initiative - testing
initiative - testing
Welcome to my topic. You are going to do the work while I'm on the beach.
Now seriously, I can't really do tests myself because heroes5 runs slowly on my computer plus I'll be away now for a few days.
Still, I think it's high time we started to collect all and any data on this matter and test out how it exactly works.
I'll summarize what we need to know in the next post.
Now seriously, I can't really do tests myself because heroes5 runs slowly on my computer plus I'll be away now for a few days.
Still, I think it's high time we started to collect all and any data on this matter and test out how it exactly works.
I'll summarize what we need to know in the next post.
So I believe the intitiative works something like this:
At any point of the combat, every creature/hero has something that I will call progress.
It determines how far is he from taking his next turn, progress in my definition will be a number between 0 and 100 and I will denote it by p.
Initiative determines how fast the this number grows. I will denote it by i when needed.
The first problem is - what happens when the combat starts. One would logically assume that units start from zero each and the first to reach 100 acts first, etc...
This is not true: there is some randomness here. Like everyone would have some starting value which is determined somehow. It is not clear how, but this randomness shouldn't be so high.
The second problem is - what if two units have the same initiative? They still do not start at the same time (they can't act simultanously), but what's even more interesting: the difference between their actions seems quite noticable. You can try it yourself - if you split your blood furies into four stacks, the 4th one will keep having a high enough delay that it makes a difference: sometimes those zombies will act between two stacks.
It is closely related to the first question: do they start with a fixed amount of p/i] and that would be the explanation? Is that so simple? Also when two units reach 100 at the exact same time, which one will act next? And the loser gets delayed? For example by one frame? Is there a fundamental amount of p?
The third one is: what is the exact effect of certain spells/actions? I think slow and haste affects i only and not p. So, for example if you start a fight versus skeleton archers and your hero acts before them (by how many p?), it's already too late to cast mass slow... they are already about 95 out of 100 so they will still act next.
What about wasp swarm? Does it affect i or p? I think Divine Guidance gives a boost to p and should boost i too, but by how much? What about stun and freezing effects? Wait? Defend? Did I leave something out?
Testing some of these won't be easy, but I think it's important, since they didn't give us a progress meter, we only get to see who acts next and such, which is not so informative when we are about to decide whether we should cast slow or a DD spell maybe to stop that deadly enemy stack from acting before our blood furies? What will happen if those stupid waters get morale? Will they get to act twice before our hell hounds do? Will we still be able to run away? (hit and run fights)... and so on...
I hope the initial setup is easy in some way, for example they start by a fixed difference of p (5 units, first starts at 100, the 2nd at 99, etc).
That should be checked first. If it's all random in the beginning, we can't really hope to use it in combats effectively, which would suck.
At any point of the combat, every creature/hero has something that I will call progress.
It determines how far is he from taking his next turn, progress in my definition will be a number between 0 and 100 and I will denote it by p.
Initiative determines how fast the this number grows. I will denote it by i when needed.
The first problem is - what happens when the combat starts. One would logically assume that units start from zero each and the first to reach 100 acts first, etc...
This is not true: there is some randomness here. Like everyone would have some starting value which is determined somehow. It is not clear how, but this randomness shouldn't be so high.
The second problem is - what if two units have the same initiative? They still do not start at the same time (they can't act simultanously), but what's even more interesting: the difference between their actions seems quite noticable. You can try it yourself - if you split your blood furies into four stacks, the 4th one will keep having a high enough delay that it makes a difference: sometimes those zombies will act between two stacks.
It is closely related to the first question: do they start with a fixed amount of p/i] and that would be the explanation? Is that so simple? Also when two units reach 100 at the exact same time, which one will act next? And the loser gets delayed? For example by one frame? Is there a fundamental amount of p?
The third one is: what is the exact effect of certain spells/actions? I think slow and haste affects i only and not p. So, for example if you start a fight versus skeleton archers and your hero acts before them (by how many p?), it's already too late to cast mass slow... they are already about 95 out of 100 so they will still act next.
What about wasp swarm? Does it affect i or p? I think Divine Guidance gives a boost to p and should boost i too, but by how much? What about stun and freezing effects? Wait? Defend? Did I leave something out?
Testing some of these won't be easy, but I think it's important, since they didn't give us a progress meter, we only get to see who acts next and such, which is not so informative when we are about to decide whether we should cast slow or a DD spell maybe to stop that deadly enemy stack from acting before our blood furies? What will happen if those stupid waters get morale? Will they get to act twice before our hell hounds do? Will we still be able to run away? (hit and run fights)... and so on...
I hope the initial setup is easy in some way, for example they start by a fixed difference of p (5 units, first starts at 100, the 2nd at 99, etc).
That should be checked first. If it's all random in the beginning, we can't really hope to use it in combats effectively, which would suck.
Hmm, some my results that have not been available anywhere yet:
http://www.celestialheavens.com/admin/pitsu/ATBbar2.pdf (if you plan to check it more than once, please save into your comp, not downoad every time again!)
It is most likely not 100% correct, and can be more complete. If someone could come over his/her lazyness and test some of the spell/ability effects over (and add shield bash :p) i would be grateful.
http://www.celestialheavens.com/admin/pitsu/ATBbar2.pdf (if you plan to check it more than once, please save into your comp, not downoad every time again!)
It is most likely not 100% correct, and can be more complete. If someone could come over his/her lazyness and test some of the spell/ability effects over (and add shield bash :p) i would be grateful.
I doubt in that. The caster creature data i had to type in sunce otherwise i could not do non-linear regression curve fit analysis. There has not been any point to type in the details other tests. The most is could send is a photo of papers with notes and calculations, but since they are not made for presenting to someone else, i highly doubt you will understand anything from them. :p sorry!
I noticed that somebody was able to create a mod that let you save/load games during combat. Given that such modification is possible, is it also possible to make a mod that would actually display the accumulated initiative tallies somehow? Obviously that would make figuring all this stuff out more useful, but it would also add a new layer of strategy to the game, since it would affect the "smartest" action in a given situation.
For instance if I'm charging a neutral shooter would I delay its action more with slow, wasp swarm (with summoning), a freezing icebolt, a stunning lightning bolt, hasting one of my own shooters etc.
For instance if I'm charging a neutral shooter would I delay its action more with slow, wasp swarm (with summoning), a freezing icebolt, a stunning lightning bolt, hasting one of my own shooters etc.
Absolutely. There are issues that may be much more complicated than i managed to see with my limited tests. An independent person with different testing and analysis methods would probably be more successfull than i finding them out.csarmi wrote:I still think we would need much more data to tell it exactly.
I would not be surprised if the initial starting position is not determined solely by a random factor, and that it has another range than 0-25. Also the "track" length 100 may not be 100 but a function of highest and lowest initiative present on the field. I just haven't got a brilliant idea how to test and verify it.
EDIT: i was also trying to find console commands or "cheats" that say anything about the ATB process, but failed.
from some limited tests i would say guidance adds +30 (could also be +25 or 33, but 30 sounds more likely, since the developers seem to prefer factor 0.3 more than 0.25 or 1/3). Shield bash seem to reduce target counter to zero.Rainalkar wrote:Hi Pitsu, do you know how much does Divine Guidance "add" according to your data (like, Tele assault gives 50)?
Hello, amazing job there Pitsu! I was wondering what happens with the Blind spell.. does it freeze targets initiative like Master of Ice effect? What is that boost (if there is a boost) the Blinded creature gets to its initiative after it has been attacked? Sometimes it gets its turn almost rightaway, sometimes I get to attack with many of my creatures... I tested it some but I just couldn't figure out the mechanism. So...if anything regarding the spell is known I would appreciate it.. I love using that spell
This file is great!Pitsu wrote:Hmm, some my results that have not been available anywhere yet:
http://www.celestialheavens.com/admin/pitsu/ATBbar2.pdf
Hello,
Where can I download this file?
It isn't here https://www.celestialheavens.com/admin/ ... TBbar2.pdf
Thanks.
ps: sorry for my english
Where can I download this file?
It isn't here https://www.celestialheavens.com/admin/ ... TBbar2.pdf
Thanks.
ps: sorry for my english
Download the semiofficial manual:
http://www.heroesofmightandmagic.com/he ... uals.shtml
Under "advanced game mechanics" or somesuch should be a nice description of how initiative works.
http://www.heroesofmightandmagic.com/he ... uals.shtml
Under "advanced game mechanics" or somesuch should be a nice description of how initiative works.
Avatar image credit: N Lüdimois
Who is online
Users browsing this forum: No registered users and 43 guests