What if combat wasn't RNG dependent?

[vincentz]

2 good suggestions whether its deterministic or RNG.

Both free healing and promotion healing worsens the snowballing, and while a "unit need to be in city and use its hammers to heal" would be optimal, the AI (as it is) would never be able to understand it.

 

[TheMeInTeam]

I'm not certain. Time is a cost unto itself. If you yank promo healing and nerf medic, a damaged unit is out of the war a long time. Even just that would make a lot of early wars not workable.

I wouldn't mind seeing a combination of "pay to heal" and "otherwise heal relatively slowly". If we nerf down healing TOO much, offensive warfare gets to the point of being impractical. I do think the game needs to nerf offensive war (and then some AI bonuses) in the general sense given how some victory conditions are consistently more viable than others by a wide margin, but there is danger of going overboard.

Although the AI would struggle to handle wonky heal mechanics, it almost certainly would flounder horribly with super-slow healing regardless. It's not even particularly good with healing in the base game.

Promo healing absolutely needs to go though IMO. It isn't a good mechanic anyway because of how drastically it alters offense/defense healing (IE it removes the difference to a fairly significant extent).

 

[vincentz]

I'm not sure either, but since the AI is horrible at healing (it will pull wounded ships into harbor, but with land units it sucks), it would be a feature (the extra healing cost) that would somewhat hurt human players most.
For me, any feature even for both humans and AI, that hurts humans more, is a good feature. (to help balance the poor AI, but without ridiculous AI bonuses).
Another alternative is to take 1 food from closest owned city * (full cost - (full cost * % damaged))
Example : a 50 unit is damaged 50%. to heal it needs 1 * (50 - (50*0.50)) = 25 from closest owned city.
a 50 unit is damaged 75%. to heal it needs 1 * (50 - (50*0.75)) = 37.5 from closest owned city.
Maybe even add "supply lines" to it, and multiply with distance to closest city, so far away offensive wars can be a costly affair.

 

[TheMeInTeam]

I'm not sure either, but since the AI is horrible at healing (it will pull wounded ships into harbor, but with land units it sucks), it would be a feature (the extra healing cost) that would somewhat hurt human players most.

I'm not sure about that. AI often will hold damaged units in cities after taking them, or retreat damaged units into cities and camp there. The only units that stay in the field after damage from what I can see are those on an offensive city attack type of script.

AI movement and usage of supermedic are much worse than its ability to place units for healing.

I like the idea of using resources to heal faster to some extent, but I truly do believe it's the AI and not the human getting the shaft there, not to mention that this is literally creating a new mechanic outright; something beyond my own current ability to do.

 

[Olson]

Hi all.

I've been thinking a bit lately about how the RNG luck-****s us in the early game sometimes, and sometimes offers great rewards. Many have considered it a necessary evil, but I'm not convinced it's truly necessary.

What if combat outcomes between two relative strength ratios were *fixed*? As in, an axe attacking an archer (neither with promotions) on a flatland *always* had the axe win, but also *always* with the same amount of damage taken? As far as I can think, this wouldn't have a negative impact on the gameplay at all, aside from not relying on lucky results or being screwed by unlucky ones . IMO it sacrifices some realism for a MASSIVE boost to gameplay viability, but that's a preference issue.

I'm not looking for preferences here. I am looking for legit reasons this would be good, bad, or indifferent to the gameplay design and balance itself. I'm imagining that once collateral is a serious element that the functional difference will be minimal, though I don't have any evidence to back up that idea just yet. The big impact therefore would be planning around the early game; IE against barb archers (always screws you unless you fortify on the right terrain) and versus and against the rush.

You could argue that the AI wouldn't handle it well, but that would be a silly argument because the AI doesn't handle any combat well ever. Thoughts? Amendments? Is there a reason such a fundamentally frustrating aspect of civ is NECESSARY?

Also, the code in civ IV blows chunks (every unit has to check if it's an air unit or some such crap constantly? Really?), but that's an issue for another time.

There are games where combat isn't resolved with die (dice), in this case an RNG, but they also aren't live/die outcomes and the modifiers were numerous. I think it's a fine way of resolving combat because even though each player would know the modifiers the randomness was contained within the PRS (paper-rock-scissors) decision each player had to make as far as which unit within a stack would he send into battle first? Since no unit could fight more than once per turn until all units had fought sometimes it was best to use one's worst defender vs. the enemy's best attacker. But, would the attacker be using his best attacker first, etc.? When it comes to multiplayer (MP) I think this is the best way to resolve combat. But can the AI simulate such tactical considerations?

 

[Kaitzilla]

RNG is all fun and games until your 37xp Combat II, Cover, City Raider III Quechua goes up against an archer and dies without scratching it trying to take a vital city.




I don't mind the death. It's not like the battle was 99.9%

I just wanted the archer to take some damage heh

 

[bestsss]

If you wish non random element you still wish something like a hidden variable. That is: courage, readiness (perhaps estimated pay level). The results are deterministic just not complete information(full-knowledge) game like chess is.
Personally, I have great dislike for any use of randomness in game mechanics (esp. not card deck drawing random, i.e. you are guaranteed to get queen of spades after 52draws), usually it's a result of lazy design and even lazier programming.

 

[TheMeInTeam]

If you wish non random element you still wish something like a hidden variable. That is: courage, readiness (perhaps estimated pay level). The results are deterministic just not complete information(full-knowledge) game like chess is.

None of the direct game elements should be RNG, in theory. The "randomness" is what your opponents do. If pushed a bit, you can get hit by all kinds of things if the AI didn't suck so much (heavy espionage sabotage, hiding units, pillage party like in kmod). What wins most games would then be empire management, barring tactical blunders made by one side. I once killed another human player when I had 75% of his power, in around 10 turns, so it's not like this game doesn't have some tactical-level considerations.

 

[Fippy]

I believe it's impossible to keep long time motivation going in games against AIs, without random elements. Another option that might come to your mind would be..
51% or better always wins, but survives heavily damaged and needs some resting time.
No resting time only in very clear battles, and only those units can continue quickly.

But..take sgotm games as example, you could discuss every outcome before it's played.
This would require AIs handling different situations much better, if a lonely Knight runs around he can still be annoying as you are not sure if your (only available there) unit wins this fight, and you might need backup plans if not. If you know the exact outcome, and the healing phase afterwards would not matter, every unit the AI sends in such silly ways would 100% be wasted.

Now lets assume better AIs, you would also need to change some game mechanics drastically.
Collateral damage would be so hugely overpowered, that attacking becomes almost impossible.
What will you do, if some outdated catapults hit your stack, and then you lose *every* battle afterwards because now all your units have 40-49% odds at best?
Nothing..cultural border defense would be unbeatable.

 

[TheMeInTeam]

Nothing..cultural border defense would be unbeatable.

I wonder about that. In traditional stack warfare you'd be correct, but the idea would be to separate units enough that the trades from suicide collateral would not be favorable...or have enough #units available that it's extremely difficult to break the giant stack.

That more like exactly how stock civ IV is than you'd care to admit. Going into someone's culture borders is suicide in MP if they have siege and even quasi parity. I've seen people with 1.5 times the power lose absolutely everything. You're right that "random AI knight in borders" would be slightly less obnoxious, but the far more meaningful stack throwaways are just a simple reality of the shoddy civ IV AI. The larger the stacks, the more precisely deterministic combat would resemble RNG combat (after all, collateral damage is pretty deterministic right now).

Note that "51%" battles aren't 51% battles at all. High strength wins. There would be no odds whatsoever.

To some extent, this is what we WANT in the game, however, because otherwise it's just a unit-spamming game. Tech lead can make the trades very unfavorable to the defender, enough to possibly break the defender...or if the stalemate goes on long enough, eventually someone will get nukes, airpower, or on many maps a big naval advantage (even frigates/galleons 20 turns before the other guy can be enough if 1/2 his cities are coastal) and then it's over.

 

[aggri1]

I wonder what the effect would be for allowing any unit with higher than 90% combat odds to automatically win, but still calculate damage as per the default CivIV mechanic (if the likely-to-win unit gets to 1 point health, just kill the weaker unit). In this way, units would still get hurt, you could still have plenty of battles where there was a random element, but when you work hard to build a force which is so much stronger than your enemy, then you are rewarded with no losses.

I'm not sure whether this would have any serious implications for the Civ AIs.

We wouldn't have one of X-hundred spearmen defeating a tank, though.

 

[Izuul]

I don't like it for the same reason that i wouldn't want to just crown Alabama the national champions of the 2013 football season right now just b/c they are favored in every game.

Or what if a poker hand was determined the moment the players turned their cards over without playing out the hand? For all that people complain about "bad beats" I'm certain that most prefer the game like it is right now.

I need the occasional upset or flukey crap to liven things up and bring excitement.

 

[TheMeInTeam]

I don't like it for the same reason that i wouldn't want to just crown Alabama the national champions of the 2013 football season right now just b/c they are favored in every game.

Or what if a poker hand was determined the moment the players turned their cards over without playing out the hand? For all that people complain about "bad beats" I'm certain that most prefer the game like it is right now.

Neither of these things are examples of random chance and thus do not make much sense in rebuking deterministic combat.

I believe many football players would be upset if one were to imply they won or lost games because of luck. When I was playing, I would have been. Players I blocked or had my block defeated against...or when I was on defense vice versa certainly didn't beat me or lose to me on a given play because of luck.

Assuming large # factors that you can't perceive all at once = random outcomes isn't accurate.

 

[rah]

I believe many football players would be upset if one were to imply they won or lost games because of luck. When I was playing, I would have been. Players I blocked or had my block defeated against...or when I was on defense vice versa certainly didn't beat me or lose to me on a given play because of luck.

They would be upset but that wouldn't matter. Quite a few sporting events have been decided by luck. A gust of wind on a gaming ending field goal. A lucky bounce. Equipment malfunction. Yes, most of the game is determined by skill but it's ludicrous to say that luck is never a factor.
Use a different argument (which you have demonstrated quite a few better ones) because this one is just plain wrong.

 

[Izuul]

I fail to see how poker is any different. The moment that two players go all-in and flip over their cards, exact odds can be determined for who will win the pot. Now they could just stop right there and award the pot to the player with the higher odds, or the dealer could deal out the rest of the community cards and let random chance come into play.

The sports reference is more complicated and might be a bit of a stretch, but there are advanced statistics out there now that can predict the outcome of games with fairly surprising accuracy. I don't know if you are familiar with statistician Ken Pomeroy who analyzes college basketball, but his model is quite accurate and gives a %chance for every team to win each of it's games. Obviously it's more accurate at predicting an overall W-L record than individual game outcomes, but this also highlights part of why i don't like your approach. If team A plays 30 basketball games and has a 70% chance to win every game, what is their expected record? Is it 21-9 or is it 30-0?

 

[TheMeInTeam]

They would be upset but that wouldn't matter. Quite a few sporting events have been decided by luck. A gust of wind on a gaming ending field goal. A lucky bounce. Equipment malfunction.

Assuming large # factors that you can't perceive all at once = random outcomes isn't accurate.

I fail to see how poker is any different. The moment that two players go all-in and flip over their cards, exact odds can be determined for who will win the pot. Now they could just stop right there and award the pot to the player with the higher odds, or the dealer could deal out the rest of the community cards and let random chance come into play.

Assuming large # factors that you can't perceive all at once = random outcomes isn't accurate.

If team A plays 30 basketball games and has a 70% chance to win every game, what is their expected record? Is it 21-9 or is it 30-0?

Their average expected record is 21-9. What doesn't make sense is why you think the outcome of any one game is random...or even all of them. There seems to be a pervading assumption that, absent of luck, an inferior player can never defeat a superior one. I find that astounding, considering that evidence points to that conclusion being inaccurate. However, I see a question exactly as the one I'm quoting posed frequently...as if any slightly better team would never lose.

Prediction is not reality. The team winning 9 games won because they scored more points within the rules than the team that won 21 times. In those 9 games, they moved better and/or made better choices. Why the insistence that such was the result of random chance?

 

[rah]

Assuming large # factors that you can't perceive all at once = random outcomes isn't accurate.

That statement has nothing to do with my comment.

You said a player would be insulted if you inferred that luck might have played a role in the final outcome. Read what you wrote back.
I just replied that luck is a part of sports events and could determine the outcome of a game.

Are you saying that luck has never determined the outcome of an event? I don't believe I ever mentioned the word random. I was talking about LUCK.

 

[TheMeInTeam]

luck [ luk ] 1.good fortune: success that seems to happen by chance
2.chance: the arbitrary distribution of events or outcomes
3.event determined by chance: something that seems to happen by chance rather than as a logical consequence

Even if a player isn't aware of all factors at once, each one is still a direct consequence of something. Indeed, even weather factors apply to both teams, and cards are going to unfold in an unknown, but pre-determined way. In poker, players are knowingly entering the situation with incomplete information in hopes that they can handle it better than their opponent. The cards are not determining the ultimate winner, but rather the players guessing based on some criteria. It's a somewhat convoluted situation admittedly, but just the game for someone who likes to play with incomplete information.

Remember, these scenarios are being used to demonstrate why RNG combat outcomes are appropriate in civ IV. However lacking foresight is not the same thing as losing by chance.

 

[rah]

So I'm getting confused.
Are you saying that luck has never determined the outcome of a game, or that there is no such thing as luck.

I stopped discussing RNG combat long ago because I've settled on it's a matter of personal preference and no amount of discussion is going to change your's or my opinion on the matter.

I just got partially sucked back in when you did the athlete scenario. All your previous stuff has made sense, (whether I agreed or not) but that one was just one that went bad.

 

[Izuul]

Their average expected record is 21-9. What doesn't make sense is why you think the outcome of any one game is random...or even all of them.

Of course the individual outcomes are random. If there wasn't a random element to it then why wouldn't a player play his best game every time out on the field/court? I played basketball in various settings for nearly 15 years and i never knew when i was going to go out and light it up or stink up the gym. I see that as randomness. And FWIW i don't believe in luck (despite using the word), just unlikely outcomes.

There seems to be a pervading assumption that, absent of luck, an inferior player can never defeat a superior one. I find that astounding, considering that evidence points to that conclusion being inaccurate. However, I see a question exactly as the one I'm quoting posed frequently...as if any slightly better team would never lose.

What i don't understand is that is EXACTLY what you advocate doing with civ combat, remove any chance at all of an inferior unit winning regardless of the strength difference. How is it realistic for a unit with 10 combat str to beat a unit with 9.90 combat strength 100% of the time?