Beyond Balance: Psychology and the Trait Trees

I hate bumping my own post, but I made numerous revisions to the original post, and I just want to say it is in it's final state now (barring me noticing a grammar/spelling error).

If you REALLY want to play the game without champions splitting XP, here is the file (put it in your mods folder, or units folder... it will override the other champion units in the game and make all the champions you play with not divide XP, no other changes... current as of 1.1

https://dl.dropboxusercontent.com/u/75549875/Fallen%20Enchantress/LH%20Mods/BURRESS/championNOXPDIVIDEv1_1.xml

I actually enjoy the "Filler" traits, whenever I played d&d I chose those "filler" traits all the time and tended to avoid taking the more "spiffy" traits, because I found more enjoyment playing with only "filler" traits than the action traits as you call them. Maybe psychologically I'm broken, but I don't look forward to these new special traits, in any of the games I play. But I'm a mathematician and like to deal with the numbers and probability changing stats more than anything else.

That is kind of you Parrottmath. I was aware of how to do it, and I tend to mod the game to how I enjoy it anyway, but it is thoughtful.

People are different and this is a generalization, so I wasn't saying this holds for everyone, and probably doesn't hold at all for people not interested in RPG elements.

I am a computer scientist, and I love math, philosophy, and psychology as well. I was thinking of some quantification of fun, with different systems/art/gameplay elements contributing to this "score", and trying to maximize the quantity contributed by the RPG element based on generalizations of how memory works, interpretations of others experiences here on the forum, my own experience, and some armchair theorizing. I think a system designed with this in mind increases the probability that a player or reviewer gives the game positive marks. I thought it was an important insight that troughs in the experience, even if they only last on screen for a levelup, have an effect out of proportion to the time of their experience in how much they affect what a person will say when asked "so how was it?". I was playing with some meta-game math.

I am really passionate that most chained traits, unexciting bonuses, and XP splitting are design decisions that cut in to perceptions of the game, whether people point to them as the reason or not. I think I have given good arguments on the game-fun level, balance is the problem stage after you decide on what is imperative (fun). I see lots of suggestions about this mechanism or that, but it is like chaotic tree-chopping if there is no consensus on what the forest should look like (what the experience of each levelup should be, what is the expectation of what the player does with a champion). To maximize fun, I think each levelup should affect gameplay and each champion should be used without reservation (which means the split should go).

On the overall about the chain of traits to get to the next good trait, one forgets the psychology of achievement to reach such a trait. There are many times when we trudge through "non-interesting" things before we find something "interesting" to get to. Once we get there we look at the path of blah and find a gem at the end of the search. Look at minecraft as an example, how much cobblestone does one get before they find their first diamond... they could simplify the process, but the process is part of the challenge and fun. Do not forget this when making a design.

The modern computer RPG suffers from too many interesting choices and not enough blah behind the scene uninteresting choices. But this is a style of how quickly do you want to play the game. Growing up without doing speed runs I would finished a fully designed game in roughly 2 hours, maybe 1. I unlocked all of MarioKart on the WII in a matter of 20 hours of play. I play these games quickly, and solve these puzzles quickly. If the game is supposed to be for the casual player, then your spot on with your assessment of "uninteresting" traits, but if you are serving an epic crowd, then there has to be "uninteresting" traits, otherwise the game will get too convoluted with variables. Mind you I have no problem dealing with an infinite number of variables, but it's called work when I do.

The game you speak of is not the XP split of champions as being the problem, it is the lack of XP on the map. When I finish writing some papers, I'll add some interesting gameplay mechanics that I think the devs missed the opportunity to add to the game.

The uninteresting cobblestone is a different psychological mechanism, that is variable rate reinforcement. That is what keeps people playing slot machines, in this game it is what keeps people going to goodie huts and makes them excited about looting lairs. It is the unpredictable chance that some big reward will come next.

The levelups are expected reinforcements which we see ahead of time. We don't want to be able to see that the next batch of cobblestone is empty ahead of time, that is no fun. This system doesn't hide diamonds, so each one must be interesting if one wants to maximize this reward mechanism.

I think XP is not a problem on the map. It is the fact there is an incentive to not use a champion, or not use them for a period of time. That makes the new champion screen less impressive if we don't fully expect to use whatever we get. They can reduce overall XP as long as they reward using champions instead of parking them. If one super-mage is better than 3 less leveled heroes, people are going to park heroes so it doesn't take 3 times as long to make a super-mage. This seems to cheapen the thrill of the very title element of the game! They could lower the XP dramatically on the map and it be fine, as long as each levelup is worthwhile and each hero can be used without reservation.

Both of you made great post, and excellent reads.  My point of view is more towards Burress opinion.  The tree's lack alot of abilities, and way to much filler is in place.  The mage example was perfect I need three levels of Prodigy to get savant, are you kidding me.    I need 2 levels of Prodigy, and four levels of Evoker to get Pyromancy.   Dont forget you need 5 levels into fire to get flame wave which is nice but not a huge deal...not 5 levels of investment.  This is way to much filler.

The assassin tree is just as bad, you need three levels of precision,  and three levels of shadow strike, to get Break.  Talk about filler. 

All these filler should lead to something great, but they lead to something so basic it should've been a sub lvl 10 talent.   All this tends to demoralize me when  leveling my champions.

Hmm Interesting ideas. However, have you tried henchman?

As the henchmans have same trait tree and divides no exp, you are very sure to have over 15 lev main hero and handful of henchmans, with same traits all togather.

So, I am very sure about if EXP does not divided in to hero numbers, Every factions will have 5 heros of over 15 lev, and the game will be "who gets the first turn game."

and the warrior type heros will be casted out because evoker + first move + haste or slow + flamedart or savant fireball or savant manablast will very surely KO opponents. Also the Plate armors will be casted out dues to initiative problems.

Additionally, these meaningless cobblestones also makes coming heroes more intersting, as many of them has end-tier traits from the begining. If each one of your champions have revenge and double strike, what would be the meaning of recruiting magnar and othollo?

I really don't understand the current idea in gaming that skill trees always have to have active abilities that are shiny and awesome and that they should get rid of passive "filler" options. Passive traits are a reward in themselves due to the way you outfit and play your heroes, they can be as powerful as - sometimes even more powerful than - active traits. If you removed all the "filler" options you get heroes glutted with active abilites of which you will only use one or two. Considering the way the mildly tactical and fast-paced combat plays out in LH, a vast selection of active traits is not needed.

I do agree though that the harsh xp split for joining heroes in an army should go. Since their power got nerfed compared to the might of regular troops, you should be able to put several of them in one army without a penalty. 

I agree with passive/ active Idea of phazonfreak, cause passive like savant is a real killer.

However, If it comes with EXP problem, I have different idea. As I mentioned, If everybody can have 5+ lev15+ heroes, close combat heroes will be casted out because mages are way to powerful. with savant manablast, anything can be one turn KO. Even epic monsters like Delin in insane level was one or two turn killed in my experience.

But I do agree with the opportunity problem. after 150 turns, there is definitly nowhere for my new hero to grow up. and I still can have about 4 to 5 more heroes. Because of this opportunity problem,  heroes after mid game become benchwarmer no matter what they have - skill, ability, item-

I solved it with adding much more of random quests, but on the other hand, I completely agree with this paradox because of the endless might of mage hero after mid game.

This is spot on.  I'm playing on insane and large maps.  I have so many lame duck heroes its worse the Bush presidency.

I didn't mean to imply that passive traits and filler traits were synonyms Phazonfreak. I think if a passive bonus exists as a full level-up option, then it must have noticeable gameplay effects. If you play any modern RPGs, this tends to be the rule. For instance, in Skyrim, a passive bonus may be healing spells +50%, +20% weapon damage, +20% armor protection, certain spells cost 50% less (any of these might work here too). Each time you use these things, the gameplay effect is noticeable. Diablo 3 is an over the top example, for 50-60 levels each level-up is a new active/passive effect or a gameplay (often visual too) change to an existing effect, and often a combination of both. 

Here is the first impression one gets to the leveling system for each class to make the point:

Level 3 assassin. You have already chosen your class, that comes with a set bonus and opens a tree, cool. Level-up again, you have two choices, +5 accuracy or +4 dodge. It is very hard to discern the gameplay effect of these changes, but you have to take one to even get started. Level 3 choices here are barriers, of themselves almost microscopic in effect.

Level 3 for the defender tree has +10 defense when not doing anything and +20 magic resistance. I think it is hard to be impressed with either. Spells are actually kind of rare to target a specific unit, and an unpredictable chance to defend against them (an extra 0% to a high % depending on the caster) is not the first thing you are worrying about starting out as a melee unit. Regarding the +defense while defending, it will only come in to effect if you don't do anything the first turn of battle or have a quicker opponent, unless you are just passing to try to soak up damage. I know the tactic of passing to be a damage wall might be effective, but it is hard to be excited about being a battle turtle (and not the teenage mutant kind!).

Level 3 for the commander is +5 accuracy for the army or -10 unrest in a city. If you are thinking of putting a level 3 hero in a city for unrest reduction, that hero is a dud. The other I guess is supposed to just add up over time, but you won't feel like level 3 made any difference for the accuracy of your army.

Level 3 for the warrior is +3 attack or +10 hitpoints. Definitely the most noticeable of all classes. It makes a difference the player notices right away. (Though +3 attack fades in usefulness over time, maybe why %weapon attack or damage might be a better way, heroes use their weapons better)

Level 3 for the mage is +10 spell mastery or +1(2?) levels to summons. The spell mastery bonus is tricky, you will never know if a spell went through because of it or not. It is just not very exciting for your very first level, and more is on the way. The other path is strange, unless you are a summoner sovereign you don't even have summons yet! This level makes either no difference or a difficult to discern difference, and it is the first choice you get in the trait tree.

Later on many branches bog down in to sequences of passive bonuses, which all together over a series of levels add up to something that is noticeable, but individually it is hard to tell the difference from level to level. +Accuracy is hard to notice in doses, +dodge has the downside that you will never know when a miss was because of the bonus, +attack to champions when you hardly ever face champions, small bonuses to army stats, +3 defense... most of the passive bonuses are too small to be noticed in and of themselves. So each level-up is not an increase in power, it is a step toward an increase in power. That takes away from the thrill of seeing the level-up screen.

*edit: I think Skyrim has a great model for passive trait bonuses for warrior classes (interesting model for mages too, but that is quite different). In abstract, the best bonuses are available right away (+%), are in the root of the tree and they are repeatable after certain requirements, which would be equivalent to levels in this game. All traits are meant to be immediately used and noticed, and most are never outdated. There is almost never a point you are taking a trait to get to another trait. The effect of this system is leveling up is always exciting, and the game sells an 8 digit number of copies (other factors contribute to this too, but just sayin).

This sounds very much like the timmy, johnny, spike idea from magic the gathering: http://www.wizards.com/magic/magazine/article.aspx?x=mtgcom/daily/mr11b.

Different people value different things from systems such as these. It seems the Spikes are going to naturally have no trouble with the idea that some stuff is crap or that some of the advantage's aren't flashy, while Timmies generally will. However, right now I see almost nothing in the trait tree that can appeal to Jonnies, which I (and perhaps Burress) am through and through

Great article.

This is something I have been pondering for a while now.  Each class needs a niche and to do that would be to have class talents be choices for active or powerful passive abilities.  Stat bonuses should be moved to the general tree.  Everyone can become stronger per se but only a warrior can know how to apply that to his/her specific weapon of choice.  I am also seriously of the opinion that the Defender and the warrior tree should be combined and add a different class to replace it either a summoner class, suggestion by others, or my suggestion being a divine class that uses holy/unholy.  I have an interesting idea for magic but its a little hard to convey it online and I have to think about how I want to word it.  That and it requires more spells and a clear delineation between the different spellbooks which is something that doesn't really happen as of right now which as you have stated makes it hard to be excited about choosing a spell trait when the choices are not that great.

Awesome thread! I change my pov after reading each post. Add me to the Johnny list.

I've read through the article and sadly... I was not represented as one of the archetypes that played the game regularly, or I did not recognize my archetype in the descriptions.

I played magic quite a bit, but I only played decks that had a lot of little creatures, nothing big, I'm talking 1 / 1 or 1 / 2 creatures. That was all I allowed in my deck. I centered around white magic mostly and had the most fun playing these games. If I won fine, if I lost great. I never played to win, I never played to lose. I played to tell a story of the adventure. It tells a story when you accomplish nothing with what you've created and that is what I find the most interesting in the games I play.

I play most games to see how long I can last behind impossible odds, Left 4 dead survival mode is rather fun for me, because I know I'm gonna lose, it's a matter of how. For the indy 500 games I've played I turned my car around and raced backwards to see how many laps I could travel before a major accident broke my car. I even broke many of aircraft simulators by seeing if I can land my f-14 backwards on an aircraft carrier (successfully), just had to stall my plane right as I was over the deck of the ship.

I can go over a few more examples of the things I have fun doing in games, but in short "I do not play to win", I play for the story I can tell.

you are a timmy vorthos. http://www.wizards.com/magic/magazine/article.aspx?x=mtgcom/daily/mr278

Oh, and anyone who reads the above article should note that the meyers-briggs personality aspects are incredibly off. The author likely meant feeling/thinking instead of sensing/intuitive.

the latter is more or less an indicator of how "down to earth" you are. Not many gamers here are likely to be sensing Type.

Close, but I'll go for ---

Melthos --- (I like the underlying rules (Mathematician) and I like the feel of how things fit together.)

Reason I like Elemental so much as the rules and the feel are darn good combination of things just work for me. The traits described above all have a place in the rules of the game and the filler traits provide more than some people think. I like the story that these filler traits allow me to tell about my character. It is the story that is more important in the game than the rewards we choose.

I suppose I can fit on the other chart if he described the negatives on the axes he has described with the 3 character types. I'm the complete opposite of spike (so much so, I make spikes mad at me)

Master of Orion did this beautifully in their research tree, in which each choice locked you out of the other options, unless you had the creative trait, in which all options stayed open to you.  Something similar here, where each choice locks you out of other choices, unless you took a trait at startup, might be fun.

I found my experience with leveling to be one of disinterest, for the most part.  During my very first games, as I learned how to play, I thought each level was great.  As I played more and realized that new levels were more important for the tiny stat boosts than the abilities, in many cases, I stopped caring.  This came quickly, and even got to the point where my first few levels were a toss-up between mediocre choices.  This was especially so for Commanders, who spend the entire early/mid game useless until transformed into ideal city babysitters through tedious leveling.

My warriors failed to outperform troops after a relatively short time.  My greatest sovereign, an Archmage in Air/Life (but not summon-focused) did not change in play style at all after the first couple of levels.  Haste and heal with the odd offensive spell.  The heals themselves became useless after a while, since restoring 26hp to a 120hp swarmed troop that just took 90 damage from a critical hit isn't all that great.  The air spells were neat, but my troops were consistently more deadly to my enemies.

As it stands now I don't even bother with champions, except to lead reserve armies that guard my back door.  All I tend to need is a single group of uber-leveled troops with a support mage sovereign to mind them, and nobody can stop me.  The maps tend to play out in such a way that there are typically only a handful of narrow passageways between my nations and the AI, so I don't have much use for a half dozen armies with mighty heroes.  I also find that, even with monsters on the most numerous setting with "Challenging" difficulty there is only enough XP to go around for my sovereign and his troops.  If I divided the map's denizens among him and a champion it seems I'd be worse off.  I suppose if I played on insane everything I might need more, but I don't tend to find stat-padding my enemies to be an entertaining challenge.  If the AI got better, sure, but an AI with higher HP is just a cheap way to add challenge.  It's like how in Civ V it just throws higher math at you to overcome, but the AI itself remains a pushover once you learn to exploit it.

To be honest, I'm far more excited about seeing my troops level up.  Seeing a spearman unit that I trained in early game hit level 15+ is more rewarding than giving my sovereign +10 to some stat.  Watching that spearman die is far more traumatic than seeing my sovereign get KO'd by something, since I know he'll just pick himself back up after the fight no worse for wear.

I don't think I ever explained the psychology part very well in the original post. Here is the kind of thing I am basing it off of. In "Thinking, Fast and Slow", Daniel Kahneman explains an experiment about pain perception. Subjects had their hands held in painfully cold water, and then were given a warm towel at the end of it. They did this twice, once where they had it in the cold water for 30 seconds and once where they had it in the cold water for 30 seconds followed by 30 more seconds where they slightly increased the temperature of the water, so at the end of the second one it was still painfully cold but noticeably less so. They were asked which they preferred (it was properly controlled for order of the tests), and it was overwhelming that they preferred the second one. This kind of experiment affirms that our memory of the event is not rational, the second 30 seconds was still in painfully cold water so we should have preferred less pain overall. Other experiments confirmed that the peak (or lowest point) and the way an experience ends are the basis for a qualitative judgement of our memory (this was Kahneman's own generalization of a bias in our memory). So it is like we have these emotional imprints, but we are not necessarily rational about them.

I think this occurs in game reviewing and player's perceptions of games quite a bit. There is an overwhelming amount of things to remember, but when asked about the quality of a game or to critique it the process proceeds a bit backwards. There are emotional marks that come up when we think about the game and then we cognitively look for explanations for them, even though it is highly unlikely that these emotional marks were caused by the explanations we come up with.  This is how I explain game reviews where it looks like the person played the game, but appeared to be bothered by something that doesn't seem to have much of any negative impact on the experience of the game. They had an emotional memory, but could not draw up an explanation that seemed to rationally fit their irritation. My posts here were an attempt to critique the rpg aspects to take advantage of memory and anticipation.

I get this entirely. It also justifies the delayed gratification idea that parrotthmath provided (though there are better ways to accomplish this). The point stands that many traits are crap, and you have to taste them everytime you level. 

By the way, have you read anything by Jonathan Haidt? His work is similar to Kahneman's, and very enlightening.

I have not read anything by Jonathon Haidt, though I can already see a couple of titles he has written I would really enjoy reading.

I think there are two features where negative emotional memories pollute two mechanics which should be associated with only enjoyment, the level-up screen and the hero arrives screen. 

The first problem arises from unfulfilling choices and what seem like barrier roads to good choices. The result is that the level-up screen is not a pure fun association. We get the good level-ups eventually, but in the evolution of RPG mechanics, you can see a move toward making every level-up fun and desirable. The old ones made you grind for the long road to glory, nowadays the realization in the genre is that the long road misses out on the opportunity for fun at each step. If the feeling of "I want something that is far away" dominates the system, then the emotional memory of that system has a downer that becomes part of the association with it. What makes it even harder is that this is a turn based strategy game, so events tend to be quite spread out in time anyway, so leveling should be trying to mitigate this situation instead of highlighting it. It's not like more levelups are needed, or quicker ones, just that the next one is always a really good one, so some relatively near event is always thrilling. 

The doesn't seem easy to fix, since there are so many different ways to go in implementing more fun trait trees. I think going through the mindset of a player, and trying to not design toward long-term wants, but instead make sure the dominant impression is always short-term thrills is the way to find the most fun design.

The second only comes in to play with game experience. The heroes themselves are fine. The problem is the XP split that incentivizes splitting heroes up immediately (instead of when it is strategically comfortable or according to tastes), so getting a hero can become associated with the hassle and long wait associated with putting an army together for their use. If someone just uses heroes together, the level-ups are slowed way down. If someone always or mostly splits them, then there can be a long delay and a logistical headache in getting them started. In addition, some people's minds just can't handle more than a couple armies (more than three seems like work to me). The point is it doesn't seem wise for their to exist downsides or hassles associated with heroes, if the hero arrives screen is to have the greatest peak and shallowest negative association possible.

This seems like such an easy fix, get rid of any hero downsides. It may not seem from a design side there are downsides, but they come out in playing the game. I have seen many people split about the XP split, but I think the ones that are still for the XP split have just become comfortable with heroes not being a blast.

I copied the file "championNOXPDIVIDEv1_1.xml" to my "...\Documents\My Games\LegendaryHeroes\Mods" folder but it doesn't work. Can anyone let me know what I'm doing wrong?

edit: Solved it. Was clashing with CoS mod so I just edited the changes into that manually. Who knew there were so many bloody champions?