posted 01 November 2017 02:07 EDT (US)   
I was reading The Cambridge History of Greek and Roman Warfare and it makes clear that many of the tactics and operational methods used in games are in direct contrast to the written and physical evidence we have. Now it's not really surprising that games aren't realistic, but a common theme throughout these is that many of these ahistorical strategies and tactics are based on the incredible amount of information and control the player has that no commander or group of politicians could possibly have had. Very few games seem to be interested in 'fixing' this because most players seem to be more interested in map painting and getting into fights than managing a country and its military forces.

Mixed forces and combined arms were both historical and represented in many games. However, actually understanding the characteristics of different military units and forming a cohesive plan was much harder than in a game where everything is recruitment lists and numbers. Also see further sections on the symmetry of deployment.
Chapter 11a, 'Land Forces' by Nicholas Sekunda
Little attempt was made to standardize Hellenistic troop types. On the contrary, the Hellenistic states reacted to contact with non-Greek military systems by incorporating yet further weapons and troop types of foreign inspiration within their lines of battle, and by devising further formations of their own.
...The principal difference between the Macedonian and Roman army systems lay in their relative complexity. Macedonian battle-tactics evolved according to what new troop types became available as new forces of mercenary or allied troops, each with their own distinctive national equipment and tactics, were incorporated into the army. Consequently, Hellenistic armies were a complex amalgam of elements with highly differing combat characteristics. The commander had to devise a discrete plan for each battle, and then coordinate all the separate tactical elements at his disposal
to achieve his aim. This called for a degree of staff work and tactical flair that would have tasked the best command elements of any army.
Strategic Reserves are a common technique in operational and strategy games. But for intelligence and maneuver reasons they do not seem to have ever been intentionally used in the ancient world.
From Chapter 12, 'War' by Jonathan P. Roth
The need to keep garrisons in various parts of an empire naturally created an ad hoc strategic reserve. When Ptolemy I annihilated Demetrius’ army at Gaza in 312, the latter was able to raise another one by stripping occupation forces from various provinces (Diod. Sic. 19.80.5). There is no recorded case, however, of forces being left concentrated in the rear of an area of operations solely to serve as a strategic reserve.
Pitched battles were usually easily avoided, and were only a small part of the actual warfare. In most games, pitched battles are the ONLY kind of battle that happens.
From Chapter 13a , 'Battle' by Philip Sabin
In the pre-gunpowder era, the advantages conferred by natural or artificial defensive positions were such that an inferior army could often deter enemy attack simply by standing on a hill or staying within a fortified camp, while relying on city walls to protect its civilian population. Pitched battles were fought only when neither side felt at a disadvantage, and so they often occurred only after months or even years of cautious campaigning. When they did occur, this dependence on a degree of mutual consent tended to give the battles a certain set-piece formality almost akin to a duel.
The other key contextual feature of ancient land battle was its extreme compression in time and space. Although the armies, each a few tens of thousands strong, were significantly larger than during the preceding hoplite era, they were small enough not to have to disperse to live off the land, as Napoleonic corps had to do. Their primitive command and communications and their reliance on deep formations meant that battlefields spanned only a few miles at most. Also, close-quarters combat was such an intense and stressful activity that it usually took only a few hours, if that, for one army to be completely shattered. I will now analyse the key grand tactical features of these sporadic and highly focused contests.
Deployment was usually symmetrical, with units fighting similar units, rather than the sort of opportunistic min-maxing that usually occurs to exploit differentials in strength. Such differential pairing usually only happened after a portion of the enemy army had actually collapsed. Also, the use of a single strong flank - which is very effective in most games - was mainly a tactic of the Greeks and most other armies preferred a strong center and double-envelopment.
From Chapter 13a , 'Battle' by Philip Sabin
The standard army deployment throughout this period consisted of heavy infantry in the centre in one or more lines, with cavalry on both flanks, and light infantry skirmishing in front. If one wing rested on rough terrain, then light infantry rather than cavalry might be deployed there, as at first Chaeronea and Issus. Elephants or chariots, if present, were usually spread out in front of part or all of the battle line. The result of this rather formulaic deployment pattern is that, in battles between combined arms forces, similar troop types tended to find themselves fighting one another – cavalry against cavalry, light infantry against light infantry, elephants against elephants, and so on. Only after their enemy counterparts had been defeated, or if the enemy lacked any similar troops of his own, were the various fighting arms able to engage dissimilar troop types and thereby to exploit the offsetting strengths and weaknesses within the combined arms mix.
The most striking overall contrast within deployment patterns in this era was between the Greeks, who tended to weight one wing more than the other, and other nations such as the Persians, Carthaginians and Romans, who adopted much more symmetrical battle lines in which each wing was usually a mirror image of the other wing unless terrain dictated otherwise...
...The result was that non-Greek battles assumed a rather more symmetrical appearance, with double envelopments being more common than the single outflanking moves which the Spartans had pioneered.
Extended front lines, very common in tactical games, were not typically used, even when massive superiority in numbers existed.
From Chapter 13a , 'Battle' by Philip Sabin
Another interesting aspect of symmetry occurred between rather than within the opposing armies. Although in Alexander’s battles it was quite common for cavalry to be deployed opposite enemy infantry, in later engagements in which both sides had large numbers of good-quality heavy infantry, the norm was for the infantry lines to be of roughly equal length even if one side had a significant numerical advantage. Armies with large numbers of heavy infantry tended to deploy their men in greater depth (as at Cannae and as with the thirty-two deep Seleucid phalanx at Magnesia)
instead of extending their infantry line beyond that of the opponent. The reason is unclear, but it probably has to do with the command problems posed by an unduly long infantry line and the difficulty of outflanking the enemy infantry in the face of intact enemy cavalry forces.
Tactical reserves are likewise something that is a much more modern innovation, and while effective in games was not practical in most real battles.
From Chapter 13a , 'Battle' by Philip Sabin
Maintaining a reserve of uncommitted troops behind the main fighting line has become an axiom of modern military wisdom, and this principle was far from unknown in antiquity. However, in this period, the principle was honoured more in the breach than in the observance. Leaving aside skirmishers, elephants, chariots and camp guards, the troops in non-Roman armies were usually deployed in a single fighting line.
The very nature of reserves is that they are held back to respond to some unforeseen contingency, but in order for this to occur, the general must first perceive the contingency and then get orders to the reserve. By the time this happened in antiquity, it might well be too late, so it was often preferable to commit the forces to the main battle line in the first place.
In many games, skirmishers, slingers and archers are very effective even in pitched battles - in Rome: Total War archers fire directly into enemies in front of your line, and are light troops are often used to maneuver around enemy flanks. Likewise in Field of Glory 2, and many other tactical games. However, this would probably be suicidal and impractical - light troops were used to screen, mostly fought each other and cavalry, and had little role in the battle itself.
Chapter 11a, 'Land Forces' by Nicholas Sekunda
The first to engage were normally the light infantry skirmishers, who screened the deployment of the rest of the line, and who duelled with their counterparts (sometimes for several hours) until their missiles were exhausted, at which point they retired through gaps in the main fighting line and played little or no role in the battle thereafter.