I have been thinking about some ways to make combat a little more interesting than just back and forth "Yout attack the foo. The foo attacks you.". Inspired by KaViR's GodWars 2, but not totally convinced that everyone is willing to abandon automated combat (manual combat can be tedious at times) I came up with a hybrid. Players may opt to toggle automated (simplified combat) on or off. Automated combat utilizes player set stances, but does not allow for individualized moves.
A character may be in a high, medium, or low stance. A high stance is the most offensive, and tends to strike at the head and chest. However, it is vulnerable defensively, so is not well suited to battles of attrition and is better used to strike decisive blows. A medium stance is a balanced stance and the default stance. Attacks launched from a medium stance tend to target whatever are is open. A low stance is the strongest defensively as it minimizes exposed areas and keeps a low center of gravity, but requires extra effort to exploit openings and so may be inferior offensively. Attacks launched from a low stance tend towards an opponent's body and legs.

Additionally, a character may be set to emphasize Offense, Defense, or neither, which may exaggerate or mitigate a stance's traits.

In automatic combat, a character will simply attack repeatedly with its main hand and not utilize any advanced maneuvers.

Manual combat is very different. A character will not launch any attacks unless directed. Characters may attack by specifying the attack type. Attacks may be strung together to attempt combinations and follow ups. A character in manual combat must be careful to watch its balance and fatigue carefully, however.

The following are attack forms an average combatant may employ:
R- Right hand. Utilizes any object held or worn on the hand.
L- Left hand. Utilizes any object held or worn on the hand.
H- Head. Attempts to butt or strike the opponent with a head attack. Risky without a helmet.
F- Feet. Attempts to kick the target.
G- Attempts to grapple the target, taking combat in closer (and making most weapons useless). If the combat is already a grappling match, G will attempt to disengage by moving out, throwing the opponent, etc. depending on skill.

Moves may be chained. So RRL would attempt to attack twice with the right hand followed by the left. If the attacker was unarmed, this would be a series of punches. If attacker held a sword in his right hand and a shield in the left, it would be two slashes followed by a shield bash. The upside of chains is that they can be devastating and can even unlock new maneuvers. However, they are also draining, very dependent on combat skill and poor combos can leave a fighter dangerously open or off-balance.

I am still working on how to satisfactorily incorporate special maneuvers like disarming, pinning, sweeping/takedowns, etc.

One idea is to get away from the auto combos and let players enter moves individually with a secondary argument that specifies an advanced components. So R would simply launch an attack from the right hand while R L D would launch a Low attack from the Right attempting to Disarm. Which probably wouldn't work very well on someone in a High stance.
KaVir's MUD supported auto-combat as well, it just wasn't recommended for use because manual combat had such a dramatic advantage in terms of utility and damage. The trouble with systems like this is that one's always going to be better than the other. Either auto-hits are going to deal more damage more reliably than manual attacks, or vice versa. If you balance them so that both are equally effective, there's no reason to use manual combat when you can set your character on auto-pilot and let them do the work.

What are some of your ideas on how to correct this issue?
I'm kinda with Evre on this one. And if the game requires a great deal of strategy to implement combat, then the style of combat kinda becomes more important. It might be interesting to develop a system where combat is automated, but players can choose an attack/defense pattern based on having encountered a similar creature before and learned its strengths and weaknesses. (For player vs. player combat, I'm not sure what would be best.) Then experience really is a matter of experience, and you could make this replayable by varying the optimum patterns against each creature for each new character.

I tend to prefer automated combat, but it needn't be boring; you can actually implement a whole fight choreography system and work that into the output, making battles epic.
Evre: I think I agree with Kavir's approach. Automated combat will be simple, but nowhere near as flexible and potentially powerful as well played manual combat. However, manual requires thought- a manual player who doesn't respond appropriately to combat situations would likely do worse than the auto system. For example, auto combat would probably just repeatedly attack with the main hand from a balanced, medium stance. Very predictable, but reasonably effective most of the time. However, against a certain opponent, it might be far more advantageous to shift to a low, defensive stance until the enemy wore himself out, then go offensive before it could recover. OTOH, a player who went all out offensive early against the same foe might find himself at a severe disadvantage.

LummoxJR: I also generally prefer automated combat. But that is because manual combat tends to be poorly implemented. I have never seen a text game that made battles look epic. Many look boring, some kind of silly, and many spammy. But I will settle for believable and fun. If I get that, I think I will be in the top .0001% of text based combat!
The reason so many text games have lousy fighting is, programmers tend to know jack about fight choreography or for that matter, describing events in an interesting way.

I read an interesting piece on Dwarf Fortress last year, where the person telling the story said they had gone adventuring and ran across a minotaur. The minotaur was way too difficult to fight with standard weapons, and the player took a terrible beating, so the player just grabbed the minotaur's leg and tried a wrestling move to lock the knee. While the minotaur struggled to shake the guy off, on his second try he succeeded in locking the knee, and in a subsequent round he broke it. The minotaur went down and was easily dispatched after that. That battle sounded epic indeed. If your combat model has this level of complexity, you can easily describe the actions taking place and then fill in with color:

The minotaur slams you to the ground. Your armor protects you somewhat, but it doesn't survive!
You rush the minotaur and grab hold of his leg with all your strength.
WHAM! The minotaur stomps firmly to try to shake you off, but you manage to just hold on!
You try to grab the minotaur's kneecap and lock it, but without success.
The minotaur squeezes his leg together, bruising your arm. You are hurt but not badly.
You try again to lock the minotaur's knee. This time you succeed! The minotaur howls in pain!
The minotaur tries to bend down and bash you on the head, but his knee makes it too difficult. He misses.
The minotaur's kneecap shatters! He falls down in agony! You are barely hurt.
The minotaur screams and flails his arms but does not hit you.
You stab the minotaur in the throat with your shortsword. The minotaur dies!
You have defeated the minotaur. With renewed confidence, you dust yourself off and prepare for the challenges ahead, feeling stronger and cleverer than before.
What about non-realtime combat? Is there any strong incentive not to allow for realtime gameplay outside of combat but have timed-turns when in combat? Then combat becomes strategic, not reflexive, although you still have only a few seconds to choose your manoeuvre before implementing it (otherwise you forfeit your turn, or simply default to a routine random strike).
Well, I don't know how real time combat is ATM. Basically, every action has an associated delay, combat or not. But I see where you are going. More of a game timed turn structure, right? I think in practice, the two aren't terribly different. The reason I am hesitant to have automoves in manual combat is that different maneuvers have different vulnerabilities and all drain stamina (even throwing a punch gets tiring), so players may want to go defensive and rest (IE not use any moves) to recover energy. Getting too tired can greatly affect performance. Tired fighters tend to leave more openings, swing more sloppily, etc.