Skill trees and the one right path problem

[GolvaeGames]

Skill trees seem to offer a myriad of customization at first glace, but they often become no different than any more linear skill progression system because there are always best paths. Some games even give you so many points you can just get everything and every character comes out the same in the end. Does anyone have any experience with games that do skill trees well? Meaningful choices are a lot easier said than done.

 

[Milennin]

This is more an issue with commercial games because people write guides for them, and have a lot of players blindly follow them. Those games are still made so that each path should be viable for general gameplay, but obviously there's always going to be a best path to pick; and it'll be found by people who play the game for hundreds of hours and write or record the guides for it.

The nice thing about RPG Maker games is that you don't have any of this. People go into them blindly, and they make all the gameplay affecting choices themselves. Then it's just up to you, as developer, to make it so that each path is at least viable to make it to the end with.

 

[GolvaeGames]

Thats true, making sure the player at least doesn't regret their choice. I have terrible decision paralysis so I overthink this kind of thing a lot

 

[BK-tdm]

Well Blizz did a good thing with Diablo 3 in my opinion, you dont have skill trees but slots where you can mix and match stuff that suit your gameplay and get enhanced by sets and legendary equipment bonuses, you can spec your wizard in one damage type or many, synergy is the key there, it still doesnt prevent crazy op broken builds that top others, but thats a balancing issue in itemization more than innate to the skill tree approach itself.

Bethesda games are pretty good too in which you get to choose and be everything but people rarely pick up the do all approach because: 1 the grind is too long and 2 playstyles differ, so even if i can do everything in Skyrim i havent bothered touching heavy armor or 2 handed weapons in 7+ playtroughs with 1200 hours on it, so letting the player do and be everything doesnt mean everyone will.

 

[GolvaeGames]

I liked how it was done in D3 too. Blizzards mmo Wow was probably one of the worst skill trees in that it was just really really bad to deviate from acceptable paths. They did acknowledge that and change it but I haven't really played in a long time so I don't know how well thats worked out.

The D3 point has me wondering if something like a skill equip system wouldn't be better

 

[TheoAllen]

When two options or more are given, it is not avoidable to have the best path among them. It can be an absolute best path or the best path of each playstyle. For example, a skilled player will not benefit from defensive skill since they have a player skill. So they could utilize a more offensive skill tree. For less skilled players will benefit from more defensive skill trees, etc.

 

[gstv87]

I don't know about Diablo 3, but Diablo 2 I did play in several different ways, and the results were interesting.
first time, I picked a Necromancer. I thought I could benefit from the horde mechanic and just swarm the enemy with skeletons, but I found it hard to control and unable to defend against wizards that'd just bypass the horde altogether with ground attacks or meteor strikes. So, I dropped it and started over.
second time, I picked a Paladin. Went into it playing as a medium warrior, with sword and bow, not really using the Paladin's core power which are his auras. I managed to beat Diablo by stacking up on potions and basically out-maneuvering him until I could land a couple of arrows. Once I beat it, I didn't really "feel" it... I was relieved for beating it, but didn't feel satisfied. We didn't really "fight", all I did was dodge and shoot out of screen.
third time, I installed Lord Of Destruction and picked up a Druid. Wasn't really convinced about the shapeshifting thing, so I dropped it and picked the Assassin.... again, wasn't really digging the 1-2 mechanic for finishing enemies, so I dropped it as well.
then I went and really thought about it, and came up with a controls structure that'd really benefit a Paladin and it's auras, so I changed the controls, and ended up with a sort of gearbox structure with offense/neutral/defense stances.
fourth time: took the Paladin again, with that structure in mind.... and boy, I was DANCING my way through. It was totally intuitive: buff up, switch to defense, tank, rush in, strike, switch to heal, etc.
I picked up the Amazon companion at the first stage, and carried her all the way to Baal. SHE killed Diablo, while all I did was tank hits and heal both of us (you know... as a Paladin would...). And for Baal, we both died several times, until I remembered I had that one thawing potion I picked up somewhere down the line and totally forgot about it, stood my ground in front of him, and using that gearbox mechanic I was changing between healing, defense, attack and buffs in between his attack runs. He froze me, I switched to reflect damage, he froze himself, I drank the thawing potion, switched to shield bash, stunned him, and hammered him down with Blessed Hammer.
That one I'll remember, because everything went according to plan: he had 1 HP and that 1 HP would cost me everything I had, because I had nothing against his freeze except that one thawing potion.

 

[GolvaeGames]

I played d2 but I don't remember the trees at all I would always quit around level 30-40 right around killing diablo or baal

 

[Aesica]

They only game I ever saw that did a skill tree system right was Star Trek Online with it's specialization system, and that's only because you could ultimately buy out every option on the board.

Skill trees are, in their very nature, a fancy way to give players only a few slices of a class kit instead of the entire pie. There is no way to make them all equal, so you basically just choose from one of these options:

• You let the player buy out everything eventually like STO did
• You say screw it and let guide-writers tell everyone else which skills to pick
• You accept that skill trees inherently suck beyond nostalgia for certain games and use a better system for yours.

Also, if your game uses skill trees and you don't offer a way to easily reset them, you're probably a monster.

 

[EmeraldSpecter]

The Sphere Grid in Final Fantasy X was brilliant because the branches were options that were needed but weren't absolutely needed. The other great "Skill Tree" I've seen in action was in Ghost of Tsushima. Basically, there were tons of trees with very shallow branches, so you didn't have to worry too much about things being left too far behind. Everything was a benefit and you eventually wanted all of the goodies because you had a need for them in the end.

Dragon Quest XI isn't that bad for a Skill Tree, either. You can't afford everything but gives you enough options to allow you customization while you play.

 

[GolvaeGames]

I haven't played STO or Ghost of Tsushima, will definitely check those out. I remember an old MMO called rift did something called purpose's (I think) that sort of gave you an outline of where to put your points

 

[Aesica]

I haven't played STO or Ghost of Tsushima, will definitely check those out. I remember an old MMO called rift did something called purpose's (I think) that sort of gave you an outline of where to put your points

It takes awhile into the game before the specialization system unlocks, so I'll just sum it up:

• Each specialization has its own tree, and there are several speciailzations to choose from.
• Of all the specializations you have unlocked, you can set one "major" (full benefits) and one "minor" (only up to the second tier of benefits is gained) as active in any social zone without restriction. Kind of like picking a main class and a secondary class.
• Like every "tree" based progression system, certain things only unlock after buying prerequisite things in the same tree.
• Every time you level up past...I want to say 50...you gain a specialization point you can use to unlock something in your tree of choice and further progress.
• The character level cap is 65, but you can still gain experience and every time you would level past 65, you stay at 65 but gain an additional specialization point. There are no limits to how often you can do this, so eventually you can buy out the entire tree.
• Spending 5 points in an available tier unlocks the next tier, and you can spend points in any unlocked tier.

Spoiler: Basically, it looks like this

Evenetually, all of this unlocks

Now this system isn't to be confused with the game's skill system--unlocked at level 10--which is a classic you-don't-get-everything skill tree system that I greatly dislike.

 

[EmeraldSpecter]

I haven't played STO or Ghost of Tsushima, will definitely check those out. I remember an old MMO called rift did something called purpose's (I think) that sort of gave you an outline of where to put your points

I haven't played Ghost of Tsushima either, I watched a Let's Play (I have time to watch a lot of stuff but not necessarily play a lot... well, I didn't have time, but I do now).

 

[marbeltoast]

I don't think the "one best path" problem is inherantly a bad thing. As a life long min-maxing munchkin, part of the fun is culling the wheat from the chaff, mechanically speaking. If the options are sufficiently different to be distinct and unique, then one will inevitably be better than the other for optimal play.

There are a couple of work arounds, though. Take stardew valley's skill tree, for example: The first pick at level 5 gives two options, and the second pick at level 10 gives 4, BUT only two are displayed, and which two is based on what you picked for the first choice. This adds an element of replayabillity for first time players, incentivising trying different, potentially sub-optimal playstyles to see the whole of what the game has to offer. The problem I had with stardew valley's system is that you weren't told this was what was happening; you would, in one playthrough, only really know about 2 level 10 skills. If the player knows there's more than what they're seeing, it encourages replayabillity.

You may worry the players will just check the wiki to figure out what all the options are ahead of time, but as other have said, most small time RPG maker games don't have massive fan followings, so we get to explore a design space shrouded in mystery in ways triple a titles can't.

Additionally, you've got to remember that as a designer, we have to build an experience that mutiple types of player can enjoy. Not everybody is a minmaxer. Some prefer role-play, others have busy lives and want simple upgrades to damage because it takes the complexity out of the choice and lets them get on with their finite free time. Skill trees are a good way to cover more ground in terms of player archetype appeal.

 

[Wavelength]

I personally like finding Skill Trees in games, so long as their presentation is simple, clear and elegant (get that Sphere Grid wannabe out of my face). They present a lot of genuine choice/expression as to what kind of player you want to be (or what kind of style you want to give a certain character, in party-based games). And unlike games which just put all the options on the table at the start and you buy them with points (which I also like to an extent), Skill Trees pace the decision-making and offer the designer more control over when the player gets really cool, game-changing upgrades.

The most exciting Skill Tree perks are ones that completely change how you play the game just by adding a small feature or conditional bonus (though admittedly, these can also be some of the hardest to balance as well).

Balance is a genuine concern, yes, but I don't think it's as severe as you suggest, at least not in single-player games. In competitive multiplayer games, skill trees do make things much, much harder to balance because there will inevitably be powerful combos that you miss which will give one player a significant advantage over others once they figure it out. But in a single-player game, unless you have failed to think through utterly gamebreaking combos of Skill Tree purchases, the difference between optimal and average use of the Skill Tree might mean... a 10% change in the game's difficulty? I find that perfectly acceptable as a designer, and audiences will have fun comparing "this level was so hard for me" versus "oh, really? I breezed through that one; here's what I was building on the Skill Tree".

As with anything else, there are a couple things you can do to make sure your Skill Tree's balance is at least good enough that it won't break the game wide open:

• Have at least a dozen people playtest your game before releasing it!! Have them record video of their playtest if possible, and see how their decisions go, but if that's not feasible, at least ask them to let you know what they picked on the Skill Tree, and square that up with their thoughts on your game as a whole ("it was so hard!!", "the balance felt wonky").
• Think through every pairing of skills that would be allowed. Could any of these break the game wide open? For example, "Vampirism" might offer 10% Lifesteal on attacks and skills but cost a character 2% of their HP each turn. "Mending" might regenerate 2% of a character's HP each turn. Are you comfortable with allowing a "free" 10% Lifesteal by combining these two perks?
• If Stat upgrades are some of the perks available on the Skill Tree, then make sure your stats are equally useful for each character in your game. This is really important, but it's easier said than done and most designers get it completely wrong. Dump stats are generally bad design! INT is usually a lot more important to mages than STR (solutions do exist, such as tying max MP to STR or simply combining INT and STR into a single PWR stat). This makes +3 INT an obviously much better option than +3 STR when choosing on the Tree. And if two equally useful and very interesting spells are gated behind the INT upgrade branch and the STR upgrade branch, respectively, well, very few of your players will ever try out that second spell because it requires spending points on an upgrade they don't need first.
• Allow your players to "re-spec" (clear and re-make) their characters' Skill Tree choices. You can allow unlimited re-spec'ing outside of combat for free, if you want players to try out everything in their play through your game. Or you can have a cost (in gold or somewhat-common items) to re-spec a character's Tree, meaning that you are encouraging the player to think ahead and make semi-permament choices that can always be undone if the player really is unsatisfied with a choice they've made.

As far as the "problem" of players eventually earning enough points to unlock every single thing in the Skill Tree, I don't think this is always a bad thing. "Accumulating" things is something that inherently appeals to certain types of players (I would go so far as to consider it one Core Aesthetic of Play), and completing the entire Skill Tree (as if it were a collection) can be a goal to look forward to completing, not to mention a motivator to keep playing your game.

Having said that, not all games cater towards Accumulation/Collection, and if an Aesthetic such as Expression is more important to your game, then you don't want players to be able to complete the entire Tree (or even a majority of it). Instead, simply make sure there are only enough points available to unlock a small swath of the tree (making sure, of course, that they can at least get to the Tree's most outer branches), and offer at least several opportunities for the player to re-spec her Tree(s), in case she purchases something because it sounds cool but finds that she hates it once she starts actually using it.

 

[Aesica]

Balance is a genuine concern, yes, but I don't think it's as severe as you suggest,

It really depends on the game. I suspect my dislike for skill trees comes from the fact that I don't have a lot of faith in developers (both triple-A and indie/RM devs) to actually know how to balance their things. If wow and its legion of developers and number-crunching, spreadsheet-wielding neckbeard beta testers can't get skill tree balance right, I don't expect some rpgmaker hobbyist and their 1-5 person team to somehow do better.

Then again, maybe that's why people like them. Skill trees let people feel clever when they see stuff like "your shadow blast now heals you for 25/50/75% of the damage it deals" and then further down, they see "your shadow blast echoes 25/50% of its damage (as a single target ability) to all enemies" and think "much synergy, very self-heal" as they sink points into both of those

 

[EmeraldSpecter]

(get that Sphere Grid wannabe out of my face)

I used to be confused at the hate for the Sphere Grid, but I have learned that it isn't so much a Skill Tree as it is a leveling system. I include it in the conversation because it's a good visual representation on how to allow for branches that offer neat tidbits to go after, not because it's a sterling example of how to organize a Skill Tree.

I tried to include a screen shot of what I was talking about with the Ghost of Tsushima before and it didn't work, so here's what I was talking about (2nd try!).

 

[Tai_MT]

I am of two mindsets on skilltrees.

The first is from a singleplayer aspect. Generally speaking, I want to be able to either get everything in the tree or create the type of character I enjoy building. I am not a "gimmick player". That is, I don't really look for all the super bonuses and crazy optimizations to become invincible. I give myself two or three goals that I think will be fun and then just try to maximize my ability to do those things.

Now, from this perspective, I feel there are ways to do this right and ways to do it wrong. Doing it wrong would be "including a skill tree with random drops". This means, the effectiveness of your skilltree is entirely dependent on how much RNGesus loves you or how much of your life you're willing to waste grinding in a game for the HOPE that SOMEDAY you'll get to have some fun playing the game. If you got random loot drops, then let me max out the whole freakin' tree. That way, my gameplay isn't so dependent on getting lucky with drops and can work with pretty much anything I pick up. But, if you have a skilltree where I have a finite amount of points I can invest, then your drops BETTER BE GUARANTEED so that I can work with the build I wanted in the first place.

Like most adults, I am not into playing video games that simulate a real life unpaid job. This isn't fun for me. It's only fun for children because they like to pretend they're adults. If you're going to create a skilltree with gameplay that requires me grind for equipment that makes that skilltree useful to me or my preferred style of play, then you better PAY ME TO PLAY YOUR GAME since it's now a second job I've had to take on.

Here's the second mindset. It's from a cooperative standpoint.

In cooperative games (no PvP at all), I love skilltrees as they allow you to have "party synergy" and allow you and your friends to drastically change the gameplay and tactics of your squad just through a few skillpoints being plonked into some nodes. I can plonk all my points into snaring enemies for long periods of time and my teammates can plonk their points into AOE damage to wipe them out. Drastically changes our teamplay and the way we tackle all the challenges in the game. Or, I can plonk all my skillpoints into just attacking as quickly as possible while my friends dump their skillpoints into distracting enemies or pulling them off of me.

From this point of view, a skilltree can create fun group moments with friends and actually create replayability within the game as you and your friend try out new things. "Hey, I'll roll exclusively traps this time, you should roll like taunts and pulls to drag enemies into the traps!".

Now, if you have a skilltree for multiplayer PvP type content...

Yeah, you're doomed. I don't recommend doing this at all. Quick way to destroy any semblance of balance within your game. PvP should be reliant on player skill rather than high end equipment and overpowered builds (or excessive levels).

Now, some advice to please me personally as a game player if you intend to make skilltrees.

I like to do weird things with my builds. That is, I like to experiment and do the things most other players aren't going to try. When I played SMITE (MOBA game), I built my character SPECIFICALLY to "tower dive". Now, Smite doesn't have a skilltree. It has equipment you can buy that drastically alters the way you play the character (and I do mean DRASTICALLY... not like LoL where you get a very small and minor percentage that doesn't mean jack and squat to the character you're playing). I turned a "jungling" character into a tower diver. Nobody builds tower divers. Or, almost nobody does. Tower Diving is stupid. It's reckless. It gets you killed. But, that's why I wanted to do it. Because this character, when utilized correctly... guaranteed no character was safe in any tower zone. I built this character for pulling you out of tower zones, stunning you at the edges of tower zones, stealing your health so I could continue pushing the attack while under fire from your tower, and chasing you down anywhere so you were not safe unless you killed me. My character (Fenrir) was designed to intimidate the other players. He was a "glass cannon". Anyone who had magic attack would melt me. But, anyone who was a physical hitter or relied heavily on towers to supplement their attack (or even other teammates) would learn what fear was.

This is a small look into the way I like to play games and make my builds. I like to do weird and quirky things. I like to disrupt other players. I like to do things the game dev might not have anticipated (not necessarily being overpowered or breaking the game, but just winning by making something weak into something heavily powerful).

I like being unconventional. I'm the guy that dumps all his points into traps and pets just because he finds it funny when the AI or other players die to them. I'm the guy that dumps all his points into turrets and walls because he finds it amusing when players have to work their way through your defenses before they can even bother with you (or you can trap powerful players with your walls and then turret them to death while they don't notice). I'm the guy that dumps all his points into Luck and Charisma because it's more fun to roll really great loot and talk my way out of every single fight than it is to just bash everyone in the face with a rock. Sure, my gun is the most powerful thing in the game and I got it 'cause my Luck stat got it to drop for me... But, my Charisma means I have pretty much never fired it, and that's fun for me. I'm the guy that dumps all his points into invisibility and then healing because I intend to just outrun everyone and play "recon" for the game. I'm the guy that dumps all his points into making drugs and then gets a tranq gun to fire those drugs at people just because it's hilarious that it impairs their vision or makes them do silly things... and then wins the game doing that.

I like doing crazy and unexpected stuff. I like surprising other players or game devs with the way I play their games. If you have a skilltree that has some options in it where I can do that kind of thing... you can bet I'll be playing around with things that sound funny.

I would rather take, "there's a 50% chance when you punch someone, it disarms them" than "20% more damage dealt with every gunshot you do". Sure, I might not be able to get up close all that much to disarm you... but it's much more fun gameplay than taking fewer shots to kill you.

I'm the guy who ran around in Battlefield Hardline using pretty much exclusively a Taser. Why? Because it was fun and funny. I love when RPG's have skilltrees that allow for such extensively weird gameplay.

The fun in skilltree based games for me is just in being able to play quirky. So, I'm always happy when there are options that just allow me to specialize in doing weird and unconventional things.