Retreat option in battle

I'm a blit split in this. I can't put my finger on what 4x it was, but it had retreats. And all it led to was endless frustration, because you could never kill anyone. Well, you could, but it meant chasing them down across vast areas, fighting again and again, as they just kept running (or get a lucky strike, chompin' them down before they had a chance to retreat). It basicly allowed masses of fundamentally worthless units to tie up arguably better units and eventually swamp you in a sea of turds.

While I'd love to see a discussion on this subject, my original stance will be a no.
I much prefer people having to think before engaging (or defending), instead of constantly "trying the waters" and winning through a tenacious trial-and-error technique.

I'm also a bit split on this issue. In Total War, retreat was a mixed blessing for me. On the one hand it was nice to be able to pull out before all is lost, hopefully to get my troops to a more defensible location or to join with another of my armies. It did create options, which was nice (retreat was not always the best option if things start looking grim), and it allowed you to save a portion of your army, which was nice. In fact, every time *I* retreated I was happy for the feature. But on the other hand, whenever the computer retreated it drove me nuts. Usually I'd just engage it again immediately until I managed to wipe it out. It was almost always a case of delaying the inevitable.

One big plus for it, though, is that on occasion it did do away with a classic flaw in games like HoMM and the like. In those games, the whoole affair almost always boils down to one big battle, where the fate of the game is decided. When the two major armies collide, one loses everything, and usually leaves the victor with enough left to keep pushing. This was prevented in two ways in total war: limited army size, preventing the bulk of a huge military to be in one spot, and to a much lesser extent the ability to retreat. If Elemental ends up like HoMM where the majority of your military will be in one army, then retreat would be a very good thing. It kinda sucks when the first battle is the only important one... On the other hand, if Elemental forces you to field multiple armies at once either by game mechanics or sheer necessity, then I'd rather there be no retreat.

It would be nice to have a way to save special units like heroes from doomed battles, though. Maybe you could retreat heroes and such, but at a cost to the rest of the army's morale? Maybe you have to decide whether or not to retreat somewhat early in the battle?

I'm with Martok on this one, I think that the ability to retreat is very important. 

The frustration factor is avoided (or at least reduced)  as long as army speeds are determined in a reasonable fashion based on what units you have in your army.  So for example, light cavalry should always be able to retreat from heavy infantry, while conversely that heavy infantry has no chance of getting away from the light cavalry.  This way, if you're having trouble resolving battles and your enemy is consistently running away from you, you need to rethink the types of units that you're bringing along for the campaign.

The ability to retreat adds another level of strategy which I always miss in the games that resolve battles as all or nothing.  If you invade my kingdom with your army of elite heavy infantry, I want to be able to harass that army (and yes, frustrate you as much as possible) with hit-and-run attacks by my cavalry, until you're all nice and worn down, ready for me to crush in a decisive battle.

I am torn on this because I can see both arguments. Retreat would need to be done carefully and not used as a "get out of battle free" card. There should be some kind of penalty and loss for doing it so that it could be a tactic, but still carries a cost making someone think about the use of it.

I've never really cared one way or another about the retreat option.  To me it's just another tool or option.  If it exists, then you can use it, and knowing it exists means you have to work a bit harder to pin down an enemy force.  If it doesn't exist in the game, then you play a bit differently, either force the "one battle for final victory" or go out of your way to keep your army split up and mobile, so that the enemy cannot force the "final battle" on you.

As long as it is implemented decently, I'll make use of it.

Some very good points and arguments here, including a couple of which (admittedly) hadn't really occurred to me before. 

pigeonpigeon probably best highlights why I feel retreat to be so important to strategy games:  As a general rule, I strongly dislike "all-or-nothing" battles where the fate of two entire empires are riding on the outcome.  Yes, I know it's happened often enough in real life, but it's usually not very fun from a gameplay standpoint.  Being able to withdraw your forces (or at least most of them) would go a long way towards insuring said make-or-break battles don't happen very often.

I completely agree.  Retreating/withdrawing your army from combat should definitely carry some kind of cost or penalty for doing so; I would never suggest otherwise. 

A couple examples:  In games like the Total War series, the losing general often loses a command star (which affects how well their troops fight in battle), and/or any soldiers under his command suffer a permanent morale penalty.  In Birth of the Federation, a lost battle (whether from your forces being destroyed or simply retreating) causes a morale drop on all of your planets (which is important in BOTF since morale affects how quickly you produce and research stuff). 

Now I'm not sure how practical it would be to apply either of these systems to Elemental.  I highly doubt heroes will have anything equivalent to "command stars" like in Total War, and we don't really have any idea yet as to whether what effect cities will suffer if their morale/happiness drops (assuming cities in EWOM even have morale/loyalty/happiness ratings). 

What I'm thinking is that armies that retreat would suffer an XP reduction, compared to what it would've been had they won the battle.  For example, an army that decides to withdraw from combat only gets 75% of the XP it would earn had it won instead. 

Obviously, any such XP penalty would have to be carefully balanced:  Too steep a penalty, and there won't be enough reason for armies to justify a withdrawl.  Too small a penalty, and armies could simply use it as a "get out of jail free" card as ABruner outlined.  I do think it could work, though. 

EDIT:  Well that's annoying.  Don't know why my text above ABruner's quote appears in blue, but I can't seem to make it go away. 

I agree that great lengths should be made to avoid the all-or-nothing battles ala  HoMM. Those battles are always epic and fun, and are usually the only ones that pose any sort of challenge at all in the late game, but I personally usually get bored afterwards and just start a new game instead of finishing.

I think to even make any good suggestions we'd need to know more about how the combat system is going to work. If it's turn-based in any way, then maybe retreat could take a certain number of turns based on their speed (faster units can retreat faster). In addition, no units can retreat from battle if they are adjacent to enemy units (this also prevents slower units from running away unless the opponent lets them, or the defender guards them). The penalties could scale with how long it takes your units to retreat. That way if you're harassing an opponent with an army of light cavalry, you could send them in for a quick strike, and then run away again while suffering only a small penalty. But if you're engaged in a huge battle and decide half-way through that you are losing, you probably won't be able to retreat all your units (sometimes maybe not very many at all, even), especially not your slow ones. And the ones that do get away will suffer a large morale and/or experience penalty. (By experience penalty I mean they get less XP than they would've if they'd won - they would never lose XP they already have).

In MOM you could always flee but depending on the speed of your troops you could loose some or all of them. Moreover being able to flee with a fast troop gave you the possibility to cast a nice fat spell and then excape before your ennemy could do anything to you. I remember being saved a few times when my armies were weak but my spells were powerful. I destroyed a few heroes that way.

I think being able to flee is very important in this type of game. It should be balance as well.

The problem with turn-based games is that it's hard to accurately represent concurrent events.  In games like HoMM, where the movement of the hero represents the movement of the hero and his army, it might be better to force an all-or-nothing fight on an opponent because an engagement represents a situation in which neither army can easily retreat.  When a unit engages second unit which has no movement points available, this becomes obvious.  That unit has used up it's ability to move, so being able to run from the battle makes no sense. 

I think this problem can be solved more easily using the overland map than anything within the battles themselves. Ensuring that units can avoid engagements before they happen is the best guarantee.  On top of that, allowing units to use abilities while in the overland map might give some units the ability to act as hit and run type units without every having to engage.  Then avoiding engagement would be the whole goal of the unit. Allowing units to move after engaging or using an ability would also allow units to use those tactics.

That said, losing your whole army simply because you can't even tell your units to run like hell doesn't seem like a particularly appealing feature.

If you loose your entire army, then in regards to 'picking your fights', you have chosen poorly.

I think echelon retreating should be an option and that the wiz should have a group retreat spell that costs a lot to use.

And why not a trait for units that allow them to double speed with a "retreat" option? Or allow them to move first, even if they are the attacker (or anything that let them attack then escape)

How about this: Something can only retreat if it moves faster than the opposing force on the strategic map. So:

If retreat is handled as "the whole army has to retreat", the option is only avaliable to whichever side has greater strategic speed (neither side, in case of a tie.)

If individual units can retreat, the ones which could outrun the enemy force on the map can do it.Everyone else has to stay. So say a human army has (these are arbitrary strategic speed figures, no idea if the game will even work like this) 5 speed infantry, 8 speed archers, 10 speed heavy cavalry, while a fallen force has 6 speed infantry and mages, 14 speed light cavalry. A unit can (probably) only move as fast as its slowest members, so the human force has 5 strategic speed, the fallen 6.

All fallen units have > 5 speed, they can all retreat.

Human infantry has < 6 speed, they're stuck. The rest of the human force can retreat.

If it's handled like this, it follows that forces can be split up, meaning the fallen would later have the option of breaking the cavalry off and having it run the human stragglers down.

Pigeons are winged rats with learning disabilities. pigeonpigeon is smart.

Silly talk aside, I thought it worth mentioning that this thread caught my attention because I've been thinking about stuff from the More than one Day battle thread. Over there, I remembered how much I'd liked being able to retreat in some games. pigeonpigeon's point just drove it home for me. I'm now really, really interested in seeing some details from a dev about how turn processing might work, especially how the 'continuous' stuff might relate to a 'full' turn (please forgive my RTS-hating ignorance).

All I care about is that I can flee if I am attacked.  I understand the idea of an ambush and the fact that the attackers would be trying to cut off retreats, but seriously.  Like if my scout riders is walkin' down the road exploring, and an army of death knights on bears attacks them, they should be able to turn tail and run away.  Black Bears can get up to 30mph apparently (and death bears might be worse) but at least a few might be able to escape provided its not a pincer attack.

It wasn't a problem in MoM or really many other games (you had some draw backs like having to re-higher the hero in HoMM or as meantioned, some slow troops might die as you run away),  But it was a problem in AoW.  In AoW, just because I was attacked meant my single dude couldn't run away.  Like, I could have more guys them them and a whole army, but my 1 weakened hero who only wants to run away has to stand there while the enemy wizard throws giant fists at him makes me very angry.  he should be able to flee.

On the other hand, I like AoW in letting my army flee on a per-unit basis.   so my weakened dudes could take off, but my tough tanks could sit there and soak up the rest of the heat.

I putting my vote in for the retreat if faster than other army option. 

I'm putting my vote for unit-by-unit, situational retreat if possible (if the battle system allow for it). An adapted version of Medieval II: Total War's system.

If my army consists of speed 5 infantry, speed 10 heavy cavalry and speed 20 griffon riders, and my opponent's army consists purely of speed 15 light cavalry, my griffon riders should be able to retreat easily, and the rest should have a difficult time of it. But, my infantry should be able to try to occupy the enemy long enough for my heavy cavalry to get away, too. Retreating should consist of more than just pushing a button - it should be integrated into the battle system so that your choices and orders determine the outcome of a retreat.

I like the idea that when you retreat, there is a chance that you are going to take some losses as you exit the field of battle. Sometimes you might get away with no losses. Sometimes you might get reamed. Maybe it could be a factor of the average speed of your opponent vs the average speed of your units. If you have much faster units, it could be a near certainty that you could run away with minimal or no losses. But if you are much slower then if you flee the odds that you are going to take serious hurt retreating should be pretty high.

I like doing a unit by unit retreat also.

And why not making retreat depend on the cavalry? They were mostly used to pursue thoses who retreated. Cavalry shouldn't be the "best" unit like it's often the case. Cavalary has some pros like being fast, being able to flank (and that's why we need flank/rear bonuses to attack and maluses to test of morale).

If the enemy army retreat and you don't have any cavalry then they can retreat with only minor loses (like some units losing health. It should'nt be leave or die. Is a unit is fleing they should make their next battle with malus of movement and/or health and/or morale)

First, movement based retreat capability.  It's handled this way in the more recent Total War games.  Your army retreats and expends it's movement.  I wouldn't let anyone retreat if they have no movement, and I wouldn't let them retreat more than once a turn either.  If the other guy has the movement to catch you, consider yourself run down.

Second, either percentile decimation of forces based on movement speeds or start people out in the middle of a larger map and make retreating a matter of walking off the edge.  Then your cavalry get away while the enemy cavalry assrape your infantry before they make it across.  If you go percentile, I wouldn't make it large unless upkeep is the primary factor in how large an army gets.

Third, the option to surrender.  Pay a fee, a large one, and they let you walk away.  Either set based on army value or determinable by the enemy human player.

Should probably give the channeler some other out method with it setup like this, a recall that takes them to the capital preferably.

If your infantry is no match for the enemy cavalry, why should that tactic do you any good? The cavalry would logically just beat the dog shit out of the infantry in front of them then run down the slower heavy cavalry.

That's only assuming someone would retreat if they're losing horribly. Who's to say the cavalry will beat the dog shit out of the infantry?It might take them a significant amount of time, enough to allow my heavy cavalry to put enough distance between them and the light cavalry that they can get away. I think that if there is a retreat system it should work more or less like in Medieval II: Total War. Just because one army has some really fast units doesn't mean they should automatically be able to run down every single slower unit of the opposition, should the opposition choose to retreat.

If a player decides to retreat, their decisions as to how to retreat (which units, when, what direction, etc) and the pursuer's decisions as to how to pursue (who to chase after whom, trajectory, etc) should determine the success or failure of the retreat. Obviously having much faster units than the opposition gives you an advantage whether you're retreating or pursuing, but it shouldn't be the only relevant factor.

Because, while that could make sense in a world in which armies move at, roughly, three speeds (Human Infantry, Horse-Based Cavalry, and Horse-Pulled Artillery), that rationale begins to break down when you add pegasi, dragons, wereleopards, quicklings, blinkdogs, etc.

In short, "You have a guy on a horse, so the Flash can't escape you" doesn't make much sense.

EDIT: Oh - forgot to add, I'm totally in favor of the TW method of retreating.

A combination of points usage might be the best. Every force would have so many movement points to use per turn. Once in range of the enemy force, movement points could be converted to attack points as well. Sorta of like the DnD idea of Move-Hit-Hit-Hit-End Turn vs the Move-Hit-Hit-Withdraw-End Turn.

That way, I can probe the enemy for weak spots but save movement points for a possible withdraw requirement. Using the "Move-Hit-Hit-Withdraw-End Turn" idea, or, if it is rather evident that I can win the day, Move all the way in and start wailing away and take your lumps on the enemies turn using the "Move-Hit-Hit-Hit-End Turn" idea.

The idea of moving to far by accident would/should not be a problem in a Turn Based environment as Allowed Move Distances (AMD's) would be shown graphically on the battle map for each formation before ever giving the Move command.