Debate: How much AI income?

[bru888]

I need to get to the bottom of this, if I can.

Gameplay balance is hard to pin down. Opinions on balance for a particular scenario can fluctuate from player to player.

But one of the most persistent comments about my projects is that it is frustrating for the enemy to be able to repair its forces.

When I was learning from Erik, I picked up the habit of awarding +1 income for each unit in play. I am still committed to that for the friendly side.

What is debatable, I believe, is whether the same standard should be applied to the enemy. I do this, and I think Erik still does as well.

Which makes our work much more difficult. Remember, this is a game. It should be fun to play for as many people as possible.

Here is an important clue: Take the scenarios from DLC into the editor and look at how much income they give AI factions. Generally, from none to a trickle.

So, I know there will be few knowledgeable opinions here but of those, I would like to know. Am I making a mistake in giving the AI parity in resource income?

[bru888]

Here is an important clue: Take the scenarios from DLC into the editor and look at how much income they give AI factions. Generally, from none to a trickle.

From New Georgia in U.S. Pacific:

Screenshot 1.jpg (175.33 KiB) Viewed 1513 times

From El Alamein in Allies Defiant:

Screenshot 2.jpg (251.74 KiB) Viewed 1513 times

[Erik2]

I'm with you and me
I use the resource income as a play balance equalizer when I test. RI may go up or down for either side.
(Same with AI unit experience by the way).
I also give the human player more income for late-war scenarios as repairs are more costly.
As a player I get frustrated when I'm not able to repair my units, at least to a large degree.
I don't like to give the AI trickle or none, I think it is 'unfair' for the AI to be denied unit repair. I like the AI to play like a human as much as possible.

[bru888]

First I will post the results of my survey, then I will comment on them.

Image0289.jpg (162.24 KiB) Viewed 1471 times

Image0290.jpg (189.59 KiB) Viewed 1471 times

Image0291.jpg (187.77 KiB) Viewed 1471 times

Image0292.jpg (173.33 KiB) Viewed 1471 times

Image0293.jpg (189.08 KiB) Viewed 1471 times

Image0294.jpg (194.8 KiB) Viewed 1471 times

Image0295.jpg (179.56 KiB) Viewed 1471 times

Image0296.jpg (161.12 KiB) Viewed 1471 times

Image0297.jpg (215.09 KiB) Viewed 1471 times

Image0298.jpg (190.55 KiB) Viewed 1471 times

Image0299.jpg (203.33 KiB) Viewed 1471 times

Image0300.jpg (226.6 KiB) Viewed 1471 times

Image0301.jpg (211.81 KiB) Viewed 1471 times

Image0302.jpg (250.78 KiB) Viewed 1471 times

Image0303.jpg (244.74 KiB) Viewed 1471 times

Image0304.jpg (199.06 KiB) Viewed 1471 times

[bru888]

As for comments, where do I begin? Let's go back to this chart:

Image0289.jpg (162.24 KiB) Viewed 1466 times

These are DLC campaigns, listed on the X-axis from earliest to most recent. The Y-axis shows the total amount of resources awarded to the enemy (not just AI, because some AI factions are allied with the human player) versus those awarded to friendly factions, in percentages.

The chart is all over the place. Everywhere from a low of 1.7% and 3.0% for U.S. Marines and Sandstorm to 44.7% and 46.9% for Red Steel and Allies Defiant. The trend seems to be increasing with recent campaigns, however.

As to why the designers (there have been several of them) awarded low amounts of enemy resources in some campaigns and more in others, who knows? Why are some scenarios within a particular campaign so variable in this regard? What was in their minds when they designed them? We will never fully know.

One thing is for sure: No one has ever even approached giving parity to the enemy factions in terms of resources, either up-front or by turns.

Yet all of these DLC campaigns are beloved (except maybe those starting with the letter "K" ) while I get comments that my projects are too hard (and too long, and too wordy ).

So here is what I am going to sleep on tonight and decide tomorrow, perhaps:

1) Stick to the +1 RP income for friendly units.
2) Unless there is a reason, no faction gets RPs up-front.
3) Enemy units to get 25% - 50% of the amount for friendly units
4) Each faction still gets +1 RP income whenever a unit is spawned.

Because it is both a strength and weakness with me that I prefer to take a formulaic approach to such matters. Less headscratching, because it can get a bit crazy trying to figure out gameplay balance.

For example, did the designer(s) of the Red Star/Steel/Storm series of campaigns plan the spike of enemy resources in Red Steel because of historical considerations, or maybe there were less enemy units in that one, or something else? If so, they have better minds than mine.

[Erik2]

Wow, that is some impressive survey...

Stick to our formula, my friend. Then adjust human/AA resource income up or down based on your play-test.
The 1-2-3-4 resource pr unit pr turn is a valid starting point regardless. We need to start somewhere...
Personally I like regular income better than a lump of resources at scenario start (except those resources needed for unit purchases).

I believe those official campaigns are beloved for other reasons than AI resource income.
I'd guess the average player doesn't give a hoot about AI income as long as their own units are not starving and they are winning

[conboy]

Awesome discussion, and thanks for the graph, Bru!

I tend to keep the AI between 15 and 25 % of the Human in the scenarios I build. This is mainly to:
1) Make one variable go away (or reduce the impact of it) so the scenario is more predictable during design
2) Eliminate the prospect of bad guys surviving and then coming back to taunt supply lines with powerful units. I hate that!
3) In scenarios where Humans attack, it's easier to make effective human attacks with limited AI RPeez, and adjust objective timelines by adding more AI foes en route, or counterattacks on the Human flanks, or more artillery to degrade and slow the Human. Even when trying to stay within tight historical bounds, this seems to work well.
4) In scenarios where the Humans are defenders, it's easier to allow the AI more RPeez and just cut back a little on the number of attacking units. But I did not take that approach in the Salerno series because there were so many attacking battalions, historically.
5) Also, in Scenarios where the AI is the Attacker, you can add more units and cut back just a little on the AI aggressiveness to calibrate the attack. In this case, with a slightly more cautious attacker, you don't need to worry so much that a highly-reinforced AI will overwhelm the Human Defender -- but in the game, the Human doesn't know what AI temper and intent are, so it provides a nice adrenalin/rage rush to see strong attackers reappear.

So, in short there are several reasons to limit AI, but the main one for me is to control a critical variable during scenario design.

wow, what a great discussion!

conboy

p.s. -- Another thing -- in scenarios where the Human is the attacker, in most cases the AI will not engage every available unit, because it is likely that many are Static or Local defense, waiting for the Human to trigger them into action. So fewer RPs are needed for the AI. But the Human will probably utilize as many units as possible during the attack, so the human will need more. I wonder what it would be like to try to attack a very aggressive AI set-up with lots of RPeez-- I think it resemble a melee more than an organized battle.

[bru888]

Conboy, I cannot tell you how timely your comments are. Many, many thanks. I was literally updating this spreadsheet when I thought to check on this thread one more time and received moral support:

Image0310.jpg (55.72 KiB) Viewed 1410 times

Yes, your approach simplifies matters. To adjust for gameplay balance, I would much rather tinker with the amount of enemy units than enemy resources.

Yes, I too have been frustrated by "bad guys surviving and then coming back to taunt supply lines with powerful units. I hate that!"

And historically in OOB DLC, a 25% ratio of total enemy resources to friendly resources is more than generous: The average of averages is 16.9% and the median value is 12.8%.

Erik, I value your opinion as always but my heart is with Conboy. Figuratively speaking, of course.

I want my work to be challenging but not a grind. Awarding resources to the enemy, point for point at parity with friendly units as I have been doing, just doesn't seem to be the way to make OOB fun as well as challenging.

The obvious retort to that is, your work is almost (political correctness ) as beloved as the game's official DLC. It also seems to be lighter and brisker while mine is a slog (all the popups don't help). I don't want to abandon my style altogether - wear your reading glasses when you call up my stuff - but if this change makes it less of a chore to play it, I am going with my heart rather than my head.

OOB should be fun to play. Period.

[bru888]

Footnote: The data shown above ignores resource point bonuses and changes to income per turn contained in scenario triggers. The amount of time that would have been necessary to hunt for and account for those would have been prohibitive and, if anything, they would tend to benefit the human player, thereby further lowering the percentages of enemy to friendly resources.

[Erik2]

Just wanted to say that I usually increase player resource income by a factor of 2 or 4, depending on how easy it is to repair units.
I don't think I've ever increased AI income. This stays put on 1 RP pr unit pr turn, to keep the variables down as conby put it.
(I tend to be much more flexible with unit exp).
Good discussion. Whatever floats your boat or balances your scenarios

[bru888]

So it sounds like you tend to approach the situation as described above, anyway. If so, it all makes sense.

[GabeKnight]

I think that conboy's answer put it best - it's the same as my opinion on the matter.

When the player's defending, it's okay to give the AI a bit more income so that its attacks do not lose steam to early - and vice versa when the player's attacking.

It depends on many other factors, too, of course. For example if the AI has many late war medium/heavy armoured units, then even 30RP/turn may be a joke. But to the contrary if the army's composed of many regular infantry or colonial units...