Horrible Combat Mechanics
Printed From: Illyriad
Category: Strategies, Guides & Help
Forum Name: General Questions
Forum Description: If your gameplay question isn't answered in the help files, please post it here.
URL: http://forum.illyriad.co.uk/forum_posts.asp?TID=3528
Printed Date: 17 Apr 2022 at 07:34 Software Version: Web Wiz Forums 12.03 - http://www.webwizforums.com
Topic: Horrible Combat Mechanics
Posted By: Gasterix
Subject: Horrible Combat Mechanics
Date Posted: 07 May 2012 at 23:21
The combat mechanics in this game are horrible. I'm sorry, there is no other word for it.
Try sending 50 Charioteers against 5 Dogs, Now open http://illyriad.varda.nl/battle/ and send 5000 Charioteers agaist the same 5 dogs. Same result.
I just received a combat report for attacking 4000 Longbows with 50 Charioteers, result: you failed, so you died. No enemy casualties. Fair enough I guess, if you have an elevated opinion of the Longbows range, penetration and firing speed; or maybe because the Charioteers were not running fast enough next to their non exsistant chariots.
Warfare has always been about fire and movement, everything else is seconday. Of the seven principles of war, terrain is just one. The principle that wins battle is combined arms, it started in 3500 bc and it's the same today. A bunch of unarmed peltasts (javelin wielding light armoured militia type troops) routed a group of Spartans because they had mobility,
You do not calculate battle winning odds by comparing stats (and your stats are a little weird). Light cavalry will try to attack missile troops, or at least keep them at range. Spears will be on the flanks ready for cavalry. Heavy cavalry will either attack the center to break formation (if they have stirrups) or try to ouflank after they have dealt with opposing cavalry. Swords and shortspears will hold the main line and/or attack enemy infantry. It's all about rock-scissors-paper with a few missile equalizers thrown in.
It takes 10 minutes to read up on this on Wiki (don't trust it all) and I bet any business analyst with respect for himself could write the pseudo-code needed in 30 mins.
So, why are we stuch with this garbage?
|
Replies:
Posted By: HonoredMule
Date Posted: 07 May 2012 at 23:45
The "stuff" we have is the result of a philosophical avoidance of "rock, paper, scissors" balance which, in large numbers abstracted to a simple algorithm, actually results in oversimplification and promoting blanket use of mixed armies.
Illyriad's goals happily sacrifice realism wherever it conflicts with the greater goal of fun and tactical diversity. And it is quite right in avoiding rock, paper, scissors balance, as doing so merely reduces tactical and strategic decisions to an elaborate 50/50 gamble or the consistent mediocrity of aforementioned mixed armies. In a mess like that, all that ultimately survives is the general principle that the biggest army wins. Many other browser-based games have in just that fashion reduced tactics to a farce, and I'm far happier that my tactical choices have actual impact that is both more diverse and more profound.
I had some sympathy for your point of view right until you spoke on what a "business analyst" could accomplish. That alone shows how little you comprehend the complexity involved in reducing realistic warfare to a scalable and efficient multi-variable agorithm that uses minimal lookups and produces near-instantaneous results without taxing a busy processor.
You need to understand the difference between simulating battle and modelling the results. This difference (and the fact that the diversity is applied through the defending unit instead of the attacking one) are the primary reasons that the games unit stats tend to differ from what "realists" expect. The results are actually quite appropriate, especially when you take into account the effect of guerilla tactics (which the presumed inappropriateness of your examples seems to discount).
Ultimately, the results do sometimes differ from what many would consider realistic. But it is a mistake to presume that realism is the primary goal to which a fantasy game should aspire in the first place, let alone do so ignoring how broadly even experts in a non-fantasy setting would differ on what constitutes "realistic."
------------- "Apparently, quoting me is a 'thing' now." - HonoredMule
|
Posted By: Albatross
Date Posted: 08 May 2012 at 01:20
Agree with HM; every calculation is a compromise on reality, and we're just modelling rather than simulating. This isn't Rome Total War (nor its successors), and for good reason: the Illy servers would simply choke if they were expected to process detailed simulations.
I think the level of detail is fine, but not overwhelming. For every combat approximation, someone will always shout that it's not realistic enough because it doesn't take account of x or y (think of silly examples yourself).
When we play a (modelled by implication) game, we accept the symbols and the rules of the game world, suspending our disbelief. We work out how to play by the game's rules and mechanics. Personally, I would only be disappointed with the modelled reality if I felt cheated by the inequality of players' understanding of it, e.g. (a) most other players understand the game better than me, or (b) I wished for a far more complex model that I was capable of understanding, while others wouldn't understand it.
Gasterix wrote:
... any business analyst with respect for himself could write the pseudo-code needed in 30 mins. ... |
I'm one of those analyst/coder types, and I agree that fairly workable combat systems could be written on a napkin. However, I wonder what proportion of those business analysts could then refine the variables and 'run the numbers' to ensure good game balance while getting 95% of cases correct-ish, and code it to run in a few milliseconds in a safe transaction on a busy server.
Ideas can be expressed in a few minutes. Fitting those ideas into a game can take 10000 times longer.
|
Posted By: Rill
Date Posted: 08 May 2012 at 03:09
|
Your understanding of the mechanics of sending 50 bows vs. 4k troops is incorrect. You were crushed so totally that you did not receive a report. That does NOT mean that there were no enemy casualties. You could scout before and after to determine the number of casualties inflicted if you wish.
Illy combat mechanics tend to be less responsive to "gimmicks" than those of many other MMORTS I've played (such as zerging with spears in Evony). When a gimmick is discovered, as with the 10-10-10 armies seen in the last tournament, the developers are proactive about closing the loophole.
There is no one "guaranteed" strategy that will always win in Illy, and that makes the game far more interesting than many of those with the allegedly complex battle mechanics you profess to admire.
You are a relatively new player whose main understanding of battle mechanics seems to come from reading the wiki rather than from actual experience. Perhaps you should build your cities and create armies, then employ them in a variety of tactical and strategic situations; this might allow you to come up with a more intelligent and articulate critique of the battle system.
|
Posted By: Diomedes
Date Posted: 08 May 2012 at 08:50
Well said, Rill. I, for one, abhor the cold analytical processes that seem to excite a small minority of players in Illy. The essence of the game is surely to experiment with what might work in practice - which, if one thinks about it, is more in tune with real life. I thoroughly enjoy the inherent uncertainties of combat, and would lose interest quickly if I had to depend on a calculator before sending troops into combat. "Go, my brave warriors - go and kill, and if perchance you fail, I'll just send some more." 
------------- "Walk in the way of the good, for the righteous will dwell in the land"
|
Posted By: HonoredMule
Date Posted: 08 May 2012 at 12:46
I wouldn't want to discourage the use of such calculators as means of experimentation and understanding gained at reduced cost. New players will especially learn a lot if they pit specialized forces of relatively similar strength against each other while playing with the choices of specialization and terrain (and which side attacks).
I've never made much use of them myself, but then I gained a feel for combat tactics before such tools were available, and my position more frequently relies on words than troops.
------------- "Apparently, quoting me is a 'thing' now." - HonoredMule
|
Posted By: Gasterix
Date Posted: 09 May 2012 at 18:56
Thanks for the feeback. I agree that the comment on business analysts was a little below my standard. Hold on there ... my standard is pretty low.
(thinking about the best way to respond to get my point across; without doing collateral damage to the Illy dev team).
|
Posted By: Rill
Date Posted: 09 May 2012 at 18:58
Gasterix wrote:
Thanks for the feeback. I agree that the comment on business analysts was a little below my standard. Hold on there ... my standard is pretty low.
(thinking about the best way to respond to get my point across; without doing collateral damage to the Illy dev team).
|
I've met some of the devs, and they are pretty sharp people. I'm pretty sure they can handle anything you might choose to dish out, although your motivation for doing so would be murky at best.
|
Posted By: Rill
Date Posted: 09 May 2012 at 19:06
HonoredMule wrote:
I wouldn't want to discourage the use of such calculators as means of experimentation and understanding gained at reduced cost. New players will especially learn a lot if they pit specialized forces of relatively similar strength against each other while playing with the choices of specialization and terrain (and which side attacks).
I've never made much use of them myself, but then I gained a feel for combat tactics before such tools were available, and my position more frequently relies on words than troops.
|
I think using battle calculators is great for many people, although it's not to my personal taste. My comment was more in reference to the complexity of planning a campaign in Illy with regard to coordination of forces, choice of troop types to build in the first place and then deploy, use of Raid vs. Attack stratagem, decision to reinforce a city to Sally Forth vs. a siege rather than using repeated Raids, that sort of thing. I think that these are the situations that can't really be captured by a battle calculator that reveal the depth of Illyriad gameplay.
My point was to encourage Gasterix to experiment with some of the complexities of actual troop use so that he could gain an appreciation for these factors, which in my mind outweigh the question of whether combat is all-at-once or round-based. My intention was to encourage engagement with and enjoyment of the game as a gestalt, the whole being more than the sum of its parts.
And Gasterix, I do hope that you will stick around and do so.
|
Posted By: Gasterix
Date Posted: 09 May 2012 at 19:18
|
quick question: is HonoredMule a Dev in disguise? (it actually rhymes)
|
Posted By: Rill
Date Posted: 09 May 2012 at 19:25
Gasterix wrote:
quick question: is HonoredMule a Dev in disguise? (it actually rhymes)
|
Nope.
And my initial reaction was that it was alliteration rather than rhyme. However, I looked up the definition of rhyme, and it can include various kinds of phonetic similarity, including alliteration, so that shows me.
|
Posted By: Gasterix
Date Posted: 09 May 2012 at 19:28
HonoredMule wrote:
The "stuff" we have is the result of a philosophical avoidance of "rock, paper, scissors" balance which, in large numbers abstracted to a simple algorithm, actually results in oversimplification and promoting blanket use of mixed armies.
Illyriad's goals happily sacrifice realism wherever it conflicts with the greater goal of fun and tactical diversity. And it is quite right in avoiding rock, paper, scissors balance, as doing so merely reduces tactical and strategic decisions to an elaborate 50/50 gamble or the consistent mediocrity of aforementioned mixed armies. In a mess like that, all that ultimately survives is the general principle that the biggest army wins. Many other browser-based games have in just that fashion reduced tactics to a farce, and I'm far happier that my tactical choices have actual impact that is both more diverse and more profound.
I had some sympathy for your point of view right until you spoke on what a "business analyst" could accomplish. That alone shows how little you comprehend the complexity involved in reducing realistic warfare to a scalable and efficient multi-variable agorithm that uses minimal lookups and produces near-instantaneous results without taxing a busy processor.
You need to understand the difference between simulating battle and modelling the results. This difference (and the fact that the diversity is applied through the defending unit instead of the attacking one) are the primary reasons that the games unit stats tend to differ from what "realists" expect. The results are actually quite appropriate, especially when you take into account the effect of guerilla tactics (which the presumed inappropriateness of your examples seems to discount).
Ultimately, the results do sometimes differ from what many would consider realistic. But it is a mistake to presume that realism is the primary goal to which a fantasy game should aspire in the first place, let alone do so ignoring how broadly even experts in a non-fantasy setting would differ on what constitutes "realistic."
|
HonoredMule - thanks for this. I don't think the technology platform this game is build on (not that I know anything about it) provides the flexibility you are looking for. Rock-paper-scissors is a good way to start in my humble opinion. Terrain bonuses is not.
I should not have mentioned the examples as I'm not really looking for realism. Just something that makes a little sense. I hammered the range- speed combat mechanics in Evony back in 2009 but at least it provided tactical options. In Illy the tactical option of "Feint" means that your army returns after it has arrived. That is not my understanding of how to feint. Just as an example.
|
Posted By: Bonaparta
Date Posted: 09 May 2012 at 19:31
|
Combat mechanic in Illy is great.
I actually like the fact that even small forces can damage huge amount of troops, which is by the way totally realistic.
Gasterix when you will gather some experience, you might change your mind about combat mechanic. No offence, but attacking with such small forces is really not relevant to large battles and especially not to alliance versus alliance battles...
|
Posted By: scaramouche
Date Posted: 09 May 2012 at 19:32
Diomedes wrote:
Well said, Rill. I, for one, abhor the cold analytical processes that seem to excite a small minority of players in Illy. The essence of the game is surely to experiment with what might work in practice - which, if one thinks about it, is more in tune with real life. I thoroughly enjoy the inherent uncertainties of combat, and would lose interest quickly if I had to depend on a calculator before sending troops into combat. "Go, my brave warriors - go and kill, and if perchance you fail, I'll just send some more."  |
Nicely put.
|
Posted By: Khèlbên Khál
Date Posted: 09 May 2012 at 20:41
|
Warfare is invariably about having more men around at any stage who haven't hid or run away yet than the other guy. What they have in their hand is almost irrelevant.
|
Posted By: Babbens
Date Posted: 10 May 2012 at 08:03
Like the Carthaginians at Cannae in 216 BC? Or the Germans at Teutoburg Forest in AD9? Or the English at Agincourt in 1415?

|
Posted By: Drydenn
Date Posted: 10 May 2012 at 11:33
HonoredMule wrote:
You need to understand the difference between simulating battle and modelling the results. |
What is the difference between these two things? Anyone? Preferably without using math that will make my eyeballs bleed? 
- Drydenn
|
Posted By: Albatross
Date Posted: 10 May 2012 at 12:17
|
Simulation takes account of all the detailed events that happen within the process; modelling guesses the results by making general assumptions.
It's really a sliding scale, rather than being one or the other: | REALITY --- SIMULATION --- GENERAL MODEL --- (no resemblance) |
|
Posted By: HonoredMule
Date Posted: 10 May 2012 at 13:15
Simulating, as Albatross implies, involves taking a battle through its paces actually processing what happens in several steps (archers soften target, then cavalry charges center, then infantry moves in, etc.--or at higher levels of detail, soldier one of 24M holds left corner of a center box formation until stubbing his toe which causes the man behind him to stumble outside the shield protection, taking an arrow...). Basically, think of flight simulator, where realism is carefully sought in real-time.
Modelling is (mathematically) based on statistical analysis, where the differences between many wildly different situations is reduced to the smallest and simplest set of variables and an algorithm is artistically produced which takes those basic inputs (troop and terrain type, equipment and training, etc.) in the form of figures which generically quantify them (terrain modifiers, unit atk and def stats, whether attacking or defending), and skips straight to an assumption of results which hopefully match what would be the statistical average result of any realistic battles that qualify as satisfying the input parameters.
To be truly modeling realistic battles the algorithm would have to also consider things like standard deviation which would take into account the massive variability of results in real-world warfare--especially when the inputs are so simplified as to leave many whole categories of input unconsidered altogether (starvation, disease, luck, etc.). But that wouldn't make for a good game mechanic. I think we'd all be unhappy if the consistency of our results regularly varied by 50% or more.
------------- "Apparently, quoting me is a 'thing' now." - HonoredMule
|
Posted By: Rorgash
Date Posted: 10 May 2012 at 14:46
|
i brought up something a while back about another browser games that has a more realistic combat system, now one of the problems that was brought up here was that a more detailed combat was server load, i dont know but this game made the combat take time, a normal sized battle 4k vs 4k troops would take 4 hours or so, so i would guess that this time was what they added to make up for the extra server load.
i dont mind the current combat system, but i would like a longer combat system where you can reinforce a ongoing battle to help your friends. but thats a huge update which im sure wont come soon :P
|
Posted By: Rill
Date Posted: 10 May 2012 at 18:19
|
Thanks for the description of simulation vs. modeling, I learned something new!
|
Posted By: HonoredMule
Date Posted: 10 May 2012 at 18:38
Calico, I stand with you in wishing battle was prolonged and thus different armies would "pile in" to the same battle with the added excitement and realism that would provide--the magnitude of their combined effect depending on the timing of their arrival as well as their strength.
In fact I was the first to make such a proposal, which I outlined in as granular and practical/thorough detail as I could muster without having access to the existing code. I'd love to see this happen.
Just to clarify, however, this does involve converting the system from modelling to simulating. In fact the prolonged version would not really take any more processor time. Each new arrival and departure would trigger a recalculation, but so does each new individual battle in the current setup. There isn't actually anything "happening" between such events, similar to how the server only knows your current resource levels at the moments when you manually observe or use them.
------------- "Apparently, quoting me is a 'thing' now." - HonoredMule
|
Posted By: Drydenn
Date Posted: 13 May 2012 at 14:31
Rill wrote:
Thanks for the description of simulation vs. modeling, I learned something new! |
I second that statement!
------------- - Drydenn, Diplomat for the Free Trade Guild & Trade Assembly member of the Illyriad Trade Union
|
Posted By: Salararius
Date Posted: 13 May 2012 at 16:44
|
I hesitate before posting this because on the one hand I don't want to contradict anyone but on the other I'm not sure this whole simulating/modeling thing is right. I always learned (as a controls engineer working in modelling and simulation at NASA) that a model is a physical or mathematical representation of "thing". Modeling was the act of building/creation/development a model. It wasn't some derivative of a simulation or a simpler simulation process.
Simulation was when you took a "model" (or models) and used it to represent a physical process (you add the time element).
Similar to the difference between a 'noun' and a 'verb'.
You model an airplane, you simulate flight. You model soldiers, weapons, terrain, weather, emotions, strength, endurance, training, etc... you simulate battle. Etc...
The detail of the simulation (ie. how many levels of models used, the complexity of the equations or the number of iterations required to obtain results) was irrelevant to the definition of a simulation. Sometimes a simple model will give very accurate simulation results (example, the mathematical model F = G * m1 * m2 / r^2 works near perfectly and is used to simulate gravity very well despite it's simplicity). Then there are things like global warming where we have many, many very detailed models that when put together do not simulate weather as perfectly.
I realize this definition doesn't help to explain what HM was trying to convey so it really won't help anyone in the context of this thread. I believed that people felt they were learning something bigger in this thread and that didn't seem to be so. Just because Illyriad uses a simple model that can simulate battle in one calculation doesn't mean it's not a simulation.
|
Posted By: Denizar
Date Posted: 23 May 2012 at 07:00
HonoredMule wrote:
Calico, I stand with you in wishing battle was prolonged and thus different armies would "pile in" to the same battle with the added excitement and realism that would provide--the magnitude of their combined effect depending on the timing of their arrival as well as their strength.
In fact I was the first to make such a proposal, which I outlined in as granular and practical/thorough detail as I could muster without having access to the existing code. I'd love to see this happen.
|
I also believe that a "pile in over time" combat mechanism would be more interesting. That's what I always loved about the Europa Universalis game system.
But, it would be tremendously more complicated from a state management point of view. They would need to rework a large amount of code. I'm talking about HUGE orders of difficulty compared to the current instant results model.
The difficulty isn't in the actual "pile in" combat code (the algorithms for resolving combat over time), the difficulties would be in managing the suspended state of the combatants. The combating armies would need to be placed in a "combat limbo" state. Doing that could be fairly easy, the tricky part is you have to rework all of the code that works around and could affect those armies, commanders and resources while they are in "combat limbo". There would be a million opportunities for interesting race conditions.
|
Posted By: Avion
Date Posted: 24 May 2012 at 15:29
Denizar wrote:
The difficulty isn't in the actual "pile in" combat code (the algorithms for resolving combat over time), the difficulties would be in managing the suspended state of the combatants. The combating armies would need to be placed in a "combat limbo" state. Doing that could be fairly easy, the tricky part is you have to rework all of the code that works around and could affect those armies, commanders and resources while they are in "combat limbo". There would be a million opportunities for interesting race conditions.
|
This brings to mind turn-based game mechanics - I hit you, then wait while you hit me, then I hit you, etc., etc., with the game recalculating hit points, defense, etc. in between turns. This approach has the advantage that it allows the possibility of a new player joining at any turn, or reinforcements being added, or a player retreating and so on. If a turn was an hour long and it took, say, 48 turns to resolve a very large battle (1 or 2 for small battles), then maybe this would simulate "combat limbo." Since a player may not be around to monitor the battle constantly, perhaps some sort of (semi)automation could be added such as retreat minimums per division (zero = kamikaze), or auto-reinforcing if numbers fall too low, etc. Alliance members could join in at strategic moments. I don't know how satisfying this approach would be but I suspect it may not require elaborate coding since events are not happening in "real" time.
|
Posted By: Innoble
Date Posted: 24 May 2012 at 15:54
|
a lot has been said in this thread. My two cents:
First: I think the combat mechanics are very simple. This is a choice by the devs. They could have made it more complex, but then it would require a lot more effort to code and require more server resources. It would also be harder to balance.
Second: Combat in Illy is deterministic. If you know all the factors (troop numbers, commander stats and terrain) there can always be only 1 result. There is no randomness. Many people like to be surprised. A bit of thrill here and there is good. For most Illy players, the lack of understanding is enough thrill. You can't really predict the outcome if you don't know how it works. If you do know how it works, then I suppose it might be boring to already know you're going to win or lose beforehand.
Personally I think there are many things that can be improved (and by improved I mean expanded) about the combat system, but I feel it should come after:
-magic being made useful -trade v2 -faction AI -pathfinding -several annoying bugs that still aren't fixed (tavern for example still doesn't work!!!)
It's all about the priorities!
|
Posted By: Rill
Date Posted: 24 May 2012 at 18:01
|
I think the key is that you only know the result if you know all the factors. What makes combat in Illy exciting is that in any PvP situation (and increasingly in PvE situations), we don't know all the factors. We know terrain, we may have insight into troops from scouting, and we can perhaps make a good guess on commander skills, but in any extended war or tournament, there's plenty of uncertainty for me.
With the minor exception of the case where one is performing a sally forth from a besieged city from which one just got an accurate scout report on the encampment, the non-terrain variables in Illy PvP can be predicted but not known.
|
Posted By: Sloter
Date Posted: 24 May 2012 at 18:33
|
Nice post Rill, i agree with you.Calculating numbers are only 1/10 of any battle betveen real players.And with new diplo changes that devs will make there will be even greater element of surprise.
|
Posted By: Ander
Date Posted: 24 May 2012 at 18:47
|
There are also situations where you do not have time to wait for a scout report.
|
|