Intro to this series here: • Design Talk: an intro Sorry for the audio ducking here, I did these first few on my laptop in a hotel room, later ones will hopefully be with a proper camera!
I think an interesting use for what you call style points would be getting sidegrades rather than upgrades. If you limit what the characters can take into each mission, those would not make the characters more powerful, only give the more skilled players quirkier options. Those could even be overall weaker, if they have some very specific use. That's a very tough balancing act, though, and I bet you already have mechanics for when you get new gear, if that's even a thing in the game at all. But I think in general, it's interesting to consider that "more stuff" doesn't necessarily mean the game getting easier. Being able to turn on more brutal risk/reward things (like in Bastion) is another option.
This could work, maybe each character can only get a single upgrade from earning style points, meaning you'll be incentivised to take each character into a mission to try and get their single upgrade by playing stylishly.
Maybe make the sidegrades completely meaningless outside of just getting more style points for doing certain things. Make an entire "style" tech tree that doesn't give you any advantage whatsoever, just nets you style points in different ways.
Left 4 Dead's AI director was amazing for this, I think. If you were doing too well, it'd increase the danger, and make items more scarce. If you started struggling, it'd throttle back, and be a little more generous. The fantastic thing about it was, I introduced a friend to the game and we played co-op, and it didn't make him feel overwhelmed. At first, it got harder because I knew how to play, but once he started getting hurt it ramped back to let him recover. Brilliant idea, wish more games copied it to be honest.
Pushing the player to do weird things is what made Heat Signature and Gunpoint really enjoyable. Getting yourself INTO terrible situations before thinking about how you're going to get out puts the game in a completely different place and makes it more enjoyable. Those style points could really help make that a more 'obvious' way to have fun. Just be careful that they don't become the definitive way to play. Even with style points, if they're broadcasted on, say, a leaderboard, then players will think they suck if they don't get them. Heat Signature's daily challenge does it fine because the leaderboard is more directly related to your gameplay but it still essentially forces an unprepared player to go without the tools they need to complete a mission if they want a chance to look decent at the game. While it does get the player more on their toes, I see more people struggling with the first mission because they need a gun than people who buy the gun because they know they'll need it (At least in my ultra-competitive friend circles).
SHORT VERSION: Instead of rewarding a good item for succeeding, reward good item that adds complex interactions. This way, if the players start snowballing, the game complexity snowballs as well, keeping players in check while not undermining their rewards. Obviously narrative consequences as well. LONG VERSION: I would say there are 2 difficulties to most games: 1) plain difficulty, like enemies having more hp, dealing more damage to you, etc. 2) difficulty of interactions, or complexity, like weapon synergies, more enemies, enemies of different types - it boils down to having X options to consider before acting From this, we build our favorite 2x2 matrix: a) A weapon that's difficult to use, and does little damage (e.g. Dead Cells Boomerang) b) A weapon that's difficult to use, but does tons of damage (e.g. Dead Cells Curse Sword, Valmont's Whip) c) A weapon that's easy to use, and does good damage (e.g. Dead Cells Throwing Knife) d) A weapon that's easy to use, but does little damage (e.g. Vanilla Sword) Usually, progression in games goes from (d) to (c), meaning you just upgrade the weapon. Rewarding (c) weapons for good performance creates snowballing. However, you can keep (d) to (c) progression but instead reward (b) for good performance, so technically, it is a better weapon, but it's more difficult/risky to use, and/or creates complex interactions. Players that will be snowballing are already finding the game a bit easy, and by proxy show good grasp of game mechanics, which means they will appreciate (or at the very least not be alienated by) increase in complexity. Another thing you can do is just provide choice for high scorers, and it could be something simple like order of unlocks, or something more juicy like alternate fire modes. In this case choice means versatility, which indirectly gives power, however increases the complexity, once again, keeping snowballing in check. Please do not take examples too literally, I hope you get that they are used for illustration purposes only.
Is there not a way to give the players rewards the ostensibly make the game harder, but more rewarding for veteran/good players? E.G. unlocking a class/weapon that is complicated to play but mechanically interesting. Good players get to try it earlier, seeing as they're all that then, why don't you try clearing a room with a MAGIC SPOON then smart guy?? HUH? ...or bonus missions? -Basically "It, uh, unlocks cosmetics" is a reward idea I hear thrown out in meetings constantly and IMO it's not the magic solution people think it is. fakeedit: unlock a magic bullet???
To me "winmore" is a term I heard allot in Magic: The Gathering which refers to mechanics or cards that will do something awesome for you if you are already winning but if you are on par with the enemy or behind then it won't do anything to pull you ahead. I think there is a distinction between that and snowballing. Also I think XCOMs high snowballing in either direction is what makes it so great. It gives you a sense of urgency in each and every mission because the soldiers (and in XCOM2 the unique equipment as well) you picked up while snowballing could be lost at any mission if it goes horribly wrong. Although I do agree the game is best when you are in the middle, not too strong and not too weak, and you are having to use all the resources you have available to you to its maximum just to get by.
For XP snowballing in games like xcom, should guys level up by staying home to create tension between using your best stuff and making you best stuff better? Even require taking a guy on a match per level to encourage variety. A typical knock-out tournament seems to have the right overarching structure; When you win a match your matches get harder, but it's not a punishment, it's progress. The question is, what can you use to motivate winning, aside from granting power?
I think the thing about XCOM's snowballing is that you're more likely to have an extreme outcome than a medium outcome: you're more likely to wipe the whole team or come back home without a scratch than you are to get a scratch and still do ok. There are problems with XCOM cause the balance is off and it gets too easy, but I don't think these extreme outcomes are part of the problem. When there's very little room for failure, there's a massive amount of tension with every choice you make and that can be really fun/interesting. I love games where the margin between success and failure is razor thin.
I recall a video/blog years ago about positive feedback loops and negative feedback loops. Snowballing is the positive feedback loop. I don't remember much but it did put across this idea that for most of the game the negative feedback loop dominates until you can break out of it and the positive feedback loop can dominate. Figure 5 of this page shows the idea (and is old enough it might be the post I read originally): www.gamasutra.com/view/feature/131426/designers_notebook_positive_.php
I refer to this as "unstable equilibrium." I consider it sometimes OK for PvE games, but it's terrible for PvP games, because one player starts to win and then they continue to win and every bit of progress makes them win slightly more. Unfortunately, it's also the kind of equilibrium most designers gravitate to, because rewarding success with power is intuitive. The opposite is what I call "Stable Equilibrium," which means doing better makes things get more difficult. I think the Homeworld games tried to do that, but unfortunately Homeworld 2 (apparently, I never actually got to play it) overdid it a bit, so that doing too well will eventually make the game completely unwinnable.
So Steamworld Heist has a fun system with it's hats. If you kill an enemy they just die, but if you put a bullet just over their head then it knocks their hat off which you can then collect and put on your own characters. Hats have no in game effect apart from being collectable and there being a bunch of styles. This also rewards risky play, body shots are easy but head shots require skill, so if you miss you'll probably get the hat instead.
When I play a difficult game with no penalty for losing, ironically I tend to fail harder than if there was one, because the game is allowing me to ram my face at it until I tilt. But obviously, it's a fine line to walk - I might not even play it if I find it too stressful. Maybe you should also give style points if you beat a mission first try, or multiple in a row if they're procedurally generated, or for using an entirely different team from the previous mission? (Assuming style points aren't a completely limited resource.)
This is great! Love the topic, as snowballing and how to avoid it has been giving me a lot of design challenges in one of my projects right now. I’m also excited to hear that Breach Wizards is still in development, as it sounds like a great concept. And finally, I can’t say how much I love the idea of style points for gratuitous defenestrations. Tom Francis is redefining window-based design! Imagine if he invented Microsoft Windows! :-D
One case where "snowballing" can be an interesting mechanic is when it's based on risk vs reward. Take Invisble Inc: i could get out of this mission now with the objectives accomplished, or stay a bit longer while the security level keeps creeping up to try to get some more loot. It's important that this loot is not just cosmetic, because I *have* to take some risks in some missions or i will be underequipped for the final mission and lose: this creates lots of interesting decisions. The snowballing effect is also limited by the limited duration of a campaign. Snowballing within one run through a relatively short game can work if i can then play another run in a more challenging mode if i feel it was too easy, or play some more challenging characters that I've unlocked in the mean time. Of course i have no idea if any of these things could be applicable to tactical breach wizards. Personally, with a few exceptions, I'm not hugely motivated by rewards that have no gameplay effects. And for the record, it sounds like i play xcom much like you, picking a mode that's just a little too easy for me and occasionally reloading if i really don't like what just happened.
Another possible solution (4 years too late), is to also provide throw away missions for farming. There are the main mission ones and the less important ones. How far a developer want to explore this idea is up to them, but for the snowballing discussion it works as a way to unfuck yourself after a terrible mission. Get the rookie team some experience before getting back on the main mission.
Snowballing can be a form of tutorialization if it is balanced by a difficulty curve. Binding of Isaac has a multitude of snowballing mechanics but the most important one is Deals with the Devil - don't get hit and you can trade health for strong items. But this is evened out by the game unlocking the more difficult parts only after the player does well. This creates a sort-of tutorial environment and a strong sense of progression once the early parts become easy. Also a neat way to invert the usual roguelite everything-gets-easier-with-playtime thing.
If I'm not mistake RE 5 has an hidden system where the better you do, lots of head-shots and things like that, the enemies become harder to kill and smarter. Maybe it's just a legend though
Is snowballing something that can only happen in a long-term campaign or run based game? I've heard arguments before that in games like Devil May Cry where you are rewarded more for playing well, that is because you are being given quicker access to more complicated mechanics and tools, and that on some level, things like scoring systems tied to reward can function as a tutorialisation method, by encouraging players to become proficient with the tools they have before getting other stuff (although I think many remark upon an apparent innaccessibility of spectacle combat games leading to them just to get used to scoring badly, and not thoroughly engaging mechanics throughout the whole game. Also that flew a lot better when it was more expected for people to do multiple playthroughs of a single campaign.)
Good initiative to try and create a common vocabulary. But it misses its effect when you use another word for an already existing term. 'snowballing' is simply a positive feedback loop, which is generally understood in game design. The opposite is the negative feedback loop, in which the game would diminish the player's capabilities when he's successful.
UGH, snowballing! I hate it! Mario game power-ups in particular usually make the game SO. BORING. The pure flight powerups in particular are BRUTAL about this, you can trivially beat like half the entire game without ever interacting with anything and then BAM, you took one damage once so now the game goes from way to easy to way to hard in a crazy difficulty spike cause you never learned to actually play the game good you just had the dumb OP powerup instead. All the Mario games with free flight powerups become so much vastly worse than the ones that don't solely because of the one powerup. There's a reason most Mario romhacks end up implementing a 'no flight powerup' ban (or just no powerup ban, or no 'advanced' powerup ban) across most of the game, removing them every time you enter a level. The propeller hat is okay. Though sometimes even having a fire-flower - the 'weakest' of the advanced powerups - feels like too much and trivializes certain bosses and enemies. Bluh. -_- The arcade-y shooter danmaku game thing where you have a billion damage until you take one hit then you have zero damage is an even WORSE version of this. Though where the line is drawn between death spiral mechanics and snowball mechanics is mostly academic. It always gives me the same bad feeling either way though.
@@TheUncommonVideo Well it really depends on the game. Some thrive with that type and some thrive with the other. For games such as XCom for isntance, I'd say it's better that way. Think about it, if it were the opposite, the game would be easier at the start then you'd end up with a game over 6 hours from now... Doesn't feel great. By doing the opposite, you make the player wary from the start then it gets easier, yes, but then the player would be prone to a silly mistake and put pressure back on them...
Pyre is good example for the kind of non-snowbally progression system you were mentioning. Defeat doesn't mean falling behind mechanically but it definitely matters story-wise.
@@randomcdude3534 I think there was something like that. Also inspiration bonuses encourages cycling through different characters. Finally winning an ascension ritual meant retiring a powerful character, forcing you to adapt and use the rest of your roster. It is like a thesis on developing non-snowballing progression system.
I feel like by default games shouldn't snowball like you are talking about, but giving those players who yearn for that challenge having a chance to play a version that does that. So like in your game maybe having an uber mode where they lose xp on defeat for a mission, but not making it the default mode.
Tom, I can not tell you how much it concerns me to hear that you are planning to put a style system into Tactical Wizards. When the recent major overhaul of Heat Signature introduced a style system there, I was immediately turned off by it. The main reason I hated it was because there was zero communication about how to earn style points. Nothing was defined, nothing itemized or reported, nothing notified during gameplay; there was zero communication about what style meant, was worth, or how it was achieved. It seemed like an arbitrary system arbitrarily put in place to intentionally annoy. It was 'Disincentivization 101'. The belief in any merit of a style system is fundamentally flawed and here's why: it's utterly subjective. Style is as ethereal and nebulous a concept as 'fun'. What is fun for you may not be fun for me and vice versa. The same is absolutely true, probably even more so, when it comes to style. The examples you specified at 7:23 & 7:34 illustrate this very clearly. If you'll bear with an analogy: your examples imply you derive or interpret a sense of stylishness from cheap action films rife with car crashes and explosions. What I find stylish in movies is a spy with unshaken nerves, a subtle cue of suspense, a clever turn of phrase by a suave personality. Bringing this analogy back into terms of the game: your examples at the aforementioned time stamps define pretty much the exact opposite of what I would find stylish in a tactical turn based game. My sense of style would be elegent efficiency, ruthless precision, eliminating a threat as quickly as possible to increase my strategic leverage or applying *exactly* the right amount of damage to fell a foe while conserving ammo (or whatever resource wizards use) as a point of personal pride. *THAT* is what *I* would deem stylish. I repeat: style is subjective. Trying to award points for player actions that are "cool" in the eyes of the developer is neither intuitive nor instructive nor fair. I can already guarantee you my paycheck for Tactical Wizards because it's a Tom Francis game and I have to try it -- that is the legacy you've esteemed in my regard since the masterpiece Gunpoint (in which style was left exclusively to the player's interpretation, like a fine piece of art) -- but I can equally guarantee you a downvote on Steam if this game includes a style points system. The only way you can make such a system attractive is if you include player-defined parameters for what counts as stylish, because that is the only way to respect the truth of such a system's subjectivity.
