Questions about Tech Tree Research Mechanics

We’ve moved over to the paradox forums. Please come visit us there to discuss:
You can still read the collective wisdom - and lolz - of the community here, but posting is no longer possible.

Home Forums Age of Wonders 3 Discussions Questions about Tech Tree Research Mechanics

This topic contains 146 replies, has 18 voices, and was last updated by  Jolly Joker 7 years, 7 months ago.

Viewing 30 posts - 1 through 30 (of 147 total)
  • Author
  • #138744

    Jolly Joker

    What I don’t like with the Civilization series is the fact that you actually HAVE a tech tree and thus can plan your scientific advance – although you shouldn’t be able to discern the fastest way to a crucial Industrial Age tech in the Old Age.

    That said, it’s somewhat difficult to plan when you have no idea how things work at all. I would like the Devs to help here, since this is a bit frustrating.
    There is a maximum of 12 Techs at any time from which to pick – which is where it starts because it actually may be less than 12 in the beginning. So that would be:

    QUESTION 1: Why may it happen that we start research with less than 12 techs available?

    Which leads to:

    QUESTION 2: How are the starting techs in your Research book picked? (There must be some restrictions on costly techs.

    Which leads to:

    QUESTION 3: How are the Techs picked you already have researched? (You start with 2 Strategic spells and 3 Combat spells already researched – how are those picked?)

    (Note: when I give Examples, the Class in question is always WARLORD.)

    So there 12 Techs to pick one from for research. But how are they picked?
    Techs are divided into 3 categories: a) Empire Upgrades, b) Strategic Spells, c) Combat spells. Specialisations seem to have a big influence on how things proceed. So to flesh this out I will make an “example list” for
    a) Warlord Class Techs
    b) Avatar Techs (techs in your research pool no matter your picks)
    I will add two configurations for specialisations, since they are very different:
    c1) Specializations: Expander, Explorer, Partisan
    c2) Specialisations: Air Adept, Fire Adept, Fire Master

    a1) Warlord Class EMPIRE UPGRADES, Tech Level:
    Produce A, B, C, D, E, F (6 production techs, based on each other: possible prerequisite to start with: Raise Militia)
    Warfare I, II, III, IV, V, VI, VII (7 CP enhancing techs, based on each other)
    Training Regimen, 1
    War Effort, 2
    Garrison’s Honor, 3
    Warrior Culture, 3
    Conqueror, 4
    Martial Arts Training, 5
    Thoroughbred Mounts, 5

    a2) Warlord Class, STRATEGIC SPELLS, Tech Level
    Authority of the Sword, 1 (may be starting spell)
    Raise Militia, 1 (may be starting spell)
    Death March, 2 (may be starting spell)
    Dread Siege, 2 (may be starting spell)
    Inspire Loyalty, 3 (may be starting spell)
    Hero Slaying, 4
    The Draft, 4
    Conqueror’s Feat, 5
    Global Assault, 7

    a3) Warlord Class, COMBAT SPELLS
    Last Stand, 1 (may be starting spell)
    Lion’s Courage, 1 (may be Starting spell)
    Berserker, 2 (may be starting spell)
    Shout of Intimidation, 2 (may be starting spell)
    Phoenix Warrior, 3
    Killing Spree, 4
    Steadfast Ward, 4
    Bloodbath, 5
    Relentless Army, 6

    b1) Avatar Techs, EMPIRE UPGRADES
    Basic Seafaring, 1 -> Advanced Seafaring, 3 -> Advanced Logistics, 5
    Greater Disjunction, 5

    b2) Avatar Techs, STRATEGIC SPELLS
    Terraforming, 2
    Resurrect Hero, 4

    b3) Avatar Tech, COMBAT SPELLS
    Dispel, 1
    Banish, 4

    c11) Specialisation Techs, Expander/Explorer/Partisan, EMPIRE UPGRADES
    Fertility Rites, 1
    Scout Training, 1
    Folk Hero, 1
    Rugged Pioneers, 2
    Off the Beaten Path, 2
    Guerilla Tactics, 2
    Expansionism, 3
    Trail Running, 3
    Swift Migration, 4
    Partisan Army, 4

    C12) Specialisation Techs, Expander/Explorer/Partisan, STRATEGIC SPELLS
    Hide Out, 2

    C13) Specialisation Techs, Expander/Explorer/Partisan, COMBAT SPELLS
    War Anthem, 3

    c21) Specialisation Techs, Air Adept/Fire Adept/Fire Master, EMPIRE UPGRADES

    c22) Specialisation Techs, Air Adept/Fire Adept/Fire Master, STRATEGIC SPELLS
    Domain of Winter, 3 (Starting spell?)
    Summon Zephyr, 3 (Starting spell?)
    Domain of the Sun, 3 (Starting spell?)
    Summon Hell Hound, 3 (Starting spell? – Yes)
    Tropical Empire, 4
    Summon Fire Elemental, 5

    c23) Specialisation Techs, Air Adept/Fire Adept/Fire Master, COMBAT SPELLS
    Suffocate, 1 (Starting spell?)
    Fireball, 1 (Starting spell?)
    Seeker, 2 (Starting spell?)
    Skin of Oil, 2 (Starting spell?)
    Fire Halo, 4
    Hellfire, 7

    What do we know?
    We know, that the Avatar Logistics, the Warfare/CP-increasing and the Produce/Summon Class trees will block 3 of the 12 tech slots, which leaves 9, and I suspect, that these are divided 3/3/3 between the 3 different tech categories.

    QUESTION 4: Is there a pre-determined division of tech slots?

    QUESTION 5: How does substitution work, when aTech is researched, in terms of tech level?

    Now, let’s for the 2 hero configuration examples (Expl, Expa, Part and Air, Fire, Fire) sum up the techs aside from the 3 listed above.

    Configuration 1:
    COMBAT SPELL, 1: 3
    COMBAT SPELL, 2: 2
    COMBAT SPELL, 3: 2
    COMBAT SPELL, 4: 3
    COMBAT SPELL, 5: 1
    COMBAT SPELL, 6: 1

    Configuration 2:
    COMBAT SPELL, 1: 5
    COMBAT SPELL, 2: 4
    COMBAT SPELL, 3: 1
    COMBAT SPELL, 4: 4
    COMBAT SPELL, 5: 1
    COMBAT SPELL, 6: 1
    COMBAT SPELL, 7: 1

    Comparing the 2 Configurations we can say that the most important differences with a view on game is that the difference in Strategic Spells is not that big – a spike is in Tech Level 3, which Configuration 2 gives us 5 instead of 1 for config 1, and 5 Techs more all in all.

    The big difference comes when comparing Empire Upgrades and Combat spells, especially on the first 5 levels. We have 18 Empire Upgrades and 12 Combat spells in config 1 and 8 Empire Upgrades and 17 Combat Spells in config 2.

    QUESTION 6: With such big differences in tech structure possible, does the tech selection for the 9 remaining techs in the research book reflect the differences or not?

    I mean, in config 1 there are 18 Empire Upgrades, 12(10) Strategic spells, and 12(9) Combat spells; In Config 2 there are 8 Empire Upgrades, 17(15) Strategic spells and 17(14) Combat spells.
    So in config 1 you SHOULD get 4 slots for empire upgrades, 2 for Strategic spells, 2 for combat spells and 1 random, reflecting the spread. In Config 2 you SHOULD get 2 slots for Empire Upgrades, 3 for Strategic spells, 3 for Combat spells and 1 random, reflecting the spread.
    However, I have a slight inkling, that the 3 specialisations simply block a slot each.
    This would make it:
    1 Avatar Logistic Tech slot
    1 CP-increasing Tech slot
    1 Produce/Summon Tech slot
    1 Specialisation slot (if researched – randomly filled?)
    1 specialization slot (if reasearched – randomly filled, if Master in same tech filled with that)
    1 specialisation slot, if no Master picked
    leaving 6 to 7 slots for the rest.

    QUESTION 7: How is this working in reality, and should it maybe work different?

    So can be bring light into this?


    Jolly Joker

    Corrections and additions:

    Pertaining Questions 1 and 3:
    You always start with the free-of-cost empire upgrades of Avatar Tech and Specializations, but with at least TWO (which may include ONE costing research).
    You start with TWO Combat Spells.
    You start with TWO Strategic Spells.

    Initially the Research book contains NINE techs:
    Avatar Logistics Tree (Empire Upgrade)
    Produce/Summon Tree (Empire Upgrade OR Strategic Spell)
    CP-Increase Tree (Empire Upgrade)
    2 Empire Upgrades
    2 Strategic spells
    2 Combat spells
    For the first 3 techs researched, you’ll get TWO new techs into your book, so that it shows 12 after you’ve researched 3 techs. I think, you get what you researched: if you start with a combat spell, you’ll get a replacement, plus a new slot filled with a Combat spell.
    This leads to ending up with
    3 slots for Empire Upgrades
    3 slots for Strategic Spells
    3 slots for Combat Spells.

    Problems with that:

    Your tech book will NOT reflect the specialisations you picked! If you picked Elemental Adept/Masteries, your tech tree contains more spells; if you go exclusively for Explorer/Expander/Partisan stuff (maybe with one Adept), you will have more Empire Upgrades, which actually means:

    The Empire Upgrades of Explorer/Expander/Partisan specialisations clutter your tech tree and may delay the appearance of your (important) Class Empire Upgrades!

    With the current situation it wouldn’t make sense to pick more than one specialization with Empire Upgrades, therefore. I can live with that, but it should be communicated.

    Also, the question is, whether a different solution might be better.

    Starting Techs include Tier 3 techs!

    That simply shouldn’t be the case! Imo, you may start with a T1 Combat spell and a T2 combat spell and with a T1 Strategic spell and a T2 Strategic spell. I mean, you do not want to start with Inspire Loyalty, simply because you don’t need to convert Mana to gold at start.



    1. You start with 9 skills because it was felt that new players might be overwhelmed with choices when they opened the skill book and saw so many things at once. It was decided to make the starting choices a bit more limited and have more choice open up later to counter this.

    2/3. The skills are marked with a flag “Can Be Starting Skill”, which is set to true for most spells at tiers 1-3. The system collects all the skills with that flag set to true that the player could research, given their class and specializations, and gives them 2 tactical spells and 2 global spells at random.

    4. Yes. The system starts off with:

    3 fixed slots (Casting points, Seafaring, Units)
    2 for empire upgrades
    2 for global spells
    2 for tactical spells.

    This quickly turns into:

    3 fixed slots
    3 for empire upgrades
    3 for global spells
    3 for tactical spells.

    As the fixed slots are exhausted (e.g. You research all the casting point upgrades), you end up with

    4 for tactical spells
    4 for global spells
    4 for empire upgrades

    The 4th slot is unlocked in that order (so the first fixed slot you clear will give another tactical slot, the next fixed slot you clear will give another global slot and the last will give an empire slot).

    Also, in each group the system always tries to guarantee that there is at least 1 class skill and 1 non-class skill (i.e. 1 specialization OR 1 avatar skill). So, you should always have at least 1 tactical spell to research from your class, and 1 non-class tactical spell to research, with the 3rd being either class or non-class.

    If the system cannot find a good skill to go in a slot, then it will randomly choose something else and put that into the slot instead. When that happens, almost anything can be chosen for the slot.

    5. When you research something, you free up a slot and the system will try and find something that will fit in the same slot. In general this means that researching global skills will unlock more global skills, and researching tactical spells will unlock more tactical spells, etc. It also means that researching a non-class skill will usually unlock another non-class skill.

    The algorithm that the system uses to unlock new tiers is a bit more complicated:

    1) The system adds all the unresearched tier 1 skills of the needed type to a list
    2) If the list has less than 3 things in it, then all the unresearched tier 2 skills of
    3) If the list still has less than 3 things in it, then all the unresearched tier 3 skills of the needed type are added to it.the needed type are added to it.
    4) And so on
    5) One thing is picked from the resulting list and added to the skill book.

    6. I’m not sure what you mean, hopefully you can answer your own question from the answers above? I’ve pretty much listed all the logic the system uses for skill picking, I believe.

    7. There’s a bug in the system right now that means that 1 randomly chosen tactical spell will disappear off the list and be unavailable (the glitch causes the system to accidentally put 13 things in the skillbook when there’s only space for 12). It’s already fixed for the next patch. Apart from that, that’s something for you guys to answer.



    This is pretty useful information for the Wiki….



    This is pretty useful information for the Wiki….

    It’s like a gold mine really.

    Starting Techs include Tier 3 techs!

    That simply shouldn’t be the case! Imo, you may start with a T1 Combat spell and a T2 combat spell and with a T1 Strategic spell and a T2 Strategic spell. I mean, you do not want to start with Inspire Loyalty, simply because you don’t need to convert Mana to gold at start.

    I usually found this kind of pointless to have the T3 Global spells available as they are not useful unless you have a really good start. But it does save you the time researching it once you can use it. Obviously, this is more useful for some T3 spells than others.

    2/3. The skills are marked with a flag “Can Be Starting Skill”, which is set to true for most spells at tiers 1-3. The system collects all the skills with that flag set to true that the player could research, given their class and specializations, and gives them 2 tactical spells and 2 global spells at random.

    Just wanted to mention that there are even higher tier skills that can be started with. It might only be one, but Hallowed Domain is still a possible starting skill for Theocrat



    If you guys see a spell turning up as a starting skill and you think it shouldn’t be one, tell me the name and I’ll see about switching it off.

    Edit: Switched off for Hallowed Domain, starting with a Tier 5 spell is probably a bit much!



    2) If the list has less than 3 things in it, then all the unresearched tier 2 skills of

    Which could significantly reduce chances of picking remaining tier 1 skills, couldn’t it? And more generally, this algorithm sometimes could make difficult for some classes to obtain low-tier skills of critical importance. Like, for example, Training Regimen of a Warlord. Is this correct?



    In before Lesser Elementals…heh.

    Actually, since we’re discussing the way skills are selected, would it be possible to add a weighting system to upgrades as well, instead of just selecting 3? I’ve mentioned before that Explorer pretty much can’t be run with Death Mastery, as its possible and often the case to pull Storm Magic before trail running.

    Also, possibly a bug but, sometimes when researching down one path exclusively for a long time, a certain tech category disappears outright. I’ve seen cases where there are no research options of one category until I exhaust a second category(which takes like 80 turns). Would it be possible to, when a skill is researched, before it finds another of the same kind, *instead* check for missing categories and then replace with one of those instead? I.e. There are no avatar globals. Even though you researched cunning retreat, it gets replaced with terraform to rebalance a spell book that is off kilter.

    Additionally, it really sucks when low techs get hidden besides high techs. Since it seems to only select from the bottom three, in some cases you’ll have to research age of deception before got get iron grip, for instance. It is mathematically unlikely if the system works as you describe, but still possible. I think if there is a gap of more than 3 tiers the spell book should say screw it and just throw up from what it found so far. If that’s only 2 techs, cool. If there’s only one tech left below tier 5, and its tier 1 make that guaranteed. Would be less random while still very random.



    It doesn’t only pick from 3 skills, 3 is the minimum number. If you have 1 T1 skill, 1 T2 skill and 12 T3 skills, it will end up picking from 14 skills, since it adds ALL the T3 skills to the list before checking the size of the list.

    Also, possibly a bug but, sometimes when researching down one path exclusively for a long time, a certain tech category disappears outright. I’ve seen cases where there are no research options of one category until I exhaust a second category(which takes like 80 turns). Would it be possible to, when a skill is researched, before it finds another of the same kind, *instead* check for missing categories and then replace with one of those instead? I.e. There are no avatar globals. Even though you researched cunning retreat, it gets replaced with terraform to rebalance a spell book that is off kilter.

    It already does this, so I guess it’s a bug. Do you have a save that demonstrates a whole category being gone? I’d like to have a look if I could.

    Which could significantly reduce chances of picking remaining tier 1 skills, couldn’t it? And more generally, this algorithm sometimes could make difficult for some classes to obtain low-tier skills of critical importance. Like, for example, Training Regimen of a Warlord. Is this correct?

    Sort of. The lower a tier a skill is, the more likely it is to be picked, since it will turn up for consideration earlier and thus have more chances to be potentially chosen. It is definitely possible that a T7 skill appears in the book before a T1 skill though, it’s just not very likely.

    We could add weighting to the system I guess, but that wouldn’t fix the issue you guys are having of a “Must Have” skill not appearing, just makes it less likely. The best thing I can think of is a fail safe like:

    – If we’re adding Tier 5 skills to the list, and there are Tier 1 skills in the list, always just choose a Tier 1 one.
    – Repeat for Tier 6&2, Tier 7&3

    At the end of the day though, this system is a min/max optimum-playstyle player’s worst nightmare. The randomness will always frustrate you if you decide you must have a particular skill by a particular turn.


    Jolly Joker

    Thanks for the quick answer, Tombles.

    I’d like to discuss this, because it would seem the crucial thing, the rest being ok with the exception, that we might discuss the feasibility of having T3 techs in the start mix – having only T1 and T2s would for example solve the Summon Lesser Elemental problem rather effortlessly. I have a couple of issues with techs either not giving me my low cost empire upgrades or feeding my combat spell slots with cheap spells when researching them would mean a massive waste of research points (you don’t want to research a 60 Tech when you have 120 Research.

    The algorithm that the system uses to unlock new tiers is a bit more complicated:

    1) The system adds all the unresearched tier 1 skills of the needed type to a list<br>
    2) If the list has less than 3 things in it, then all the unresearched tier 2 skills of<br>
    3) If the list still has less than 3 things in it, then all the unresearched tier 3 skills of the needed type are added to it.the needed type are added to it.<br>
    4) And so on

    What would happen, if “Less than 3 things” would be “Less than TWO things”?
    The tech tree would have the tendency to include the lower-tier techs and come up with the higher techs later. Would that be better? I guess so. Here are the reasons:
    If you look at the Configuration 2 player in my initial post, the specialization picks will leave Empire Upgrades Class only, except Greater Disjunction, but you will VERY soon have ALL Empire Upgrades in the list, which you actually don’t want – you can’t research Greater Disjunction or Thoroughbred Mounts early.
    In Strategic spells the specializations boost T3 techs which is solid, since this will somewhat delay T4 and higher techs to a time when it’s right.
    In Combat spells you have a lot of low level 1s which is ok as well.

    However, if you pick one or more specializations that come with Empire Upgrades, these are all LOW level. Say you start the Warlord with Explorer plus the two Fire Specials. In that case, Empire Upgrades for initial Techs will be picked from
    Training Regimen (1), Scout Training (1), and since that’s only 2 the system adds War Effort (2) and Off the Beaten Path (2). You will get 1+1, which leaves 4 possibilities with equal probability:
    a)Training Regimen + Scout Training; b) Training Regimen + Off the Beaten Path; c) War Effort + Scout Training, d) War Effort + Off the Beaten Path.
    This leaves only 2, so the moment you research any of those/ 3rd slot is opened, the T3s are added, Garrison’s Honor, Warrior Culture and Trail Running. My chances to get the remaining T1/2 Class tech now are pretty low, actually: If I research Warfare I, that will open the 3rd Empire Upgrade slot, which will give me a random tech from the 5 remaining.

    If however a new tier is opened up with ONE remaining tech only, things look differently. In this case, your initial pick will consist of 1 T1 tech as a matter of course. After that the 2 T2 techs are added, and another Tech is drawn. This gives
    Training Regimen + Scout Training (50% prob)
    Training Regimen + Off the Beaten Path (25%)
    War Effort + Scout Training (25%)
    Also, since two are left, if I research Warfare I, the free slot will be filled with one of the 2 remaining, after which T3s are added!

    Now, this is just an example, obviously, but it’s true for all the classes, and I would like something like a “test run” with that setup plus leaving T3 techs out of the qualifies as starting tech loop.

    Would that be possible?




    I don’t think I have one handy. Its rare in typical play, but I know the circumstances where it is most common, which is when you turn up research to ludicrous speeds and sit in one city only researching one category. I had weird things happen when playing with abilities doing that. I’ll see if I can replicate it tonight.



    What would happen, if “Less than 3 things” would be “Less than TWO things”?

    I think we’d need to have a lot more people say they want it before we’d do it, to be quite honest. As you’ve shown, it would have quite a large effect on how skills appear in the skill book, essentially it would become a fair amount less random, which is something that I worry many people would dislike. It would also slow down the appearance of high tier skills, which might not be popular either.

    I’ll have a look at what T3 things are actually marked as starting skills. I said it without actually checking if it were the case, so I might actually be wrong about them all being available.



    Thanks Garresh!


    Jolly Joker

    as you know, randomness is just another word for “playing the odds” – but that means, you have to know them – which actually happened now for the first time, since you can now calculate what effect the research of a specific skill will have.
    HOWEVER, and this is a really big however, the problem isn’t so much that higher tier techs appear, because even if they appear early you will have to research them first, giving a trade-off between early access and long research time, the problem is that lower tier skills do NOT appear, because often of the time their use EXPIRES or decreases in later game phases (Courtesan Ambassadors, for example), while others are DEARLY needed.
    Keep also in mind, that this would change only Early and Early-Mid game, since there is only one Specialisation of the 3 you pick that will add T4 and higher Techs, which means 2 of them will add T1-3 Techs, which you obviously want to have, otherwise you’d be better off without specialization. The second disadvantage of getting lowtier tech late is having to waste research points to research them.

    Anyway, when I say “test run”, I mean something like a closed beta test over a couple of weeks with some testers, comparing things, probably even with 2 game versions installed, if that is possible. I wouldn’t want to change this for everyone before it was actually tested and played.


    lesser elementals is definitly a candidate for “not allowed for free at the start of the game”
    whenever I start with it as a sorceror its GG, I dont have to research any class summons at all from this point, since lesser elementals are so rediculosly powerfull and easy to evolve to the T3 version, that I spam them alitte -> focus research and mana a bit-> go for empire upgrades and item forge -> now its turn 20-30, I have a super fast flying mounted mainhero with 3-4 T3 elementals on gold, super high vision, the whole stack invisible and stealth detection

    and all of this is only possible because of early “lesser elementals” which renders all other summons obsolete when I get it fast



    I’m inclined to agree with Jolly Joker(is hell freezing over yet? Actually maybe. Frostlings lol). This is the first game in the AoW series where we’ve extensive passive effects and where some techs are literally iconic and mandatory for a particular class. Even where passives existed before, they had their own side book and everything, so you could take casting specialist at start, then pick up channeler around 50 cp or so. I like the randomness. I’m one of those guys who prefers the regular RMG over settler because accounting for luck is part of strategy. But sometimes you get really screwed, and it reaches the point where certain strategies are mutually exclusive on the basis of how the research book reacts. And it actually somewhat hampers and limits strategic options, which is not food. Warlords especially know how much it hurts to not get key empire upgrades til later. I believe I remember BBB telling me a story once of him not getting the 20% cost reduction tech until after throughbred mounts, which meant he was screwed well through midgame. A rogue who doesn’t get iron grip has no real economic boosts if he doesn’t go for lesser stalkers treasure raiding, although bardspam is still okay synergy there. Anyways, there’s a lot more techs available this time around than in AoW2, and it makes the book feel very cluttered and erratic. I’d almost say bringing back a separate page just for passives would be a good idea, but that seems like trying to so surgery with a hammer. Still, part of the issue is that the percentage of all your researchable skills available to you at any given time is *much* smaller, and it creates some weird statistical irregularities more often than most players like.



    Ugh I’m writing this on a phone so I really don’t have time to figure this our, but I bet if you took the total number of research options available and divided by the total visible at any given time, and compared between AOW2 and AOW3, you’d see something odd. I’ll do math later.



    Very interesting this. The one issue I have with the spellbook is the fact that those lowlevel spells refuse to show up( like Training Regimen).

    Which is a real curse if it happens.

    For the rest the randomness is nice, once you know how to work with it.



    Low tier Key researches are to random for my tast and when they show up to late that isnt fun or does the system feel more exciting beacuse of it. A change would be nice.



    I’m a bit torn on the issue – in one hand, I’d almost want to eliminate all randomness, and for example be able to pick all “allowed” techs (conditions could be anything from “only 3 lower tier techs left in this category” to “research 2 techs of the previous tier, and it opens”). On the other, I understand that the current system is pretty good for its simplicity, which is always a good thing.

    As a note to all later posters: Lesser Elementals have already been untagged from the starting tech’s list.



    Yeah I don’t think any longtime fans want a removal of randomness. It may not be purely competitive, but you can really tell the good players based on how they react to changes and times when plans or prepared strategies fail. And for that matter, how they react to opportunities players who just memorize strategies might miss. Removing randomness is definitely bad in a game as deep as this, but its the extreme cases which cause concern, like when you don’t get your key passive until like turn 65.



    , but its the extreme cases which cause concern, like when you don’t get your key passive until like turn 65.

    or are unable to research dispel magic turn 80? I’ve had it. Seriously. And let me tell you, that tech is important.



    As to not completely remove the randomness, why not just change the probability distribution to favor the lower tier skills in certain situations.

    For example, once all of the T3 skills are added to the list, why not weight the T1/T2 skills left slightly more, and possibly increase the weight the longer they are in the list without being chosen.



    The minmaxing thing is something which I think is contributing to sorcerer and archdruid superiority, though. One of those classes can simply do well researching CPs and summons, so they’re fairly randomisation-proof, while other classes are more at the mercy of randomisation. So it probably is worth considering doing something to ameliorate this somewhat for some classes.



    Ha! So I sat down and tried to replicate some of those issues I was talking about. I never explicitly broke the skill tree, but I did find myself in a situation where I was sitting there with Air Elemental in my book before I got Freeze Water. Now I’m sitting here considering this, and this is the same problem I had with Storm Magic eclipsing Trail Running. But now that I think about it, they’re only 2 tiers apart. Interestingly enough, it seems like many of the most “crucial” and “iconic” kills for a given class or specialization are tier 3 or below. I mean, Corrupted Killers is only Tier 2, but Urban Cover is Tier 3 and costs 3x as much. If that comes first, it completely shuts down or heavily delays scoundrel based openings. Looking at Warlord empire upgrades, it’s easy to see how something like war effort can get pushed back almost indefinitely.

    That being said, I fired up Shadow Magic and compared total tech available between games, and was pleasantly surprised to have made a total fool of myself. AOW3 and Shadow Magic are almost identical in total number of spell options available at any given time. In fact, they’re so close it’s basically negligible, so changes in book structure aren’t really to blame for this. If anything, it’s more a case of so many “vital” techs being blocked by more expensive techs. I haven’t crunched numbers on how much costs vary between games, but I may look into that some more tomorrow.

    Perhaps leaving the same structure, but having an exception in the logic so that as long as one of the techs in the current listing is Tier 3 or lower, techs above tier 3 can’t show up if it costs more than 2.5x the RP of the lowest tech in the current selection? Basically, until you clear out the Tier 3 and lower skills, no tech will show up that costs more than 2.5x the lowest cost tech available to you. Once you finish this “crucial” techs, midgame becomes much more wildly random. In essence, it would still be quite random, but it would “squeeze” the available options down significantly, so if you’re digging through empire upgrades, eventually the game throws you a bone and says “FINE. I’ll give you your stupid war effort before thoroughbred mounts” But then it goes off and complains to its friends about how entitled these warlords are and how they shouldn’t invite them out anymore. It’s not ideal, but a lot of the issues with classes, particularly warlord, stem from these mandatory techs. I mean I get headaches enough when I try to mix Explorer with Death Mastery. I can’t imagine how warlords who play MP a lot feel about going with passive-heavy specializations…



    OK, since everybody seems to agree that this is an issue, I’ll implement a failsafe system. The algorithm will work as above, but will detect if the pool it’s creating to pick from contains both Tier 1 and Tier 4 skills. If it does, it will be guaranteed to pick a tier 1 skill. Same for tiers 5 and 2, 6 and 3, etc. So, the system will detect if you’re being unlucky with a particular low tier skill and give it to you.

    This will of course have to be tested, I’m not sure how well it will work in practice.

    The issue with most of the other suggestions is that they don’t fix the problems you guys are having, they just make it less likely to happen. Lower tier skills are already weighted to turn up more frequently, weighting them more won’t make the issue go away, it just means it will be less frequent.


    Jolly Joker

    For everyone – Tombles included – not wanting to read the details, the bold print is important. 🙂

    With the new Knowledge in mind, I fired up a game yesterday evening to try and “play the odds” with the Tech tree.

    Since I’m somewhat stuck with this class now 🙂 I fired up Draconian Warlord (the fact that Garresh talks about people playing ELF Warlord is an indication to me that people are on the wrong track with Warlord, especially when comparing it with Sorcerer).
    Anyway, for no other reason than avoiding all Summons 😉 I picked Explorer and Creation Master in my usual setting, 6 players L with UG, King diff, few towns, Many sites and stuff, with settling (settler paradise for AI), a bit less arctic and volcanic terrain, all classes present, all races different.

    I start out with AotS and Dread Siege on Strategic and Bless plus Shout of Intimidation on Combat.
    Starting Techs in the book: (Raise Militia – of course), Scout Training, War Effort, Death March, Terraformer, Holy Cure and Lion’s Courage.
    USUALLY, at this point I’d have started to play without thinking much, but NOW I wanted to know what would happen with the Tech tree – that is, what Techs were, at this point, in the pool for me.
    On Combat it was Last Stand, Berserker and Phoenix Warrior – which would mean, the moment the 3rd slot will open I will get one of those and T4 spells are added to the mix which will be Condemn Killing for Creation plus Killing Spree and Steadfast Ward for WL. That in turn means, when I eventually research a spell – Holy Cure comes to mind – I will get Condemn Killing next, while, if I go for either Lion’s Courage or the spell I will get in the new slot, will net me a 50% chance to get a T4 spell as well.
    Since it will take some time to reach 3-digit numbers in research with this setup, spells are certainly something that will have to wait, even though my spells are not great at the moment.

    On Strategic I have AotS and Dread Siege, and I can research Raise Militia, Death March and Terraformer. The other way round would have been better, obviously. What do I have in the pool? Inspire Loyalty for Class and Domain of Life plus Cleanse the Land for Creation. Since I started on temperate with Draconians I will need to change something here. When I research Terraformer I may get Domain of Life which will make it cheaper to make my people happy in terms of Mana, but I may get Cleanse the Land as well, which would be of no use at this point. Also, when the new slot is opened, T4 Strategics will enter the pool, and since I definitely will research both Death March and Terraformer early – what will happen? It will add Temperate Empire for Creation and Conqueror for WL Class, but that knowledge doesn’t gain. The new slot may be filled with a Creation tech, and when I research Death March, the replacement will be a 50/50 between Inspire Loyalty and Conqueror, the latter being out of reach for some time to come.

    Now for Upgrades, most importantly. Here the Explorer hits, and my Pool consists of Training Regimen, Warrior Culture, plus Garrison’s Honor for Class and Off the Beaten Path plus Trail Running for Explorer. I will get two out of the Pool. The additional slot randomly.
    I’m not entirely sure which new slot will be opened when I research Raise Militia, I hope that it’s an EU, but I’m not sure.
    When I start playing and find, that I have Scrolls waiting I decide to change the plan; Raise Militia isn’t THAT good at the moment – while I have enough Mana after finding a stash, I still miss Death March for fast Scouting, and the production of Militias will delay my town growth and put a strain on my economy. I would research Warfare I with Raise Militia and Death March, but in the given situation I decide to research War Effort. This will net me 10 gold each turn, but will also guarantee at least one Class tech from the pool to replace the tech.
    Since I don’t have anything better to cast I bang AotS onto my town, which nets me +20 pop for a Bountiful melon (I gain a node as well, cancelling out the cost; when the villiage becomes a town a gold mine will follow; at city I can cancel the spell).

    The next turns show that my position isn’t good. I’m on a peninsula; when I leave it into the only possible direction I find a hostile human town – which is not too bad, actually, because it means, I can actually use Dread Siege which costs 60, so I fire it up immediately.
    Meanwhile I research War Effort (3 turns) and get Garrison’s Honor for my effort. The new slot opens for Trail Running. Bit of bad luck involved, since I had a 50% chance on getting Training Regimen (2/3 for not getting it as a replacement for War Effort and 3/4 for not getting it in the new slot, 1/2 for not getting it in both). Still, nothing is added to the pool, yet, so I will have another shot at it.
    However, there is no way to force it. The way I set things up INITIALLY gave me a 2/3 chance of getting it into the book after researching the first two EUs, but that chance has halved now to a meager 1/3. Now, if I had taken Explorer/Partisan I’d have more cheap EUs, but now, I have only 3 more EUs not yet in the pool. I will get more 1/3 chances. Since Trail Running is in the book, researching Scout Training will see a random replacement (not necessarily an Explorer tech), so I will go for Raise Militia now and follow that with Scout Training and Warfare I.
    Raise Militia opens a new slot for Strategic that is filled with Inspire Loyalty. Meanwhile I take the town which is under Dread Siege and since it’s on temperate terrain I decide to absorb the Humans.
    I need a replacement for my stack since I lost a unit in the process, so I cast Raise Militia on the town waiting for its fate to be absorbed which actually works – bug or intended?.

    Scout Training opens a 3rd Combat slot and I get Last Stand. Also, Scout Training is replaced by Training Regimen, which is great. I do have the Lab in my main, and the conquered town has it built, so Research is at 44 now with the extremely happy humans under my domain, and when I make first contact with a Dwarven Arch Druid and his main stack consisting of 5 units 2 heroes included, I declare war and kill that stack. Things look good now. I will get Training Regimen in 2 turns and I can research the 140 techs in 3 turns (setting one town to producing Knowledge): Death March, Holy Cure, Terraformer.

    Now. What would have happened in that setup, if the new Tier would be activated for the pool with less than TWO techs remaining?

    For Combat spells the pool would be Last Stand and Berserker, Phoenix Warrior being added to it, when the new slot opened. In the actual setup you could get Phoenix Warriors followed by a T4 spell once you’d research Holy Cure. With the suggested setup it would be Berserker followed by Phoenix Warrior at “worst”, which seems better.

    For Strategic the pool is the same, so the new slot will be filled with a T3. At that point, though, T4s are added, so there is no way to make sure I get Inspire Loyalty. With the suggested setup, though, I can guarantee this: since there is nothing added to the pool, if Inspire Loyalty does NOT appear in the new slot, there will be a Creation tech. If I then research Death March as my first Strategic (after Raise Militia), this will be replaced by Inspire Loyalty, before the T4s are added.

    And what happens in EUs? Here the pool is different. It consists only of Training Regimen and Off the Beaten Path, so here suddenly the question arises, what does the system do, if it has to pick TWO techs, but will have to add to the pool after the first? Which one is picked first: when I research War Effort, both a replacement must be found and a tech for the new slot. If the replacement is found first, the system MUST pick Training Regimen, before adding T3s to the pool (this might happen with the actual setup as well, mind you), picking one for that at random. This would be the logical thing to do.
    In this case, researching War Effort first (a very unusual move) would guarantee me an early appearance of Training Regimen, while researching anything else would leave the new slot at random while going for Scout Training would guarantee Off the Beaten Path to appear at that point.

    Anyway: Tentative sum-up for Tombles, if this has been too long a read: The last bold print shows why I’m in favor for a test-run on reducing “less than 3″ to less than 2” (and to put T3 techs off the list of possible start techs as well).
    With the CURRENT setup, no matter how thoughtful your play is, you cannot guarantee anything; what is worse, imo, is the fact that you cannot tweak probabilities somewhat more in your favor. Playing for best odds, a 2/3 probability isn’t that great. If considerate play could tweak a 50% or slightly worse chance to an 80% or something I would be content – satisfied even.

    The suggested change on the other hand rewards thoughtful play: you get what you want, provided you play for it, which is as it should be.

    That would leave the question, whether planning would be too easy then, and not enough randomness involved – which is why I would like to actually test that: keep in mind that you still need a clear vision of what tech you want in every stage of the game to take advantage of the possibility to make halfway sure the techs you want WILL appear, while the current setup can easily screw you, even IF you know what you want.
    The Warlord example (people playing Elves) shows that it’s important to have THE RIGHT plan as opposed to getting the techs you need for your plan to succeed. IF THAT’S THE CASE, while a little randomness is ok, especially for replayability, randomness shouldn’t be TOO big, then, since there are enough chances to err anyway – you don’t actually NEED randomness to screw you – you can do that all by yourself. 🙂

    Lastly, when I pick my techs to research with the current setup (at least it was that way) my main considerations were: a) DO NOT WASTE RESEARCH POINTS and b) Try to research only techs you can get within 3 days; with the additional info, this changes, obviously, but I’m not really sure whether it’s really worth the effort to do so – which would be rather unsatisfactory.

    So I repeat my question: is it possible to set this up for a couple of testers the judgement of which to trust?



    I think the issue JJ is that the problem you have is different to the one everyone else has. The main problem I’m getting from the most people is that they don’t like that low tier skills can turn up extremely late due to the randomness of the system, and the solution I just proposed is designed to counter that issue. If we tried what you suggest, then the issue the majority are complaining about would still be present (though less likely).


    Jolly Joker

    @ Tombles
    I don’t think, the failsafe is a good move.
    Why is that? The failsafe actually makes it unnecessary to plan AT ALL: you can simply pick your techs with two points in mind: a) don’t waste research points and b) don’t research techs that would need more than X days to research.
    If you follow that, the failsafe will make sure the low-tier techs appear at some point, so there is no need to actually try and press something out of the system.
    The system actually plays YOU.

    Instead a hope to get a system where it actually pays to sit down, mull over the situation and the probabilities, decide for techs you want and need – and have plays to get them.

    This is SOMEWHAT possible with the current system, although, as I said, I’m not sure the result is really worth the effort. What I AM sure of, though, is, that it will definitely not be worth the effort with the failsafe you suggest. That will just allow the system to play you even more than now.

    So I would rather leave it as it is, instead of the failsafe – keep in mind we were missing all info, and the info allows better play.



    Would that actually address the key complaints though? Storm Magic is only tier 5, so it could still block trail running. I believe martial arts training could also show up and block a lot of low tier techs as well for warlords. If that doesn’t, thoroughbred mounts definitely could. Not in front of a computer, but some of the current skills are either very cheap or very expensive for a given tier, so…

    If that is to work, it may involve some tier shuffling too. Like raising storm magic and martial arts training up a tier. Possibly the thoroughbred mounts?

    Edit: I like the failsafe idea. I’m just not sure that basing it off tier, with such a wide gap in rp costs between tier 3 and tier 5, would really help much.

Viewing 30 posts - 1 through 30 (of 147 total)

You must be logged in to reply to this topic.