There's never anything concrete, but I'd say that in general early game might be the first twoish years and lategame begins somewhere around five. However, when you hit early and late game are more characteristics of what you're doing than the time, and they're really vague anyway. Making sure your mages are always doing something useful is probably more practical than setting timing targets.
That makes a lot more sense than some of the timelines I've been seeing. I suppose I'm just trying to get a handle on when a given task becomes the most useful thing to do, operating on the logic that ideally one would want to start putting the infrastructure (boosting, blood hunting, etc) for the next thing together before whatever is currently working stops being adequate.
These things all depend almost entirely on your nation and your specific strategy. Within that context, one way to think of it is to consider what strategy will allow you to win the game, and then rush for that while only focussing on low-investment strategies to keep you from losing before you get there. This is the way a lot of mid-game nations play, with early game nations forgoing the infrastructure development almost entirely and just throwing niefel jarls at any problem that comes up. But if you play this way, you have a core goal and your timeframe is "as soon as possible". Of course, this can get silly, like with the fellow who put Burden of Time out on turn fifteen back before it was nerfed, but as long as you add just enough reasonable moderation to the mix that you don't get obliterated first, it'll win you a war or two. And when it stops working (if you can't win first) then you move on to a new game-winning strategy and everything else becomes a holding action until you get there. This is kind of a simple way to go about it, but it's still basically the flow of the game.
For example, by the time my early/mid armies are usually running into trouble, I've often got a number of mages en route that could have spent those turns researching or site searching or whatever, had I known they would be obsolete (or inefficient) by the time they arrived.
A mage is never really obsolete, they don't have to memorize their spells ahead of time. But if your armies are winning already, moving mages to the front line is definitely a waste of research turns. And if moving mages to the front lines is a big time investment, you should consider building or taking more forts nearer to where the fighting is happening, or possibly going for cloud trapeze/teleport/etc if those are available to you.
Maybe my entire conceptual framework for this is off, but it seems to me that the best way to think about getting a counter to your opponent's turn-X army out on turn X is to plan on building that counter on turn X-Y based on an ongoing Bayesian model of their likely strategy, for which a general timeline is a useful prior. Or am I overthinking this massively?
You're not overthinking until you've thought twice as much as all the other players combined. Dominions is a game that's won or lost based on planning. But if you try to be predictive like that, there's you should definitely be wary of committing too much unless you're really certain you know what they're doing, because you can definitely waste a lot of resources on them if you miss-guess. Specific counters are great, but versatility also shouldn't be underestimated, not only because it can deny your enemy specific counters to your stuff, but because you'll want to be prepared if an additional enemy decides to get involved.
tl;dr: Plan ahead for your own game-winning actions, but don't plan too much for specific enemy actions.
Try not to engage a large hostile player army if you don't have any idea of what it is capable of. But them attacking a 1 PD province would give you that intel.
With a bit (ok, a lot of) familiarity with the spell list, you could figure out their rough research levels from the spells they cast, and checking their mages for gems would give you a rough idea of what 'big spells' they're bringing to the party.
If it's a game where the score graphs are on (or you have a site that reveals the research graph) you could get a vague guesstimate from their research score, though.
Generally, I'd say it's a question of: What does he have; what do I have to bring to the table, what can I research in the next turn or 3 that would have a significant impact, do I have enough ablative territory to accept territory losses to allow me to blob/tech up just a bit more, or am I just able to smash it without expending too much effort?
But that's just me, everyone has a different style/idea of what is fun in this game, I enjoy having to sit back and do extensive sifting through the spell list to sort all of those things out.
I wouldn't say you should avoid engagement entirely, just avoid engaging with any army you'd mind losing. It can be effective to harass a big army using thugs or battlefield wiping spells, even if you haven't seen much of their script yet, just as you might roll over a small army without expecting them to have a particularly nasty surprise waiting for you.
The point about ablative territory is also a good one. In a war, don't be afraid to retreat for a time.
That's kind of my problem: being able to balance adaptability with the mechanics of force multipliers or synergy requires some concept of the atomic components of a given strategy, and that's wrapped up in my timeline question, since parts of a given idea come online sooner than others.
Just from this, I can't conclusively say I understand what you're talking about, but I'll deal with your specific examples more.
1. indie commanders leading crossbows & shields around
Always do this, and do it for as long as it works. Even when it stops working in pitched battles, do it for raiding. This doesn't cost you anything except gold, and gold is also what it gives you.
2. Iron Priests carrying crystal matrices casting Wind Guide, Flaming Arrows, sundry earth buffs and some of the E/F magma evocations, using Illuminated Ones as communion slaves and crossbows as meat shields/more arrows.
Avoid doing this as much as possible unless it'll win or lose the game for you. This costs mages and gems, which are what you need to win. You can get by in the mid-game with a few of these, and hopefully win a war with as few as two or three armies like this, and not that many iron priests or slaves per army. Especially if you rely more on the buffs than the evocations, though evocations are always nice as well. But if you've got enough thrones in close range, secretly prepare a few more armies like this and take them, and take the game. This could win you a mid-game victory, but it could just be the thing that keeps you alive until the late game. If it's going to be the latter, don't invest any more than you need to in order to win the wars your in. And of course, never be in less than one war for long, though that's a slightly different issue.
3. Vampires. Big blobs of vampires with boosted vampire lords casting Death and Blood buffs on them, with the Counts handling blood hunting and summoning and supplementary casting.
By the time you can do this, do this. Vampires are an excellent defensive strategy since they're immortal. Get enough vampires that you can destroy any army before they reach something important, and take whatever's in your dominion. Without blood sacrifices, you can't push your dominion as well as some nations, but it's still more than worth it.
4. any of the high-end Blood summons leading demons around.
You'll want these as well, for offensive play, but as with the mid game armies, have just enough to win unless it's the last war of the game or you're sure you've got enough vampires for the foreseeable future.
But those are composed of smaller pieces that can be shuffled around and are in some cases mutually exclusive; for example, communions big enough to make 2 work are going to delay researching to high-end Blood/Construction for 4 (and buying the S1s in the first place to get all the evo/thaum/alt needed for 2 quickly is going to slow down 1, which is partly needed to grab provinces for the blood needed for 3-4). Thus why I'm trying to figure out what needs to be in place to make a given thing work and what just makes it work better, which partially depends on being able to project, in a rough way, what my units need to not die on any given turn.
After the first year, you should pretty much never let anything get in the way of your mage recruitment. If you've snapped up all (or almost all) of the indy provinces in the expansion phase, don't pay for any crossbows that'll keep you from recruiting mages. If you have to lose research, it'll be more effective to deploy those mages than to get something besides them anyway. The research issue is a stickier problem, and that's why it's often good to plan according to your game size, aim for a midgame victory on the back of strategy 2 if there's less than maybe eight players. In a bigger game, you still probably want to to some degree, but once you've got all of the research that's strictly necessary for that, dive into blood. The delay on your blood lategame will be less from researching other things than the gain will be from taking other people's land, assuming you win your wars. It's still better to switch to the late game strategy too early than too late though, because you can always switch back to the late game trees to make rapid progress, and anyway you'll want those support spells on your late game armies as well.
Author
Topic: Dominions 4: Thrones of Ascension (Read 524275 times)