Game Design: Avoiding Min-Maxing


  • Pitcrew

    I'm not a coder. I'm never going to be a coder. So this is purely a thought exercise that I have been kicking around this past week rather than something that I am actively working toward. But I thought it might make for interesting discussion so here goes!

    Now, I haven't played on a lot of games so maybe something like this already exists. But something that has always bugged me is how, with forward-facing stats, players invariably become driven to min-max their sheets. I'm a player who tends to role-play rather than roll-play and even I min-max almost out of reflex; like, I see numbers and I naturally wanna optimize how they are utilized even if I know I will hardly ever lean on them. And with +rolls showing the amount of successes, it becomes this race to see who has the most ridiculously high exceptional success roll.

    Anyway. So. Back to this thought I have been kicking around.

    I was thinking of a sheet that, instead of having a number of dots assigned to attribute/skills/whatever, it would be something more broad:

    Low - Weak - Average - Good - Great - Excellent

    So, going off WoD attributes, instead of dots, it would look more like this:

    Intelligence: Good
    Wits: Average
    Resolve: Average

    Strength: Average
    Dexterity: Good
    Stamina: Weak

    Presence: Average
    Manipulation: Great
    Composure: Good

    (I didn't worry about proper starting-dots balancing for this example; I'm just dropping in random designations)

    Staff side of the sheet, there would be numeric associations but it would be a range, not just one number for each category (low =1, weak = 2, etc.). So something more like this:

    Low: 0.01 - 1
    Weak: 1.01 - 2
    Average: 2.01 - 5
    Good: 5.01 - 8
    Great: 8.01 - 9
    Excellent: 9.01 - 9.99

    This way, players have a sheet that is less about numbers and more just a guideline of 'you are good at this thing' and 'you are not so great at that thing' and can use it to steer their RP. +rolls would just tell them that they failed, they succeeded, they exceptionally succeeded rather than spitting out exactly how many (I'm still trying to work out how it would work when you are rolling against another player; I'm guessing it would involved MATH)

    I dunno. Would people just HATE this? Would they still find a way to min-max? Do people like min-maxing and would be endlessly frustrated if they couldn't? Should I just go to bed and stop typing at 2:17a on a work night?



  • What exactly are you trying to prevent?

    Possible answers:
    Players putting all their eggs in some basket, and having the least in everything else.
    Players putting all their eggs in some critical seeming basket (combat or investigation rolls), and having the least in unused baskets (social rolls).
    Players ignoring stats etc that should be important, but never really are in play.
    Players ending up with an advantage because how they created their character (eg in WoD putting all 1s in stats to get as many 5s as possible, because that is worth more total XP).
    Players knowing how many dice they have, and so knowing their odds of success or failure.
    Players knowing what they rolled, and so knowing their likely success failure and the likely range of abilities opposing them.
    Players playing like they know exactly how good their character is, and where others are weak(+sheet telepathy).

    What do the number ranges mean (are they perhaps the total of a skill and stat?)

    I have addressed some of this for WoD many times:
    convert stats or skills for each block (physical stats, social skills whatever) into whatever pattern of high and low you want to se on average.
    (I use min-maxers delight, mentioned above, as many high as possible stats and skills for 4+ dots, but you can do sane things like set all stats at 2 base, and gives a reasonable spread of skills like 1 1 1 2 2 3 4)
    Figure the value of those spreads in XP
    Use XP in chargen.

    This means there is no advantage to one way over another, no matter how you place the starting xp, 100 xp later the two characters are still worth the exact same number of XP. Using the dot system makes that almost completely unlikely, rewarding the focussed character concept/power builder, and punishing the realistic or generalized concept player.

    Players will always figure out your numbers, so there is no mystery there. If you want mystery have +code assign an adjective to a given range of success, and never let the players see how many dice are rolled or their results. A lot of work for little payoff, unless you do a great deal more with that behind the scenes.



  • @sockmonkey FS3 does this to some extent, though it's a 1-1 range instead of a variable range, if I'm reading this right.

    It's something about @faraday's system that I really like, and may shamelessly steal, because I think it makes a good guideline.

    I've considered some things like this as well, with the range examples. I am forever on the fence, though, because I weigh these two things:

    1. 1-1 as numbers is easy to understand and you can have a standard template of 1 = Bad, 2 = Average, etc. (more or less WoD style) that could display the words on the sheet either alongside or in place of the numbers. This is much easier to learn and comprehend if it's a completely new system to the player.

    2. [range] is more realistic, and could cover more incremental raises rather than saving up to buy something on a point by point basis, which FS3 also does. It's less inherently visible player-side, though, in a way that will make things easier for some, and confuse and frustrate others. FS3 has a 'progress toward next level' marker of sorts that helps prevent this, since 'word levels' cost different amounts to raise up to.

    I decided to stick with #1 for the time being for my dev stuff, with flat raise costs (CoD style), even if I like #2 and see a lot of merit in it.

    I went with a much broader range of more specific attributes in cluster groupings as my means of reducing min-maxing to some extent; sometimes the specific attribute is going to apply, others, the average of the attributes in the cluster will. I picked this ultimately because I'm working on this stuff specifically for this hobby, rather than tabletop or a more automated setup. (I actually think what you're describing would be completely amazing for a very automated setup.)

    The design goals I am focusing on are these: 1. needs to be easy to pick up and understand since it'd be new to literally everyone, and flat costs and basic 1-1 ranges reduce confusion; 2. allowing for specificity creates more niches for the larger number of players in a persistent world vs. tabletop -- you aren't just 'smart', you can define in what way you're smart, etc. -- because having some additional distinctions there sheet-side is a helpful tool, and gives people a little non-destructive built-in 'snowflake' factor with the broader range/ability to be more specific. (There's basically enough of them that even with a pile of XP, it'd be hard to be good at everything, and everything that's there is going to matter in some way.)


  • Coder

    I've played around with skill systems a LOT - both with FS3 and other things, so I'll share my experiences.

    The #1 reason people min/max IMHO? Players have a vision and your system gets in the way.

    They know what they want their character to be - either immediately or down the road if they envision a "Hero's Journey". If your system prevents them from achieving that vision - they'll rail against it in the form of min/maxing to get as close to that vision as possible.

    Ironically, many systems try to limit chargen points to prevent min-maxing, but this ends up just encouraging min-maxing instead.

    For example: On one game I wanted to play a good, veteran archer. I went through chargen the way I thought my character should be, well-rounded and all, and was way overboard on points. It then became an exercise in "what can I live without?" And guess what? A good, veteran archer can live without Persuasion. They (quite literally) can't live without combat skills at a decent level. Is that min/maxing? Absolutely. It's also Common Sense. I am not a min-maxer by nature; I would have been happier to make my character more well-rounded. But I'm not going to hamstring her in relation to other characters by doing so.

    You can try to combat that by forcing people to be well-rounded. Make them take background skills or Persuasion at a minimum level, etc. I'm all for that, actually. It's what I did on BSGU. But it doesn't fix min/maxing, it just curtails the degree by which they are able to min/max.

    I did an interesting experiment once of having a chargen that was unbounded by points. Players were free to pick whatever skills they felt were appropriate to their characters. Of the few dozen people who went through chargen, I had one that went overboard and needed to be told to tone it down. In contrast, there were several characters who I felt had undervalued their skills based on their background and I suggested raising some. There were way more 'useless' skills on sheets than you'd find on typical games because they were free to take them without essentially penalizing themselves.

    That was one game and one very extreme experiment (that I don't really recommend for other reasons), but it supports my main point: I really don't think most players inherently min/max. Some do, sure, but I think most of the min/maxing we see in the hobby comes from people chafing against arbitrary system constraints.

    One other tip from my experience: Players Ares Confused By Stat Descriptions. Ask ten players what "Good" means and you'll get 20 different answers. The fact that different skill systems are wildly different compounds this problem. "Good" in D20 gives a very different success chance than "Good" in WoD which is again different from "Good" in FUDGE. I once had a system that used only descriptions - no numbers whatsoever, and everyone was super-frustrated because they didn't know what levels they should take.

    With no universal agreement on the words, players fall back to choosing stats based on their perceived value. What can I do with "Fair" Riding and do I think it will come up a lot?


  • Admin

    @faraday said in Game Design: Avoiding Min-Maxing:

    The #1 reason people min/max IMHO? Players have a vision and your system gets in the way.

    I'm a min/maxer. I'm also pretty unapologetic about it - it's like a minigame for me, I just simply enjoy optimizing stuff, having spreadsheets with numbers and tweaking them around until I get more bang for my buck. As such I wouldn't think a system is getting in my way when I do this but rather that it attracts me.

    In other words if I don't like it, if it's too complicated or if it gives me insufficient options to customize then I won't get into it, which in my eyes that's a failure to engage me.



  • @sockmonkey Deliberate obfuscation of numbers has some strong and weak points, and it is something that's existed for a long time. I've played games that didn't use any numbers at all, hiding them entirely to staff side while there was still automated systems.

    Dedicated min-maxer types try REALLY hard to reverse engineer the formulas and numbers. Really, really hard. I was one of those people, and as a much younger person I ran a whole lot of tests on MMOs the embraced some numeric obfuscation to reverse engineer the numbers on mechanics so I could create optimally efficient paths. Just having a character spamming abilities thousands of times in a row, recording results and so on. Figuring that stuff out was basically a mini game.

    Now in some games, like RPI type games that forbid ooc discussion, that is partly because they want to ban that behavior, of people talking oocly about mechanics in a way that would let them game the system. What happens then, of course, is that people that really, really know how to abuse systems become at a privileged position in the game with them and their friends. If the game has any competitive aspects at all they become extremely dominant, since new players don't have any way to access that knowledge with performing the same kind of exhaustive tests themselves.

    So imo the way to stop min-maxing is to make strong, effective ways of designing characters be very, very, very intuitive and what someone would do anyways.


  • Politics

    @apos said in Game Design: Avoiding Min-Maxing:

    So imo the way to stop min-maxing is to make strong, effective ways of designing characters be very, very, very intuitive and what someone would do anyways.

    In my opinion, the way to stop min-maxing is by providing multiple avenues of "attack." In a competitive PvP game, give people more than one way to knock their opponent on their ass.

    In order for this to work, though, you have to guarantee some sort of advancement cap, or else people will increase their stats until there is no weakness.

    In the Chronicles of Darkness, capping stats at 35 XP is reasonable. That allows for min-maxing if you want, but doing so is going to open up vast holes in your PC. If you go combat-masher, you're going to get capped badly by sneaky (filthy hobbit) types, politico-make-you-look-dumdums, and double-up-ungh-ungh seducers.


  • Admin

    @apos said in Game Design: Avoiding Min-Maxing:

    @sockmonkey Deliberate obfuscation of numbers has some strong and weak points, and it is something that's existed for a long time. I've played games that didn't use any numbers at all, hiding them entirely to staff side while there was still automated systems.

    This is a dangerous road to tread, though. Let me give you an example using Arx specifically.

    When I first played there I knew nothing about the ins and outs of the system - things were rolled and results were given out, but I - by design - had no way of knowing the specifics of what factored into what and how strongly. So far the obfuscation worked, and whatever the purpose of obfuscation was (which can be argued whether it's successful or not), it was achieved.

    However then I was talking to people who started revealing more about the ins and outs of those invisible system. Maybe they were reverse engineered the way that you described (through trial and error) or someone knew a guy who knew a guy, but the point is the formulas were better known by some players more than others.

    At this point there's a fundamental imbalance - a non-systemic factor not intended by the original design - where part of the playerbase has more access to optimization than the rest. In other words, some people could min-max better than others could. At least in an open, published system everyone has the same opportunity to make that choice for themselves.

    It's something to take under consideration.



  • @arkandel You can kinda address that with a hybrid method, though. Keep the benefit of the 'tooltip' descriptor, but instead of:

    • 4
      or
    • Good

    have:

    • 4 (Good)
      or
    • Good (4)

    That's the direction I lean on it, anyway. You get the benefits of mechanical transparency and RP-reference guidance that way.


  • Pitcrew

    One thing to consider is that a lot of players, when considering what makes an 'effective' character, aren't comparing themselves to an outside metric that measures their PCs against the NPC hordes, but rather directly measure themselves against other PCs. So that 'great' rating doesn't mean much if every other PC is also 'great'. This creates a pressure to min-max the areas where the player wants their character to stand out - not against an NPC baseline, but between other PCs.

    This also tends to create staff pressure to up the level of challenges facing those PCs, which also creates min-maxing pressure, because it soon becomes that 'great' just doesn't cut it in a typical staff-run challenge or against other PCs, so you have to have 'excellent' to even be considered effective, much less stand out.

    I think one way to combat this might be to encourage width rather than height in character design. Have a number of alternate progression paths that help characters stand out and be somewhat unique among other PCs without creating too great of power disparities. Fighting styles theoretically are meant to do that, but WW/OP has never been good at making them balanced between styles or between style/no-style players.



  • @ganymede said in Game Design: Avoiding Min-Maxing:

    @apos said in Game Design: Avoiding Min-Maxing:

    So imo the way to stop min-maxing is to make strong, effective ways of designing characters be very, very, very intuitive and what someone would do anyways.

    In my opinion, the way to stop min-maxing is by providing multiple avenues of "attack." In a competitive PvP game, give people more than one way to knock their opponent on their ass.

    I agree with the Multiple Avenues of "attack". I know on BitN I would have took my PC into a total combat focus with Min/Max used to optimize combat. I had a friend that ST'ed a lot of scenes for me and a small group that would make sure we needed to use a number of skills to succeed. This meant to be successful I had to have a well rounded character which kept me from Min/Maxing.


  • Pitcrew

    Fate style skills (I think Fate, certainly DF RPG) address this by their idea of a skill pyramid. In order to be upper tier of one thing, you still had to have a supporting base


  • Admin

    @pyrephox said in Game Design: Avoiding Min-Maxing:

    One thing to consider is that a lot of players, when considering what makes an 'effective' character, aren't comparing themselves to an outside metric that measures their PCs against the NPC hordes, but rather directly measure themselves against other PCs. So that 'great' rating doesn't mean much if every other PC is also 'great'.

    Yeah, quite agreed. In fact this is a very common staff pitfall where they offer things (thematic secrets, theoretical niches, whatever) which end up being completely irrelevant in-game to their players, and are then surprised when they play out the actual results of their implementation than their intention.

    For example: "In our world healers are treasured and highly regarded."
    Actual play: "One out of every three PCs is a healer, so it's so saturated they're stepping all over each other's toes".

    The same principle definitely applies to skills. It means nothing to be 'exceptional' at something compared to NPCs since I don't play with NPCs; I play with PCs, who are all as 'exceptional' as I am at $skill.



  • @apos said in Game Design: Avoiding Min-Maxing:

    @sockmonkey Deliberate obfuscation of numbers has some strong and weak points, and it is something that's existed for a long time. I've played games that didn't use any numbers at all, hiding them entirely to staff side while there was still automated systems.

    Dedicated min-maxer types try REALLY hard to reverse engineer the formulas and numbers. Really, really hard. I was one of those people, and as a much younger person I ran a whole lot of tests on MMOs the embraced some numeric obfuscation to reverse engineer the numbers on mechanics so I could create optimally efficient paths. Just having a character spamming abilities thousands of times in a row, recording results and so on. Figuring that stuff out was basically a mini game.

    Now in some games, like RPI type games that forbid ooc discussion, that is partly because they want to ban that behavior, of people talking oocly about mechanics in a way that would let them game the system. What happens then, of course, is that people that really, really know how to abuse systems become at a privileged position in the game with them and their friends. If the game has any competitive aspects at all they become extremely dominant, since new players don't have any way to access that knowledge with performing the same kind of exhaustive tests themselves.

    So imo the way to stop min-maxing is to make strong, effective ways of designing characters be very, very, very intuitive and what someone would do anyways.

    I agree with most of this.

    And will go so far as to flat out say obfuscation does not work.

    People will figure it out and it just leads to a massive gap between the people who have figured stuff out or have friends who figured it out and people who are less interested in that stuff or are new.

    I've done the RPI circuit, I've even played MUDs with full eq systems/etc that didn't have numbers. Weapons/armor had things like "minor" or "greater" enchantments, skills went from 'not learned' to 'mastered', attributes were all words too. People figure it out and it drastically widens the gap between the competitive players and the filthy casuals.

    You can have all the "don't talk about it" rules you want, people will just do it in discord (or IM services of yesteryear).

    Edits :

    For the record, I personally don't even mind obfuscation. I spent years playing games that did it.

    But, basically, obfuscating the numbers just increases the problem of min-maxing, because some people will still know how to do it, and everybody else will have a harder time understanding things like "how big of a difference really is there between Good strength and Great strength?"

    TL;DR Obfuscating does reduce min-maxing OVERALL, but it doesn't remove it, and in exchange for a reduced 'amount' of min-maxing, gives an insurmountable advantage to the people who bother figuring it out.

    I say this as somebody from PK/RP corpse-looting MUDs who has played games where 1 max level char who 'understands the game' can kill and corpse loot groups of 6+ max level/skill-maxed "perma noobs" who are running around in what seems like End-Game eq, but doesn't really stack up when you dig into the numbers.


  • Pitcrew

    @tempest I'll have to agree with this. It's just a matter of knowing players, and recognizing that there's a pretty significant minority of players for whom system mastery is a part of the game that really excites them. When you obfuscate the details from these players, you don't discourage them, you instead turn their focus with laser-like intensity towards working out the numbers, and they will use every resource at their disposal in order to know them. Not because they're evil or want to break your game, or anything - just because system mastery is part of the fun for them, and puzzles are inherently attractive to people.


  • Coder

    @arkandel said in Game Design: Avoiding Min-Maxing:

    As such I wouldn't think a system is getting in my way when I do this but rather that it attracts me.

    I think we're talking apples and oranges. I'm talking about people who come in with a particular character vision in mind - "I want to be a badass archer who grew up on a farm" or "I want to be a doctor with an interest in child psychology" or "I want to be a fighter pilot who played the violin when they were younger". You might be surprised how freaking hard it is to make well-rounded characters who don't suck (or are at an extreme disadvantage compared to their companions) in some games, especially when systems throw in a lot of skills that everyone should have to some extent, like Athletics, Persuasion, Awareness, Driving, etc. You never have enough points to get all the things you "should" have, so you are basically forced to min/max.

    Your approach of making a numerically-optimized character just for the fun/challenge of it is a completely different approach so my statement doesn't really apply to you. Nevertheless, in my experience the former has been far more common.


  • Pitcrew

    @faraday said in Game Design: Avoiding Min-Maxing:

    especially when systems throw in a lot of skills that everyone should have to some extent, like Athletics, Persuasion, Awareness, Driving, etc.

    Wait, why should everyone have these, though? People don't all have these skills. Like driving I get, imo you should be able to be fine driving a car baseline but maybe you aren't a stunt driver or a getaway driver or a street racer unless you invest in that skill.


  • Coder

    @kanye-qwest It depends on the system obviously, but I've seen countless cases where one person will say: "Well I'm not a stunt driver - I don't need Drive" and someone else will say: "Well, you don't have the Drive skill so you never learned how to drive."

    I remember on B5MUSH back in the day - if you didn't have Swimming it meant, literally, you couldn't swim. We had a lot of people who min/maxed that away (since how likely was it you'd actually need swimming on a space station game) and had some lulz when the garden flooded from a water main break and nobody could swim.

    This issue is compounded by the systems themselves. Many systems have specific descriptions for what the different levels mean and people generally just ignore them. This actually penalizes people who actually pay attention to them and try to make their character "fairly".

    For example, I'll pick on @Seraphim73 (because he knows we're pals :)) and the 100's skill descriptions, where there were things like:

    Resolve:
    1 - Someone who might turn down a knuckle sandwich.
    2 - Someone who can go hungry for a few hours without complaining.

    Alertness:
    1 - Someone who notices when someone close to them shaves their head.
    2 - Someone who notices when someone close to them gets a haircut.

    Deception
    1 - You told a lie once.
    2 - You told a lie once — and it was believed.

    Seriously - can you imagine a single MUSH character that wouldn't have at least Resolve-3 / Alertness-3 / Deception-3 ? I can't.

    I tease 100 but it's by no means unique to them. Has anybody actually taken time to read what the dot levels or skill examples are in WoD or other systems? If you actually follow what they say, people would have low level skills in a lot of things.

    But most people think "If it's not on my sheet, it just means I'm not particularly good at it" not "If it's not on my sheet, it means I don't notice when my BFF just shaved his head and literally cannot tell a lie to save my life."



  • @faraday I often pose my characters missing something that is otherwise obvious. I know a fair amount of people who do as well. So, I don't think it is really accurate or fair to say they are 'most have' skills.

    I think what skills people think they must have on their sheets is about the way a game presents itself. For example, on Arx, it doesn't feel like I have to have.. um.. 'brawling' to get into fisty cuffs with someone. Yeah, it means when code combat happens I am more likely to utterly fail.

    In my opinion, at least, it is all about the OOC culture that a game breeds more than players because the way they think on a game is based on what the game breeds for that.


  • Pitcrew

    @faraday said in Game Design: Avoiding Min-Maxing:

    You never have enough points to get all the things you "should" have

    If that is actually the case then you are talking about poor game design. However, in my experience what you are usually looking at is more often a factor of 'poor understanding' and your example probably falls into the second group. Most games are designed around the idea that if you don't spend points on a particular skill then you are 'average' at it. Not poor or unable to do something (unless the average person would be unable to do the activity because the activity is highly specialized such as brain surgery) but just 'average'.

    For instance, if I don't but Medicine then I have the 'average' level of medical skill. I can put a bandaid on someone with some neosporin if they get cut. If they get a burn I can put an ice cube on it. I don't just sit there like a complete moron without the slightest clue what to do when faced with such trivial challenges.

    In other words, most games are designed around the idea that if you 'should' know how to do something you already do. You don't lack the points to buy everything you 'should' have because you've already got everything you 'should' have for free.

    Now you can design a game around the idea that people need to buy even these common skills and then give them enough points to afford those skills but that's usually terrible design. You tend to not end up with well rounded characters as much as a wandering party of idiot-savants who are all world leaders in their particular talents but are unable to read or tie their shoelaces.