Mabuse

forget AOM - take AOK-AI as standrad - there the AI know nothing, unless scouted Ideas ? Plenty. I could go evenin detail about some things But no hope that they will be implemented, or somebody listen to me most imortant is that we (we= interested scripters) get a raw frame how the AI-language looks like, first basic commands, some simple example scripts, etc. show us what you have so far. then we can add more and more commands and details how things should/could work - i could start right now working/developing CONCRETE Ideas

awesome btw, since my newer machine can take, you might as well add some more POP probably by selection (as in AOC) or with new techs or whatever -

btw, for the sake of a good and blanced gameplay, i would not nessessarily FORCE a Roman Player (for example) to repeat the failures of history - if there is automatic defeat when a possible persian enemy can use its cavalry it is probaly historical accurate (at least if you want to replay the battle of 'Karrhae' or whatever it was) but it is probably not good for gameplay historical accuracy is good, but gameplay is more important

But i know that the reason of all this "i want to have a learning ai" because the people are not interested in scripting and making thoughts on their own, instead they think they run a fe games with ai and then its perfect. man if this owuld work like oyu think, then somebdy would hve scripted this learning ai and it would be used in every RTS game, becasue the ai could simply learn it - a MOD with new units or differnt stats is nothing less than a differtn game. and if its not completely different than you would just need to ADAPT your current ai - but oyu imagine to play a few game with your ai and then its getting better and better - forget it

its anyway not working like you imagine

or better said, what you deccribe is atndard in almost all ai. my ai train archers and attack the enemy (perhaps with few units to scout) - and see that enemy trains Cavalry - then my ai adds spearman to the mix (assumed spears do good versus cav in the game of course) nothing special, nothing learning.

sounds strange. surely we miss each other. what you describe fit into my imaginaion of smart scripted ai, and not a "learning ai" script it for yourself and we are fine. it has nothing to do with learning things like: ---------------------------------------- "It will also have learned earlier that planes are good against men" ----------------------------------------- should be known to everyone before the ai is scripted so there is no more need for the AI to learn this. but no need to discuss it anymore, surely we completely miss each other

i don`t think you know what i am talking about - hell, i guess yu even don`t know what you are talking about. but anyway - feel free to explain it more - yeah, oyu want that the AI can learn, for example it planes don`t work it should use ships q1: how does the AI know that planes didn work ? q2: (after q1 is answered): how do the ai know that a cerain thing was responsible for a defeat q3: how can the ai calculate something more effective ? q4: how can the ai perform it ? q5: why does the ai then not play the right game from beginning on ? q6: what are you good for then ? and so on and so on - don`t know what you mean with laerning - i can only imagine some very basic parameters to be saved if the map is restarted. that would be learning that i support and can imagine to be implemented in short time

and so on and so on ... == and change mimiltary buildings, economic distribution, upgrades and so on - and i say it again: I cannot imagine how this should work anyway

At least i hope that this learning feature can be disabled, becasue i don`t want my AI learn things, and the next time i start it, it do things i did not tell it

i cannot imagine how this should/would work in reality , but maybe i have just too little imagination and beside that: Why does the scipter of the AI no know that ships don`t work vs Invsion from Ships (assuming your example is right) how will the Ai recognize that and so on and so on ...

I just hope, when it is finished, i just don`t feel that there is a real lack of playability and balance and i hope that the designers of the game make a goof job in terms of playability and balance - and play RTS (or have expierience with the play of RTS) on a high level, so they know what the important points of an RTS are

btw, i cannot think of too many things that would make many differences thogh starting buildings can be ALWAYS checked per MAP-Cheat (or in AOC cc-cheat) to evaluate the starting situations, something that a HUman anyway knows - if he plays an SCN always. so starting situations of the enemies and othe rplayers is of course somehtign that will influence your startegy and knowing them early (as a result of "learning") will help to set up a successful strategy (an Ai specially written for a cerain SCN will of course also take that into account automatically - so if i say that this cold be of use - i mean of course for an "all purpose ai") and then position of certain things like resources will help to get them if they are needed. also "choke points" could be somthing to be remembered. (anyway just optional)

I am very sceptically looking at the "learning aspect" not becasue then i have nothing left to do, instead, because i have expierience these "my ai will learn" thingy is more complicated than "simple" scripting. Why not make the AI use a (several) strategies that "work", and also try to react upon certain things. Then a nice diplomacy sytem that make 8 Player ("free for all") a bit more interesting and it is ok. To make an AI that really learn for itself will not happen within the time-lmit this game should be finished There are some things like it the SCN/MAP is restarted the AI has "saved from the previous" Game the position of the (warning AOC example) "sheep" or other things that are required as soon as possible. to make AI "adapts" to the tactics of the Human it plays against is either: nonsens, or simply a good AI that is able to react upon certain situations nonsens, becasue a working strategy wil already take some things into account, and the things that could "counter" it should be evaluated and if they happen the ai need to react upon that. a learning AI (what will the AI learn ? How to play the game ? Why not tell it how to play the game ) anyway i am sceptical upon learning AI Usually scripting beginners hope to make their wok shorter by let the AI do all the work they do, but unfortuanately (for them)- the AI are far from being able to "learn" anyway like i said, some things could be worth to be "remembered" (saved) when the Máp is restartet (so we need to think about which information would change our way we play if we would knew them in advance) but is only an OPTION no nessessity. First basic things need to work.

Hi !!!! realizing the Danger of getting on your Nerves (ope that makes some sense in english) i just ask about some Progress or just a kind word like "everything is alright" i think it is not too easy what you are aiming on and though i am sure that "too many cooks will ruin the soup" (or something like this ) - i am of course willing to give suggestions or judge about some things (to judge is fun - ) also of course of interest is how the scripting-language (if there is one) wil look like - something like - "it will look like Javascript" makes no sense for me as i don`t know "Javascript" so if somebody could slowly introduce the syntax of the language and maybe present some Ideas - well, i would find that entertaining - of course nothing need to be complete or something, just few notes etc etc and we ask then stupid things - that would be nice

i think it is a good decision, btw i mean someone may doubt that this lead to anything, but that OS is the answer and leads to something (ok - something == the ame that 0.a.d. should be) is, errm, very optimistic however, i want to mention that i probably not agree with all things in the desgn document - but may need to take a fresh look at it to express this more losely, to be honest

btw, i would nevertheless make the caalry strong, btw. or better said appropriate to their strengh it had. That the french knights were slaughtered at agincourt in muddy terrain with their heavy armored Horses and men was an exception and does not mean that hvy cavalry was weak in the middle ages. Maybe some Terrain will be more difficult for Cavalry And the armor was superb. Not the bow or crossbow ended the knight area, the gun-power weapons which did damage through heaviest armor made the hvyknights (but not cavalry) useless. now i don`t know much about the time around 0.a.d. perhaps the cavalry in these days was usually more or less LIGHT acvalry, not very armored and vulnerable to many things

no. would like to check it out though. But don`t ahve the game. However - in my suggestion there is no stamina that go down while moving - instead there is a charge-value that go up while moving - and is a certain Factor that is added to the attack Vlaue (perhaps) if Full 1.5 (or even 2.0 or higher depend maybe also on the class hevier units get more )) - if zero 1.0 - for example) the charge move is maybe restricted in a certain direction, so changing direction by 180 degree may set the Valaue to 0 (1.0) again of course this can be done also easier - in any way should the attack out of movement provide a nice bonus for cavalry. then while standing they fight with normal stats. if the player uses (extensive) micro he may use this to his advantage, and even master this certain unit type. if it is worth it to care less of other areas (eco, other units) in this situation

btw, as i just read this i remembered that the Lt Cav (unarmored) in AOK lose vs Swordmen at least if it has no Bloodlines whatever ... about my moving Cav-Bonus - i meant it that cavalry should have a general (charge)bonus if it moved a certain Time and deals damage then. after dealing the damage the (charge) Bonus is used up and the Cavalry can stand and fight or move to fill this Bonus again. how this may look in practice i cannot judge atm, maybe the Cavalry charge, and the player pulls from time to time some cav out, with the time the players get used which distance they have to chose which makes the Horse get the bonus back, but still near so that it will join combat on its own This is then a form of micro which makes the cavalry better - or the whole Cav is chosen and retreats to charge wit full strengh again but maybe this is also crap At least i offer my abilities for the unit "balance" section - maybe i can help in the planning stage a bit

"easy" solution for "moving Cavalry" : simply force the player to move them on their own btw, if there is a charge formation for cavalry (should be avialable for all cavalry this charge formation should have certain adantages for Horsemen - but this formation will be dismissed after a certain time when the formation had contact with enemy units - or is activated by default whenever a mounted unit hits a Unit while moving - crush damage ?) after that the caalry will fight with their regular stats which are weaker. so for anohert successful attack, the caalry have to move on (retreat or ride thrigh the enemy line, reformate and charge again) keep em moving, if the enemy unit group is weakened enough after the charge, the cavalry may mob them simply up of course. sometimes after the charge the cavalry wil be sourrounded, so that retreating or riding through the enemies will be simply to much casualties, so that stand+fighting is better solution also may depend on sitiation, the best attack of cavalry, the hvy charge shopuld in any way bring great bonussess with them. perhaps other weapons like spearman, may neglect this cavalry ability it will be intersting to simulate that. But i agree with TitusUlthor that the use of cavlary may strongly depend on Situation- so if your goal is to simulate this, you should try to implement something in this direction keep em moving, if possible nice to think about it. btw, pls try to go for another POP cap than 100 or 150. I mean why unpopluar 3d in a RTS instead of thousands of units ? We have the Year 2005, and if your 3d does not support immense Details, it shoul dbe able to handle a few more units, and if even this is not possible by the game-engine - i must say to be honest, i would have gone for 2d instead. 2d offers enough advanatges, and real RTS Fans will prefer more Units and 2d

Sure. But i can tell you also many other Situations where hvy Cavalry owned the Battlefield Which was quite more often than Agincourt. Also in later Times, the defense of Cavalry from ranged units was always important. Ranged units have not the ability stopping Cavalry that charge at them (if not extreme firepower or significant superiority in numbers) Also one thing about Armor, the fact the knight were mounted allowed them to wear Armor that would be impossible for Foot-Units - even the Horses were strongly armored in Cases. And due to that much more weight and higher trample damage. Now in the Time-Line that 0ad features that may be something else though. For example RomanCavlary may be vulnerable to archers, but if the Cavlary can reach the archers they may do great damage to them. and due to the fact that they are FAST ....

btw, if archers hit always i could also live with that. i have no problems with some restrictions in reality-factor. compensate it with lower fire-rate and thats it and now i apaologize for posting 3 times a row EDIT: ups it was 4 times. Now i revealed that i can only count to 3. EDIT2: Btw, before now someone tells me that in reality some cavalry was vulnerable to archers (if they were not over them of course ) then i agree and just say that in AOK the archers use theri bow also in Melee and thats also not realitic, or that Horses can`t trample theri opponents which is also not realistic - so many unrealistic things and i still love the game, while other went straight into the dustbin

now thats only my opinion - most of the things are standard anyway, just like attack bonus

My coments: - Line of sight is a useable concept - Infantry may have formations which make then more resistant - standard and useable - archers don`t automatically hit heir target, but may traget the point a possible target would be at when moving with same spee in same direction ? come on, its standard - arrows realistic affected by gravity ? they may shoot in a bow idf the target is more distant - thats it. - height advatabge ? standard. - forests with advanatges and disadantage ? Nice Idea - but anyway standard, if the forest are walkable (btw you should álo not be able to make certain formatiuons if the place is not enough for them - standard ) - Bonusses for cavalry while in Formation. Standard for any formation. if it won`t be a bonus then they would be useless - and in fact formations were used becasue they offered advanatges - standard - Morale ? Now you want it all realistic, but hate Mass-Panics (which occured IMO), now you have to decide if you want to insist on TOTAL REALISM or not. If yes, then mass pAnics are also STANDARD - wow, now you hate MASS-Panic, but Horses standing still in Melee would be complete unrealistic or whta. Lets agree about a compromise - we cut out the Mass-Panic and then you just imagine that a Horsemen hacking at its opponent and standing still, would be riding around in realtity, but just not displayed in game. agree ? OK.

Now in AOK (which is a very balanced game, and due to that very good) a spearman don`t win "always" vs a knight for example - in fact they lose vs it in 1v1 - put are able to beat em with 2 or 3 vs 1 while being cheaper. also Cavalry has no Attack-Bonus vs Archers, they have instead a basic Pierce-armor. (and are fast) Btw, the Pierce-armors and bonusses are excellent balanced in AOK (ok i repeat myself here ) In my Opinion an Attack-Bonus for vcertain UNits vs certain UNittypes is IMO no bad solution, becasue you now may give spearman high crush armor, so that they can stand vs cavalry (if cavalry make crush damage), but this would basically the same . Not much difference if i give spaerman an attack bonus so that they are better vs knights than vs other units, or i give em a high Armor vs cavalry attacks - nope ? (in fact it would be unrealistic to give em armor instead of attack-bonus ) So basically i don`t mind that "simple" RTS concepts, especially if i think about that spearman have in fact an attack-BONUS vs cavalry - and surely not crush-armor The Spear is good vs cavalry (damage), not the fact that a spearman carries a spear make it more resistant to crush damage anyway, if i see people talk too much about realism it makes me flee anyway