DS Visual Bestiary

Battle simulator is a nice idea, but we're missing one crucial piece of data to make it relatively accurate. I don't know which is applied first, dea or ipa or species down or guild down. Given that it is adding or subtracting a percentage, order does matter.

Also, thanks to bbr, the tool in my sig is also up to date, so I can focus on making various improvements to it.

Bbr, you're awesome, that's the bottom line.

We could observe the battle logs and gather the data. It might be the creature with the highest bp goes first with the attacking team getting to use their skills first.

The battle log tells us what happens and in what order but we need a descent sample to figure out.

Not entirely, there's two ways that the skill stack can be applied. For instance, if battle log says "ipa, dea, psycho down, demon down" the effects could be applied either in the same order or "demon down, psycho down, dea, ipa". Depending on the order, final attack would be different. Think of it as a stack of documents in your in tray at work vs a line for the bathroom.

This weekend I'm going to start work on a web application which will have the look, feel and functionality of the current visual bestiary but will not be limited by the ever-frustrating google docs issues. I also believe I can create a scheduled process to update monster data on a daily basis without any human interaction, though I haven't fully explored that yet. I'll continue to support the existing googledocs sheet and give updates on the web app but will be sooooooo happy when I eventually delete that pain in my a$$ googlesheet.

I'm planning on using the Google App Engine service (I know, hopefully better than docs lol) with Java to code the WebApp. To parse monster data I'l be using a virtual android device loaded with dark summoner which will be imported into a google cloud sql DB. I'm going to run some tests on the speed of google's cloud service, if it's too slow I'll use an alternative.

PS As everyone else has said, thanks to BBR/BigSeanDog? for the data parsing method, no more manual entry for the community hurrah!

Armitaage wrote:Not entirely, there's two ways that the skill stack can be applied. For instance, if battle log says "ipa, dea, psycho down, demon down" the effects could be applied either in the same order or "demon down, psycho down, dea, ipa". Depending on the order, final attack would be different. Think of it as a stack of documents in your in tray at work vs a line for the bathroom.

As of now I don't think we can realistically emulate Dark Summoner combat, for the reasons that Armitaage gave. If we did come up with theories on how all aspects of combat works, we still couldn't reliably validate it against the battle log.

Armitaage wrote:Battle simulator is a nice idea, but we're missing one crucial piece of data to make it relatively accurate. I don't know which is applied first, dea or ipa or species down or guild down. Given that it is adding or subtracting a percentage, order does matter.

Also, thanks to bbr, the tool in my sig is also up to date, so I can focus on making various improvements to it.

Bbr, you're awesome, that's the bottom line.

No problem folks - though donations are always welcome


I've just finished updating the individual monster pages on the wiki last night as well.
Every single monster now has a detailed page.

Still uploading some images, but after that everyone can simply focus on stuff other than monsters on the wiki. And the automated script should be able to keep the monster pages updated.

Eventually ill get the program in such a shape that i can release it to a select few people to keep their stuff updated, but until then, the data you've got now should go a long way at analyzing the best formations.

TigerStrike wrote:

bbr wrote:Alright, so ,.. the program I'm using to parse the monster DB currently loads the "entire" list of all info about every single monster in under a second.

I tried adding a progress bar - It just wasn't worth it. Beat that google docs.

In theory, the program could be expanded to show monster info runtime, get some formation tools, even an automatic formation builder if you so desire.

My question for now would be:
* Would there be an interest in such a tool?
* Would you mind that it's a standalone .exe? (no .net or creepy dll files needed)
* What would be your demands on functionality?

It's not like It'll be finished in under a week or so,. but heck - if there are people here with delphi programming language skills, it shouldn't be too hard to expand on the current functionality if there is enough desire for it.

I updated the bestiary last night thanks to BBR. Thank you for the data!

I think creating a tool would be great. If we could make it a web application, that would allow people to access it more easily.
With the information provided, we could create a battle simulator based on ideal conditions.

A web app will always have to rely on a database or large xml file somehow ,. not my division so ill leave that to you lot. But I can change the export to only output the required fields. It should be able to go under about 200KB, rather than the 2.3 MB that it is now.

Armitaage wrote:Not entirely, there's two ways that the skill stack can be applied. For instance, if battle log says "ipa, dea, psycho down, demon down" the effects could be applied either in the same order or "demon down, psycho down, dea, ipa". Depending on the order, final attack would be different. Think of it as a stack of documents in your in tray at work vs a line for the bathroom.

No. That is not correct. The final value is the same no matter which is applied first.

Example
Attack of 1000
IPA 10%
DEA 40%

IPA goes first and DEA goes last.
1000 + (1000 * .1) = 1100
1100 - (1100 * .4) = 660 Final

DEA goes first and IPA goes last.
1000 - (1000 * .4) = 600
600 + (600 * .1) = 660 Final

Both final values are the same.

If you wanted something more complicated.
Attack 1000
IPA 10%
DEA 40%
Psycho Down 20%

Sub are the subtotals.

Att IPA Sub PD Sub DEA Final
1000 0.1 1100 0.2 880 0.4 528

Att IPA Sub DEA Sub PD Final
1000 0.1 1100 0.4 660 0.2 528

Att PD Sub IPA Sub DEA Final
1000 0.2 800 0.1 880 0.4 528

Att PD Sub DEA Sub IPA Final
1000 0.2 800 0.4 480 0.1 528

Att DEA Sub IPA Sub PD Final
1000 0.4 600 0.1 660 0.2 528

Att DEA Sub PD Sub IPA Final
1000 0.4 600 0.2 480 0.1 528

This is all the different order combinations. Unless my math is wrong, which it can be since it has been 20 years since I had a math class, the final values are the same.

We just need to know the effects applied to each monster and we have a battle simulator. Which of course would only be ideal situation since we don't know the calculation of miss, critical hit, or any other unknown factors that they use.

Also, the new Party skills will not work as above since they give you an exact value and not a percentage. We would have to see what the order of those are and use them but any percentage base skills will work for our scenario.

TigerStrike wrote:

Armitaage wrote:Not entirely, there's two ways that the skill stack can be applied. For instance, if battle log says "ipa, dea, psycho down, demon down" the effects could be applied either in the same order or "demon down, psycho down, dea, ipa". Depending on the order, final attack would be different. Think of it as a stack of documents in your in tray at work vs a line for the bathroom.

No. That is not correct. The final value is the same no matter which is applied first.

Example
Attack of 1000
IPA 10%
DEA 40%

IPA goes first and DEA goes last.
1000 + (1000 * .1) = 1100
1100 - (1100 * .4) = 660 Final

DEA goes first and IPA goes last.
1000 - (1000 * .4) = 600
600 + (600 * .1) = 660 Final

Both final values are the same.

If you wanted something more complicated.
Attack 1000
IPA 10%
DEA 40%
Psycho Down 20%

Sub are the subtotals.

Att IPA Sub PD Sub DEA Final
1000 0.1 1100 0.2 880 0.4 528

Att IPA Sub DEA Sub PD Final
1000 0.1 1100 0.4 660 0.2 528

Att PD Sub IPA Sub DEA Final
1000 0.2 800 0.1 880 0.4 528

Att PD Sub DEA Sub IPA Final
1000 0.2 800 0.4 480 0.1 528

Att DEA Sub IPA Sub PD Final
1000 0.4 600 0.1 660 0.2 528

Att DEA Sub PD Sub IPA Final
1000 0.4 600 0.2 480 0.1 528

This is all the different order combinations. Unless my math is wrong, which it can be since it has been 20 years since I had a math class, the final values are the same.

We just need to know the effects applied to each monster and we have a battle simulator. Which of course would only be ideal situation since we don't know the calculation of miss, critical hit, or any other unknown factors that they use.

Also, the new Party skills will not work as above since they give you an exact value and not a percentage. We would have to see what the order of those are and use them but any percentage base skills will work for our scenario.

actually, they are different, though you won't see the difference using 1000. try it with something like 12500 and you will see that the two values are off by a point. yes, the two are extremely close, but it does matter. note: the sample below rounds DOWN, rather than up. if it rounded up, the difference would be greater than 1 point.

	atk
ipa -> species down	12500	species down -> ipa
15625		7500
11718		6000
9374		4500
5624		5625

	atk
ipa -> species down	9900	species down -> ipa
12375		5940
9281		4752
7424		3564
4454		4455

i used a spreadsheet to make the work easy, and i round down because ATeam merely uses integer variables rather than float for everything and in integer division, 5/2=2 is a true statement, at least as far as how computers do it behind the scenes.



also, the difference between "health up > party health plus" and "party health plus > health up" is 500 points more for the latter

What percentage values did you use for each steps?

on the left hand side i have
1.25, .75, .8, .6(for ipa, dea, imp down, demon down)

on the right hand side i have
.6, .8, .75, 1.25(reverse order of the left hand side)

all percentage values i used are what's left after the effect is applied at max(dea is -.25, so .75; imp down is -.2, so .8; demon down is -.4, so .6; ipa is +.25, so 1.25)

On the topic of building a battle simulator, has anyone quantified any of the following?
% chance of a miss
% chance of a crit
Damage dealt by crit?

Retired, use the DS Wiki or Armitaage's Calculator

Thanks HORG. I for one really appreciated the HARD work you put into the Visual Bestiary!