
Thread: Initiative Weirdness  This thread is pages long: 1 2 · NEXT» 

Gidoza
Famous Hero

posted April 19, 2020 08:42 PM 

Edited by Gidoza at 20:43, 19 Apr 2020.

Initiative Weirdness
Excel Sheet mentioned below here: https://mega.nz/file/8l8GGAAK#2a2YjxQaecUx5g6vXl3Yp6R5ByeA8hW69CI1jWJBU
Hey there,
So I'm continuing work on an old project where based on another poster I came up with an Excel sheet that estimates unit power levels pretty accurately for use in unit balancing. It's not perfect by any means so make what you will of it.
In any case, I have one particular conundrum and I'm hoping that someone with skill in either math or Excel can help me out with it.
Pay particular attention to the stats I've given to the Plague Zombie. The initiative on the unit is only 6, and I've ended up with an offensive multiplier of 0.35.
In short, a unit's raw power is calculated as a factor of Hit Points plus the Defence multiplier, multiplied by the Attack plus the Attack modifier, multiplied by Initiative.
It would be a simple equation really if Initiative didn't exist or if everything's initiative was 10. While just saying "HP times damage" is of course overly simplistic  for the most part, this system works.
Initiative, however  is weird. While there is a technical truth to the idea that a unit with 20 initiative is twice as powerful as a unit with 10 initiative, in practice this doesn't work: it comes down to the fact that highinitiative units can tie down more opponents, or get more hits before a unit can retaliate again, or back off out of the range of an opponent; on the other hand, a lowinitiative unit has the opposite problem...a unit with 6 initiative, for example, might theoretically move less often than a unit with 10 initiative, but in practice can sometimes never have a turn at all when it can reach or do anything useful.
I find that the issue doesn't seem to be as bad with increasing initiative above 10 as it is reducing it below 10, seeing as trending to 0 has an escalating effect  50% more than 10 is 15 (a 166% factor), whereas 50% less than 10 is 5 (a 200% factor). For each initiative drop below 10, the injuriousness to the unit escalates to the point where it just can't go further and be a useful unit. Consequently, I can't justify the use of anything less than 6 initiative on a unit.
However, here's the real problem: how can this be  or can this be at all  represented somehow by an equation? I initially drew attention to my Plague Zombie setup here (special abilities removed to be a plain unit): I ended up giving this unit a 0.35 multiplier (meaning the unit is 300% more powerful than it "should be") because any multiplier higher than that left it as a unit that could not accomplish anything: despite 6 initiative, which gives a 70% otherstats bonus compared to units with 10 initiative, the Plague Zombie just can't accomplish anything of value at 6 initiative. The scaling towards 0 just isn't enough to represent the actual power of the unit. So in the end, I have a 6damage, 40HP Zombie that still spawns upwards of 20 per week, and that's what it takes at 6 initiative for the unit to strike me as somehow valuable in a battle.
Can anyone tell me what system, formula, logarithmic scale, or philosophical insight is representing my experience? I hate having that 0.35 multiplier there because it's essentially arbitrary even though I know it's necessary because it represents my testing experience. I just can't understand *WHY* it's necessary in a mathematical sense  because the weirdness of initiative affects things other than just damage, like how many times a unit can retaliate in a turn. Perhaps it can never be represented accurately. Let me know if you have any insights.
P.S. In case it helps your maths, I should also note that the low (3) movement speed on the Plague Zombie is also influencing that low multiplier. Other units with slow speed (but not low initiative) also have a low multiplier, but nothing as low as this.
Here's the Excel Sheet:
https://mega.nz/file/8l8GGAAK#2a2YjxQaecUx5g6vXl3Yp6R5ByeA8hW69CI1jWJBU


dredknight
Responsible
Supreme Hero
disrupting the moding industry

posted April 19, 2020 08:57 PM 


Take what I say with a grain of salt as I dont have currently time to invest into math but having in mind the folllowing
 20 init unit act twice as fast as 10 init
 10 init unit act twice as fast as 5 init unit.
This if 10 is base initiative of a turn:
 +10 increases turn actions twice ( actions*2)
 5 decreases turn actions twice (actions/2)
This is why going down seems quite more impactful then going up the initiative scale.
Regarding formula estimation for power those are easy to calculate:
 Defense number delivers certain % bonus to hp
 Attack number delivers certain % bonus to damage
 I have not thought in depth about speed and Shooting ability but their addition to power is either a multipler of the power by certain flat value depending on the value of speed (or the type of range  no penalty? melee preference etc?)
As you say initiative stay aside from all of the above.
So as mentioned earlier more initiative gives the ability to act more, thus more opportunities to deal damage actively and passively through retaliation.
Have you thought of the variant that, after you calculate power value of all things except intiative, then to put that Power value on power of X.
Creature_power^x
Where X is value that depends on initiative. Some rough examples:
 10 init  1
 20 init  2
 5 init  0.5
So unit with 1000 power and 10 init will be 1000^1 = 1000 power
Unit with 1000 power and 20 init will be 1000^2 = 1000*1000 = 1000000 power.
Those numbers are absolutely ridiculous but I use such big numbers just to point out the exponential change which is also how initiative affect unit potential.
I am not sure I do explained it correctly but feel free to think over that and give your iteration if possible .
____________
Join our official discord channel  NCF Utility Beta


Gidoza
Famous Hero

posted April 20, 2020 01:54 AM 


dredknight said: Take what I say with a grain of salt as I dont have currently time to invest into math but having in mind the folllowing
 20 init unit act twice as fast as 10 init
 10 init unit act twice as fast as 5 init unit.
This if 10 is base initiative of a turn:
 +10 increases turn actions twice ( actions*2)
 5 decreases turn actions twice (actions/2)
This is why going down seems quite more impactful then going up the initiative scale.
Regarding formula estimation for power those are easy to calculate:
 Defense number delivers certain % bonus to hp
 Attack number delivers certain % bonus to damage
 I have not thought in depth about speed and Shooting ability but their addition to power is either a multipler of the power by certain flat value depending on the value of speed (or the type of range  no penalty? melee preference etc?)
As you say initiative stay aside from all of the above.
So as mentioned earlier more initiative gives the ability to act more, thus more opportunities to deal damage actively and passively through retaliation.
Have you thought of the variant that, after you calculate power value of all things except intiative, then to put that Power value on power of X.
Creature_power^x
Where X is value that depends on initiative. Some rough examples:
 10 init  1
 20 init  2
 5 init  0.5
So unit with 1000 power and 10 init will be 1000^1 = 1000 power
Unit with 1000 power and 20 init will be 1000^2 = 1000*1000 = 1000000 power.
Those numbers are absolutely ridiculous but I use such big numbers just to point out the exponential change which is also how initiative affect unit potential.
I am not sure I do explained it correctly but feel free to think over that and give your iteration if possible .
Thanks for the reply. You listed a number of considerations I've already made, but the one I couldn't really decide on was using the power value  so your comment confirmed that. I think that's the right way to go about it. How exactly that will pan out, though...I'm not sure. It's still pretty difficult and will require a lot of testing.
Regarding your instinct about ranged units  I'm actually sorry that I didn't save the original version of this, where I input all the original H5.5 unit values into here. It's from a combination of those and Vanilla values that I was able to weed out the multiplier of 1.4 for ranged units: basically, I looked at the relative power of units, and it was roughly a value of 1.4 that enabled a ranged unit to get the proper power values as established in the game. Quite frankly, more input from more people would help to refine a lot of the values I've used  sadly, many are arbitrary, because there's not exactly any objective way or sometimes any way at all to measure the effects of certain abilities, and a variety of numbers could be used to properly represent them. So while neither Vanilla nor H5.5 nor this can ever reach a state of perfection of balance because that is a perfect myth  nevertheless we could develop something against which to test numbers in a more thorough way.
Anyways cheers and thanks!
G


thGryphn
Promising
Famous Hero

posted April 20, 2020 03:24 AM 


What you should do is experiment with powers of the initiative value.
If you believe having initiative of 10 is more than twice as valuable as having an initiative of 5, you should include Ini^X as a multiplier, where X > 1.
Experiment with X=1.5, X=2, X=2.5, X=3, etc...
For example, with X=2, a creature with initiative 20 gets a multiplier of 400, one with initiative 10 gets 100, one with 5 gets 25.
In another look, this approach, assuming all other stats are equal, would set the power ratio between two creatures to (Ini_1/Ini_2)^X, instead of just (Ini_1/Ini_2).


dredknight
Responsible
Supreme Hero
disrupting the moding industry

posted April 20, 2020 09:06 AM 



Gidoza
Famous Hero

posted April 20, 2020 01:12 PM 

Edited by Gidoza at 13:13, 20 Apr 2020.

dredknight said: I would also suggest the NOIADE project by Nargott.
As far as I know it is the most comprehensive power calculator up to date.
Unfortunately the program just displays a ton of question marks for me???
Quote: What you should do is experiment with powers of the initiative value.
If you believe having initiative of 10 is more than twice as valuable as having an initiative of 5, you should include Ini^X as a multiplier, where X > 1.
Experiment with X=1.5, X=2, X=2.5, X=3, etc...
For example, with X=2, a creature with initiative 20 gets a multiplier of 400, one with initiative 10 gets 100, one with 5 gets 25.
In another look, this approach, assuming all other stats are equal, would set the power ratio between two creatures to (Ini_1/Ini_2)^X, instead of just (Ini_1/Ini_2).
Something I've noticed is that reducing initiative below 10 is much more severely sensitive than increasing initiative above it. Unless the factorization above 10 is mitigated somehow, it would cause some really weird results.


thGryphn
Promising
Famous Hero

posted April 21, 2020 01:08 AM 


Gidoza said:
dredknight said: I would also suggest the NOIADE project by Nargott.
As far as I know it is the most comprehensive power calculator up to date.
Unfortunately the program just displays a ton of question marks for me???
Quote: What you should do is experiment with powers of the initiative value.
If you believe having initiative of 10 is more than twice as valuable as having an initiative of 5, you should include Ini^X as a multiplier, where X > 1.
Experiment with X=1.5, X=2, X=2.5, X=3, etc...
For example, with X=2, a creature with initiative 20 gets a multiplier of 400, one with initiative 10 gets 100, one with 5 gets 25.
In another look, this approach, assuming all other stats are equal, would set the power ratio between two creatures to (Ini_1/Ini_2)^X, instead of just (Ini_1/Ini_2).
Something I've noticed is that reducing initiative below 10 is much more severely sensitive than increasing initiative above it. Unless the factorization above 10 is mitigated somehow, it would cause some really weird results.
I don't know why you're stuck with number 10. Don't you see that number 10, in your statement, is arbitrary?
When comparing two creatures to each other, which is what a power rating should be about, "10 initiative" has no importance.
There are no real "turns" in H5. Just initiative. Just as a 10 initiative unit moves twice as much as a 5 initiative unit, a 12 initiative unit moves twice as much as a 6 initiative unit, and so on. the only thing that matters is that the number of times a unit acts is directly pproportional to its initiative.
However, like you said, it is easily conceivable that a unit's power increases faster than the number of times it acts. How much faster is up for debate. That is why I said you should experiment with the exponents of the initiative value...


Gidoza
Famous Hero

posted April 21, 2020 01:29 AM 


thGryphn said:
Gidoza said:
dredknight said: I would also suggest the NOIADE project by Nargott.
As far as I know it is the most comprehensive power calculator up to date.
Unfortunately the program just displays a ton of question marks for me???
Quote: What you should do is experiment with powers of the initiative value.
If you believe having initiative of 10 is more than twice as valuable as having an initiative of 5, you should include Ini^X as a multiplier, where X > 1.
Experiment with X=1.5, X=2, X=2.5, X=3, etc...
For example, with X=2, a creature with initiative 20 gets a multiplier of 400, one with initiative 10 gets 100, one with 5 gets 25.
In another look, this approach, assuming all other stats are equal, would set the power ratio between two creatures to (Ini_1/Ini_2)^X, instead of just (Ini_1/Ini_2).
Something I've noticed is that reducing initiative below 10 is much more severely sensitive than increasing initiative above it. Unless the factorization above 10 is mitigated somehow, it would cause some really weird results.
I don't know why you're stuck with number 10. Don't you see that number 10, in your statement, is arbitrary?
When comparing two creatures to each other, which is what a power rating should be about, "10 initiative" has no importance.
There are no real "turns" in H5. Just initiative. Just as a 10 initiative unit moves twice as much as a 5 initiative unit, a 12 initiative unit moves twice as much as a 6 initiative unit, and so on. the only thing that matters is that the number of times a unit acts is directly pproportional to its initiative.
However, like you said, it is easily conceivable that a unit's power increases faster than the number of times it acts. How much faster is up for debate. That is why I said you should experiment with the exponents of the initiative value...
I'm perfectly aware that there are no turns in H5  but you're omitting a really important factor  actually several  and I'm surprised that you're omitting them.
First is a Hero's initiative, which is generally 10. I suppose this is arbitrary as well, but you need a baseline of some kind.
Second, you have abilities like Stunning Blow that knock out a unit's progression. That's not arbitrary, because how much is lost isn't proportional in the same way that unit initiatives are proportional to one another.
Third is units' ability to retaliate. "Technically" the rate is half at 5 and double at 20, but the real results are completely different in a battle.
If you don't have some kind of baseline, it's not possible to account for any of these variables and the term is meaningless. Initiative does more than just change how often a unit moves.


thGryphn
Promising
Famous Hero

posted April 21, 2020 03:03 AM 


Gidoza said:
thGryphn said:
Gidoza said:
dredknight said: I would also suggest the NOIADE project by Nargott.
As far as I know it is the most comprehensive power calculator up to date.
Unfortunately the program just displays a ton of question marks for me???
Quote: What you should do is experiment with powers of the initiative value.
If you believe having initiative of 10 is more than twice as valuable as having an initiative of 5, you should include Ini^X as a multiplier, where X > 1.
Experiment with X=1.5, X=2, X=2.5, X=3, etc...
For example, with X=2, a creature with initiative 20 gets a multiplier of 400, one with initiative 10 gets 100, one with 5 gets 25.
In another look, this approach, assuming all other stats are equal, would set the power ratio between two creatures to (Ini_1/Ini_2)^X, instead of just (Ini_1/Ini_2).
Something I've noticed is that reducing initiative below 10 is much more severely sensitive than increasing initiative above it. Unless the factorization above 10 is mitigated somehow, it would cause some really weird results.
I don't know why you're stuck with number 10. Don't you see that number 10, in your statement, is arbitrary?
When comparing two creatures to each other, which is what a power rating should be about, "10 initiative" has no importance.
There are no real "turns" in H5. Just initiative. Just as a 10 initiative unit moves twice as much as a 5 initiative unit, a 12 initiative unit moves twice as much as a 6 initiative unit, and so on. the only thing that matters is that the number of times a unit acts is directly pproportional to its initiative.
However, like you said, it is easily conceivable that a unit's power increases faster than the number of times it acts. How much faster is up for debate. That is why I said you should experiment with the exponents of the initiative value...
I'm perfectly aware that there are no turns in H5  but you're omitting a really important factor  actually several  and I'm surprised that you're omitting them.
First is a Hero's initiative, which is generally 10. I suppose this is arbitrary as well, but you need a baseline of some kind.
Second, you have abilities like Stunning Blow that knock out a unit's progression. That's not arbitrary, because how much is lost isn't proportional in the same way that unit initiatives are proportional to one another.
Third is units' ability to retaliate. "Technically" the rate is half at 5 and double at 20, but the real results are completely different in a battle.
If you don't have some kind of baseline, it's not possible to account for any of these variables and the term is meaningless. Initiative does more than just change how often a unit moves.
Like you said, the fact that Hero initiative is 10 is also arbitrary to the power rating of creatures, so I don't know why you think it matters to anything on this.
Stunning Blow reduces ATB value at the same "amount" for any creature, by 0.25 to be precise. Which means it actually impacts high init creatures more, but they also recover back faster. Regardless, what does it have to do with initiative value of 10??
Units retaliate once between their own actions, not according to any "turn" system based on any specific initiative value. So, high init creatures actually do retaliate more times over the course of a battle.
All in all, I don't see why you think any of this matter in determining the power of creatures. There is no need for any baseline when comparing creatures to each other.
All that matters is this fact: a creature that has 2X initiative takes twice the number of actions (including retaliations) as a creature with X initiative. But again, it doesn't mean 2Xinit creature, assuming all other stats are the same, is twice as strong; it is probably more than twice as strong. That is, I agree that as initiative increases, the power of a creature increases faster than linearly. How much faster? I don't know.
If it was increasing linearly, including Initiative value directly as a multiplier in the power calculation would make sense. If you agree that "s initiative increases, the power of a creature increases faster than linearly" then you should include "Initiative value to the power that's larger than 1". But again, how much larger, I don't know. Hence, I said, experiment.
You asked for how to reflect mathematically the phenomenon you mentioned, and this is how. Out. Cheers.


FirePaladin
Supreme Hero
Undercover modder

posted April 21, 2020 10:20 AM 


To be honest, I noticed the troops retaliating only once per turn, no matter the initiative. It's easier to see this when sieging a town.


thGryphn
Promising
Famous Hero

posted April 21, 2020 04:21 PM 


FirePaladin said: To be honest, I noticed the troops retaliating only once per turn, no matter the initiative. It's easier to see this when sieging a town.
Not sure what you mean by turns, but to be specific, there are no "static turns" in H5. Each stack retaliates once between two consecutive actions of their own.


Gidoza
Famous Hero

posted April 21, 2020 07:07 PM 


Quote: Stunning Blow reduces ATB value at the same "amount" for any creature, by 0.25 to be precise. Which means it actually impacts high init creatures more, but they also recover back faster. Regardless, what does it have to do with initiative value of 10??
I strongly disagree with this statement. Because high initiative creatures recover faster, Stunning Blow affects them less (I'd rather use Ice Freezing on high initiative units, and Lightning Stun on low initiative units). On the other hand, a lowinitiative unit if it is being stunned regularly can end up never seeing a turn again. You can otherwise rewrite a unit's initiative (based on 10) with stunning blow as x0.25. For a unit with initiative 10, that's a 25% loss; for a unit with initiative 20, that's a 12.5% loss, assuming it was near moving. The proportionality ends up skewed against the lower initiative values here. So I don't get why you'd say that having a placeholder value like 10 is arbitrary: all the initiative values relate to one another but especially to special effects in unique ways (e.g. Decay will have more serious effects on high initiative units because they get more turns). High initiative units get a "pretend" Unlimited Retaliation ability if there are few enough units on the field, while having a "pretend" No Enemy Retal against units with less Initiative. Again  you can measure Initiative merely as a measure of how many turns a unit takes if you want, but that's not an accurate measure  many more things are happening.


Elvin
Admirable
Omnipresent Hero
What if Elvin was female?

posted April 21, 2020 07:40 PM 


Initiative is pretty hard to quantify and transform but I believe that it should always be considered as a multiplier of damage. Trying to find a correlation with hp, defense or speed is kind of pointless. I also prefer to keep unit init close to 10, with below average units around 89. Any less and the unit will become a pointlessly boring slowpoke. Similarly, messing around with init values above 13 can cause some pretty extreme scenarios with mass haste and/or 5 morale. Not that it can't be balanced, just requires more care and is more likely to go haywire ^^
____________
A sea without waves is no longer a sea. It's a pond.


FirePaladin
Supreme Hero
Undercover modder

posted April 21, 2020 07:52 PM 


Elvin said: Initiative is pretty hard to quantify and transform but I believe that it should always be considered as a multiplier of damage. Trying to find a correlation with hp, defense or speed is kind of pointless. I also prefer to keep unit init close to 10, with below average units around 89. Any less and the unit will become a pointlessly boring slowpoke. Similarly, messing around with init values above 13 can cause some pretty extreme scenarios with mass haste and/or 5 morale. Not that it can't be balanced, just requires more care and is more likely to go haywire ^^
Yeah, that's what balances Phoenixes (at least recruited ones). And as you said, initiative simply just allows the unit to strike more often, so it should logically be a damage multiplier (the reason why Sylvan dragons deal pretty low damage). Units with very low initiative, under 8, are mostly kinda meant to be unused though (I mean, Zombies, they stink in all HoMM games, quite literally). And Hydras too, but their health pool is insane and they can strike more units than in HoMM3 especially because of how the battle grid is in HoMM5.


thGryphn
Promising
Famous Hero

posted April 21, 2020 10:53 PM 


Gidoza said:
Quote: Stunning Blow reduces ATB value at the same "amount" for any creature, by 0.25 to be precise. Which means it actually impacts high init creatures more, but they also recover back faster. Regardless, what does it have to do with initiative value of 10??
I strongly disagree with this statement. Because high initiative creatures recover faster, Stunning Blow affects them less (I'd rather use Ice Freezing on high initiative units, and Lightning Stun on low initiative units). On the other hand, a lowinitiative unit if it is being stunned regularly can end up never seeing a turn again. You can otherwise rewrite a unit's initiative (based on 10) with stunning blow as x0.25. For a unit with initiative 10, that's a 25% loss; for a unit with initiative 20, that's a 12.5% loss, assuming it was near moving. The proportionality ends up skewed against the lower initiative values here. So I don't get why you'd say that having a placeholder value like 10 is arbitrary: all the initiative values relate to one another but especially to special effects in unique ways (e.g. Decay will have more serious effects on high initiative units because they get more turns). High initiative units get a "pretend" Unlimited Retaliation ability if there are few enough units on the field, while having a "pretend" No Enemy Retal against units with less Initiative. Again  you can measure Initiative merely as a measure of how many turns a unit takes if you want, but that's not an accurate measure  many more things are happening.
Look bro, you say many things but nothing you say makes number 10 special in terms of initiative.
Also, you should read my sentences in full. You seem to disagree and argue against the first half of my statement. There is also a comma and a second half that starts with "but"


Gidoza
Famous Hero

posted April 22, 2020 12:01 AM 

Edited by Gidoza at 01:36, 22 Apr 2020.

FirePaladin said:
Elvin said: Initiative is pretty hard to quantify and transform but I believe that it should always be considered as a multiplier of damage. Trying to find a correlation with hp, defense or speed is kind of pointless. I also prefer to keep unit init close to 10, with below average units around 89. Any less and the unit will become a pointlessly boring slowpoke. Similarly, messing around with init values above 13 can cause some pretty extreme scenarios with mass haste and/or 5 morale. Not that it can't be balanced, just requires more care and is more likely to go haywire ^^
Yeah, that's what balances Phoenixes (at least recruited ones). And as you said, initiative simply just allows the unit to strike more often, so it should logically be a damage multiplier (the reason why Sylvan dragons deal pretty low damage). Units with very low initiative, under 8, are mostly kinda meant to be unused though (I mean, Zombies, they stink in all HoMM games, quite literally). And Hydras too, but their health pool is insane and they can strike more units than in HoMM3 especially because of how the battle grid is in HoMM5.
For me, the idea of a pointless unit or a unit that isn't intended to be used is purely insane. Zombies are going to be useful in my mod, I tell you.
Quote: Also, you should read my sentences in full. You seem to disagree and argue against the first half of my statement. There is also a comma and a second half that starts with "but"
I read your sentence in full  I don't find that your "but" mitigates anything about the statement apart from making the statement meaningless? The fact that some units recover faster than others has precisely been the point of the initiative discussion from the beginning.
And, as I've said several times  there is something about initiative 10 that makes it nonarbitrary to us: it's Hero initiative, and it's the average initiative. I didn't pick the number arbitrarily: the people who made the game did, and now it's a good standard. Please explain to me why it's a problem to have a standard and work with it  having baselines and standards isn't arbitrary.
EDIT: Allright, how about we go about this a different way. How about I just say that 10 is totally arbitrary no matter what and that I want to balance the initiative around this arbitrary number 10. That's the given. Now, the question is how to do the math around it. One suggestion is that it's just a multiplier of damage. I'm not convinced of that personally, partially because that's exactly how my Excel is designed and I'm not satisfied with the results. If you think this isn't worth your time to work on, that's cool then and don't; on the other hand if you want to help me out (because I'm still going to be chugging on at it), I'd welcome the input. Cheers.


Elvin
Admirable
Omnipresent Hero
What if Elvin was female?

posted April 22, 2020 02:32 PM 

Edited by Elvin at 14:34, 22 Apr 2020.

Technically, one could set hero initiative to 20 and use that as the new point of reference. It would also allow units to have a wider range of initiative  Like 21 would be 10.5 in the old system. But that would not matter much unless one introduced a less random starting atb through hex editing. (Naturally the static init boosts like battlerage, commanding presence etc would have to be doubled as well.)
I too hate the idea of 'useless' units. Zombies below 8 init are meatshields that almost never move. Hydras below 8 init will never have interesting gameplay like their H3 fortress counterparts and will only shine with teleport assault or other trickery. But if they were more mobile, that would also make the might dungeon class more versatile.
____________
A sea without waves is no longer a sea. It's a pond.


FirePaladin
Supreme Hero
Undercover modder

posted April 22, 2020 02:38 PM 


Elvin said: Technically, one could set hero initiative to 20 and use that as the new point of reference. It would also allow units to have a wider range of initiative  Like 21 would be 10.5 in the old system. But that would not matter much unless one introduced a less random starting atb through hex editing. (Naturally the static init boosts like battlerage, commanding presence etc would have to be doubled as well.)
I too hate the idea of 'useless' units. Zombies below 8 init are meatshields that almost never move. Hydras below 8 init will never have interesting gameplay like their H3 fortress counterparts and will only shine with teleport assault or other trickery. But if they were more mobile, that would also make the might dungeon class more versatile.
My only little initiative complaint is the Hydras; they really never get to move.
And I have no problem with useless units; they either exist to balance the faction, or to defend the castle, or to be somewhat "realistic", etc. Who sees Zombies acting often in a game like this? And Necropolis has the skeleton hordes and good hero specials.


frogo
Adventuring Hero

posted April 22, 2020 04:31 PM 

Edited by frogo at 16:38, 22 Apr 2020.

I have only read a few parts of the current discussion, so bear with me if i repeat something
Creating a strength estimator for creatures is a tough task, and given that a creatures stength depends on the situation it is in, its practically impossible without first defining a situation.
A few examples:
You claim zombies are weak units, but when playing necro magic, i tend to invest my DarkEnergy into zombies over all the other creatures.
Zombies are a strong choice, because what a magic hero needs is durability, chances are the units itself wont move out to attack anyway, but just defend.
To push this to an extreme example: suppose a unit had 0 intitave. With your current formula of hp*dmg, where initative is a multiplier in dmg, the units worth would be 0 independent of the other stats.
However, i m pretty sure a unit with 0 initative but 1000000 hp would win me every battle, as my hero can do dmg forever.
So to specify the situation, maybe we should take the hero out of the equation. The resulting values wont really represent how much i gain from a certain creature when i have it in my army, but maybe for creeps.
But again, situations can vary to much to define a creaturs strength. Our 0 ini, 1000000 hp unit is pretty worthless on its own, but its a premium blocker.
Suppose we had 10 stats to distribute on dmg and hp, meaningg for example if we gave it 3 hp and 7 dmg (and disregard any overkills or whole integer implications), our units value would be 3*7 = 21.
The strongest unit we could create distributing our 10 stats would be a unit with a worth of 5*5=25.
Now if we had 2 stacks instead and unitss are able to block, the strongest team (regarding our calculation) would obviously 2 of these units, with an overall value of 25+25 = 50. However, a team that consists of 1 unit with 10hp and 0 dmg combined with 1 unit that has 0hp and 10 dmg, so a combined value of 0 + 0 = 0, would actually be the strongest team. This is because while with each hit the 50value team takes, its dmg output is reduced, while the 0value team keeps up its dmgout till the end. Both have a durability of 10hp.
Once again, defining a strength value in these kind of situations seems impossible, so lets once again simplify the situation:
Only 1 creatre stack vs 1 creature stack.
Now, obviously a stone papaer scissor situation can be created, even when disregarding overkills/whole number implications.
For example:
A ranged unit with high dmg, low hp could defeat a meele unit, but lose to a ranged unit that is balanced in hp/dmg, which in turn would lose to the melee unit.
We could disregard this and say that despite there being stone/paper/scissor mechanics, you could somehow create a ranking based on average performance.
Certainly doable, but even in this overly simplified scenario, where the result would hardly have much to do with a units actual value, calculating the values would be incredibly difficult.
For example:
You wonder about the worth of initative. Lets disregard retaliations for now (suppose its shooter vs shooter). Higher initative in Homm5 is worth more than just the ini value times dmg, because it not only increases your dmg, but also decreases the enemies dmg.
When you shoot faster and first, not only is your dmg higher, but the enemy stack loses troops before it acts.
How big the decrease in dmg out put is depends on how much was lost in that attack, so on the faster units dmgoutput.
For example:
a stack does 1 dmg per 5hp (like a unit with 1 dmg and 5 hp, only disregarding the whole number/units stuff for simplicity. This can also be achieved by letting the amount of units in the stacks converge towards infinity).
Example:
If the stack had 85hp, it therefor would doe 85/5 = 17 dmg.
Now, lets say it has 1 dmg, 5 hp and 5 initative.
Its enemy:
1 dmg, 2hp, 10.00001 initative.
Winner: [calculating...] Its the 1 dmg, 2hp, 10.00001 initative stack!
Now if we change the ratio of dmg to hp (equally for both) by multiplying hp*10, then first shot becomes less important, as the relative dmg reduction from shooting first is less significant, and therefore initative becomes less important:
new match up:
1dmg, 50 hp and 5 initative vs 1 dmg, 20hp, 10.00001 initative
Winner: [calculating..] the 1dmg, 50 hp and 5 initative !
Notably: When the dmg/hp ratio converges to 0, the strength of initavie congerges to its value (say, a unit with 8 ini could be estimated with (strength_estimation_without_factoring_in_ini)*8, just like you did in your current formula). For any higher ratio of dmg/hp, the worth of initative is higher.
Having this in mind, its easy to conclude, that importance of iniative depends on the dmg/hp ratio, and therefore can not be logically implemented into your formula when the initative multiplier is not dependend on this ratio.
Same principle goes for shooters. A flat 1.4 multiplier for being a shooters makes little sense, as a unit with a low dmg/hp ratio benefits way less from it than a unit with high dmg/hp ratio.
Its a honorable project and if you are just looking to kill some time while doing some math puzzles: sure thing!
But in order to get any usable results, the effort would be immense even for the single stack vs single stack with no hero scenario.
____________


FirePaladin
Supreme Hero
Undercover modder

posted April 22, 2020 04:49 PM 


I agree. By useless I meant "acting rarely", but Zombies are really good cannon fodder (not trying to be rude, lol).



