Well i did contribute to that discussion with in a more general fashion as well, earlier on.
But did want to add the souls to the mix since they did spring to my mind as a solid option / counter after testing them in game and thus noteworthy here. (and because souls are exclusive to Sidar i don't need to take into account other civs to that special discussion.)
Also noteworthy might be that a one-landmass game like pangea with everyone attacking you is about the easiest you can get for Sidar (huge ammounts of XP---> Huge ammounts of Shades---> easy game due to insane outputs). As seen in the way the survival challenges play as Sidar.
So lower difficulty because of erebus-mapscript is hardly an argument for such maps (one could argue the sidar do worse on an erebus map, if not extemely! exploiting it, than in an open one-landmass one.).
It should still work on many maps (most! mapscripts actually feature heavy terrain. Even though some quite popular ones like inland-sea and pangea don't. Multiplayer is another matter...). It just should works exceptional on erebus and since i don't play much other maps i can't judge very well on the viablity on other much different maps (like pangea).
It should! still be decent on enough mapscripts nontheless so saying its limited to a rare few mapscripts might not be fair. On Survival you might Field Tier 3-Tier4s when lots of Pyres start to appear (though this might have changed with 0.34 admittedly the way Zechnophobe describes it...)
I don't use chokepoint-exploits to a large extent (and the AI has the annoying habit of blocking of the barbs lately when i need them most / have got the base-setup complete so its not actually all that much of a pure advantage to my gameplay. I also overcrowd the map lately so it has its drawbacks...).
And since its a clear AI-failure which I'm sure will! be fixed (shouldn't be soo hard), i don't take this one to the limit. And wouldn't advertise it. So this is not part of my considerations outlined above. (doesn't change that the attacks come from a certain direction not from a random side. Which as i mentioned i rather like... No need to agree here.)
When given the choice of 35% withdrawl (regular + homeland) with a good chance to win but still quite a chance of dying or a whopping 75% chance of Withdrawl / win at the worst (regular + homeland + Flanking 1 & 2.) I'd chose Withdrawl and medium chance of winning (one can disagree on that, its very much down to the real situation in question and different experiences i guess). If i don't have a good number of horsemen i won't stand all that good of a chance anyways in such a situation. I won't have soo much time to go back to a city and heal that early usually unless i see their stack coming and DoW on them. At least the way i play.
On Rangers: An additional worker at an early point (unless conquered) might mean 1 or 2 shades less which is possibly a huge disadvantage. And if playing sidar i don't count in raw hammers but in turns to build.
No need to build a cheaper unit when it takes 1-1,5 turns to build such a unit. Faster than a unit a turn is not possible anyways (and those fast-builds do add up with Heroic Epic, Command Posts and the likes....) that's why that equation is a bit different for sidar.
Also those situations would happen with the timeline outlined by you (even with 2 workers getting roads in large parts of the empire would either mean delaying improvements or taking some time indeed.) If you give target-turns i might comment if rangers-ghosts are a real option in my experience (or even tougher units).
Also what i even did forget is that at the time i field rangers / ghosts the units from my main city usually get an additional free promotion from Shrine of the champion. Further tipping the odds quite visibly.
On defense of mounted units though i have to say that my usual beline for Tier 4-units is War-Chariots so mounted line isn't strictly bad in any way. Just not the early focus.
Why would I ever want to get Walls as Sidar? (except perhaps during a survival-challenge.)
Defending means less XP than attacking which means less Shades (unless the unit doesn't make it of course. But better lose the odd one and get your first shades fast than taking your leisure and losing the game because of it.). I want to kill them in the field.
Be it barbs or enemy superstack.
Also building the wall itself means less units, which translates into less shades (for quite some time non % yields (libraries and the likes are still very much worth it) from exconomic buildings which are rather expensive, so except perhaps! markets and councils, just don't cut it in comparison to more fast-built units. Much less so for things like Walls. Its just Military, Military, Military with a few wonders and military-enablers or +% buildings and other utter essentials sewn in.).
Only instance i can think of where i would want masonry early would be if marble is near (cheap early Great Library (+ some other cheap wonders on the way) really is hefty for Sidar which have a little wonder-focus anyways. No doubt about that. If bronze + horses are near as well it may really pay off due to early powerful Chariots/Rathas... but that happens so seldom in my experience that its negligable for at-least for such a strategy-tip.
Someone trying to win deity-level / coping with such a situation should know that by himself / herself anyways.)
Trade can wait for Sidar (because they can generate enough beakers to get their techs the hard way unless the game went bad.) until the most essential techs are done (all it does is helping to fill out some of the more problematic gaps. At the time you get it usually you can get a tech lead in certain branches by belining which still includes trade but much later in comparison to the essentials.). And again: What i said only counts for Sidar (obviously
).
Also some AIs tend to get Trade early enough at deity. Trading with those is always an option. Might even get you trade without getting it the hard way (unless you do play with no tech-trading or to a lesser extent no tech-brokering. Which i don't currently...).
That tech path of belining Military Strategy and KotE is something i got down to do regularly, after experimenting alot with the Sidar and trying most paths (including a focus on mounted-line). Couldn't find a better path yet (Naturally variations are possible because of different circumstances in a given game).
Also thanks to OO beeing interesting to sidar for various reasons sailing is another possibly interesting way to go for trade for them.
Sure there are some variations (like if lots of livestock is around) and horses might! be viable (Because you realize a horse-resource is near as well as soon as you get AH.).
Believe me Cats just suck for Sidar (no shading, no benefit unlike any single other military line. Not counting arcane and divine which still have some benefits ) more than for almost any other faction (and they are truly important only to relatively few civs even though i still advocate them as far from useless. You can see this just in this thread.).
which the Sidar boost to a certain extent each thanks to homeland withdrawl for mounted and recon, glaringly obvious recon-focus, early enchanted Blade for Melee, relatively early flaming Arrows for Archers (Archery-Line still sucks.). The Sidar might not look like a militarily-focused civ on first look, but thanks to their mana and favored theme they are.)
And even taking enemies down to low HP for mopping up sucks because it means better odds which translates into less XP from battles which translates into less shades. (Highly promoted units shortly pre level 6 usually have high enough odds on their own.)
The point about tech costs is not directly comparing raw costs but to compare the value of those techs in situations which happen in most games.
And be-lining construction with Sidar is about as useful as be-lining a Sphere-Tech or even Sorcery with the Khazad (i mean it!) in comparison to the alternatives.
Horses as outlined above might be another matter but for a general strategy they are unreliable at best for reasons allready mentioned (i won't even want to go AH early unless essential Livestock is near or i want some animals caught.).
KotE is a rather early tech (sometimes even pre bronze-working if no copper is near, enchanted blade + courage really is that valuable in my experience and adepts make outstanding shade-material.) and Hunting should be as well now thanks to divided Souls (even though i still have to teach myself to prioritize it somewhat thanks to divided souls. But that won't last long i reckon. With the power the souls have.).
On cities
thats the best part:
Those are really toast with CR 3 Axemen with enchanted blades (in the time I'm ready for a bit of conquest they are city Raider 3 after fighting one barb. And thats easy to come by usually.).
If copper is near (best would be right on the tile of your first city which i had a few times) even warriors suffice.
And a stack of City-Raider 3 copper-Axemen with enchanted blade + courage can even toast a city defended by Tier 3s (thanks to being fast-healing and power 13! when attacking cities.
Need i say more?) without much fuss or losses (save those defended by Tier 3 Archery like longbows but those usually come much later at a time i allready field War-chariots. And those really like units of the archery-branch... Mix in a few Axemen with different promotions or a few souls and the stack defends just fine as well.).
I have killed Acheron with Sidar pre turn 100 with such a setup not encountering much resistance (sacrificing a scout for taking down stoneskin first of course.).
That's a real goody if you can pull it off that early. No more great bards needed to pop those new cities + some neat income and resources and an easier less micro-heavy upgrade to enchanted blade for your melee units. Simmilar for Orthus spawning near you.
You should still better catch an AI with their pants down like right after vanquishing a Superstack or backstab when they are clueless if you are on attack. immortal/deity-AI often is rather unforgiving.
So if the AI can build up a massive defense (like multiple pyre-zombies for each city) you already made something wrong / the events turned out not so well...
Remember, i get Warfare-tech rather early most games for Form of Titan + Conquest and because its required for Military Strategy.
Those actually are usually there before many Divided Souls appear (even though i might change this after extensive testing for games where early copper is unavailable.).
On the final verdict I'd fully agree. Signed (for most civs that might be the most viable counter to go horses + Cats / Cats + Horses / Cats + Warriors. Especially for dwarves of course. Both factions. For different reasons.).
Nothing wrong with pointing it out still. (since there is no option to extend it to other civs / a more general situation anyways, it should be rather obvious.)