Site Searching Strategy
 

A recent post on expansion strategy included a post by KissBlade that had the following link: thread

This excellent post has a paragraph on the end on site searching strategy that is in all likelihood excellent advice but I am still balking at taking this advice (please forgive me my noobishness). As the post suggests I do find acashic record very tempting. Maybe this because of my latent OCD, or my lack of experience in MP games, but I do seem to have a hard time adjusting to the idea of not using acashic record.

Now, most of my experience is with EA T'ien Ch'i, so that's my reference point, but I'd be interested in comment from the point of view of any nation. I've read posts indicating acashic vs the individual spells is a tradeoff of gems vs mage time. That makes perfect sense, but we are talking about a lot of mage time, here, right? I mean wouldn't that equate to 8 casts of the individual spells vs 1 cast of acashic? That could translate to an awful lot of research or mage combat availability.

Also, given that you're going down the single spell road, that still means you're not hitting the holy sites. Now this could easily be a reflection of my OCD, but it just bugs me that in all probability I would be missing out not just on holy, but in the case of a non-blood nation, also blood and maybe some of the other sites, as well.


Having said all the above, here are some detailed questions:

1) One thing that surprises me is that in the post it says that almost all of your site searching is done with spells. Would this apply even to TC, which seems to have such an excellent cadre of mages for searching? In SP testing, I do find myself going through a phase in the early game where I do a fair bit of searching with 2-3 Masters of the 5 Elems, and/or 2-3 CMs. Is that just plain bad mage use?

2) Does this question of the relative cost of acashic vs single spells change when site frequency changes? Say 35% vs 50%?

3) Same question, but say you have a 20% conjuration bonus site, does this dramatically change the balance in favor acashic?

4) I imagine in practice, that experienced players using the spells don't actually cast 8 different spells on every province. Obviously, if you already found 4 sites you're not going to cast more spells, but would anyone care other rules of thumb on how much work they put into casting site searching spells?

5) If you're not using acashic, how much work do you put into finding the holy sites? Just send an indy priest around, or do you take your best national priest? Or do you just skip this holy site thing, for the most part?

Re: Site Searching Strategy
 

Everything has pros and cons...

(1) Manual searching means search, move, search, move. Spell searching gives you search, search, search, search. So as soon as you can do spell searching its best to switch to it. Your mages will also be in the lab if you need to summon or forge instead of search. Drawback, if you find a province that needs a mage to enter a site or you need a lab to recruit then you have to send a mage from home.

(2) Not really.

(3) Good question. I will leave that to people more math oriented than I

(4) in practice most set a group of mages of each type to casting. They all cast on the same province each time. Sometimes I will leave off some because I want (possibly incorrectly) to increase the chance of finding an earth site before filling up the slots with nature sites

(5) I tend to try and keep priests up near the moving armies anyway. And anytime they have to stop then I search.

There are times I use Acashic. Such as finding a prime chokepoint and I want to everything about it before investing in a fort there. I might get a free fort, or surprised later by a site that causes disease. I rarely have every mage-type so I can check for every possible site using other spells. Im not sure if thats spread-sheet thrifty but I do it.

Re: Site Searching Strategy
 

Well, one keeps learning. What are holy sites? Even though I've had priest/magi search provences before, I don't remember uncovering any of these.

Re: Site Searching Strategy
 

1) Not necessarily, no, since 5 Element Masters and Celestial Masters can each search several paths worth of sites depending on what picks they get from randoms. The lv 4 sites are so rare as to be almost nonexistent and they only appear in certain kinds of terrain (e.g. waste for fire, plains fore air, forest for nature etc) and lv 3 sites are, aside from a common one for each path (Iron Cliff, Well of Pestilence, Cave of Clouds and a few others) rare compared to lv 2 sites. So a mage with multiple lv 2 paths is far and away the most efficient searcher early on. Even lv 1 paths help find some really good sites, like the metal order towers.

2) Yes. Somewhat. The higher the frequency, the bigger the payoff since there WILL be more sites with the higher frequency. Of course, searching by mage will also be more efficient.

3) Depends, but astral gems are so useful for a lot of ther things too that it's still somewhat on the high side. Now, if you were to find The Ultimate Gateway (conj. bonus 50%), that would be a different story...

4) Too situational, depending on available paths, gem income and other needs to make any hard and fast rules.

5) Rather little. They are rare and aside from a couple of unique ones (which you are unlikely to find, all other things being equal), they aren't much good to justify the effort.

Re: Site Searching Strategy
 

Holy sites are sites that require priest levels to find. They act like any other sites otherwise.

Re: Site Searching Strategy
 

So a holy site is another source of gems?

Re: Site Searching Strategy
 

Some are, some let you recruit priests, some strike undead in the province with holy fire. They are different holy sites, just like any other path.

Re: Site Searching Strategy
 

This explains what was happening to me in an SP game of mine. I had the PoD who was generating these dead guys and I kept getting messages saying something about holy fire or something was knocking some off every turn.

Guess I'll do more priestly searching.

Re: Site Searching Strategy
 

Thanks for the answers, this does help!

I seem to recall from Dom2 days that there was an unholy type, and that that was different than holy. However, in the site DB, I don't see anything labeled as "unholy" - does this exist any more?

Also pulling from quantum mechanic's original post, we have the following list of tendencies in type of sites vs. terrain types:

nature in forests,
earth in mountains,
water in mountains and forests,
astral in plains,
fire in wastes,
air in mountains and wastes.

Is this list complete? I would certainly have guessed that water sites were more common in sea provinces (for example).

Also, death isn't on this list and neither is the terrain type 'swamp', is there no terrain type particularly good for death sites and no site type more likely to be found in swamp?

Re: Site Searching Strategy
 

"Unholy" sites don't exist anymore afaik. (...and I miss them!)

Not sure about the completeness of the sites vs terrain list.

(btw, I use acashic record with a vengeance when I have access to astral.)

Re: Site Searching Strategy
 

Some disorganized thoughts...
If you've got no mages to use the gems produced by sites uncovered by Acashic Record, you can alchemize them to get more Astral (or, even less efficiently but still occasionally useful, empower somebody so that you do have mages to use the remaining gems). However, anybody who takes the province from you gets the benefit of those sites too, so there's a small potential cost to exposing them. It's also just about the only way to search for sites that you don't have mages for, which makes it potentially valuable. How expensive is it, if you do have mages otherwise capable of casting the little radar spells? You'll probably stop casting them on a province after about 5, on average, or about ten gems and five mage-turns per province. Compare that to 25 gems and one mage-turn. Acashic record will net you slightly more than the individual radar technique, but not much. So it's only a good deal if you really want to make sure you can find every damn site, if you want lots of gems that you can't actually use very efficiently, or if you value your mage time at 3+ gems per turn. It sure is fun and easy, though... Oh, and if you only have 2-3 paths for which you can use radar spells, and Acashic Record is an option, you can expect (at normal site frequencies) to make back your gem investment vs. radar spells (by alchemizing those useless gems that start rolling in) in about 20 turns. So if you're playing a nation with strong Astral but not much else and no options to change that (is there such a nation? If so, it better have a lot otherwise going for it) it wouldn't be a hideously bad bargain.

Re: Site Searching Strategy
 

Hellboy, take a look at the site DB for Dom3 (link in sig), that should give you an idea of what sites to find in which terrain. A lot of sites are "Any land", but in general the tendency is:

• Fire: waste, mountains
• Air: mountains, plains
• Water: swamp, other non-waste
• Earth: mountains
• Astral: any, slight emphasis on plains
• Death: any, slight emphasis on mountains
• Nature: forest, sea
• Blood: any

I'll be updating the site DB in a few days because I now have all of the info I could possibly need, thanks to DrPraetorius. His efforts have given me accurate terrain masks, site frequencies and quite a few other bits of info that helped me track down some errors and add missing things to the DB. I'm also going to shamelessly steal his formulas for displaying the specific terrains and uniqueness of sites (using the terrain masks as basis) to make the DB more readable. That way the DB will be easily accessible and only the hardcore enthusiasts will need to try and decipher his data dumps.

Re: Site Searching Strategy
 

Do border mountains count as mountains for the purpose of magic sites?
I haven't seen a single occurance of regular mountains in a map generated by Dom3 random generator, so that may skew the distribution and availability of magic sites quite significantly.

Re: Site Searching Strategy
 

Border mountain provinces count as mountains.
To get regular mountains in a random map, uncheck the border mountains box prior to map generation.

Re: Site Searching Strategy
 

Well, these types of questions are very situational. I agree that path specific radar spells are the better choice in most situations, but certainly not all. Basically, you've got 3 choices 1) Manual search 2) Search each path remotely 3) Acashic record. Here are some factors which have a significant impact on what you do.

1) Do you have strong astral with not much diversity, while also having conjuration as one of your first research goals? Bandar Log comes to mind, using acashic record sparingly when you first get it will give you a good multi-path gem income that your nice summons can use. Not a bad idea to go for arcane probing, then straight into conjuration.

2) Is the research required for radar spells at cross purpose to other research you need done early on? Mictlan comes to mind. With astral, fire, nature, & water mages all of path two, I often end up manually searching because I'm focused on researching construction and blood.

3) Do you really want gem incomes for paths you've only got first level mages in, or no mages yet (perhaps you're expecting to summon something to take advantage of them)? Expecting to need to bring a fight underwater and need water gems? etc.

4) Site frequency (including era, LA has less sites I think) absolutely effects the choice because the payout is higher so the more important factor is mage-turns. Conversely if you've got low sites, and consequently a low gem income then even the path specific radar spells may have a low ROI and it makes more sense to manually search.

5) Are there some higher level sites that would have a huge impact? For example there is a level 3 blood site which gives you a 60% blood magic discount, if you're playing a blood strategy its worthwhile to make sure you search for level 3 sites which often means remote searching.

6) Do you have a lot of swaps/wastelands in your kingdom? It really doesn't make much sense to cast acashic on farmlands, I usually only cast it on swamps/waste/forest/mountains with no sites showing yet.

7) How valuable are off-path gems to you? If you've got no use for a path unless you empower somebody twice gems of that flavor aren't terribly tasty.

Re: Site Searching Strategy
 

Using Acashic Records usually gets you lots of gems for paths you can't use and wouldn't find otherwise, which you can then alchemise into pearls, and then cast Acashic more with those pearls. 2->1 ratio isn't that bad.

Over time you'll get more pearls than you'd save by not using it, and it saves mage time and reduces micro. I like to use it when I can.

Re: Site Searching Strategy
 

That's good to know - forgot about that!

Edit: Ooops, that was a dumb question, never mind.

Re: Site Searching Strategy
 

Not having regular mountains is pretty bad for nations like Agartha who then cannot build cave fortresses.

Re: Site Searching Strategy
 

I almost never use Acashic record, even when I have access to it... too expensive, I find, and not worth the payoff.

I tend to begin manual site searching (move, search, move search) very early with several mages to get my gem economic going, particularly if I have access to diverse paths and will need many types of gems. Once its later in the game and I have more mages and more gems, I can stop some mages from researching and have them cast spells to search instead... I find this method gets me a very large income of gems starting early in the game.

Re: Site Searching Strategy
 

Again (why do I keep doing this) you might notice that the game style affects the answers. Everyone is correct, for their game.

The advantages of using Acashic Record makes it balance out in long games on larger maps.

The advantages of walking around would tend to be obvious for smaller maps like most of the blitz games play on.

Re: Site Searching Strategy
 

The common wisdom seems to be that level 3 and 4 sites aren't worth searching for because they're too rare, but the preference seems to be for remote searches anyhow. I don't follow how this is a good choice if you have national mages with anything approaching decent magic paths.

If you can move and search for 2 level 2 paths you're getting 2 searches every 2 turns (move, double search, move, double search), just as with a remote spell, and saving 4 gems in the process in return for not finding level 3 and 4 sites. You give up some flexibility since your mage is away from home, but as long as you script a retreat order they're still pretty safe (I've even gotten a free mage turn out of it when my mage runs to an unsearched province after a random indie attack). If you have 3+ paths on your mages the benefits of manual searches seem even more obvious to me. I can understand remote searches for single path mages, or ones with 1 map move because it's much harder to path them effectively to avoid wasting turns, but I don't see how it's a better choice for most nations at the start of the game.

Are the shots at level 3 and 4 sites really worth burning the gems to do level 9 searches, or is the preference there just because it avoids MM issues? Or is there some other reason I'm missing?

Re: Site Searching Strategy
 

The site searching by traveling mages is most effective in the early game. Most good searchers would probably be capital only and have a lot of important uses elsewhere too, so it's a situational balancing act. But that's basically how it seems to be done. *shrug*

Re: Site Searching Strategy
 

I do tend to use my mega-mages close to home. Such as the perimeter around the castle. But I dont tend to want them to wander too far from a lab.

Re: Site Searching Strategy
 

What's the consensus on Voice of Tiamat? I think it's a waste of gems; finding all elemental sites sounds good until you realise that fire and air sites are pretty rare underwater.

Re: Site Searching Strategy
 

Voice of tiamat is pretty commonly used or atleast it was. I for one definitely still use it.

Re: Site Searching Strategy
 

I really like voice of tiamat its great for water nations, I believe it gives them a gem adavantage over the land nations and saves alot of mage time. Also its 8W gems which I always tend to have loads of anyway.

Its an halfway house for water nations between Acashic record and individual site searching. Well air is rare, the water/fire/earth more than make up for it. As soon as I get it, I mass cast it on every water province. You usually get 1 or 2 free castles this way so saves you a load of money as well.

Great spell.

Re: Site Searching Strategy
 

Voice of Tiamat also has the advantage of being a conjuration spell, so it's in the same school as Dark Knowledge and Voice of Apsu and it covers all of the elements for seas, giving some damn good research optimization. There are no blood sites underwater, so that leaves Haruspex, DK and Arcane Probing. AP comes from evo which everyone needs sooner or later.

Re: Site Searching Strategy
 

I prefer to send my mages out, and when one can't move into a province that hasn't already been searched, I have them build a lab and then they can cast/forge/research right there.

Re: Site Searching Strategy
 

I do belive that Earth is most common under water. In all my games as a water nation I have had huge earth income.

Re: Site Searching Strategy
 

Voice of Tiamat does tend to reveal those underwater castles which are very useful, as well as saving cash as Meglobob mentioned. It is valuable.

Re: Site Searching Strategy
 

Saxon, isn't that Haruspex? The UW forts almost all come from nature sites, namely Kelp Fortress giving a fort of the same name. Academy Underneath and a maybe one other UW fort site are non-nature and they are unique.

Re: Site Searching Strategy
 

Academy Underneath seems pretty common though. Seen it in both games where I was using Voice of Tiamat extensively.

Re: Site Searching Strategy
 

Looking at the site DB there are 3 possibilities for underwater forts from magic sites:

W1: academy underneath, uncommon freq. (NOT unique)
E2: firbolg fortress, Unique, may also appear on land, uncommon freq.
N1: kelp fort, non-unique, sea only (not deep sea), common freq.

I think I got that right, although last night I was making a lot of dumb math mistakes talking about growth scales, so feel free to double-check.

Re: Site Searching Strategy
 

Ah, okay. My mistake then, thanks for pointing it out, Micah. I'm not quite as current on the uniqueness in Dom3 as I was in Dom2 since I just copied those stats and assumed them to be accurate until DrPraetorius did his code diving expedition into the realm of magic sites, at which points I just copy-pasted a few columns and a few formulas to display things and drag-copied them downward without specifically going over each item separately.

I just came to work and I'll probably check it anyway when I get home, to get it out of the way. http://forum.shrapnelgames.com/images/smilies/happy.gif

Re: Site Searching Strategy
 

Academy Underneath can't be unique, I have had in 3/4 games where I have been searching water provinces.

Re: Site Searching Strategy
 

EDIT: Academy Underneath is unique. It's an uncommon site, so it will appear in most games, especially since the number of possible underwater sites is much lower than the number of sites that can appear on land.

This much can be said of ALL unique sites: They can have any frequency, meaning that there are unique sites that are common. In most games of any size, especially if the map is large and site frequency is high, they will be found. If you found several of the same site, it was not unique.

The rare uniques are another matter, because rare sites only appear if the reroll for site determination also comes up with a rare site (all rare sites are rerolled on first occurrence).

Note that unique sites may appear more than once if they were scripted in the map file. The reason is that sites are randomly assigned first and only after that the map file scripting is added.

Re: Site Searching Strategy
 

Thanks Edi, thats very interesting. I have been wondering a bit about the concrete mechanics of magic site distribution.

Could you - or somebody else perhaps - tell me:

1) What is the difference between common/uncommon/rare sites? You mention someting about a reroll being required for a rare site to appear?

2) How often does the different site types appear? I would guess (from experience) that at least 70% of the magic sites are common sites, but perhaps I have just not looked hard enough for the rare (and often high-level) sites.

3) The graphics icons of the magic sites differ - even among sites of the same magic type (fire, death, earth, etc.). Does th different icons indicate the rarity of the site (common, uncommon, rare) or the magic path level (1-4) you need to find the site?

Re: Site Searching Strategy
 

1) There are 4 possible slots for sites per each province. First slot is checked against site probability. If site appears, it is randomly determined. Then next slot is checked for site probability. As soon as a site slot comes up as empty, it and all the slots following it will remain empty. So if the first one comes up empty, no sites at all in the province.

Common sites have a greater likelihood to appear than uncommon sites, which again have a greater probability than rare sites. If the random determination results in a rare site, that result is rerolled and the second result will be the actual site that appears. In other words, a rare site appears only if the reroll also comes up as rare.


2) The exact percentages of common/uncommon/rare are not known, but I'd throw an off-the-cuff guesstimate of 60%/30%/10%, give or take something. It'd need some statistical analysis of a fairly large sample size with controlled terrains to get reliable results, and even those would be rough results.

3) The relationship of site icons to their level and rarity is not clearly established, but there seems to be some correlation, in some cases at least. The correlation seems to be more closely to the site level than to rarity.

Re: Site Searching Strategy
 

Is this really so? I usually play with 30-35% sites, and according to what you said, the site frequency for, let's say 30% sites would be:

0 sites: 70% provinces
1 site: 21% provinces
2 sites: 6.3% provinces
3 sites: 1.9% provinces
4 sites: 0.8% provinces

Even though I haven't done any statistical testing, I am fairly confident this is not what I'm seeing. Even for 50% site frequency (say, the abovementioned 30% plus terrain bonus) the numbers would be 50 (0), 25 (1), 12.5 (2), 6.25 (3), 6.25 (4).

As opposed to a binomial distribution, which for example for 30% sites would look like this: 24 (0), 41 (1), 26 (2), 8 (3), 1 (4).


K.

Re: Site Searching Strategy
 

I'll take your word for that. I barely scraped through statistical and probability math at high school and failed every course on the subject after that.

It's not that I have a problem understanding the results of statistical analysis and probabilities, but trying to figure out the actual process from raw data, I screw that up every time.

So if the binomial model looks like it fits the observations better, then it's probably more accurate.

Re: Site Searching Strategy
 

Is there any chance for Capitals to have magic sites other than those you start with? In other words, do you search your capital for magic sites?

Re: Site Searching Strategy
 

No and no.

But they can get a site via a random event.

Re: Site Searching Strategy
 

Can someone please make a suggestions for a new algorithm - that make a better default target for a site search spell? Then we can hope that Johan will implement it and include it in the next fix...

Make it prefer areas that:
- is'nt close to enemy borders (if possible)
- has'nt got any located sites yet
- start with terrain that will favor your capitals mages
- has been searched by other magic types, but not located any sites yet
- etc.

A better algorithm would save a lot of boring checking time, especially on large maps...

Re: Site Searching Strategy
 

I'm quite sure this is incorrect; they are just plains, as to site frequency, and I believe as to allowable sites... ie. a site that could only occur in a mountain province could not occur in a border mountain province. I am not sure about the latter, though.

Re: Site Searching Strategy
 

I think they count as mountains in terms of allowable sites. No idea if they have the same site frequency as mountains. They count as plains for movement.

Re: Site Searching Strategy
 

border mountains, or "edgemount," is its own terrain type. If mountain allowable sites can occur there, the database should be updated with at least an annotation.

I still suspect edgemounts count as plains, though.

Re: Site Searching Strategy
 

Dunno. I find a lot of Earth sites in border mountains.

Re: Site Searching Strategy
 

If province has "Mountain" mark in description - it`s mountain, sure :) If there`s mountains on map, but no "mountain" terrain type in description - you may surprise, but it`s...right, it`s NO mountain :D
Also I always check first - with any type of easy scan, which is possible to my mages (e.g. Auspex, Gnom Lore etc.), and then (late, with good gem income, which I can trade to pearls) scan with Acashic Record. Some rare and powerful locations possible to detect only with ARec - and just one site with recruitable Trolls, Crystal Castle, Loremasters or powerful magic bonus can change balance to your side in no time

Re: Site Searching Strategy
 

There aren't be any sites only possible to detect with Akashic record.
There may be sites your nation doesn't have the paths to find without using it, but that's different.

Re: Site Searching Strategy
 

So what I must do if my nation cannot detect those sites, or my mages have limited skills, or skilled in just 2-3 paths of magic? Strategy of searching "Cheap scan in early game > ARec in mid-game" is quite easy and effective
First you get just 2-3 types of gems, but you can use them right now. Late - with ARec you receive different types, so you can improve your own mages or hire neutral specialists (for me - now Marignon longplay MP - it`s Jade Sorceress/Enchantress for Nature and Water paths, Necromancers for Death and I`m lucky to find Silver Order so I can use Astral and Air magic)