User blog:Windyplains/A Troopless Tree System

Quoted: ''The troop system is being reworked in two ways. Firstly, two factions are being added (earliest is v0.16), and troops are being redone for the existing factions. Dawg of War is leading this design and balancing effort. Secondly, the troop tree system is being removed (upcoming in v0.15) to allow for more units with more uses instead of the two or three "top tier" units in each faction who are superior. (I may be wrong with my understanding, so I'd suggest waiting for Windyplains' explanation.)''

Diessa has stated things correctly here. There are a lot of plans I have for the future of Silverstag's design which is why it has the 0.14 version vs. 1.14. The mod stays in a generally release ready state so that I can address issues quicker and with smaller, incremental changes, but the overall shaping of the mod has a long way to go yet. My preference is to let everyone play it as it is developed in what would best be described as an open alpha so that you can help me weed out the bugs and guide the balance of the work. This does mean that certain features remain unchanged until I get around to them and that some features are continually getting revamped as they are developed in stages.

One of the goals I've had for the mod stems long ago from my dislike for having to upgrade troops all of the time. I knew that I wanted to change the system and I had some ideas how, but I had not yet quite put my finger on the why. That reasoning was better stated by Hanakoganei last year and helped to pinpoint a few issues with the upgrading design:
 * AI & Player use different troop recruitment systems. While it is often a necessary evil for finding the balance of a game it is usually an annoyance to players. The other problem is that for a feature to exist it must have an AI version and a player version which means twice the work for me.
 * Faction identify and troop focus is fixated at the top tiers. Folks may use troops of lower tiers, but only as stepping stones towards the "top tier" troops in a tree. Player parties run around with all knights while the AI simply wasn't designed to deal with that kind of opposition which leads a lot of players to feel the game becomes too easy.
 * It doesn't make sense. As you fight battles it makes sense that your troops would learn from these engagements and be able to apply some of this newly acquired experience in becoming better soldiers. What doesn't make sense is that they also change out their gear and/or acquire a horse along the way in mass numbers. An army may have supply personnel, including smiths, along for the ride, but not on that kind of scale.
 * It is a grind. After most fights you'll see the "some of your troops may be able to upgrade" message and so you begin the process of going through your party panel and clicking away to upgrade each troop to its next link in the chain. I always found this to be more of a chore then enjoyment after its initial novelty wore off.
 * The tree system requires extra troops simply to support branching. If you want to have a troop of a given type then it needs to exist on the tree's branches somewhere. Typically it makes sense that it upgrades from something that shares similarities with it so that it appears your troops evolve into this later troop type. The problem is that if you want to have more viable troop types then the tree has to have more branches and more filler troops early on to support these branching points. What if I want to give you an option for Rhodoks to have a bowman, but just a relatively weak one? Should this be a branch that leads off early and dies or should it be a predecessor to their crossbowman which makes less sense?

The answer to this is simply scrapping the upgrade system entirely and making the troops individual classes. If you take the time to read through Hanakogenai's original thread that details a desire to do this very idea then you'll see a fairly supportive stance from me and this is why. My approach will be different and focuses on a different reasoning, but the genesis of the idea is the same.

I have less issue with wanting to force a more tactical game vs. wanting to allow more troop diversity and situational access to troops. The new system I've developed is part of the overhaul to how you'll deal with towns and villages and looks a little like this:



What you're seeing is the new recruitment screen for how you'll acquire troops. Each village, castle and town will have a pool of peasant recruits to draw from that you can pay to directly train into a role. That means you can take a raw recruit and train them to become a man-at-arms directly as most of their upgrade lies within the equipment. Being able to do this, however, means that they'll cost considerably more as you are not paying for upgrades or having to keep this troop alive in order to get that far up in the upgrade process. Also this list will allow me to setup unique troops available in certain locations, do so easily and in a way that the AI uses just as the player does.

Here's a few ideas on what is easily possible with this new design:
 * Knight orders could be created to show on the list if the player has completed prerequisite quests or met a certain relation value.
 * Bandits can be made to show up on the list if the player has a negative honor.
 * Unique troops that are just a flavor of the area can be made to only show up their respective towns

These are just a few examples for now as I've run out of time before work. For now I'll have to run, but I'll add more later and as you can see from the screenshot there are other pages yet to cover.

In the previous post I covered some about the generic change itself and it mostly centered around the reasoning for the concept so this time I'll try to explain a little more of the nuts and bolts on how the concept will function. Hopefully that will alleviate some of the questions on how this can shape things.

Generically we use the tier term to describe how many upgrades it took to get from Troop A to troop D and this gives us a rough estimate of how tough they are. Just how rough an estimate really depends on how well the modder's eye for balance was paid in making sure that T4 troops are better than T3 troops. They may not be better in the same thing, but they should generally represent an improvement in overall quality. The reality is that I can make a T3 troop that is vastly superior to a T4 troop, but is just a lower level and they'll work out just fine in the standard upgrade system. I'm sure everyone has seen examples that at least loosely fit this situation. Since we've moved to a system without upgrades then the question of how a player determines the general usefulness of a troop comes into play.

Since you can't compare the abstract tiers of an upgrade system, I've instead instituted a rating system that figures out useful your troop is in combat. Let's review the picture of our Swadian Recruit in that last post:



He has a lot of information displayed here in very generic terms so I'll try to break it down into a bit more detail. With the release of v0.15 I fully intend to have wiki pages built covering this kind of info. For now I'll keep it a bit more generic as specific numbers are bound to get adjusted as we test the mod prior to release.

"Level 4 Faction Troop (T1)" - Obviously, he's a level 4 troop. More specifically he's a member of what is considered a faction's troop list which has implications for your Party Unity score. This is why that is made a generic reference. He could be labeled a mercenary or an outlaw as well. The (T1) information labels him as a tier 1 troop. Since we won't be limited by how many jumps it takes to get here the term tier will be used to reference the general strength of a unit. This will be on a scale of 1 to 10 and the tier is dynamically set based upon the strength of the unit as a whole, their rating, and I'll cover that lower. For reference sake a Swadian Knight weighs in at T6 on that 1 to 10 scale if that gives you an indication of where we're going.

"Armor Rating: 160 (+253)" - The mod already has an algorithm for assign a relative usefulness value to items based on each of their factors. For a shield this means it compares the shield's health, resistance, size and speed comparing these to a general "baseline" value and comes up with a total "relative usefulness" score. You've already seen these scores whenever your companions want to upgrade from one item to another as it is shown in 's following the item name. This rating shown here is the combination of the average item rating for each body + helm + foot + hand armor. So if a troop has 3 body armors then his "body score" is the average rating for them which is then added to the other average ratings. This poor fella at 160 is a very cloth wearing individual. A native Swadian Knight on the other hand is sitting around 755 by comparison. The (+253) represents that he may carry a shield and what the value of that shield's rating is. Shields have more factors so they tend to get an inflated comparison score and they are separate so that someone like this fella in cloth doesn't have a rating of 413 where most of that is his shield and looks as good as someone with that same rating that doesn't have a shield. I felt using the (+253) would be easier to read instead of having it say "Armor Rating: 160-413".

"Melee Rating: 456" - Similar, but not the same to how armor ratings are calculated. Here the melee weapons of a troop are considered and then they are compared against their proficiency to use that weapon. This score is boosted by having a stronger weapon or a higher proficiency with using the weapon you have. If a troop does not carry a weapon of a given type (such as two-handed) then it will not be considered and its proficiency in that type is ignored. These one-handed, two-handed and polearm ratings are then averaged to see how strong is the average melee strength of this troop when one is generated. So a troop with a higher melee rating is generally always going to be more effective in melee than one of a lesser rating.

"Ranged Rating: 417" - This is calculated exactly like the melee rating only it is looking at bows, crossbows and thrown weapons. This factor is ignored if a troop does not carry a ranged weapon.

"32 denars (-6%)" - This is the training/recruitment cost of the troop (described below). The -6% represents the discount you receive in cost based upon improvements you have built and your trainer skill (described below).

"Peasant Recruit - This covers basically every standard troop in the game and they work as a form of currency for buying troops. When you visit any location the list of troops shown will always include every troop type currently in your party, but if they could not be hired here then they will only be able to be dismissed. If you dismiss a troop you will not get back the money you spent training them or purchasing gear, but you will gain back a peasant recruit at that location as this value is stored individually for each fief. These recruits are refreshed as time passes much like how village recruits are in native warband.

Since troops no longer upgrade based on experience earned this awesome skill has lost its usefulness entirely. I consider the purchase cost of a troop to represent the training & gearing of a new recruit so I have redesigned this skill to provide a reduction in that cost that is [i]application in every location[/i]. This reduction is 2% per point invested in the trainer skill. Some improvements will also lower these training costs, but they're specific to their location. With this change the trainer skill has shifted from being a (personal skill) to becoming a (leader skill). It does not stack with your companions in any way. I may eventually reconsider that stance, but if it did the stacking would be very limited.

This doesn't, however, mean that leveling up the trainer skill on your companions is useless. Companion advisors directly take usefulness from the trainer skill. Since the AI also receives this reduction benefit then so do your companions that have been promoted to vassals. A promoted companion with a 10 in the trainer skill can enjoy a significant reduction (-20%) in purchasing costs which will make his option for troops much better.

Final note on the trainer skill is that it will still have both a readable and reference book available for your benefit, but the cap will still remain 10.

The native cost for troops always hinged upon recruits costing 10 denars and then you paid to upgrade them through each tier. The hidden cost was that they had to survive the process before they could become whatever end troop you had in mind. So let's analyze that cost really quick... So if I were then to tell you how it is now going to cost you closer to 1100 denars for that very same knight you might be understandably confused. Except in this exchange you desired a top tier troop and you received one without any risk of investment loss in-between due to death nor having had to pay weekly wages for that troop while awaiting it to upgrade. This price for purchasing a recruit is dynamically generated based on their equipment, skills, attributes, proficiencies and if they are ranged or mounted. I'll speak more on the economy and its role next...
 * +10 denars to purchase the recruit.
 * +20 denars to upgrade to a tier 2 - level 9 troop.
 * +40 denars to upgrade to a tier 3 - level 14 troop.
 * +120 denars to upgrade to a tier 4 - level 21 troop. I even get a horse in this upgrade.
 * +160 denars to upgrade to a tier 5 - level 28 troop.
 * 350 denars total for a tier 5 swadian knight.
 * 350 denars total for a tier 5 swadian knight.

Funding The Silverstag economy of v0.14 is very easy with tournaments, prisoners and enterprises you can develop a solid income that can fund a top tier army all of the time. In v0.15 my goal to shift budget back into being part of the challenge you face in developing your realm. While there are many cash sinks being added to the mod in the form of improvements, better equipment and buying troops, I want these choices to be a choice of what is best for your goals vs. an easy "I can afford it all".

Thankfully your enemies are facing a similar downturn in the economy so you might just find that stocking your army with 5 T3 troops for the same cost of that 1 T6 troop might just be a good idea. It is my hope that as we redesign the troops you'll see the lesser "tier" troops for what they are. Cheaper, but less trained and equipped manpower, but ideally suited for helping fill out your army. If you wish to ride with only the best then there probably won't be that many of you riding, but that is a call for you to make.

The end result is that you should be able to wage war on an economic level as well as the battlefield and so should your enemies. Smash an enemy party in the field and he'll live to rise up again, but it'll cost him a pretty penny and he may not be able to afford those shiny troops this time around.

Manpower Native warband tends to treat people as a very disposable commodity. You hire these poor villagers with claims of glory only to feed about 90% of them into the grinder while the few that make it through get upgraded. Most armies of the past could not have afforded to fight a war of attrition with that kind of recklessness. People simply do not multiply like orcs in a fantasy story. Instead I'd like you to consider is it worth the risk of losing men in that engagement? Certainly you'll have to take losses to defend and expand your territory, but since you won't specifically have to grind enemy parties for the experience anymore you can be a little more careful with when and how you fight.

Deny an enemy fresh recruits and his choices will become limited. Raided villages will supply no recruits. The AI use a collective pool that can be whittled down if they're heavily pressed and only rebuilds so fast. Think of it like the garrison recruitment system of v0.14 only it happens everywhere all of the time.

Events The final lock that prevents you from recruiting just anyone is what you've done within the game. The below examples are just that as they may or may not exist in v0.15, but they are places I intend to go with this system.
 * Did you piss off this faction? Well no recruits for you then.
 * Did you help this guild of merchants out? Maybe they'll let you hire a merchant advisor troop that helps ease your wage costs by showing you how to improve efficiency in your warband. Try not to let him get killed as they may not appreciate that.
 * You actually spent the cash to build a mercenary chapterhouse in your town? Maybe they'll let you hire their mercenary types within that town. Helped that poor village out?
 * Announced your town as a safe harbor for thieves? Maybe you can live out that bandit king story by hiring only bandits from that town.

This won't be in v0.15 as I want to make sure the setup works properly first and try my best to iron out the kinks while I give Dawg time to flesh out his idea, but I will give you a little insight into how this system fits with factions. We're not getting rid of "non-top tier" troops so you're not going to see faction lists of solely knights, sergeants and sharpshooters. Instead factions will have so many troops of a given tier. For the sake of an example I'll use a spread like this...

Faction A gets...
 * 1 - tier 7 troop options
 * 2 - tier 6 troop options
 * 2 - tier 5 troop options
 * 3 - tier 4 troop options

Each faction will have the same number of options available to them. Where they choose to develop troops is what sets their military culture. So Nords are still going to favor infantry and khergits will still favor cavalry, but now they'll be able to have new options that wouldn't have fit on their troop trees in the native setup. Troop types that fit within their faction and look like they belong as a part of that faction, but are limited in how strong they can be by their tier.

That may seem intentionally vague, and it is, because this is Dawg's territory and I'd rather let him showcase it. Hopefully that answers the basic question of faction design though.

Okay, folks, I'm pretty much exhausted so I'll wrap it up here. I hope that answers folks questions on the vision I have. Only time will tell how well we can pull it off and I hope you all will continue to give us the same patience that you've shown so far as we tweak, balance and restructure the mod to arrive at this destination.