This is an interesting question in the context of early Ireland, because we don't really know how battles worked. Our historical sources tell us when battles were fought, where they took place and who fought them, but rarely give us accounts of what these fights actually looked like, in terms of scale and mechanics (there are a couple of exceptions in the Fragmentary Annals of Ireland, but these are Irish descriptions of Scandinavian battle-customs).

We can glean a few facts, like that the Irish mostly fought with a spear and small shield, their nobles with swords and perhaps leather armour. Curiously, chain mail was known to them and is listed in the Lebor na Cert as a tribute to be given to kings, but it appears that they preferred to fight unencumbered by armour. Anglo-Norman descriptions of the Irish going into battle 'naked' shouldn't be taken at face-value, and most likely represents their authors' dismay at the Irish practice of fighting with little-to-no protection. Warfare could encompass conflicts as small as inter-communal raids with forces of a handful of men to major dynastic power-struggles between regional hegemons and hundreds, or possibly thousands, of warriors. It also appears that the Irish practiced a form of Christianized military-magic, and believed that invoking regional saints as though they were pagan deities could bring them victory against all odds, especially if their saint was regarded as more powerful.

One account in the Fragmentary Annals about the Battle of Almu between the Laigin and Uí Néill does provide actual numbers, which is the reason why I started writing this post; the description includes seemingly impossible troop numbers. The entire entry is rather terse, to the extent that the writer infuriatingly states that:

The combats of the Laigin and Leth Cuinn warriors would be excessive to relate

...not that helpful, right? The gist of the entry is that 9000 Laigin warriors defeated a host of 21,000 of the Uí Néill, and then lists a few dozen kings (in the context of early Irish history, a king could be anyone from a community chief to a provincial overlord) who died in battle. This entry interests me for two reasons: the first one is the same question that you are concerned with, and second is its unusually large list of 'royal' casualties.

First off, these sources usually have a propagandistic flair, as they were compiled by clergy to enhance the prestige of their relatives who commanded armies and fought for political dominance. These authors might have inflated the number of combatants to make the victories of their ancestors seem even more impressive, which would confer even more prestige on their dynasty. Annal authors would also legitimize their contemporary political arrangements by projecting contemporary notions of kingship and governance into the past. It is possible that if the Fragmentary Annals were composed at a much later date, its author might have also projected contemporary standards of warfare into the past; the Battle of Almu is said to have taken place in 722 AD, and I think that it's unlikely that kings from such an early date would be able to muster such huge hosts. What's more likely, I think, is that the author inflated the number of warriors on both sides of the battle for two reasons: they might have done so for propaganda reasons and because they were projecting their contemporary perception of warfare, where armies in the thousands might have been the norm, into Ireland's military history.

This second factor is not just me guessing as to why troop numbers might have been inflated; Irish military historian Marie Therese Flanagan has theorized that Irish literature focusing on military conflicts such as the Cogad Gáedel re Gallaib is more reflective of methods of warfare employed when the books were actually composed than that in the periods they are written about. It isn't too much of a stretch to then conclude that an Irish author, writing the the High to Late-Medieval period, might project contemporary army sizes into the distant past, because that's what they would have been familiar with.

Well, most of warfare isn't big, pitched battles. The kind of battles that Total War depicts are all-the-marbles affairs, which even if they won't end the war will have a very large impact on it. Most warfare was sieges and smaller actions between parties off of the main armies, such as attacks on foraging parties.

That's not to say that these big battles with tens of thousands of soldiers on each side didn't occur. For one, Rome only fought tribes towards the north and the west. To the east, Rome fought kingdoms. Even the tribes could put together large armies under charismatic leaders to fight the Romans.

The battles wouldn't have been so... Bloody as Total War depicts them. They would usually drag on for hours, with the lines making contact, inflicting a few wounds and kills, then breaking off, over and over again. The real killing started when one side's order began to falter and a rout started. Then, there was no real protection and the losing commander would usually find himself without an army. Reserve forces could be used to cover the retreat, but even then it was hard to break contact with an enemy army.

So yeah, I'd go with battles certainly being less frequent than Total War would lead you to believe, but the big ones certainly did occur with enough frequency to decide a lot of wars.