How should Upgrades work?

[Hans Lemurson]

I recently discussed with AIL some inconsistencies with how the Upgrade system worked. It was possible to save a significant amount of money by doing upgrades in two steps (Armor/Weapons first, then Chassis) rather than in one go. This was clearly an oversight, and AIL was able to figure out a way to make the prices consistent.

He also changed the Togra bonus so that it was no longer cheaper to Upgrade a unit to completion than to rush-buy it.

However, the result is that across the board, Upgrade-costs are higher, and maybe that's not a good thing. So I thought it might be a good idea to discuss a fundamental question at the base of all of this: How SHOULD upgrades work in the first place?

Let's even back up one step further: Why upgrade at all? Why would you ever want to upgrade a unit rather than just building a new one? Experience and Time.

• Experience: If you have a unit which has levelled-up a lot in fighting, then it is MUCH more valuable in combat. A rank-11 unit is 2x as powerful as a rank-1 unit, but has the same price. Upgrading gives you "more bang for your buck" than building a fresh unit.
  
• Time: Upgrading a unit is equivalent to rush-buying. If a city is threatened, and you have a lot of money in the bank, then upgrading all it's defenders to Top-tier technology is a good way to keep it safe. It's a lot faster than trying to build new units from scratch.

Other than that though, there isn't really any strong reason to Upgrade. It's just a form of rush-buying, and so it only makes sense in the same situations that rush-buying makes sense in. This is what makes adjusting the price of Upgrades so difficult:
-If it's a lot more expensive than a rush-buy, then it'll never be worth it.
-If it's cheaper than a rush-buy, then you are better-off building Blank White-chassis units and upgrading instead of rush-buying (as is the case with the default Togra).

However, talking about the Togra brings up an interesting point: Because the Togra's upgrade costs were halved, and upgrading a blank chassis costs approximately what a rush-buy does, this means that the Togra's special ability is actually half-priced rush buys for units. Half-priced military units sounds totally OP, doesn't it?

But tell me: Has anybody here who has played Togra, and made use of their "upgrade trick" ever actually found this to be horribly overpowered? How is half-priced military NOT overpowered? Because it's not really half-price! Credits are scarce, and it's just not practical to build your entire military force using nothing more than credits (at least in the early/mid game). Also it's a real pain in the ♥♥♥ to micro-manage, keeping a few dozen empty hulls around at all times to upgrade into a military as money is available, but just because something is a pain doesn't make it strategically poor.

What are your thoughts on Upgrades, their pricing, and the availability of Credits?

[Ail]

I have reworked it today compared to the last upload and it is a lot more consistent now.

The exploit is way too effective and the biggest problem about it is: One usually uses it without even knowing!

In my test as Togra I could get a Tangora with Gold 1-Armor, Gold-Cannon and defense-device for a total of 796 credits. The normal price is around 3400 credits.

The reason for all this is that the price-tag of weapons, armor and devices is a percentage of the price of the hull of the unit instead of a fixed value. So when you upgrade armor and weapon first you get it for the price of the old hull and in a second step only pay the price of the hull.

I have now made it so that upgrading the hull will consider the currently used other things even if you don't change them and calculate the price-difference.

This is the same for all factions. Togra-Bonus was changed to: getting back 100% of the previous price. So for them they can turn something into something completely different for the same price without paying anything.

One thing that makes upgrades less valuable is the golden barracks. Then you can have new level 15 recruits while your old units rarely will be higher than level 10.

I want to test some more with that and also some other things. I might want to change the AI to adopt the new prices.

The scarcity of credits early on is no longer existant later in the game. In my last game where I won with economic victory I had an income of 8000 credits/turn. And it is a big differenc if I could make 8 new units from that via the exploit as togra or only 2.5 with regular buyout.

[Hans Lemurson]

Yeah, the golden barracks is hilariously powerful. It almost makes battlefield experience obsolete! It certainly makes "Field Training" obsolete.

I guess another use for upgrading is that it lets you get past the "1 unit per city per turn" limit by letting you perform a rush-buy on all of your currently existing units.

It's funny that by the time you have enough credits to easily upgrade everything, technology has made upgrading nearly obsolete.

[Ail]

Well... i got an idea for that... When you upgrade a unit that is in a city with a barracks and the upgrade costs more than 0 credits and the units level is less than what it would have through the barracks, it could be put to the level of a new unit. This way upgrading will stay viable even if it's more expensive.

[Hans Lemurson]

I like the direction that's going. You would be able to "re-train" your troops so that those poor Colonial Troopers you built at the beginning aren't useless forever. You also are trying to prevent the exploit of just having a single barracks in your empire and have every fresh unit drive itself back to the capitol to get upgraded.

However, you'd just be changing the cost of the re-training from 0 credits to 64 credits, and the situation would be the same, just slightly more hassle. Zero-cost upgrades are rare though once the White Missile/Machine-Gun are obsolete, so I don't think the restriction would even apply most of the time.

I actually think maybe that the Golden Barracks might be OP. Instead of the barracks line giving +2/+4/+8 experience points for units, +2/+3/+4 might be better. Fresh units will come out at level 10, experience will be harder to gain overall, and any units stronger above rank-10 that will be worth preserving.

[Ail]

Your way of thinking is amazing! I suggest something and you immediately see how it could potentially be exploited. :D

Another idea would be that the cap for the field training could be max(10, barrackspotential) so that once you get all 3 barracks the field-training becoms better aswell and allows you to get your old troops to level 15 aswell.

Also I kind of agree that the later barracks-tiers might be too strong. And I also don't like field-training. Both waters down the effort it takes to level a unit by killing stuff.

However, I'm not so fond of major balance-changes. If they are necessary to fix an exploit, then yes but if they only serve to better fit personal taste then I'm somewhat reluctant of changing ways that everyone got used to in the bast.

[Hans Lemurson]

Yeah, it's generally best to stay conservative with your changes unless there is something seriously broken you're trying to fix.

Yeah, I have mixed feelings about field-training. In one sense, it's nice to be able to invest in your troops, but in another it means that rank-10 becomes the new "average" for everybody, and it's a bit of a hassle to crowd all of your low-ranked units together into one city to train them. It ends up being mechanically similar to having only one barracks in your empire and training everybody there.

Upgrades themselves paradoxically seem really expensive, but are right on the edge so that reducing their cost any further makes them become an exploit.

[Zak0r]

This question first:
Would it be possible to let a unit rest for more than one turn after an upgrade?

If yes then there could be a totally different approach to the issue. In short: The more credits you save by upgrading the more you pay with time.

So if you want to save money by upgrading in steps you will have to wait a lot longer thus rendering your unit useless for that time. I could elaborate on this but I want to have my first question answered first. :D

A second possibility would be that an upgraded unit loses a bit of its health and thus the upgrader loses some time this way or has to fight with a weaker unit. This could be the fix if the waiting more than 1 turn doesn't work.

A third possibility would be losing experience for upgrades. If you lose say 1 experience level per upgrade then you might rethink what's best in your situation. Saving some credits to get up-to-date units but then having to retrain them or paying a higher price but keeping the experience - thus having a stronger unit.

[Ail]

The second possibility would be the by far easiest to implement.
However, I already implemented something and it should be tested first before neglecting it outright.

[Zak0r]

Sure, we can test it in our next MP game.

[Hans Lemurson]

Originally posted by Zak0r:

This question first:
Would it be possible to let a unit rest for more than one turn after an upgrade?

If yes then there could be a totally different approach to the issue. In short: The more credits you save by upgrading the more you pay with time.

So if you want to save money by upgrading in steps you will have to wait a lot longer thus rendering your unit useless for that time. I could elaborate on this but I want to have my first question answered first. :D

A second possibility would be that an upgraded unit loses a bit of its health and thus the upgrader loses some time this way or has to fight with a weaker unit. This could be the fix if the waiting more than 1 turn doesn't work.

A third possibility would be losing experience for upgrades. If you lose say 1 experience level per upgrade then you might rethink what's best in your situation. Saving some credits to get up-to-date units but then having to retrain them or paying a higher price but keeping the experience - thus having a stronger unit.

Saving money by upgrading in steps is just an exploit (cheaper to upgrade an empty hull than rush-buy a fully formed unit), and I don't think that adding hassle to the upgrade process is going to help. Turning it into a pain-in-the-ass isn't going to change the amount of money you would save, but it would make upgrading normally even less appealing.

I think what I started the thread for was to consider the question of: "Now that we are no longer accidentally saving money on upgrades, they are even more expensive than they were before. How do we make sure that upgrades are worthwhile and not a waste of credits without opening up exploits?"

[Ail]

To be honest I think they never were intended to be cheaper than rush-buying. As long as you upgrade anything except the hull it still is exactly as it was before-cost-wise.

Upgrading still is cheaper than rushbuying but you have to consider what is better: Paying slightly more to keep the old unit and get a new one or getting your current unit to up-to-date-tech. Except for Togra, where it still should be a no-brainer to upgrade everything.

I think it simply was an oversight that you could get it so cheap.

What I now wonder is what actually is better in which case. AI currently tries to upgrade everything if they have leftover-credits. But depending on the level of the unit in question to upgrade compared to what you get for a new one it might or might not be wasteful. It's probably better to not pay 2600 credits to get a white level 3 unit to a current design when you could buy a new level 15 unit for 3000. It is a bad military-power per credit-ratio.

Hmm... I could calculate the military-power-per-credit-ratio when considering an upgrade and compare it to the constant ratio of building new units and disallow it when it is worse.

[Zak0r]

Well if we fix the difference so you can't get cheaper upgrades that would be the best solution imo. Upgrades will still be useful even if they're more expensive than now. I honestly never calculated the price when I decided to upgrade. That's probably why we needed Hans to find that problem. :D I personally upgrade when:

1. A city is in imminent danger and an upgrade will make it safe.
2. I have a high exp unit and it will be the champion of my army that gets to clean the enemy cities and keep the enemy from attacking my forces because he is too strong. In the early game this will be around lvl 7+ and later rise to 15+ when all those exp buildings are built and researched. I really love those early lvl 20 units you can get from the water aliens. :)
3. I got a new unit tech and I'm now starting my timed push. Rushbuying units would be too slow because I specialise my cities and often have only barracks in my production city. So upgrading gives me like 10 of those new units instantly while I would have to wait 10 turns when I would want to rushbuy the same amount.

So you see the price is never really considered, it is simply paid because it's needed.

[Hans Lemurson]

Upgrading still ends up as an expensive rush-buy, though. Preserving the experience of veterans is really the only good reason I can think of upgrading versus rush-buying a fresh unit. And once the golden-barracks comes out, then even that becomes worthless since your fresh units will have more experience than your Field-trained regular army and all but a few of your veterans.

I think the best way to make upgrading worthwhile is to weaken the Barracks line of buildings. EXP is too cheap.

[Ail]

Originally posted by Zak0r:

So you see the price is never really considered, it is simply paid because it's needed.

Well, there was no need to consider the price as it was extremely cheap anyways.

I am 100% sure that it was not intended to work as it did. It must have been an oversight. We all used it not knowingly to get units much cheaper. But we didn't realize it and thus didn't exploit it as hard as possible.

I was used to getting all my white units blue within a few-turns after I got the blue design. That cost like 384 or so per unit. 384 is the equivalent to 48 production. That's a total steal considering how much power per credit you get!

The cost to upgrading a white tank to a golden one was, as long as you didn't change anything else, 1152 credits or 144 in production.
Ridiculous when you consider a new one cost like 5 times that much.

Now it is completely different and you can only do that if you are filthy rich. There won't be a massive powerspike just for getting the new designs anymore. It imho feels much better.

The advantage of the theorethical possibility of upgrading all units at once remains but it is theoretical and not practicable. Armies where new and old troops are mixed will be much more common.

For the AI I plan to calculate a power per credit-value in order to compare whether upgrading or rushbuying units is more cost-efficient. This will come down to the level of the upgraded unit vs. the level that the units would get when built with the current barracks-tech level.

@Hans: I'm not sure that upgrading must be worhtwhile afterall. We just all got used to it as a must-do because with the old prices not doing it simply was stupid. It was not really a decision you had to think about.
From now on it will be.

[Zak0r]

Of course the strategies will have to be adapted when the price is higher. But I wanted to say that there will always be a use for upgrades even if the cost is the same as rushbuying.