Dune Wars 1.9.4 Patch Feedback

[ChrisAdams3997]

I don't see this going well for the Fremen, devastators are fun

Spoiler :

This has been a fun game so far, obviously Harkonnen, Beast Rabban, Monarch (wanted to see how it differed from Emperor), epic speed, standard size map.

As you can see from the mini map, I'm doing well and have taken the whole polar cap. I'm friends with the Corrino to my North, have been at war almost constantly with the fremen, though I haven't invaded beyond the polar region as my hands have been quite full, and I'm also currently fighting the Tleilaxu to the west (just took their capital).

The eastern part of my territory was originally the Ecaz, who attacked my early on, and... didn't do so well in the end (though they put up a good, fun fight). The Atredies (Leto II) also conquered the Ordos (the NE of their territory) early but have been pretty quiet since (other than one stalemated war with me). It's only been in the last hundred turns or less that I've really put up a tech lead on them, as they lead or were neck and neck with me much of the game. At this point though I've just built the Guild research facility and University of Arrakis and I have the Choam Directorship, and a lot of spice, so I'm starting to put some distance between us.

observations on air combat

Before launching the (ongoing) invasion of Scytale, he had roughly two wasp fighters in every city, which is pretty good, and if I didn't have Locusts I wouldn't have bothered trying to get air superiority, and even as it was it took some time to really take them all down. Something I've seen is that the AI fighters love taking interception% promos, which made sense before when it actually affected air combat. But it doesn't do them much good against my AA missile promoted locusts, I'll be taking a look at rebalancing their promo priorities to correct this.

As a general rule, it was pretty easy drawing their fighters out by lauching air strikes with my fighters, in which my combat1/AAmissiles1/AAmissiles2 locusts (9 + 50% = 13.5 strength) where able to shoot down their generally interception% promoted wasps (so 6 strength) in most of those engagements, while about 10% of the time the wasps would get away nearly dead. If there was a general parity in strength, it'd be very difficult to actually shoot down many enemy fighters since most engagements would result in a draw and both sides' damaged units would be healed before risking another fight.

After shooting down all their fighters I moved 5 firefly bombers into carriers in range of their capital to coordinate with an attack force moved in on carryalls. They had 5 missile troopers in the city for my bombers to contend with. The plan was to hit the city (air strikes) with the bombers till the interception rate came down enough that the last bomber or two would have a fair chance at getting through and inflicting some collateral on the city, and indeed the first four were intercepted, and at last the final bomber got through.

If I recall correctly the general survival chance was somewhere in the area of 80-90% in those attacks, and none were killed by chance. Should this tactic maybe be a little more risky, say work it out where we could expect twice the risk of death under the given circumstances (so a general small increase in lethality of ground interceptions in general)? My thoughts is I probably should have been risking on average one or two deaths out of the five attacks.

 

[Ahriman]

Whether there is a problem or not, tile bombing affects the human player as well.

Yes it does. If an AI player you're at war with gets bombers before you have interceptors, then they will destroy your infrastructure too.
It's not really fun, because it is so dependent on a single tech.

In my latest test game (on Monarch) the AIs have been fairly aggressive in keeping fighters in most cities near my borders during war

The biggest problem is that there is often a long period between when the human gets bombers and when AIs start to build interceptors; they might not have air power tech.
Even once they have the tech, it's not hard to eliminate the entire enemy airforce, or to get to places where they don't have defenses.

To put it simply, this is a fun thing to do sometimes. I don't use it often, but when I do, it's because it's fun to be able to hurt an enemy economy using air superiority

I don't think it's much fun to have an ability that the AI can't deal with very well. We have made a big effort in Dune Wars in general to try to limit new mechanics to mostly be those that the AI can handle.
I think balance is inherent to fun; if one play path is much stronger then others, then it tilts optimal play.

That's a much better approach than destruction of buildings. It represents more the temporary disruption in infrastructure without setting the city back 20 turns or whatever from one successful bombing run.

Perhaps we could implement something like this then, rather than the ability to bomb imrprovements or destroy buildings?
It seems low-MM, low-fuss, and easy for the AI to do and to counter. And you can probably ignore Dale's code?

I'll be taking a look at rebalancing their promo priorities to correct this.

Sounds good.

As a general rule, it was pretty easy drawing their fighters out by lauching air strikes with my fighters

If there was a general parity in strength, it'd be very difficult to actually shoot down many enemy fighters since most engagements would result in a draw and both sides' damaged units would be healed before risking another fight

This sounds good, working as intended. If you have locusts and they have wasps, you should be able to annihilate their airforce, but otherwise you should mostly use your fighters to suppress the enemy aircraft and make way for your bombers.

Should this tactic maybe be a little more risky

Not sure. It does seem like the interception chances are good. If they're 80% survival chance that seems reasonable, 90% might be a little high, I suppose this could be tweaked slightly, there should be some risk of death.

 

[ChrisAdams3997]

A little Mech maintenance

I got around to doing something I've been tentatively planning for a while. I've been putting off playing a proper Ixian game for a while because of the way the mech line of units was set up. The idea that there are different kinds of mechs with different abilities and uses starts off on the right foot, but then they all upgrade in a very illogical line.

If you are talking about scorpions or melee, each upgrading to the next works just fine because the next unit is generally just a stronger version of the previous, their general role stays the same. In the case of a Crawler upgrading to a Walker though, you are moving from an anti-melee to an anti-vehicle/air unit. This is as weird as if the only upgrade for bladesmen was to missile troopers.

This means that if you are upgrading your mechs, at any point in time, you really only have one of those roles filled by the different mechs present at one time through most of the game, so all the effort to differentiate them is largely wasted. This is frustrating for a player, and is also not good for the AI since it's not going to know to switch Unit_AIs when the best use of the unit changes upon upgrading. It relies on the upgrade path being set up logically for units to stay in their same role as before the upgrade.

Of course you always have the option of not ever upgrading your crawlers (for instance), but then they just fall too much behind in strength as technology progresses and they become useless.

So what I've been wanting to do is expand the mech units and reorganize their upgrade paths so that we keep several general 'roles' with upgradability within that role. To accomplish this I've created 'upgraded' versions of several of the mechs, representing how the armament and whatnot increases with technology, just like a medium scorpion is obviously an improvement on a light scorpion.

This boils down to the 'anti-personnel' line and 'anti-vehicle/air' line at first. anti-personnel of course starts with the crawler, which is unchanged, and upgrades to the strider. The striders keep the 13 , but instead of +50% guardsmen, I wanted to keep them more similar to the crawler's strengths, so I've given it +25% versus guardsmen and melee and ignores building defense (like crawlers). The Strider MkII has the same bonuses, but boasts a strength increase for the late game to keep them relevant. Walkers also have a late game MkII unit that simply has increased strength.

The Cymek is an interesting idea, but in practice it acts like a scorpion with a few possible first strikes and no malus for city attack/defense. At just 15 , they are easily outclassed in usefulness by later scorpion units. Again with the idea of keeping them relevant into the late game, they have a MkII unit with 18 and an extra chance first strike.

The only change for the taratula (which was before the only late game mech) is that nobody upgrades to it, they must be built from scratch. I think that none of the others really logically do enough of the same thing so I preferred to keep them in their separate 'class'.

Spoiler :

The MkII units are all meant to be quite late game units. For now, they all require stravidium (which means you'll need plasteel even though it's not a prereq for any of them). You'll also notice that the icons for them all have a little 'thinking machines' icon on them. I really wanted to tie these into requiring the thinking machines religion, as there's not a whole lot of reason to actually adopt it at present. It makes enough sense to justify this link that in order to build these more powerful mechanized units that you'd not only need the technology for better guns/armor, but also the use of thinking machines.

I'll also note that I'm really pressing these to fill roles different from standard vehicles available to everyone, and of course keep those different roles available throughout the latter part of the game. They should play differently, be something you'd want to get for the tactical roles they give you, while also enhancing the flavor of going a technocratic route.

I've balanced them as well as a first pass could. I've generally reworked the hammer costs of the late game ones to be more in line with other vehicle/thopter choices within the same tech group. The way I'd describe them before was less powerful than other alternatives (particularly against the most powerful scorpions) but cheaper in hammers. For most players, they'll gladly build a more expensive unit (in hammers) if it's more powerful since you're economically limited in how many units you want to field. In the late game, fewer, more powerful units is generally considered a smarter play than a bunch of cheap but less useful/survivable units. As they were set up before I don't think for a moment I'd be better off using mechs over scorpions, they just didn't offer anything a scorpion couldn't do just as well.

I'm willing to hear any suggestions on further refinements to the mech line, what techs to require, bonus prereqs etc. Nothing is set in stone, but I did want to get the general layout set down before hand so we have a good base to work off of.

 

[keldath]

ChrisAdams3997,

hi to you,
i havnt touched dune for so long.....

anyways - the reason you got mech in dune and scorpions, is because in my few first release, ack n the day i was the dune team, i intended to have an all land map, and mechs was suppose to be an all terrain, so now i guess scorpion and mechs, probably have the same purpose.

you can make the mechs be like some special limited units in addition to scorpions, that can e like a realty good support unit but one that you cant have over say 5 at all time.

i dont know how todays game play goes, i left dune ions ago,
but i plan to play with your recent patches soon to see how the mod feels once more.
ill give feedback later on i hope.

plz keep up the work on this, makes me proud

 

[KHM]

A little Mech maintenance

The idea that there are different kinds of mechs with different abilities and uses starts off on the right foot, but then they all upgrade in a very illogical line.

If you are talking about scorpions or melee, each upgrading to the next works just fine because the next unit is generally just a stronger version of the previous, their general role stays the same. In the case of a Crawler upgrading to a Walker though, you are moving from an anti-melee to an anti-vehicle/air unit. This is as weird as if the only upgrade for bladesmen was to missile troopers.

This means that if you are upgrading your mechs, at any point in time, you really only have one of those roles filled by the different mechs present at one time through most of the game, so all the effort to differentiate them is largely wasted. This is frustrating for a player, and is also not good for the AI since it's not going to know to switch Unit_AIs when the best use of the unit changes upon upgrading. It relies on the upgrade path being set up logically for units to stay in their same role as before the upgrade.

Of course you always have the option of not ever upgrading your crawlers (for instance), but then they just fall too much behind in strength as technology progresses and they become useless.

So what I've been wanting to do is expand the mech units and reorganize their upgrade paths so that we keep several general 'roles' with upgradability within that role. To accomplish this I've created 'upgraded' versions of several of the mechs, representing how the armament and whatnot increases with technology, just like a medium scorpion is obviously an improvement on a light scorpion.

Great. I came to the same conclusion in an earlier game but simply decided to change upgrade paths so that mk2 upgraded directly to mk4 (Tarantula), not mk3. This eliminated the biggest problem, since mk2 imho is more useful than mk3, and far cheaper. Anyway, this was more like a workaround than a solution.

Keep up the good work!

 

[Ahriman]

It's certainly worth thinking about the mech units and doing a balance pass here. I'm not confident that their unit strengths are correct given their place in the tech tree, and I'm certainly open to some tweaking. However, I ask that you understand and respect some of the original design decisions that went in here.

However, I do quite like the current system of specialization at each level, but different specialization types across levels.
This was a deliberate design decision to make them different. At each level, they are an extreme specialist, but the specialization changes. So you are faced with an interesting strategic decision; you can either play up their short-term advantage by giving them promotions that fit their current load-out (eg: anti-melee promotions on an anti-melee mech, anti-guardsman promotions on an anti-guardsman mech, etc), or you can play up their long-term advantage by sticking to their generalist power (with combat or drill promotions) but at the cost of missing out on extreme short-term specialist.
I think it would be boring to split them into separate tech lines and just make them replicate existing infantry/guardsmen/scorpion lines. All the other unit lines already work like that. I would like the mechs to be different. So I would oppose your general design approach here.

There is also a strong design approach to have a limited number of unit tiers, and then to have the top-level units get enhanced by the late-game super-promotions. The late-game units (Fedaykin, Kindjas, Sardauker Noukkers, etc.) are supposed to stay as the highest tier units. These are the thematic and iconic units; we had a deliberate design decision to make sure that these high-end units stayed useful right until the end of the game, and that we did not want to add a separate tech layer on top of them.
The top end Mechs should be on par with these guys. There should not be secondary upgrades of these.

Another important thing to remember is that the Cymek is "different"; it is one of the main features of the Thinking Machines religion and tech-path. So it needs to be quite strong. I think it should probably be good for city assault; this helps keep it different from the scorpion line, which an Ixian player will also be using.

The Thinking Machine religion should be about the Automated Factory and Computerized Laboratory, and then the Cymek. (And the shrine, which might need a boost.) It is supposed to be primarily about production, not military.
The Mechs are for Ix. They are the Ixian UUs.
We had a very big long series of discussions about this over many many years. The Ixian mechs are not robots; they are piloted vehicles. The Cymek is different.
Please do not change this. A lot of blood was spilled to get to this compromise. We want there to be some synergy between Ix and Thinking Machines, but the religion should be open to anyone, but should not be clearly more valuable than other religions.

 

[ChrisAdams3997]

There is a certain danger that comes from being too deep into the design of a mechanic that we are all susceptible to. Since you know, and possibly even came up with, the deep reasoning for going a certain route it becomes harder to see problems with it. In particular you lose the ability to see it from the perspective of one who doesn't know that they are supposed to be this or that, or that it's supposed to encourage this type of game play/strategic decision.

In such instances, the proof of wether it was well designed to meet the intended goal or impact on gameplay is wether players with no knowledge of what it is supposed to do or represent actually get it and naturally follow into that intended pattern. If it actually, in this case, feels like it's making for better strategic decisions. As such, I hope you'll listen to someone who wasn't there during the original design discussions when I tell you it doesn't make as much sense as you might have thought, and instead comes across at best as an unnatural restriction on how to use mechs as opposed to any other unit class in the game, and at worse makes them look (even if erringly) like an afterthought that was never fleshed out right.

 

[Ahriman]

There is also a danger in tinkering with a mod that has been under design for years. Never take down a fence unless you know why it was built. I note that you didn't address any of my actual concerns.

Certainly I will listen to feedback and suggestions. And certainly the values can be tweaked. But I hope you will listen to existing designers and not just decide that you know better.

I think it is *good* that the mechs work differently from the other unit classes.
Each faction is supposed to have a unique mechanic that will make them play a little differently.
Mechs are the Ixian mechanic. They should make you play differently in the same way that Fremen sandrider or BG Kwizatz units and so forth change the game. To me, they do not do this if they end up just replicating existing unit lines.

As for whether the design intention is realized or not; do you not find that with mechs you face a tension when choosing how to promote them? Do you really always promote them the same way? What promotion do you give them?

The mech lines get very large specialist bonuses inherent to each chassis. I think this is fun.
This is why we do not want it to be easy or wise to use promotions to specialize them further.
It is not a good idea to have a unit with a +50% vs guardsmen to be able to get another +40% vs guardsmen with no downside, because this leads to over-specialization. Specialization is very dangerous and powerful in Civ, because of how the best unit always defends. With large stacks of units, over-specialization tilts the game too far towards the defender and away from the attacker.
So, there is a deliberate mechanism with the mech units to discourage you from over-specializing; you can do so and will gain short-term benefit, but at a long-term cost.

It sounds to me like you are frustrated that you cannot just choose to boost the specialization of each mech further, and that you cannot keep each mech purely specialized in a single role as it upgrades. Good! If you were able to do this, they would be too specialized.

If we adopted your model, then every game I would give the anti-guardsman specialist extra anti-guardsman promotions, I would give the anti-vehicle version extra bonuses vs vehicles, etc.
There would be no tradeoff; I would choose the same promotions every game.

Here is a rule of thumb we could adopt; With ~20% of base strength modifiers (from terrain, combat promotions, or whatever), you should be better off to promote the unit than to not promote it even vs the specialized target.
So for example, a strength 8 unit with +40% vs melee and +20% modifiers is strength 12.8 vs melee.
A strength 11 unit with no anti-melee specialist and +20% modifiers is strength 13.2 vs melee. So you are always better off with the upgraded version than not. I think we could review stats with this design in mind.

The other thing to note is that this it would be bizarre to have 7 Ixian UUs. This is far more than other factions. The design goal is not for Ix to use nothing but mechs, or for mechs to obviate the existing unit lines. The design goal is for them to incorporate mechs into a combined arms approach that still uses other units. In particular, there is a nice synergy between generalist scorpions and specialist mechs.
If you had an anti-air/vehicle line of mechs, why would you ever build the missile trooper line?

And again, a distinction between the Ixian mechs and the Thinking Machines Cymek is important.
I am very open to changing Cymek stats, I have no strong opinions here, but I think it should remain a single unit. Each religion has 3 buildings or units that they give you.

 

[Ahriman]

I should also note; the fact that picking the anti-unit-class specializations isn't a great idea for mechs tends to increase the relative value of the unique Ixian vehicle class utility promotions (self repair, etc.). These are part of the Ixian flavor, so it's important that they are worth using sometimes.

 

[ChrisAdams3997]

It sounds to me like you are frustrated that you cannot just choose to boost the specialization of each mech further, and that you cannot keep each mech purely specialized in a single role as it upgrades. Good! If you were able to do this, they would be too specialized.

My problem is more so that at point A in the game, mechs are for one thing, and at point B they are for an entirely different thing, and at point C for something else altogether again, but I know that that doesn't bother you. My position is that players don't see this carefully designed system as you see it. If you know nothing about the theory behind the current design as you've just described, the mech promotion line instead comes across as very disjointed and illogical. At the same time, especially in the late game, they are too weak to compete with other unit choices.

Now, I really didn't expect this to be a major point of contention, and I want to let you know that even if we still disagree, I'm not changing this over your head so to speak if you disagree that strongly, it's not worth it to me. To be honest with you though, it's one of the things that I've felt was most, to use too strong of a word, 'broken' in the game design, to the point that I never bother wanting to get Ixian technology. By contrast, I'm happy to take a trade for Sardauker coop or Water Debt at the drop of a hat, even if I can't get the best units they offer. I don't want to play the Ixians because the one thing that is most special about them feels... off. There are reasons for that that I find hard to get past, and they are inherent in the current design.

So the biggest point here is that it's not about what you or I think about the system, it's how player's react to it, so I'm hopeful we will get some more thoughts on how people think about the current setup and if they find it to work well as designed or not. If you haven't heard me say it before, I always reserve the right to be wrong, no matter how adamant I seem on something , but right now we have very little to go on except our own opinions on the system.

The one thing I think we do agree on is that the Cymek at present isn't all that 'special' of a unit. It's basically from a player's standpoint a light scorpion without the city malus... That comes later in the tech tree really than a medium scorpion... And requires you to alienate yourself from everyone.

As a side note, Thinking Machines in general is such a late game technology to get to in relation to every other religion, it really needs to have more 'bang' so to speak. Right now it's more of a whimper, and there is no real reason to switch except maybe just for role-playing. Even a much better cymek feels too weak to substantiate it. It really should open up more options in my opinion, hence the direction I went. Not the most important aspect amongst everything discussed here in my mind, but I like the idea of it.

Hopefully we'll get somewhere on this, hate to see the discussion turn too negative. I am a nice guy at heart after all .

 

[ChrisAdams3997]

I also realize a lot of the reason you want to keep the current setup is as a deterrent to over-specialization, would you be open to other approaches to dealing with that? I think there could be ways of doing it other than a disjointed promotion tree (I really didn't mean to use a negative sounding word there, just all that popped into my head).

 

[Ahriman]

My problem is more so that at point A in the game, mechs are for one thing, and at point B they are for an entirely different thing, and at point C for something else altogether again

Can I ask: why is this a problem? I don't really see a reason or an argument. It isn't what you were expecting... ok... but why is it a problem?
Your reaction is negative, you don't like units whose upgrades change roles. Ok. Why?

Do you disagree with my point that overspecialization is a potential problem given the civ combat engine?
Do you disagree with my point that mechs are boring if they just duplicate existing unit lines?
Do you dislike the idea that it can be interesting to have a tradeoff between short-term benefits from boosting existing specialization vs taking generalist or utility promotions that boost long-term unit value?
I don't mean to be a hardass, but you're not giving me much to work with here.

It would be great to get feedback from other players; I've never seen any negative feedback on the mechs before. And as I said, I'm happy to tweak their stats. But I would prefer to keep the core design where they are specialists, but specialists that alternate focus depending on their chassis.

The one thing I think we do agree on is that the Cymek at present isn't all that 'special' of a unit.

Ok. What do you propose then?
I think it should be a powerful unit, most likely a good unit for assaulting cities.
But ideally it should be different from the existing units. Loading it up on first strikes seemed like one way of doing this.
Should it do some collateral damage?

As a side note, Thinking Machines in general is such a late game technology to get to in relation to every other religion, it really needs to have more 'bang' so to speak. Right now it's more of a whimper, and there is no real reason to switch except maybe just for role-playing. Even a much better cymek feels too weak to substantiate it. It really should open up more options in my opinion, hence the direction I went. Not the most important aspect amongst everything discussed here in my mind, but I like the idea of it.

I'm fine with strengthining the Thinking Machines religion, but its benefits need to come through the two buildings, shrine, and Cymek. So the right way to boost the religion is to boost the value of those buildings and unit - and have the other Mechs remain an Ixian mechanic.
One possibility I had in mind was that the robotic factory and computerized research lab should give +X benefits rather than +X% benefits. This could help enhance the flavor feeling that its automated things operating on their own rather than with the people. They could get sizeable values. With the unhappiness they give, this could also enhance a lots-of-small-cities strategy, as opposed to a super-cities strategy with Qizarate.

We could also make the buildings quite cheap, so they were easily spammable. This would also encourage you to spread Thinking Machines to lots of cities.

The other possibility would be to change the benefits you get from having the religion in a city. So, Qizarate could provide 2 happiness and 1 culture, and Thinking Machines could provide say +2 hammer +2 beaker just for being present in the city.

 

[Ahriman]

I also realize a lot of the reason you want to keep the current setup is as a deterrent to over-specialization, would you be open to other approaches to dealing with that? I think there could be ways of doing it other than a disjointed promotion tree

I'll definitely listen to whatever ideas you have. I think the most important thing is for mechs to work differently than the existing unit classes.

As I hope I have made clear, in general I support the vast majority of the changes you have made, the mod is greatly improved since you started working on it.

 

[ChrisAdams3997]

Didn't want to forget the other comments in the midst of the discussion.

ChrisAdams3997,

hi to you,
i havnt touched dune for so long.....

anyways - the reason you got mech in dune and scorpions, is because in my few first release, ack n the day i was the dune team, i intended to have an all land map, and mechs was suppose to be an all terrain, so now i guess scorpion and mechs, probably have the same purpose.

you can make the mechs be like some special limited units in addition to scorpions, that can e like a realty good support unit but one that you cant have over say 5 at all time.

i dont know how todays game play goes, i left dune ions ago,
but i plan to play with your recent patches soon to see how the mod feels once more.
ill give feedback later on i hope.

plz keep up the work on this, makes me proud

I wouldn't mind letting them be pretty powerful units to have, but limit the total numbers you can field like other special units. That would make balancing them easier without worrying they become too powerful/specialized.

Great. I came to the same conclusion in an earlier game but simply decided to change upgrade paths so that mk2 upgraded directly to mk4 (Tarantula), not mk3. This eliminated the biggest problem, since mk2 imho is more useful than mk3, and far cheaper. Anyway, this was more like a workaround than a solution.

Keep up the good work!

Appreciate the support

 

[Ahriman]

I wouldn't mind letting them be pretty powerful units to have, but limit the total numbers you can field like other special units. That would make balancing them easier without worrying they become too powerful/specialized.

This seems sensible.

I should note though that the units, like the tech tree, civics and religion, have all had a total redesign since keldath's incredibly important original work on the mod, so the current Ixian mechs have nothing in common with the original ones except the name "mech" and the art.

 

[Jester Fool]

I'm hopeful we will get some more thoughts on how people think about the current setup and if they find it to work well as designed or not

IMO, the current design is somewhat *unique*. I do not like the fact crawlers lose "ignore city building defenses" upon upgrade. However, I am doing just fine utilizing mechs in my Ix game. I think the overall design is good. You just have to know how to use each of the different mechs. An important point to consider is the totality of the units you will be using at each stage of the game. For example, crawlers early but then shift over to shield fighters with walker support once crawlers become obsolete. Ix is actually a very fun faction to play (with appropriate mechanic abuse).

 

[ChrisAdams3997]

Can I ask: why is this a problem? I don't really see a reason or an argument. It isn't what you were expecting... ok... but why is it a problem?
Your reaction is negative, you don't like units whose upgrades change roles. Ok. Why?

Do you disagree with my point that overspecialization is a potential problem given the civ combat engine?
Do you disagree with my point that mechs are boring if they just duplicate existing unit lines?
Do you dislike the idea that it can be interesting to have a tradeoff between short-term benefits from boosting existing specialization vs taking generalist or utility promotions that boost long-term unit value?
I don't mean to be a hardass, but you're not giving me much to work with here.

It's like what I said in the first post about if the only upgrade for a melee unit was an AA unit, or lets say a scorpion that could only upgrade to a siege weapon. If you are facing an opponent in the early-mid game when you have crawlers who fields a lot of infantry, it's a good idea to build some crawlers to fight them. Because they ignore building defenses, they can also be used for attacking cities.

Then you research mechanization, and it's time to upgrade. Any other unit in the game upgrades to a stronger version of what it did before. This gives continuity of your armies composition. If I upgrade all of my mechs, I've got something else entirely, even though it's still graphically another mech, and it's totally out of the player's hands. I realize part of the design is to make them work differently from all the others, but at this point, I've lost the choice to have mechs that are good at fighting infantry. Now the only choice I have is mechs that are good at fighting vehicles and thopters. When the next upgrade comes, I can now only build guardsmen specialists.

So it's not only that the upgrade is different, the build possibilities are limited in a way that just feels odd. Sure I can choose to not upgrade the ones I have, but as I said before, that's only an option for so long if your opponents are teching decently, and the AI will totally fail to see the difference or why it maybe shouldn't upgrade them.

After getting the tech for strider, what if I still wanted to build a vehicle fighting mech, I can't. It runs so counter to any other unit combat in the game, it changes uses in a way that is entirely out of the player's hands, that it can be hard to get a handle on how to use them in the first place. By comparison I know if I build a scorpion I can upgrade it without worrying about wether it'll have useful bonuses after the upgrade. It stays basically the same, just stronger. Breaking from that general design idea makes them much less desirable.

The somewhat less important 'realism' argument I'd make is that having for instance designed vehicle/AA specialist mechs (walkers), why would Ixian engineers abandon the idea of improving on the basic design, just as you'd improve on scorpion design over the course of the game, and instead go in a totally different direction. This is, as usual, of course less important to reconcile than the gameplay arguments (and of course 'realism' arguments can run in circles depending on whose version of 'realism' you are following), but I think this is a valid argument in a players head (I know it is in mine) when looking at it and plays into immersion.

Coming back to the AI for a second, depending on what UnitAI it built the original mech as, the new one might not fit that UnitAI role at all, and it's not going to change it. It has no idea that it should maybe reevaluate what UnitAI to use. They probably often fall into 'counter' unitAIs if they are better attack/attack city units available, but if crawlers are the best attackCity unit available and the AI builds a bunch of them, I don't think attackCity Walkers will be very scary. Just as an example.

 

[ChrisAdams3997]

An easy alternative way of dealing with overspecialization would be to simply not give mechs access to the specialized promotions (e.i. AArockets, shock, etc.), though it might feel a little over restrictive.

I think for now my brain is tired and I'll revisit all this tomorrow

 

[Jester Fool]

Coming back to the AI for a second, depending on what UnitAI it built the original mech as, the new one might not fit that UnitAI role at all, and it's not going to change it. It has no idea that it should maybe reevaluate what UnitAI to use. They probably often fall into 'counter' unitAIs if they are better attack/attack city units available, but if crawlers are the best attackCity unit available and the AI builds a bunch of them, I don't think attackCity Walkers will be very scary. Just as an example.

You may have a point about this. I would never use walkers to assault a city. I do think crawlers and walkers should be the same "type" of unit (both with "ignore city building defense").

 

[Ahriman]

This gives continuity of your armies composition.

I understand why continuity is valuable; I don't understand why lack of continuity is fundamentally problematic. It lets us make the units more powerful individually without making them overpowered as a whole.
The Mechs are designed to be incredibly powerful counters for their tech level. The only way that can also be balanced is if you don't have a ton of control over what exactly you are countering. If you could always choose what kind of counter you wanted, then they would either be too strong, because you could always counter anything the AI could throw at you, or each unit would not be super strong at its role anymore. This is the price you pay for getting superior specialist units. You can then prioritize or not prioritize the tech for the next upgrade level depending on which version you wanted.

It's different from the existing lines, sure, but different is good, IMO. It encourages you to play in a different way; you have super-powerful counters, but you have to adapt the composition of the rest of your army around them, or change your playstyle as the game progresses.

The somewhat less important 'realism' argument I'd make is that having for instance designed vehicle/AA specialist mechs (walkers), why would Ixian engineers abandon the idea of improving on the basic design

The idea is that the various mech frames are each a fundamentally different chassis, with different advantages. A walker creature with sharp blades might be really good vs melee. A unit with good shields and ranged weapons will be good vs guardsmen. Just like the fact that a modern tank isn't necessarily better at dealing with helicopters units than a ~1970 mobile anti-aircraft unit, which isn't necessarily better at dealing with entrenched enemies than a WW2 flamethrower tank.
And, as I said, a good rule of thumb would be to have each successive version at least as effective vs the specialist as the lower-strength preceding version.

Coming back to the AI for a second, depending on what UnitAI it built the original mech as, the new one might not fit that UnitAI role at all, and it's not going to change it. It has no idea that it should maybe reevaluate what UnitAI to us

Are you sure? I thought that unitAI was dependent on the current state of the unit, that it doesn't care where it came from or what the previous versions were. In vanilla, a longbow unit promoted up to infantry won't necessarily only act as a city defender, right?

An easy alternative way of dealing with overspecialization would be to simply not give mechs access to the specialized promotions (e.i. AArockets, shock, etc.)

I could live with that, though it still feels a little blunt. I prefer to leave the option open to the player; they *can* overspecialize their unit now, but at the cost of having a promotion that is not really useful in the long term.

I am worried that your design (specialized lines where you can always build the specialist of any type you like) makes the mechs pretty boring. They function like every other combat line and end up just replacing them because they are superior. Why build a missile trooper if you have a mech that is strictly superior?
If you can come up with a design that doesn't have this problem, I'll consider it.