power vs. infantry slightly.
It might be worth testing to make sure the rocket soldier vs. trike match-up is fair.
I thought it might be neat to make grenadiers powerful vs. light vehicles, so that they could theoretically destroy trikes (as long as they are protected long enough to shoot).

I made some balance tweak suggestions in a spreadsheet for D2K a while ago if you have any interest in some of the things I proposed.

[Unit158]

As was said above, this was meant to revert d2k to a similar state as the original's balancing.

[Mailaender]

I don't recommend that.

Part of the problem is that Westwood attempted, but failed, to balance the game.

http://www.gamespot.com/reviews/dune-2000-review/1900-2540099/

[Unit158]

That is what I've been told to do, and for all my intents and purposes - the reviewers at gamestop are liars. Each of our attempts to rebalance from scratch failed miserably. We could at least go with the balancing that has the larger user base, then ask at the d2k fan forums for help with balance.

Also, lots of gamestop's reviews are complete bullshit.

On 2014-08-17, at 1:09 AM, "Matthias Mailänder" notifications@github.com wrote:

I don't recommend that.

Part of the problem is that Westwood attempted, but failed, to balance the game.

http://www.gamespot.com/reviews/dune-2000-review/1900-2540099/

—
Reply to this email directly or view it on GitHub.

[Mailaender]

Okay. This needs a rebase and as it is half a year old and we did lot's of engine changes already, you may want to start from scratch, but still using this as a template. Resolving merge conflicts in MiniYAML is very time consuming, annoying and error prone.

[Mailaender]

My rebalancing idea was to use the new BuildAtProductionType to have a combined light/heavy vehicles queue, but production acceleration enabled #5956 so it makes sense to have more factories, but still avoid the super spammy gameplay of the original. You can still get an extra queue with the starport.

[huwpascoe]

Skimmed this rather long topic, here is my quick 2 cents

The world of Dune is all about the spice. The worm is the spice, the spice is the worm. Get the spice correct first then everything else will start making sense.

The harvesters aren't long range hunter gathers, they are slow vulnerable feeders that have a very close relationship with the carryalls that transport them to and from the refineries. Now the sandworms aren't fond of all this harvesting and the more vibrations that the harvesters cause they will attract worms. If you have 3 harvesters working a patch then they're certain to get wormed. This interaction is what limits production. the worms also gobble up armies of spam that sit there on the sand, so you can't just wait outside enemy base pinning them in.

Being on sand is risky business, units should not be speedily driving around as if it were pavement. Dune itself is the true antagonist in the war.

[UberWaffe]

To pitch in my little bit.

I'd say play to what makes D2K different from RA and TD.

• Rock and Sand distinctions
• Three sides

1. (Rock and Sand) It usually isn't possible to build your refineries right up against your spice deposits, so harvester raiding should play a larger part.
   • I'd recommend making light vehicles good at murdering harvesters, but comparatively terrible at most other jobs.
   • I'd recommend slowing harvesters down and giving them a larger capacity, so that the income rate is unchanged overall but they spend more time out in the danger zone.
   • Arguably this is mostly a map design thing.
2. (Three sides) Currently near identical, I think a distinct strategic approach needs to be defined for each side and then balanced accordingly (from rifleman to superweapon). I suggest:
   • Artreides: Trench and Bombardment warfare.
     • Long range, area weapons.
     • Strong base defenses.
   • Harkonnen: Brawler warfare.
     • High hitpoints.
     • Short range high DPS weapons.
   • Ordos: Guerilla warfare.
     • Fast but fragile.
     • Burst damage (High damage, burst weapons, long reload between bursts. Allows hard first-punch, emphasising ambush tactics).
3. (Three sides) I'd mix up the support powers a bit more as well.
   • Atreides:
     • Outpost: Reveal Area
     • Upgraded High-Tech: Airstrike
     • Palace: Orbital Beam strike
   • Harkonnen:
     • Outpost: Upgrade Target Armor
     • High-Tech + Starport: Combat Tank Paradrop
     • Palace: Death Hand
   • Ordos:
     • Outpost: Grant Target Cloaking
     • Upgraded High Tech: Infantry Paradrop
     • Palace: Saboteur Paradrop

[huwpascoe]

Agreed, UberWaffle. This is the direction the game should be developed.

[Unit158]

I'm going to have to ask you to please leave your comments at the forum. Your ideas are great, but I'm not implementing them in this PR.

[pchote]

@Unit158 mentioned in IRC that he doesn't have time to work on this anymore, so closing again. I would really like to see this completed at some point.

[psydev]

I like some of UberWaffle's ideas for support powers. I think unique support powers for each faction would go a long way to improving gameplay by giving each faction a more distinct flavour and playing style.

Atreides:
Upgraded High-Tech: Airstrike
Palace: Orbital Beam strike
Instead of Orbital beam strike (since there were no orbital cannons in the original D2K), maybe Fremen drop instead? Or perhaps an electronic weapon that temporarily reduces enemy units' sight range to 2 in an area of the map, giving Atreides' long range tanks a bigger advantage.

Harkonnen:
High-Tech + Starport: Combat Tank Paradrop
Palace: Death Hand

Ordos:
Upgraded High Tech: Infantry Paradrop
Palace: Saboteur Paradrop

His suggestions about "granting target unit an armor upgrade" sounds nice but someone would have to code that feature in. Also not sure if people really want that in D2K.