[Discussion] Tactical battles vs. Auto-resolve

This post is a discussion between "Tactical battles" (battles where you command your troops yourself) and "Auto-resolve" (battles where you leave it to the AI to generate results without your participation).

My thoughts
________________

Tactical battles

Pros:
-Tactical battles are fun because you are able to fight battles, and win fights using your troops. For many players, the enjoyment comes from being able to win difficult battles.
-Others might enjoy the feeling of proving themselves better than the next guy.

Cons:
-Some battles can become very repetitive. In such situations, it remove the sense of enjoyment of fighting difficult battles (maybe because after the first five, there is very little left for you to figure out).
-Repetitive battles can also feel like a chore over time. If your strategy always works (the opponent can't figure out how to defeat it), then you might start wondering if there is any point to these battles. You might ask yourselft, shouldn't I be able to hand things over to the AI and you go worry about something else?

________________

Auto-resolve

Pros:
-Solves the problem with repetitive battles by handing things over to the AI to determine who would win (and with what results).
-Saves game time so you can worry about other stuff.

Cons:
-Auto-resolved battles may get very different results than what you could get if you fought the battles yourself.
-You might quickly determine that the AI is out of touch with reality, causing you to lose fights you could easily win had you fought them yourself.
-You might decide that your computer has something against you and is cheating to cause you pain.

________________

Possible solutions:

1) If you can win 5 or more battles under the same conditions (such as terrain, and troops, etc...) using the same strategy, you should have the option to save the results. There after, every such battle you fight, you could choose to auto-resolve those battles getting the same results as the average of those 5 or so battles (maybe with a +/- 5% or 10% difference each time). This will ofcourse have no effect on any battles that the computer determines to be significantly different enough that the outcome might turn out to be significantly different.

1a) If the battles starts as you expected, but then the AI does something different than the last 5 times, should you be allowed to enter the fight where things start occuring differently? If not, would some players in multiplayer intentionally lose the first 5 battles in hopes that in the next battle, his opponent decides to auto-resolve the battle so he would fight against an AI that he knows he can beat, and not the player?

2) Create an AI to handle the battles for you. For those who have played Dominions 3, that is what I'm kind of thinking about, except you might be able to make better strategies for your AIs to use.

3) Have a tactical ability stat for your civ (like diplomacy or research). Having a high value in this stat will give you an edge in battles where you choose to auto-resolve things. You might have some variations on this stat to reflect different battle conditions like an open field, or castle siege.

3a) This stat is researchable like most other civ stats.
3b) This stat can be be boosted by fighting tactical battles every so often to show the game how good you are. Likewise, not fighting these tactical battles will cause this boost to gradually fade with time.

________________

What are your guy's thoughts?

17,610 views 41 replies
Reply #1 Top

your solutions seem very specific.  Obviously number 1 is only for the cons of the tactical battles.   I'm sure the tactical battles will be interesting, so I have little input to give.

 

for 2:  I haven't played Dominions 3, so could you explain it?

 

3.  I don't think there should be a stat for auto-resolve.  Because then nobody would use the stat, and nobody would use the auto-resolve.   Or they would maximize the auto-resolve stat and always use it.  (it would depend on how broken the auto-resolve tactical stat was)

Reply #2 Top

I think the most sensible solution is 2.

In Dominions 3, you can create battle scripts for each individual unit in an army (advance x meters, shoot when enemy gets in range, target priorities, cast X or Y spell and so on). The battles play out by themselves but you have the possibility to view them and analyze their outcome to refine your tactitcs.
It's rather obscure and difficult to master but it works out pretty well and is actually the only viable solution for a game mainly designed to be played in PbEM.

But, that would be a game in itself and I doubt somewhat that Stardock will be able to get into the level of details of Dominions.  Maybe it can be simplified and still bring a good alternative to tactical battles. After all, you will only auto-resolve battles which outcome is fairly obvious.

Reply #3 Top

Looks ok, except for the stat: it just doesn't seem like it fits within a game.

Reply #4 Top

Ok, possible solution #3 might not be the best of ideas. I was throwing it out there to see what people thought.

As for Dominions 3, you can find the demo for the game here (if you care to try it):
http://www.shrapnelgames.com/Illwinter/DOM3/DOM3_page.html

________________

It occured to me shortly after making this post that, how would you determine the victor of an auto-resolve battle for this game? We can design our own units, not only make it difficult to determine the overall effectiveness of the unit, but also their overall tactical value. What about combat spells? How would you measure the effectiveness of a well placed fireball in the middle of your enemies formation?

How about modding? Lets say you made a spell where you could not only kill a bunch of units, but also turn them into undead already equiped with the stuff they had when they were alive moments before. Do you take the value of your enemy units and add them to your military might? What about units with exotic movements like flight, jumping, teleportation, demon portals (open it in the middle of the enemy formation, and pull some random guy in to meet your demon friends), etc...?

Reply #5 Top

Having an AI that doesn't suck monkey nuts, thus doesn't lose spectacularly with 2-1 odds and then cream you in the auto resolve, is the solution to auto versus tactical.

Reply #6 Top

I’ve suggested something similar to below a while back:

On the strategic map, whenever gamer highlight an attacking stack, hover the mouse pointers over an enemy stack, the game will shows the % chance of winning.   This Win% is determined by running a few AI Auto-resolves behind the scene.  It will also show the average # of unit lost and % of HP lost for both sides.  If your computer is slow, the game uses a formula to calculate all these instead, using their relative stack strength.

If the gamer like the odds, he can choose to attack; and he has the choice of either using TC or Auto-resolve.

Reply #7 Top

Auto resolve should only be implemented on MP games if you don´t have the option to replay the battle yourself. That would take too much of precious MP time by waiting the PC to finish it´s calculations and then replay again if u r not satisfied. 

The idea of keeping record of previous combats is good, altough it would be hard to determine whether a battle is or is not the same as a previous one. 

I have no clue how the AI is goin to handle user´s made spells, abilities or even creatures. Since the dev team is goin to filter what community content is goin to be on the game, they might "teach" the AI to use these new features. What do you guys think?

Reply #8 Top

Quoting Climber, reply 6
I’ve suggested something similar to below a while back:

On the strategic map, whenever gamer highlight an attacking stack, hover the mouse pointers over an enemy stack, the game will shows the % chance of winning.   This Win% is determined by running a few AI Auto-resolves behind the scene.  It will also show the average # of unit lost and % of HP lost for both sides.  If your computer is slow, the game uses a formula to calculate all these instead, using their relative stack strength.

If the gamer like the odds, he can choose to attack; and he has the choice of either using TC or Auto-resolve.
End of Climber's quote

If I remember correctly, HoMM V used a system like this where the computer would tell you if a battle with a stack of creatures/hero/town would be easy, normal, or hard.  I think it was only based unit numbers but I'm not entirely sure about that.  I dont know if I like the idea of seeing # of units lost and % of health lost by mousing over another stack, it seems ripe for some abuse to me at least.  I would say it should be made into a skill for heroes to use, like "scouting" or something.  I also think that the victory % should only reflect the auto-resolve battle, but the scouting ability should show you the enemy army composition (numbers, stats, equipment) to help with deciding on tactical battles.

Reply #9 Top

Quoting Houlio, reply 8

Quoting Climber, reply 6I’ve suggested something similar to below a while back:

On the strategic map, whenever gamer highlight an attacking stack, hover the mouse pointers over an enemy stack, the game will shows the % chance of winning.   This Win% is determined by running a few AI Auto-resolves behind the scene.  It will also show the average # of unit lost and % of HP lost for both sides.  If your computer is slow, the game uses a formula to calculate all these instead, using their relative stack strength.

If the gamer like the odds, he can choose to attack; and he has the choice of either using TC or Auto-resolve.


If I remember correctly, HoMM V used a system like this where the computer would tell you if a battle with a stack of creatures/hero/town would be easy, normal, or hard.  I think it was only based unit numbers but I'm not entirely sure about that.  I dont know if I like the idea of seeing # of units lost and % of health lost by mousing over another stack, it seems ripe for some abuse to me at least.  I would say it should be made into a skill for heroes to use, like "scouting" or something.  I also think that the victory % should only reflect the auto-resolve battle, but the scouting ability should show you the enemy army composition (numbers, stats, equipment) to help with deciding on tactical battles.
End of Houlio's quote

 

Yes, there´s a HoMM5 add on that shows you how hard a fight will be. And there´s scouting too. IIRC,  Heroes without scouting can see only what types of monsters are wandering somewhere. Their number appear as an "?" . With basic scouting u can see a range like 1-4 and with high level u can see the exact number of monsters and even know what troops are inside an enemy building. =) Pretty good system!

Reply #10 Top

I don't think tactital stats should be this complicated. Either you fight your battles or either you don't.

 

I'm pretty sure that any or all close battles we will do ourselves. For the easy battles that we are sure to win you auto resoves. Usually they AI is good enough that you will not loose any battles that are way in your favor.

 

In Glaciv2 it's auto resolves and most battles turn out the way they are suppose to.

 

So why complicate the plumbing?

Reply #11 Top

So why complicate the plumbing?
End of quote

Because a lot of people think tactical battles are fun. It all comes down to my balance of complexity:

If a game is too complected, the players get headaches, which are not fun. If it is not complecated enough, the players get Pong, which is not fun. If you can avoid both headaches and Pong, you have a good game.
End of quote

The problem is, one player's headache is another player's Pong, so devs need to be careful: hence, my new idea:

Four "positions" for the tactical combat setting at the beginning of the game: Tactical is for the TC junkies: all battles are tactical. Auto is for the lazy/impatient people: all battles are auto-resolved. Choice is a balance between them: you can selecty tactical or auto-resolve before the beginning of each battle. Replay is something a little different: all battles are auto-resolved, but at the end you can choose to re-fight the battle tactically if you think you can do better. Obviously, this last one probably wouldn't work in multi-player.

Reply #12 Top

So why complicate the plumbing?
End of quote

I don't know about 'complicate,' but I'd just like to be able to start a battle with whatever they call manual control, do some basic moves/orders to help ensure that my advantage isn't wasted or at least that my knowledge of enemy weaknesses is well-used, and then turn over to auto-resolve when the 'hard parts' are addressed if I don't feel like running the whole thing myself. That's how MoO2 felt for me, and that's the only thing I regularly missed in GC2 combat.

Reply #13 Top

The biggest problem I have with solution 1 is what the computer decides are "similar conditions". At best, I think if you choose autoresolve, then it looks at your win/loss record and gives you a SLIGHT edge (or disadvantage) based on your win/loss. All this traccking of previous fights, determining what is a roughly equivalent fight, seems like a lot of overhead. To me autoresolve is you are turning your fate over to the AI and are relying on a superior force for victory.

I like option 2, I have played the Dominions demo and I like the simple method you can script unit behavior.

I like 3 as well, it is kind of what I was referring to by giving an edge based on your win/los to that point. I prefer 3b, I don't really like 3a much although it could work if done right.

 

Reply #14 Top

I love a good tactical battle and rarely do I use auto-resolve. I think auto-resolve however should be well done for the people who like it even though I’ll probably never use it. If one of your solutions was to be used I’d go with 1 over everything else. Although, in reference to number 2 I wouldn’t mind having an AI supported ally / supporting armies. 

Reply #15 Top

Quoting GW, reply 12

I don't know about 'complicate,' but I'd just like to be able to start a battle with whatever they call manual control, do some basic moves/orders to help ensure that my advantage isn't wasted or at least that my knowledge of enemy weaknesses is well-used, and then turn over to auto-resolve when the 'hard parts' are addressed if I don't feel like running the whole thing myself. That's how MoO2 felt for me, and that's the only thing I regularly missed in GC2 combat.
End of GW's quote

You might like how Age of Wonders SM handles it. When a fight starts, you get the option of auto or tactical. If you choose auto, stuff happens and you get a result.

If you choose tactical, you go into tactical combat and make all the decision. Once you don't want to do that anymore, there's an "auto" button you can press that turns control over to the computer again. So you can play the first couple of turns, win the hard stuff, then let the computer automate the mop-up.

I think that's the best way of handling it. Sometimes I want to control fights myself (big, important fights). If it's an Elder Dragon vs a guy with a pitchfork, I probably want to just let auto-resolve do its thing so I can get on with playing the game.

Not a fan of an option 2 style of combat, though. I want to control things, not script them before the fight.

Reply #16 Top

Looks good, but one thing I mentioned before was the ability to re-fight an auto battle in stactical mode if you lose.

Reply #17 Top

I would also like the ability to re-fight an auto battle in tactical mode - even if you win. So many times I have pressed auto-combat when I outnumber my opponent 5:1 and have better quality troops to boot - and then I get the results screen and the AI lost as many of my own troops as my opponent's. Those are actually one of the reasons why I often fight even the most trivial battles, because I don't trust the AI to be competent. With a feature to fight tactically anyways after seeing the auto-combat results, I could at least give the AI a shot and only have to take over when it really screws up.

I also really like the idea to be able to watch the auto-combat and issue orders if it looks like the AI is about to do something dumb. And also the idea to start out the battle yourself, and then let auto-combat take over.

Reply #18 Top

Four "positions" for the tactical combat setting at the beginning of the game: Tactical is for the TC junkies: all battles are tactical. Auto is for the lazy/impatient people: all battles are auto-resolved. Choice is a balance between them: you can selecty tactical or auto-resolve before the beginning of each battle. Replay is something a little different: all battles are auto-resolved, but at the end you can choose to re-fight the battle tactically if you think you can do better. Obviously, this last one probably wouldn't work in multi-player.
End of quote

Being new here, I don't know how much my opinion matters, but I am in favor of this. It seems to pretty much appease everyone, no matter how they play, especially the "choice" option.

Reply #19 Top

Actually, I did not include refight-if-you-win becuase I thought nobody would want it. With this response, it seems like it would be a good idea.

Reply #20 Top

 

I thought Age0fWonders provided a good solution by offering both the tactical option and the auto-resolve.   I recall Stardock has already commented the same will be for Elemental, but I can't find the statement.   This solution I think is best because some battles need the tactical guidance and others such as 1 Dragon verses a single unequipped Scout  don't need the tactical guidance.

Reply #21 Top

Any kind of replay or do-over option to take auto-resolve results and then do the same combat again in tactical would be a nice touch, for single player. That would need to be turned off in multiplayer I think.

Reply #22 Top

Quoting pigeonpigeon, reply 17
I would also like the ability to re-fight an auto battle in tactical mode - even if you win. So many times I have pressed auto-combat when I outnumber my opponent 5:1 and have better quality troops to boot - and then I get the results screen and the AI lost as many of my own troops as my opponent's. Those are actually one of the reasons why I often fight even the most trivial battles, because I don't trust the AI to be competent. With a feature to fight tactically anyways after seeing the auto-combat results, I could at least give the AI a shot and only have to take over when it really screws up.

I also really like the idea to be able to watch the auto-combat and issue orders if it looks like the AI is about to do something dumb. And also the idea to start out the battle yourself, and then let auto-combat take over.
End of pigeonpigeon's quote

 

I agree 101% =)

Reply #23 Top

Yeah, a lot of this stuff would not work well in multi-player, especially online: you would have to vote or whatever on which option you wanted, which would be a big headache: LAN and other systems where you could actually talk to the others face-to-face woudln't be as bad, but..... if the battle do-over option is implimented, how should that work?

My initial thought is that you get a re-match only if both fighters agree. However, I suppose some argument could be made for it happening if only one agrees............... or not.

Reply #24 Top

I just wouldn't allow it in multiplayer at all. It'll lead to abuse.

In the lobby, give the option to enable tactical combat or not. If not, all battles use auto combat. If the option to use TC is on, both parties are given the option to use TC. If either does, it goes to TC. If both choose auto resolve, they don't get a do-over.

Reply #25 Top

Quoting Tridus, reply 24
I just wouldn't allow it in multiplayer at all. It'll lead to abuse.

In the lobby, give the option to enable tactical combat or not. If not, all battles use auto combat. If the option to use TC is on, both parties are given the option to use TC. If either does, it goes to TC. If both choose auto resolve, they don't get a do-over.
End of Tridus's quote

Far and a way the best way to deal with it. Although I would also venture that there should be an option (in multi- and single-player) for all combat to be auto-resolved, no tactical combat whatsoever.