If autoresolve plays through the battle completely, which is what it sounds like it's doing, than having it be as good as possible is an important test for the tactical AI.  If the AI is to be a challenge then the auto-resolve needs to be as good as possible.

That's irrelevant. In multiplayer, such an option would just need to be disabled. The ability to replay a battle that you auto-resolved is not a bad idea for single-player, IMO.

I agree with the apparent consensus that auto-resolve should be very good at winning with minimal casualties vs. significantly weaker opponents. Personally, I enjoy playing out big and/or tough battles myself. Autoresolve, on the other hand, is a wonderful tool for quickly getting through the small skirmishes where the outcome is a forgone conclusion.

The most comprehensive auto-resolve would be to have the AI control both sides, but run through it quickly in the background instead of showing it to the player. The downside is that it might result in a lot of waiting, especially for slower computers (which is when auto-resolve would probably be most-used).

There is no way, though, to make an autoresolve that will rival the actual player taking control. For one, everyone plays differently and there is no way of mimicking that. Secondly, players will always be better than AI so the only way to make auto-resolve give similar results would be to give the player a massive, cheating advantage in auto-resolved combat.

So yeah, IMO the auto-resolve should focus on winning with minimal losses in uneven combat, and just focus on winning whatever it takes otherwise.

You know, I don't think I've ever, EVER, seen a comp AI use auto-resolve and come anywhere near as close or better then I can do on my own playing out the battle. That would be a very impressive accomplishment if you can pull it off,  Frogster. Good luck.

Well stated.

You're looking at a very difficult proposition. Make auto-resolve too good and there is no point to doing a tactical battle. Make it too bad and everybody will feel obligated to use tactical battles.

You can't have both "be the best" and the AI can never replace the human's capabilities because every person is different and may even have different tactical abilities day to day (get some sleep, then battle!).

This is an equation. The human player is a variable, the AI is a constant. Inevitably one is going to be better than the other most of the time.

If you really want both to be viable, you'll have to implement a second variable into the equation which I don't advise. (I.e. one has a better average casualty outcome while the other has a better average loot outcome)

I agree with this post. I think this is something that has to be done, but I'm not expecting success. It's just a knife edge balance that is impossible to get exactly right. You still have to do it, just be ready for lot's of complaints.

There was one game I was so bad at that I was much better off using the auto-resolve. I think it was 'Conquest of the New World'.

First of all I think auto resolve will never replace tactical battles completely. It is just so much more fun to play the battle for youself instead of auto resolving.

On the other hand you do not want to fight each skirmish in the end game. If I think of the total war games, how often I would have like to skip small battles to focus on the large interesting ones, if just the auto resolve would have been good enough.

In Empire:total war (ETW) the auto-resolve is so good that there hardly is any point inplaying out the battles. I therefore do not play them, which makes the game very very dull. Also the campaign map AI of the total war series is very poor, even the developers do not like it. in ETW the developers said that they made the AI so accomplished that it took everything into account like a human would, which resulted in the AI being letargic and completely frozen most of the time. So in the latest patch they basically broke the AI, causing it to make rash decisions that are completely stupid and nonsensical, although this resulted in the AI at least doing something, which apparently was quite an accomplishment. So now they pretend to be proud of their AI even though they already admitted that it does not work as intended and that they are unhappy with it. Creative Assembly are just a bunch of liars like that.

And I love Stardock for their open policy, at least the people here do not keep on the 'correct' side of doing things which is refreshing after being on the ETW beta and being lied to there by the developers even as I and many with me devoted countless hours on a game as crap as ETW.

So in the end, every auto-resolve that is too good like the crappy one in ETW would be very very bad. I want to run the risk of losing when I should and winning when I should.Wether or not the AI would lose more units then I would when playing the battles is of little concern to me, as long as it is within the borders of what is reasonabe. A unit of 5 peasants killing 10 of my battlehardened veterans is pushing it, those veterans killing 100 of my peasants is not.

Edit: foul language removed.

I have been waiting to ask this very important question that many developers never think of, but im sure the old boys at StarDock already have decided. In a larger multiplay will everyone be able to watch the battles (ensuring prolonged interest for all players) or will it just be those near with a line of sight (realism). The former would make it much less likely that boredom will make others call it quits, whereas the latter would perhaps lend to the idea of players being able to do strategic planning and maybe even some limited maintenance or tech pathing and analysing during other people's or fallen's turns.

What do you guys think

The point about watching other battles seems abit iffy to me. If I am not involved, or have LOS, then I could simply gather inportant Intel on the enemy forces I would not otherwise be privy to. Not good. As to the Auto -Resolve question.

Would it be possible to segment a large force and have the AI run the Left/Right Flank portion while the Player runs the Tactical fighting segment in the critical Middle section?

I know it sounds complicated but would be a real boom for all players who like Tactical but wouldn't mind giving the AI some of the action when and if the Armies begin to get bloated.

If auto-resolve is "too good"  Tactical battles will be as useful as they are fun.  I don't see a problem with that.

If the tactical battles are fun, people will do them for fun.  If they're not fun, I'd rather have the autoresolve be good enough that I don't have to waste time on tactical battles.

One important key: make sure no unit/ability is gimped in one mode and all powerful in the other.

It's possible that for multiplayer the engine could be made smart enough to run multiple battles simultaneously.  If Player A and Player B are fighting, and Player C and Player D are fighting, why not run them at the same time?  The only catch is that results of combat--retreats, moving out of supply, etc.--need to happen simultaneously, or the engine needs to be smart enough to build a reasonable dependency tree.

It's worth noting that AoW's auto resolve isn't running through tactical combat. It uses a fast combat resolver and the results are sometimes drastically different then if you go into tactical combat, then hit the "auto" button to let the computer play it for you. (That was done because on a big map AI players do a ton of combat, and the fast combat shortcuts dramatically reduce the computational load. AoW was originally played on far slower computers then what we have now.)

Presumably, Elemental will handle auto resolve by actually running the tactical combat at super fast speed and without the graphics, so the results won't deviate so much due to strange shortcuts in fast combat.

The surrender feature in AoW:SM is horrible!   Players have reported on horrible results on the heavengames forum. One guy had 8 Crossbowmen in a town and they surrendered....Sure the attacker was more powerful but even if he had 8 Red Dragons one could have died from 8 Crossbowmen. The surrender feature was hated to no end and got removed in the next patch.

Why would i ever surrender?

I am a guerilla warfare specialist. Units don't live past their usefulness, its a fact I've learned to accept. I think having a surrender option would go unused or just muck up people's early game strategies. I would need to see some example combat to be sure.

I do.

A game like this is made up of components which interact with each other.  Part of the fun of tactical battles is influencing the rest of the game.  You get to see how the soldiers you built and the monsters you've summoned hold up, you get to try your battle spells, you get to take advantage of your general's combat bonuses, you get to deal with whatever you've done to the terrain, and so on.  When the battle is over, you benefit from the result -- your army preserved, the enemy's army diminished or destroyed, experience for your units and their leader, and whatever strategic goal the battle had (continued control of a node, capturing a city, killing a dangerous enemy hero, etc.).  The same goes for the other side -- you don't want to let the enemy get their hands on resource X because of what it will do in tactical battles, and you have to live with the tactical-battle consequences of letting them get resource Y.

When the connection between playing the tactical battles and playing the rest of the game is broken, both are diminished.  Not being able to take meaning out of manually resolving the tactical battle and apply it to the rest of the game sucks.  Not being able to derive meaning from the rest of the game based on what you'll do in tactical battles sucks.  The auto-resolve absolutely should not restrict the usefulness of manual tactical battles to "fun."  The real fun needs genuine usefulness to happen.

so you are willing to play out every single battle of your 10,000 Doomy Knights of Doom against the 20 Peasants of Frailty during the 150 turn mop-up phase even though you know you'll win?

No, as my earlier posts should indicate.  I think that autoresolve is great for one-sided battles but should not replace fighting difficult engagements manually.  I suggested a "sanity check" autoresolve which can provide an instant, overwhelming victory to an obviously superior force*, but objected to an autoresolve which yields equal or superior results to manual control in combats with serious threats.

*according to either a simple metric like total force cost or a more sophisticated one like threat points based on how the AI calculates its reactions to stacks

I like getting to see the details but with that level of detail wouldn't it get completely overwhelming in a 1000 vs 1000 man battle?

I had a proposal for auto-resolve combat I've been thinking about but never got around to posting.  In the past I played PBM games where battle resolution was handled by the computer.  Instead of just telling you who won or how many points of damage were done the battle was broken down into a small story which gave details of the battle in an interesting way.  I think this would be a great addition to a game like this, especially since I keep seeing how you guys are empasizing how tactical battles won't be necessary.  It's been too long since I played these and I'm no story teller myself but I will try and dig up an old battle report to give an idea of what I'm talking about.  Reading these battle reports was one of the highlights of each turn.  You can still give the gory details in another tab but with a nice battle report you could convey most of the important details in a more consise and entertaining manner.  Yeah it would take some work but I think it would add some flavor to auto-resolve and make it a worthwhile option.

Here is a sample battle report.  Not the most exciting battle but it gives a better idea of what I'm talking about.  I'm sure Brad and company could do a lot better though.

BATTLE BETWEEN THE 3RD DWARVEN ARMY GROUP AND THE 4TH GIANT FULL ARMY IN  
 THE PLAINS OF AREA IJ:                                                    
                                                                           
 DWARVEN                                               GIANT               
 GENERAL II THARIS                      COMMANDER           WARLORD      
 KEVINMOR                                                                  
 POWER-1 WIZARD KISVOR                RANKING WIZARD        POWER-2 WIZARD
 NERAFANA                                                                  
 20952   LIGHT/MEDIUM/HEAVY/GUARDS         INFANTRY             11907     
 GUARDS/MEDIUM/HEAVY                                                       
 485     MEDIUM                          CAVALRY              0           
 3201    EXPERT/BOWMEN                   ARCHERS              729         
 EXPERT                                                                    
 7                                KINGDOM-NAMED BRIGADES     3           
 0/5                             RECRUIT/VETERAN BRIGADES    0/0         
                                   OTHER BRIGADE TYPES       6 OGRES     
 GOOD LEADERS                              OTHER            HIGH ATTRITION /
 HIGH MORALE / EXCEPTIONAL HEROES                                          
 DETERMINED ATTACK                  TACTICAL SELECTION          DETERMINED
 ATTACK                                                                    
                                                                           
                                                                           
 AT THE FIRST LIGHT OF DAWN GENERAL II THARIS COMMANDED THE 3RD DWARVEN ARMY
 GROUP TO BEGIN THE ATTACK UPON THE ADVANCING LINES OF THE 4TH GIANT FULL  
 ARMY LED BY WARLORD KEVINMOR.  AMONG THE GIANT TROOPS WERE BRIGADES OF    
 OGRES, WHICH NOW LEAPED AHEAD TO THE FIRST RANKS!  AS THE ATTACKERS       
 ADVANCED THROUGH THE HIGH PLAINS THE BATTLE BEGAN. BOTH THE DWARVEN       
 DISCIPLINED CROSS-BOWMEN AND THE BOULDER HURLING TROOPS OF THE GIANT FORCE
 IMPOSED MINOR LOSSES TO THEIR OPPONENTS LINES.  THE DWARVEN KISVOR (POWER-
 1.0) TRIED TO AVOID AS BEST ABLE THE ENEMY'S ATTACKS.  THE DWARVEN GUEGRO 
 (ADEPT) SUMMONED NO NOTICEABLE MAGIC DURING THE FRAY.  THE GIANT POWER-2.0
 WIZARD NERAFANA GLOWERED INTENTLY AND ISSUED INDECIPHERABLE UTTERANCES.  HE
 THEN CREATED A POWERFUL WARDING, WHICH WOULD PROTECT THE HEROES AND WIZARDS
 OF HIS FORCE FROM THE BLACK MAGICS OF ANY OPPOSING SPELLCASTERS.  THE GIANT
 HERATON (ADEPT) CAST NO SPELLS TO AID HIS FORCE IN THIS BATTLE.  WITH A   
 LUSTFUL CRY DWARVEN GENERAL II THARIS CARRYING THE AXE OF FARIN PLUNGED   
 INTO BATTLE THEREBY INCREASING THE DETERMINATION AND ENDURANCE OF HIS GROUP
 AND AIDING HIS CHANCE OF ENCOUNTERING THE GIANT LEADERS   QUICKLY GENERAL 
 II THARIS BECKONED THE DWARVEN CAVALRY DOWN UPON THE GIANT POSITION.  THE 
 FEW CAVALRY DID NOT APPRECIABLY EFFECT THE OUTCOME OF THE BATTLE.  NOW THE
 INFANTRY OF EACH SIDE WAS COMMITTED TO THE BATTLE, AND THE FIGHTING WAS   
 HARD AND CHAOTIC.  INFANTRY FROM EACH SIDE INFLICTED APPRECIABLE DAMAGE.  
 HOUR AFTER HOUR PASSED, YET NONE COULD TELL WHICH SIDE WOULD FIRST BREAK  
 AND WHICH WOULD TRIUMPH.  DESPITE ALL THAT HAD PRECEEDED THIS MOMENT, THE 
 BATTLE OUTCOME WAS STILL IN DOUBT.  NOW ALL REMAINING TROOPS OF ALL TYPES 
 HURLED THEMSELVES INTO THE THICK OF BATTLE TO SETTLE IT IN BLOOD RED      
 FASHION.  DWARVEN LOSSES WERE SAID TO BE 17900 TROOPS.  DWARVEN LEADERS   
 BATTLED COURAGEOUSLY.  ALL DWARVEN LEADERS SURVIVED THE BATTLE.  THE      
 DWARVEN WIZARDS KISVOR AND GUEGRO SURVIVED THE FRAY.  THE DWARVEN GROUPS  
 MORALE WENT UP EVEN THOUGH THEY RETREATED DUE TO THEIR FORTITUDE.  GIANT  
 LOSSES WERE PUT AT 6900 TROOPS.  GIANT LEADERS BATTLED COURAGEOUSLY.  ALL 
 GIANT LEADERS SURVIVED THE BATTLE.  THE GIANT WIZARD HERATON SURVIVED THE 
 FRAY.  THE GIANT WIZARD NERAFANA WAS SLAIN IN COMBAT.  THE CONVINCING     
 COMPLETENESS OF THE GIANT FORCE'S VICTORY HAS GREATLY INCREASED MORALE.  
                                                                           

I could go for that.

That's a Great Idea. I'd like to see them do that as well.

"Your feirce Amazon Warriors got stuck on a rock formation."

"Oh no! An enemy has attacked your Amazon Warriors burtally!"

"Oh no! An enemy has attacked your Amazon Warriors burtally!"

"Oh no! An enemy has attacked your Amazon Warriors burtally!"

...

"Your Amazon Warriors, alas, have died in a storm of glory but managed to kill [0] enemy forces!"

For a summary that's not bad minus the fact you spelled "Brutally" wrong 3 times in a row.

It's better then...

You Won!!! You have taken (0 of 100) Casualties.

or

You Lost!!! You have taken (100 of 100) Casualties.

.... I'd take your example over that any day.

Where's the "I WIN" button?

Woe is me!