FPI - limiting unit levels a contradiction?

[Blackbeard]

I can't understand why graphics is given as the main reason for limiting unit levels, but not a problem when creating new unit types.

I'd rather see more graphic effort spent in adding levels to a few unit types than creating hundreds of types with little or no advancement prospects.

I think it contradicts the game's purpose when developers freely expand on unit types, but limit unit levels.

My first thought when told about graphics limiting levels was,

* 'Hey, why not use the same graphic for all levels and change the hp bar and orb?
* Use the computer to change the unit's graphic colour and use a fixed palette for each faction?
* Change the orb to different shapes (triangle, pentagon, etc) and colours for each level range?
* Programming in a level-to-palette mapping (alpha blending and/or transparency), allows the same unit graphic to be rendered differently for each level.'

Why not:
* Give each faction a fixed palette (range of colours)
* Give each level a fixed palette modifier (change colour/blend/transparency)
* Use a fixed gray-scale to represent colours
* Use a configurable gray-to-palette mapping
* Produce a gray-scale master from a coloured graphic
* Just create level one graphics set and generate all the rest.
??

Briefly, this means light-gray will be rendered as yellow, orange, and red for Elf levels 1, 2, and 3, but mauve, brown, and blue for Dwarf levels 1, 2, and 3. A unit graphic will undergo a colour change on leveling. Not all colours will change at each level, but the colour change should be marked enough to make the graphic appear suitably unique at each level.

A single gray-scale for all graphics makes it easier to map palette colours into the graphic. Also, the reverse, mapping coloured graphics into the gray-scale allows existing graphics to be converted.

Yes, this will require coding resources coders don't have...but, it need only be done once, and coding the actual colour mappings isn't that tough. Real effort is needed to define the faction colours and level modifiers, but once the artists have reached agreement, coders can produce a near perfect program in no time .

Yes, I do like a new graphic for each level, but not at the expense of the game's primary objective. Unit leveling should continue all the way through a campaign -- this is what I believe the developers implied by 'unit development' --, or at least up to the last few scenarios. There is not much to gain by levelling at the end, so there is some enjoyment in playing out the campaign with (as some think) super units.

By focusing creativity to a level one graphic and generating the other levels, artists can quickly check and correct any problems. And players can get more levels.

What am I missing here?

[Christophe33]

Some time ago, I played at making lvl 4 and 5 version of half a dozen units(including some on the enemy side) to play test them and it worked fine. In some cases I did a quick color changes in the clothes or weapon for the fun. It would be certainly interesting to individualised troups, maybe including a logo applied on shield (when there are any) or a prefered color for each factions.
I would also be possible to stretch the change in graphic by makeing significant changes only every 2 or 3 level.
The main limiting factor is actually to find a different name for each lvl. Maybe you can go over it using number (1sst class, 2nd class...) or just keep the same name all the time.
Anyway I think the real limitation is the imbalance it would create and the choice not to make it looking too much like a RPG.
Both could be argued about...but probably in vain.
However, what happened to the idea to give units some bonus (HP, attack...) for experience after they levelled up?

[Blackbeard]

The idea of branching units on leveling is a good idea, but I think it only works well when the number of unit types is small. By branching, you can generate more variations to a base unit (such a as mage or archer). If you start out with many unit types, then branching doesn't do much more than add to the already large number of unit types.

So you can start with 20 level one unit types and, via branching, develop 20 new level 2 types, and then 80 level 3 types, and so on. This is cool. But starting with 100 level 1 types and then developing 300 level 2 types, is too much...

Branching is a drawback for novices because they don't know which branch is better suited for the current campaign. Further, it isn't possible to defer the descision until they have read up and understood the benefits of each branch. With a large number of unit types, only experienced players will be able to make informed descisions.

Branching is irreverseable, making the 'wrong' choice wastes the effort spent in levelling the unit. A single development path doesn't allow new types, but isn't a drawback when there are so many types already.

Campaign developers may find it more difficult to balance their scenarios with branching because players can choose to branch their units all one way, or another. Some combination may make a scenario easier to play than all others.

I don't equate branching with leveling or unit development. I do see it as producing new unit types. There is a subtle, but important, difference in being able to continue levelling a unit and creating several types after a certain level without further progress. The difference, is the fun.

[Neoriceisgood]

Well I realy prefer leveled units to be more unique, and if you look at the "Elvish cataprath" or whatever it's spelled topic you see that although some units have graphics; it takes them alot of time to make it in the game for reasons other than graphics. And I realy think that graphics sometimes make you much happier with an unit than if they were just the level 1 version recolored. not to mention that some units have much larger changes than just more strenght while leveling units; and at the moment there's only two extra leveled versions in the game suggested(the new "outrider" wich made in in cvs I believe, and level 3 revenant) and one has graphics in progress and the other graphics made.
perhaps the only problem is that new leveled versions are suggested and approved in ideas to get in the game.
(just remembered, level 3 bloodbat and dwarven bezerker are also being made I believe)

[Blackbeard]

The main limiting factor is actually to find a different name for each lvl. Maybe you can go over it using number (1sst class, 2nd class...) or just keep the same name all the time.

I thought it might be interesting to give each faction their own language. Then each faction would give rank names for each level. Since the names are 'foreign', they could sound exciting, but be nothing more than level 1, 2, 3,... in that faction's language.

The existing practice to rename each level is excellent, but is a problem for 10 or so levels. However, many MUDs have over fifty levels (although far less classes than Wesnoth), so finding good names will take a bit of research. Looking up military and other discipline ranking names can be very helpful, too. It isn't too difficult to think of 10 names for each faction. Unique unit names is admittedly a huge problem, about 1 000 names?

Anyway I think the real limitation is the imbalance it would create and the choice not to make it looking too much like a RPG.
Both could be argued about...but probably in vain.

I'm not sure what you mean by imbalance, since you play tested high levels without balance problems? I don't see how increasing levels causes imbalance. The purpose of the game is to gain levels, so not doing so causes an imbalance. (several newbies have remarked at having to restart a campaign because they failed to develop units)

I cannot grasp that the whole development effort is aimed at allowing players to advance units and then make that the most limited part of the game. It's .. uhm .. insane, almost. I mean, there is more effort spent in creating new unit types, different attacks, better maps and graphcis, and nothing about increasing levels. Hello? What am I missing? Campaigns are about levelling units, not fancy scripting tricks and strange objectives. Even multiplayers want unit advancement, and rightly so. Getting obssessed about village gold and placement is understandable, considering unit advancement isn't <em>really</em> possible after several scenarios. Players are asking for all kinds of features because the main objective is being ignored by the developers.

I don't know what a good maximum level is, or whether there should actually <em>be</em> any. But creating more unit types and then figuring out the 4 levels, kills the point of the game. Yes, it is fun to do and rewarding in its own right. But as Dave's sig notes, stripping code down to its minimum is the real art. The same goes for games, too. Let players develop more unit types, and developers focus on expanding levels.

No one higher level can imbalance a game simply by being numerically greater than another number. The same goes for extra hp, damage, etc. It only causes an imbalance because it is made to. Levels are relative. Two level 10 units attacking each other with 1 000 hp each is no more unbalanced that two level 4 units, with 80 hp each, etc. As the campaign progresses, recruiting becomes less of an option and recalling more important. It should. You can't easily replace a level 8 unit in scenario 20, or a level 4, for that matter.

The FPI gives branching as a reason for not increasing levels, but doesn't say anything about RPG. In any case, whether it appears RPG isn't an issue to me, but levelling units throughout a campaign is.

However, what happened to the idea to give units some bonus (HP, attack...) for experience after they levelled up?

You mean bonuses to units that have maxed levels? Heh, just add levels, it comes to the same thing. I don't really care if a unit gets nothing than more hp and/or attack at higher levels. But I do care that it can gain levels.

Shoot me for suggesting this, but if 4 levels is a max, then it should take a substantial part of a campaign to get. That means each unit must gain large amounts of experience to level. I don't care how many levels are used, but I lose interest when units stop levelling way before the campaign ends.

[Neoriceisgood]

one question, have you ever recalled a pack of about 5 level 3 units and attacked the enemy straight on while he recruits flocks of units? if you're smart enough to recall a strong healer as well you can totaly anniolate him in a few turns with minimum damage, level 3 units can kill most level 1 units in 1 turn (grand knight+paladin+knight can kill orcish assasins in 1 hit even) now the problem with your suggested "bring them up to level 8!" thing is that at a certain point; units become gods. now defaldor is the highest leveled unit, and I believe he can kill a level 2 troll in 1 turn if all hits are succesfull; and in the campaign leveling units is important; but a level 8 unit as you called WOULD be unbalanced for two things, if you had it; you had the entire campaign to level it and it could pretty much kill every enemy with one blow. but if the enemy had one; you'd have to recall an entire army and loose most of them fighting it, and you can't use any of your heroes since if the thing kills them you loose.
and frankly in either cases, it slightly ruins the game to make these "god-like" level 6+ units.
the only thing that could work for this if you had to defeat one unit of amazing strenght within a certain amout of turns like a dragon or some sort of demon, but at certain points units reach to a level they can't surpass; and in wesnoth that limit is set around level 4/5

[Blackbeard]

And I realy think that graphics sometimes make you much happier with an unit than if they were just the level 1 version recolored.

A absolutely agree with you. I am very impressed that the art developers agreed to produce unique graphics for every level. The first time I saw my unit level (an Elf archer, version 0.6.1), I was delighted at the way it changed. Awesome!

My main objection to unique graphics is that it kills ading levels. I don't believe any feature should override the main goal, and levelling units is a main goal. Having fun playing is THE main goal.

My second objection is that the each graphic is produced almost from scratch. And each colour combination and unit placement is debated at length. Standardise, simplify, automate. Yes, great artists will produce excellent graphics that way. The Wesnoth artists are outstanding. With minimal auto-genrating tools, they can proto-type rapidly and focus on design, not mechanics.

[Sangel]

Well, as an artist and a player, I have to say that I like the new graphics for each level. I like being able to swiftly recognise higher level units, and know that they're higher level because of how they look - and not because of some arbitrary colour change. Who says Green is more advanced than Blue?

There's also the trouble of increments between levels. Dave has already stated that if ever more levels are added beyond 3, then the power levels will be adjusted to make sure that the highest level unit is no more powerful than a level 3 unit is now. While I can see this working with, say, 5 levels per unit, once you get above that the increments between levels will become so small as to be inconsequential. Currently levelling a unit is a major and exciting event, as there are usually dramatic changes. Under your proposed system, units would be constantly levelling, which reduces the excitement.

Basically, I feel that the current levelling system is a lot of fun, and doesn't need adjustment (save for adding second and third levels to those poor units without them).

[Blackbeard]

... know that they're higher level because of how they look - and not because of some arbitrary colour change. Who says Green is more advanced than Blue?

Yes I agree a new unit look at each level is better than a rehacked graphic. But I don't believe that should change or limit the game's objective. If it can be done with less manual effort, then it is much better than recolouring a unit. My idea is simply to solve the extra work in creating more levels.

Colour schemes are debated for each unit. By debating a faction's colour levelling up front, there is less need to debate that for each unit, but rather good design and quality animation.

There's also the trouble of increments between levels. Dave has already stated that if ever more levels are added beyond 3, then the power levels will be adjusted to make sure that the highest level unit is no more powerful than a level 3 unit is now.

Excellent! That means each hp and damage is really important, and any small bonuses will really count. I don't care if a level 10 unit has 60 hp and hits 4-5, or 2-10. Capping hp etc is very good.

While I can see this working with, say, 5 levels per unit, once you get above that the increments between levels will become so small as to be inconsequential.

I don't see it that way. If the max hp is 50, then a level 1 unit has about 5hp and hits 3-1. Level gains can be scaled, or fixed. Whatever increment is used, a level 9 unit can have 40hp and hit 3-9. A level 10 unit has 50hp and hits 3-10. A level 8, 35hp and hits 3-8... for example. Resistence to attacks (pierce, blade, etc) only take effect after a few levels. This not only eases newbies into the game, but adds interest to higher level units.

Currently levelling a unit is a major and exciting event, as there are usually dramatic changes. Under your proposed system, units would be constantly levelling, which reduces the excitement.

Different units will be constantly levelling, yes. I don't want level 4 units soaking up exp, while level 1 units get killed trying to level. A campaign should mean that all units are near their max and recruiting is pointless by the final scenario. Scenarios progressively limit the level range, while some may only allow recruiting. A MUD builder guru remarked, 'A good builder will design an area that controls a player's action, while giving players the impression they have freedom of choice.' I believe that is applicable to campaign developers, since they, like MUD builders, create a world for players to fantasize in. And campaign developers should control unit advancement, just as builders do MUD characters.

Basically, I feel that the current levelling system is a lot of fun, and doesn't need adjustment (save for adding second and third levels to those poor units without them).

It is fun, until units stop levelling. If the number of unit types was reduced, and their attack skills merged into a few types, then branching and extra levels will be more exciting, imho. With 4 factions, each with 5 unit types, you can easily end up with 100 level 10 types. So the constant levelling also creates oppotunities for new unit types. A mage might branch 3 times and offer 7 different mage types. Same for a fighter. So each unit becomes more powerful, but limited in range of attacks / terrain / defense / resistance. Others might gain many abilities, but less effective than a specialist unit.

For multiplay, unit levelling will become a reality, and not just once, but perhaps 2 or 3 times. Naturally, it will become more important to keep units alive near the end of the scenario, so players who don't level some units won't last to the end.

I am not objecting to the current levelling used, merely pointing out that it doesn't extend throughout a campaign. No one would comment on this if units could be levelled to the very end. The number of levels is unimportant, but that units stop levelling is. That, to me, is a contradiction between the game's objective and the game's design.

[Blackbeard]

one question, have you ever recalled a pack of about 5 level 3 units and attacked the enemy straight on while he recruits flocks of units?

My response is, if a campaign developer allows this, its not much of a campaign. Using level 3 units to kill level 1 won't get you much exp anyway and stops your level 1 units from advancing.

A good campagin developer knows that a player has high level units and will make the computer players recruit similarly high level units. Now, put my level 8 unit up against several level 7, some level 8, and a few level 9 and 10 units -- suddenly it looks pretty ordinary.

This is no different to having level 3 units running among level 1.

[Dave]

I think that a unit getting a new graphic every time it levels is a very important part of the game. Some people may reject this as simply 'eye-candy', however I think that it shouldn't be under-estimated as one of the key things that makes the game fun.

If artists were to show a commitment to making graphics for more levels, I think it'd be okay to add them. If we were to add one more level, I think it'd be best to insert it where level 2 is now -- you'd have a new 'Elvish Warrior' [1] that is a little better at fighting than an Elvish Fighter, and who advances into a Captain or Hero.

You'd have a 'Lancer' who advances from a Horseman, and can advance into a Knight. And so on, and so forth. If this were to be done, our current 2nd level units would be made into 3rd level units, and made a little more powerful than currently, and the new 2nd level units would be made somewhat weaker than 2nd level units currently are.

I do think that perhaps we have the resources to extend the game to 4-5 levels in this way. I don't think we have the resources to extend it any further, and even if we did, I'm not sure it'd be a good idea.

If the game got to move levels than this, I think we'd just end up with all advancements being miniscule skill improvements that'd get boring.

Alternatives:

- improving after max-level systems that have been already proposed.
- 'half levels'. When an Elvish Fighter advances it becomes a 'Veteran' Elvish Fighter. I.e. an Elvish Fighter with some improved stats. Only when it advances again does it become an Elvish Hero.

David


[1] all names here are of course just examples

[turin]

what are you missing?

it seems to me to be this: graphics are not the main reason new levels are rejected/set aside. the main reason is 1) it will take forever to balance them and 2) what is the point?

i am focusing on 2) here. why do we need more levels? do you want to finish the current unit trees, so each unit can get up to level 3? this seems fine to me, but just recoloring the image won't work.

or, do you want to get up to 4, 5 or even 6 levels? this, to me, is a bad idea. they wouldn't be allowed to get more powerful; level 6 would be just like current level 3.
and, this would make the game more like games where you get slightly more powerful as you gain XP. i like wesnoth as it is, where you can only advance 2 times, but you get major bonuses when advancing. this is different, in a good way, from games like HoMM, where it might be for every Xp you gain you gaim 1/10th of an HP, and this just continues.

in short, i like having few levels, it makes the game distinct, and adding more levels will make the game just like all the other ones, and it will lose the part of the game that made it fun.


just my $.02

and dave: no half levels. please. this goes against the whole game system.

[Blackbeard]

I think that a unit getting a new graphic every time it levels is a very important part of the game. Some people may reject this as simply 'eye-candy', however I think that it shouldn't be under-estimated as one of the key things that makes the game fun.

Yes, it definitely is. But when the that key thing stops for some units, the fun stops too.

I do believe artists can play with blend, transparency, and colour changes to produce a new image for each level. The question I'm asking here is: 'Can artists use a single image to produce 5 or 6 distinct ones?' I believe artists need to rethink what they're actually being asked to do .. not how they get the result. What they're being asked is to make each level a stunning event and the new level image better looking than the previous one. If that can only be done by creating each image manually, so be it..but I'm sure less labour intensive methods can be equally effective.

To say it another way, rather then artists spread their effort over many graphics, focus their skills on designing a core set of reusable ones. The emphasis is on using artist tools to design an image that decieves the viewer into seeing several unique images, rather than going through the same mechanical process to create them. I am certain this issue isn't unique to Wesnoth and animation artists, or whatever, have some good alternatives to replace grinding out each frame.

If artists were to show a commitment to making graphics for more levels, I think it'd be okay to add them. If we were to add one more level, I think it'd be best to insert it where level 2 is now

Fantastic! A 5th level will make a huge difference and keep players more interested in advancing units throughout campaigns. Reducing the gap between level 1 and 2 will also benefit multiplay, won't it?

[Darth Fool]

I don't like half levels and I don't want too many levels. Advancing units is a major part of the game, largely because advancing a unit makes a big difference. If there are too many levels, one of two things happen, the unit advancement becomes nearly meaningless (ooh My unit now has 1% more HP, let's throw a party...) or the higher level units become so strong that 1st level units are not able to take them on. I like that in Wesnoth, The most advanced units when surrounded by level 1 units is pretty much guaranteed to be dead. It forces you to be careful with your best units, because they are not invincible, but they are a significant enough improvement over level 1 units that you don't want to throw them away...This is good.

I do agree that a campaign should not go on much past where you can reasonably be expected to have enough high level units and gold that you don't need to recruit level 1 units anymore. There are ways that campaign-developers can remove these from you, but they typically are too contrived to not be irritating. This desire may be in conflict with the planned plots for some of the campaigns . In such cases I would encourage the campaign developers to consider splitting the campaigns into two seperate campaigns and approach the second half of the plot from a different main-characters perspective.

[turin]

blackbeard: no, i'm pretty sure a good effect cannot be achieved by reusing the same image and re-coloring it. not that i'm likely to try, since i dislike the idea anyway. but, in my experience, recoloring an image always leads to either the skin looking wierd or the clothes looking wierd or both, and it never leads to anything that looked nearly as good as the original.