Figuring base attack of 817 -16%(plains) -32%(boar spear) +240%(boar spear) = 1586 attack

475 Kobolds base 9 -16%(plains) = 3591 attack

5177 total attack against 5172 Defense per Digioso calculator

after the attack i only did around 88% damage or 4554 damage

4554 - 3591 = 963 damage done by my commander

so 817 -16% = 686.28 +40% = 960.8

so the base boar spear is +20% (+40% roughly with elite status) or is my napkin math wrong

That what im wanting to know is if the math is right or missed a variable mainly. I was told the +animal bonus weapons were off thats why i ran the numbers before and after.

total attack = basic attac * general multiplier * crafted multiplier,

where

1.

a) Basic attack of a unit is basicly a basic attack of a unit (eg knight's BA = 65, stalwart's = 45),

b) commander basic attack = UBA * (1+ HL*6), where

UBA - unit's basic attack  - the attack score of the unit the commander is recruited from;

HL - Heroism skill level;

2. General multiplier is comprised of every bonus out there but crafted:

Division bonus attack skill

Terrain;

Biome;

Prestige;

the multiplier is = 1 + (DBA% + TB% + BB% + PB%)/100

3. crafted multiplier is  1 + (armor bonus% + weapon bonus% + mount bonus%)/100;

or when elite: CM = 1 + (armor bonus% + weapon bonus% + mount bonus%)*2/100;

so lets assume a classic situation: you deploy an army of 61 knight troops lead by 5 elite knight commanders on plains, with only commanders equipped with pains spears and plainsman armor and heavy horses, no biome and pres bonus on,all commanders have heroism 10 and charge level 15;

so each of your commander's attack will consist of:

a) basic attack: 65*(1+ 10*6) = 65*61 = 3965

b) general multiplier consists of: terran 30%, charge 15%, biome 0%, pres 0%, total of

GM = 1 + (30 + 15 + 0 +0)/100 = 1.45;

c) crafted multiplier: plains spear +60%, plains armor +60%, heavy horse +5%, all elite;

CM = 1 + (60 + 60 +5)*2/100 = 3.5;

so your total (knight) commander's attack will be:

TCA = 3965 * 1.45 * 3.5 = 20 122.375;

5 commanders will do: 20 122.375 * 5 = 100 611.875;

61 knight troops gonna deal: 65 * 61 * 1.45 * 1 (unequipped) = 5 749.25;

and your total army's attack is:

TAA = 100 611.875 + 5 749.25 = 106 361.125.

I've checked this script on virtually thoudands of attacks during last 5 years and it did work all the time. But I haven't been too active during last few months so unless the devs have changed the rules recently it all should be true.

PS A bit tipsy atm and English is ot my mother tongue so sorry guys if it's not coming too clear. tried my best you know, hope it helps.

From: System

Subject: FW: Victorious battle against NPC units's forces at [@l=219|-1388|41|113|7] by Sene's forces from Acoron
Received: 31 May 2018 19:49
Original Message:
Two opposing forces clash against each other.

Heavily wooded terrain spells difficulty for ranged units and cavalry. Infantry units find they have a large advantage here.

Like the attackers, these thick forests do not provide much assistance for anyone except your footsoldiers, who take great pleasure in the extra cover afforded to them.

Attackers:	Unit:	Quantity:	Casualties:	Survivors:
Commander: Yozz	[@i=5|61] Stalwart	1	Damaged for 100, 200 health remains.
Troops:	[@i=5|61] Stalwarts	1501	1501	0
Commander: Leshiy	[@i=5|61] Stalwart	1	Damaged for 100, 200 health remains.
Troops:	[@i=5|61] Stalwarts	100	100	0
Commander: Sohatiy	[@i=5|61] Stalwart	1	Damaged for 100, 200 health remains.
Troops:	[@i=5|61] Stalwarts	100	100	0
Commander: Voron	[@i=5|61] Stalwart	1	Damaged for 100, 200 health remains.
Troops:	[@i=5|61] Stalwarts	100	100	0
Commander: Koteiko	[@i=5|61] Stalwart	1	Damaged for 100, 200 health remains.
Troops:	[@i=5|61] Stalwarts	100	100	0

Defenders:	Unit:	Quantity:	Casualties:	Survivors:
Troops:	[@i=5|146] Spitting Crawlers	804	804	0
Troops:	[@i=5|147] Crawler Queens	607	607	0
Troops:	[@i=5|144] Small Poisonous Crawlers	9700	9700	0
Troops:	[@i=5|145] Adult Crawler	4303	4303	0

Your victorious troops retrieved:
0 [@i=4|1]
from the slain foes.

I'm sorry to disagree but you do NOT multiply the equipment bonus by the terrain bonus. I've also checked this over long periods until I correctly predicted kills in advance. Take a human Knight for example. Its basic attack is 65. The correct terrain gear adds 60% for the weapon and 60% for the armour - this comes to +39 for each REGARDLESS of terrain. Ignore all other bonuses for the moment.

So on plains 65+39+39 + 19.5(terrain) = 162.5 (not 185.9 as if multiplied)

and on large mountains 65+39+39 - 19.5 = 123.5 (not 100.1)

Note that the difference is not as big as you might instinctively think and it is NOT insane to send cav up mountains! if properly geared.

Remember also that a terrain spear costs the same to make no matter which troop type you put it on - so the decision to use gear is much easier for a human knight than it is for 'in-between' troop types and nearly impossible for a one million kobold army. 

There are a lot of misunderstandings about equipment and charts that draw conclusions that are only valid if no equipment is used - my advice to you is to analyse your battle reports until you can do the calculation yourself - experiment - predict - analyse. You will find that you can get EXACT results.

If you have a battle report and can tell me the equipment used, I can do the calculation for you a couple of times.

PS: - this really matters when you use equipment on ALL your troops, not just commanders.

If you have deep pockets, then yeah. equipment makes sense.... if you don't then it should be used sparingly when it will have the most impact.

... but the bigger the score you're augmenting the more the economics work for you. Remember that the bonus is not just bigger kill ratios but smaller armies too - equipment has no maintenance cost - and faster recovery times - you can make equipment in parallel to building troops. Kill and craft your own stuff - keeps you busy while waiting for the paint to dry.

A knight at current centrum prices costs 3237

A mountain spear costs 8200 and gives a 60% bonus

Therefore a mountain spear removes the need to spend 1942.2 on knight building equipment.

This leaves 6257.8 in cost from the spear.

Now, assuming you have a 50% cost reduction on troop upkeep, a knight costs 2 per hour in upkeep.

Therefore a equipping a knight with a spear is worth it if the knight you would be equipping it to was around for 3129 hours or more.... which is over 130 days.... 

If you do the same but with elite divisions used, then that figure falls to just under 90 days.

Not really worth it imo.

You may have faster recovery times but if you are at war then you can't be spending troops hunting meaning you need to buy the ingredients to make the equipment... meaning you need deep pockets.

I'm not too sorry to disagree with you, Hucbold, for the debates give birth to the truth:) I take it you will not believe my words if I just say "you are wrong". So let's post our battle reports and see whose method works better. I'd love to see my fails if there are any, but up to this date i've made virtually thousands  victorious hunting runs with 1 survivor, about 10 with zero survivors, which showed calculations worked perfectly right.

Let me start wtih posting this one, made the day before yesterday. I'll make few more and you make some too, lets compare them

In hunting crafted is only used on commanders, troops are never equipped. Not sure about PvP but I'd say it must be some extremely desperate situation to use thousands of equipment units to loose in a single battle. Well, unless one is defending a town. But generally I guess no one casres to equip troops too, far too expensive. Remeber "Brave Heart"? "Send Irish first: arrows cost money, Irish cost nothing". Well, this cynical approach rules Illy too: troops cost pennies, whyequip[ them you you could just use more troops

That's a fresh one. 4 commanders, 3 elite, 1 non-elite. Pres bonus on top to make calcs a bit more complex, but no biome. PLainsman gear, heavy horses, Heroism 10, Charge 15 for all. Ifyou could provide your calculations Hucbold I'd be most curious to learn them. And if you could provide one of your battle reports that illustrate your outcome prediction skill - I'd be both most curious and most grateful too:)

Great lucks and good hunting, pals

Totally true:)

TBH, I only really looked at mountain spears because those were what was being talked about in the original post, I have by no means studied this in depth.

Which ones would have a better payback period?

I also hadn't thought of the use of a non-involved account for the purposes of gathering ingredients for crafted items.

The actual payback periods are ~217 days for non-elites and ~75 for elites.

The rats details:

I lost 10 knights (kill ratio of 175 to 1). To achieve this my attack score needed to be in the range 113,022.95 to 124,983.06.

Here's my table

pres applies to the total - all others add (not sure about biome college bonuses but will check). The total is in the required range.

Putting figures on the elements - so you can gauge the value of each element

and as a percentage:

Note that the terrain effect is only 6.9% not 30%. The terrain bonus does NOT multiply the equipment bonuses. Rather the equipment counters and overtakes the terrain effect. QED!

Had I sent 65 knights, I would only have lost 7.

Alternative workings that produce a value in range will be entertained.

It seems we haven't met before so no, I did not know you are Irish. Kinda thought it was going without saying that those words belong to the baddiest baddie of the film and the phrase was to demonstrate how rotten and filthy he was. In any case, my sincerest apologies, pal, if you took it personally. I am a strong rasism-hater myself, got a couple of Irish friends here, got no problems with no races or countries. Russia is my neck of the woods and believe me, I know how it feels to be looked down on and to be blamed for all the world's sins.

Anyway, down to business if you don't mind. I take it you 1) equipped all the troops and and all the commanders with plainsman gear, commanders had heavy horses in addition; 2) commanders all have Hersoim 10 & Charge 15; 3) You have prestige attack bonus on.

I'll try to make my cacls but quick word straight away:

1) You heavily oversent there, man. Each army must contain of no more than 61 troops distributed between 5 divisions within elite limits;

2) Rats are so easy that 61 knights is enough to kill most of the myriads, not mentioning legions and furthermore hosts;

3) In your caclulations you clearly missed the elite bonus that doubles crafted bonus up, which implies your calculations might be seriously misleading;

4) Last, but most important to me. Your attack was 'in blind", you just sent the amount of troops that was obviously sufficient and then tried to speculate the result out of it; now I posted 2 reports proving I knew in advance the exact minimum number of troops needed to prevail in the battle. Don't you think the latter speaks a bit LOUDER than the former?:)

Great lucks and good hunting meanwhile, I'll try to post something meaningful a bit later.

Knight's basic attack:

Knight commander basic attack: 65*(1+6*10) = 3965; 

General modifier: 1 + (15Charge + 30Terrain + 10Pres + 0Biome)/100 = 1 + 0.55 = 1.55;

Crafted modifier for commanders: 1 + (60spear*2 + 60armor*2 +5heavy horse*2)/100 = 1 + 2.5 = 3.5;

Crafted modifier for troops: 1 + (60spear*2 + 60armor*2 +5heavy horse*2)/100 = 1 + 2.5 = 3.4.

Now 1 commander total attack: 3965 * 1.55 * 3.5 = 21 510.13, all 5 dealt 21 510.13 * 5 = 107 550.63;

1 knight (troops) total attack: 65 * 1.55 * 3.4 = 342.55; all 100 dealt 34 255.

Total army attack 107 550.13 + 34 255.0 = 141 805.63.

Casualties rate must be 11 840.5 / 141 805.63 = 0.08 or 8% or 8 knights. Hmm... It's either your data wasn't too correct or one of us is mistaken:)

Now you see those your precentage figures can't be taken serious because of one of your early statemetns: "To achieve this my attack score needed to be in the range 113,022.95 to 124,983.06" (not sure how you got it btw, i got different range here). This means only one: "I do knot know what my exact atack score was". Now if you don't, all those QEDs are but mere speculations. Sorry, man, no offense.

You see in the example your commas combined dealt 107550.63 attack. HAd you not equipped the troops they would have dealt 10 075 attack - less than 10%. The addition of 200 equipment pieces (of which 20 you lost on the battlefield, of which only 3-4 are harvestable) you gained only 24 180 attack, and the overall perfomance went up from 117 625.63 to 141805.63 - by mere 20.5%. And it only allowed to save 2-3 guys. Now if you attak a bigger stack with casulties rate of 80% or more, you will loose around 160 pieces, of which 130-140 pieces will go for good, others will remain on the battle fiels. Now is it really worth it? Everyone is to have personal opinion here, for me cannon fodder is mich cheaper

The range is because you can't kill half a soldier - so anything between 9.5 and 10.4999 rounds to 10. I wasn't playing the 'get the exact' number game! The exercise was to demonstrate the calculation not get a perfect kill. Now you admit you got the wrong number of casualties. You overstated the attack force (but corrected slightly by pulling the 10% pres inside the calculation). Here's the battle report

OK  I'm not too sure how Illy rounding up algorythm works so I took it from 0.9 to 0.11 just in case. But so be it, let's assume you are right for it follows general math rules.

Hucbold wrote: I wasn't playing the 'get the exact' number game! The exercise was to demonstrate the calculation not get a perfect kill.

Hucbold, there's one good attribute of any solid knowledge: it enables us to get precisely expected outcome over and over again. What does your example demonstrate? You killed a host of rats and then tried to work back the attack values. Well, anyone can kill a host of rats and anyone can come up with some version of what was going on there. With all my greatest respect to you efforts your attempt wasn't a big deal. And I would like to emphasize it once again: I showed you I knew EXACTLY what was going to happen, and it did happen. In order to make a 1 survivor battle one must put exact number of troops. Well, there's an errand of 3-4 soldiers when attacking a big stack, but gessing the number would still be like winning a lottery. And I would like to see if you could do the same thing. 'Cause if you can, we would face quite an interesting situation here.

Hucbold wrote: Now you admit you got the wrong number of casualties. You overstated the attack force (but corrected slightly by pulling the 10% pres inside the calculation)

I never admitted having wrong value, I said one of us is mistaken. It could be due to me miscalculating the outcome or you not giving the correct picture. Now I'm not saying you are lying, you could have mistaken too. Unfortunately, the bonuses can't be seen from the battle report so you must agree both scenarios are feasible.

Cheers

I find it interesting that you disregard the math - what's your explanation for the disparity in your result?

giving a kill ratio of 0.9222 (better)

and here's yours

giving a kill ratio of 0.11045

Hucbold, let's be honest: this doesn't prove anything. Let me make it clear: it's not that I do not believe you, it's justt I prefer to know for sure. Hence I only believe facts. While I do not have proof you lied I by default assume you tell the truth, but it still leaves the chance of the opposite. Forum post is not a fact for figures can be easily altered while during posting.

Hucbold wrote: I find it interesting that you disregard the math

That was the second time you had blamed me for something I've never said and done. A friendly advise if I may: refrain from making up quotes and deeds unless you want to be regarded as a mere troll. Let's keep to facts. Your words were: "my attack score needed to be in the range 113,022.95 to 124,983.06". After that you provided the calculations that showed the result was in the range. There is no actual proof it was the real attack score. It could've been 115k, 120k, 123k. How am I to know you were right? No facts, man. No facts - no result, sorry. On the contrary, I presented the rep that showed next to 100% preciseness. In the first rep i presented the def score of 15 414 NPCs, it was 279624.1, and my attack score of 1901 units and 5 commas, which was 279 628.4. The difference was  0.00156% (and you never tried to comment it in the first place). 1 per miller per cent, man. One. Per mille. Per cent. Now you come up with a rep where attack score is 10 times bigger than a def score and claiming you succeeded there - ah, c'mon! Are you serious??? Let's be fair: who's more precise here: me hitting a dime 300 yards away with an arrow or you hittting a 100x100 yards square with a nuclear bomb and crying "I hit it!!" Well, you did, man... 

It's not me who is to be asked. Your perfomance was different from your words, how am I to know what was wrong. The wild guess it that you forgot to equip some or all of your troops, but who knows. It's either skills, or terrain, or crafted or biome issue. Pres is less likely, its input is not that significant

First, that's the third time you've made it up. Second, it's utterly funny that by mere accident you've got completely correct results. It was because of terrain penalty value neautralizing the charge one, -15 + 15. As a result, the general mofidier consited of nothing but pres bonus, and you correctly multiplied it by the crafted one. Here is the calculation:

The def score is 6630.8, no objection here.

The attack score:

1) Commaders: basic 3965, 5 dealing 15860. General modifier: 1+ (15charge - 15terrain + 10pres + 0 biome)/100 = 1.1; crafted one: 1 + (60*2spear + 60*2armor)/100 = 3.4. Total attack = 15860* 1.1 *3.4 = 59316.4;

2) troops: basic 65, Gm is the same, crafted = 1 + 60*2spears /100 = 2.2. total attack = 65 * 80 * 1.1 * 2.2 = 12584;

3) total attack 59316.4 + 12584 = 71900.4

Casualties rate 6630.8 /  71900.4 = 0.0922 = 9.22% At least something wew agree on:)

Now it's either you do some serios battle trying to match your attack with the NPC def, or I loose interest in this arguing. Let those guys reading the forum make the decision on who's right here.

Lucks there

Your latest post clarifies your formula for me - I had misunderstood it and we agree on the figures, because the 15% charge bonus cancels the -15% terrain bonus. We both get 71900.4. I've done a spreadsheet -

We both believe there are two multipliers. 

I believe the first multiplier is (1+weapon+armour+horse+charge+terrain) and the second is (1+prestige). I don't know where the biome college bonus goes.

You, I think, believe the first multiplier is (1+weapon+armour + horse) and the second is (1+charge+terrain+pres+biome).

In this case they come out the same. Now lets go back to the first example and heres the table:

The point here is that I am correctly predicting the kill number whereas you are out. Repeat the experiment - let others fill in their numbers for their equipment and kills and lets see which has a better predictive ability. I'm not crying 'I hit it', I'm saying that my math predicts the kills correctly.

The other point I'm making to you is that if you are always or usually using infantry (positive terrain bonuses) and only equipping the commanders, then both our results will come out very close. However with big negatives like -30% for cav on some terrains and fully equipped armies, the answers will be very different and that all affects the argument whether it is economical to fully equip a 50k knight army - which is the subject I actually want to address.

ATM I seem to be surrounded by hosts of rats and packs or smaller of other things but I am looking for reasonable size kills on forest and mountains to support my argument - whatever degree of overkill or underkill does not matter - even if I lose a battle my formula will predict the NPC survivors too.

Mistakes do happen (in your analysis you describe 5 commanders but correctly multiply by 4 - a typo but I'm making no deal of that) but to suggest that I am lying or must have made a mistake because my outcomes differ from your predictions is demeaning. I will produce multiple repetitions of the exercise enough times to  verify the experiment as quickly as it is possible to. In fact on one I will misequip a division to further verify.

Well, another one of your fancies about me doing something I didn't. I will try to explain myself for the very last time: dear Hucbold, I'm trying to be purely, perfectly rational. Please correct me if I'm wrong, but being rational involves: 

- dealing with facts, not speculations;

- leaving out emotions, reasoning withing 'cause-consequence' logic (not sure what it's called in English);

- accepting all reasonable possibillites rather than "right/wrong" cliches.

For the last time: I've never said you were lying, I said you could've been wrong and so far you haven't prove that's a wrong statement. And honestly, 'lying or mistaking?' issue doesn't bother me much. It's "are you right and what proof is there?" instead.

Now is there a possibility you were wrong when posting all the battle circumstances? Yes, there is. It's only a possibility, but it's there. Are there facts showing you calculated EXACT attack value, not some vague range? "No, the arn't. Both reports produced range values, with ranges being quite large to make the errand reasonably serious to question the outcome of the calculations". So is there a possibility the actual attack value was different from what you calculated? 'YES, it's quite probable'. In both cases. So is it possible you were wrong in your assumpitons? The answer is "YES, it's quite probable" again.

No let's estimate my 2 reps: is there a possibility I was wrong when posting the battle cicumstances? "Yes, of course". Are there facts showing I calculated the EXACT attack value? the answer is "YES, beyond reasonable doubts. It's been proved twice by attack value exactly matching the def score with both values being reasonably big in order to make them highly unlikely to be guessed". So Is there  possibility I was wrong in my calculations? "The probability is extremely low, next to impossible".

Please tell me where my logic fails. I'd be most gratedul if you could point my mistakes.

Hucbold wrote: You're the biggest killer in Illy and... I'm only the 15th...

We both know it matters not.

Hucbold wrote: The point here is that I am correctly predicting the kill number whereas you are out... I'm saying that my math predicts the kills correctly

The point here is that "pre-" in "predicting" means "before" or "in adavance". What you did is wrote a couple of 'post-mortems" that - again - only produced values fitting withing the large errand range. Ther is no proof your assumptions are correct (though it's possible they might be) or you might've known the exact outcome before the attack. Sending an army approximately 10 times or more stronger than the defender and saying "I knew it would win" is also a kind of "prediction", but doesn't bear any significant knowledge to me.

Let's rationally estimate our input:

S: presented 2 battle reports proving exact attack values. Estimated 2 rival's reports, one estimate lies within the possible values range, one is presumably out.

H: presented 2 battle reports showing wide range values, estimate fits in the ranges in both cases. Didn't try to estimate rival's reports.

Nuff said. Please show something more solid to get back in the discussion.

Hucbold wrote: The other point ... is that if you are always or usually using infantry...only equipping the commanders

That could hardly be an issue. The battle algorythm deals with pure digits, races or troop types are only different attack values and terrain modifiers, but governed by the same formulas. Thus deathpacks and marshals show absolutely identical results both being cavalry and having the same attack value of 59. Personally checked. The only difference is when using racial mounts 

Hucbold wrote: ...but to suggest that I am lying

...sigh...

Hucbold wrote: ...must have made a mistake because my outcomes differ from your predictions is demeaning

1. No "must", just "might". Hucbold, do you always believe people who say something different from what you know? Never question their statements, never try to find out who's wrong, no?

2. Demeaning. You are heading into some emotional fields I am totally not interested in. I am sorry if yo ufind my words demening or humiliating but if you demand your opponent to believe your because presenting proof is 'demeaning' than honestly we have nothing to discuss.

Regards,

S.

1. the devs have not told us exactly what the algorithm is.

2. What you do get is a battle report that tells you what was killed.

3. So one can conjecture what the algorithm is - using the hints the devs have given us.

4. If a conjectured algorithm gives the correct kill ratio it is a candidate worthy of consideration.

5. If it does not it must be wrong.

I believe that is rational - a standard scientific approach.

On the two samples I provided, my conjecture worked. Yours 'failed' on one of them. To not reject your conjecture we need to come up with an explanation as to what was wrong with the experiment . Your suggestion is that the setup I claimed is not, for whatever reason, what I sent out. Okay. What I claimed was that I sent out all troops fully equipped with terrain gear and in addition a horse for the commanders and applied pres. Assuming your formula to be correct, the 'alleged' setup would have an attack score of 141805.6 and the max score that would give the correct kill ratio would be 112484.8. So my error would have to be greater than 17598.76 (12.4% of the total) up to 29320.85 (20.6% of the total). That's a big error. I can still check the pres bonus - its expiry date covers the date of the attack. So different equipment? Well the returned army is still there. Its commanders have the alleged equipment and its troops have the equipment and are fully equipped. The terrain type is as I stated. So I cannot see the error.

If I have misrepresented what you said, I am sorry. We appear to use language differently and in some places I do not find your exposition clear and have tried to interpret as best I can. My point was that sending any army against any other army is a valid experiment  and that a workable hypothesis has to predict the kills correctly in all cases. BTW I avoided boar spears as there are other people saying the figures on those are wrong.

I'm going to sign off at that. Others can do their own experiments. Good luck with your hunting - not that you need luck.

Well we seemto mainly disagree on point 4: in order to minimize error probability, the difference between attack and defence score should be minimized. The closer losses ratio to 100%, the more precise the outcome is. I'd still say I'm not convinced by your figures, they leave too much space for speculations.

Hucbold wrote: So I cannot see the error.

Since I have no means of checking the input data - ok, let's assume my method somewhy didn't work in that case.

I offer a tie:) Thanks for the conversaiotn in any case.

Ten Kulch wrote: [QUOTE=Sene] These factors disappear in any battle of meaningful size, but they will serve to muddy your results data on the micro scale..

We talk hunting, you talk PvP. I guess it goes without saying those are different matters, arn't they? While the basic algorithm, formulas and stuff are the same, those modes demand completely different approaches. Hunting is single player, while PvP is a multi-player campaign. i can't really get your point here

P.S (PreScriptum in this case:)

I'm still struggling to get your point, Ten Kulch. Since I'm not a born English speaker, I am asking you to be as 'plain words' as possible. Please imagine you are talking to your 8 year son and he is not a young einstein (I'm times older than 13 in case devs are reading).

In previous post you stated that (plz correct me if I'm wrong) proper experiment should involve:

1) both armies being of reasonable size;

2) one of them should be several times bigger than the other.

While I completely agree with 1), that 2) is totally escaping me. 

Now I'm not sure If you are familiar with that algorithm hunt. Hucbold precisely defined the basics in his 1-5 points of his last post. 

The intention - the HolyGrail of Illy attackers - (the way I see it) has always been to find the connection between all the factors influencing attack/defense score. Innoble in his post 6 years ago presented a clear formula that included terrain and divison bonuses - everything that had been introduced by the moment. Since then biome, pres and crafted were added on and the task was to squeeze them all in without loosing the formula's working algorithm.

In the battle outcome/attack score/def score equation the latter part is easy to calculate - well, at least when it comes to hunting. No crafted, biome, pres or divisional bonuses for the poor NPC fellows, they are only ifluenced by terain. Hence all the experiments first aimed to work the actual attack score back from the def/kill rate, and we are brothers in arms with Hucbold here. Just like him or many other guys I spent years attacking NPCs over and over again trying to extract attack score form the def one. Hundreds of email with guys like Naigoizati who is sadly abandoned few months back were exchanged discussing how to do it the proper way.

The problem there was that casualties rate is shown with integer number, whilst calculations regard it as float. The fractional part isn't shown and it has been quite a headache for it produces what I call errand range (just checked the dict and it seems i musunderstood the word, it should be just 'error range'. My apologies here).

That's the way the story goes for those who arn't too familiar with the issue. The reson I'm telling you all that is to clarify the following 2 points: the way I see it to narrow that error range one has to 1) use big armies - the bigger, the better and 2) keep the attack as close to def possible,  the kill ratio will be very close to 1 and def/kill ratio is as close to deff and hence to attack as possible. And yes, 1 division is a better case but it would mean less commanders hence less attack score hence more trrops to compensate hence much more expensive experiments troops-wise, which is a serious issue for single 'researchers' like me.

So I agree on 'big armies' point Ten Kulch but can't see the "different size". You say many divisions result in fractional casualties, and that is partially true. Partially firstly cause all troops are added up when outcome is being calculeted, secondly this factor is diminished when using big troops numbers. Now that was my main point in the discussion with Hucbold: I used 2 scenarios with 1901 troops / 100% casualties rate and 837 troops  100% - 1/837 casualtie rate. While Hubold presented 100 troops with 10 CR% and 80 troops with 9% CR.

So please let me repeat the question: why would you propose clashing armies with troops numbers differentiting greatly? Wouldn't that increase the error/deviation?

P.S. (proper 'post scriptum' now) I get a bit graphomaniac sometimes, please let me know if do. Cheers.

So even when say classic 61 elite army is deployed with 5 commanders  and 57-1-1-1-1 troops allocation, and they win the battle with 45.5% or 27,76 rounded up to 28 casaualties, the divisions will suffer 26-0-1-1-0, with 2 single lost unit divisions being chosen at random. I'm not too sure if It could be 27-1-0-0-0 or even 28-0-0-0-0 either (this one is unlikely I'd say from my experience), but in any case that will not alter the total casulaties namber and rate. That's why I would upvote you 'big armies' point to eliminate fractional error, but thumbs down for single division. Or thumbs sidewards rather: while it does make life a little easier computing-wise, it makes no impact on the outcome.

You do have a good point, Ten Kulch, in proposing 'suicidal' (for attacking side) battles, hypothetically that is the best way to take computational error away. I did use failed battle reports to get the precise attack score, but in practice the bigger issue was not finding attack itself, but rather finding calculation model that gives the same result. Since I did found it and I'm sure beyond any reasonable doubts its true, I ceased any experiments and focused on precalculating battle reports ranging from 61 to thousands of units deployed. I think my maximum was around 3,5k stalwarts and 2,5k runeriders deployed, with 1 or 0 survivors left, can't remember for sure. Not that you have to believe my words, guys.

Lucks,

S.

Ah, it must've been "... but I'm ready to provide battle reps if ncessary".

PS Checked some battle reps, seems like "26-1-1-0-0" allocation in my last post's example is the only real case. 1-1 still randomly chosen. The same goes to survivors, actually. When casualties ratio is close to 100%, some divisions will loose all guys and few left troops will be randomly awarded to the remaining ones.