Okay, I originally posted this on rpg.net, but I thought most of the points listed here are relevant in CRPG''s. Note though that most of these points are more valid towards turn-based combat systems as opposed to real-time systems. So here goes: 1. Skills relying on one attribute In game terms when this happens, it tends to mini-max characters such that they concentrate all their attributes in whatever skills they think they''ll need. So several useful attributes get abused, usually Agility (and its derivatives) and Intelligence. The problem is not just unrealistic, but it creates game effects that causes a failure in roleplaying because of this minimaxing. Solution?: Have skill systems that rely on more than one attribute or aptitude. 2. Initiative is all equal, or grossly unfair Most systems either let every character act once a turn, or they have some kind of phase system in which certain characters get to act at different times according to their speed. Either way it''s unrealistic and unbalances (or doesn''t give enough balance to) the game system. When everyone gets one act a turn, even if slower characters get to act last, it still gives them an opportunity to act at all. Moreover, too many systems base speed solely on physical prowess and not on intelligence or awareness (see complaint #1 above). It''s possible for someone to be very fast, and yet freeze up, or not know what to do. Conversely, you can have someone who''s physically slow and yet can react mentally to situations very well. Solution?: Have actions have "costs" to perform....initiative is determined by a combination of factors and creates a "pool". The actions the character performs subtracts from this pool, and the highest number in the pool goes first. 3. No Seperation between pain, shock and physical trauma Many games have concepts of hit points, or life pools, or damage tracks or whatnot. But very few seperate the actual physical trauma inflicted by a wound and the pain or shock that it inflicts. For example, some wounds could be very lethal, and yet they don''t affect the character greatly (for example some poisons, or internal bleeding). OTOH, you can have excrutiating pain which isn''t damaging at all (certain joint locks or stubbing your toe). Some games have "stun" or fatigue tracks, but this is still slightly different, as it tracks your "Consciousness" level or your ability to perform. Solution?: Have two seperate factors that track both the physical (life threatening) damage as well as a pain/shock factor (which can be life threatening, nut not always). 4. Making your to-hit roll very well, and then rolling for the worst damage (and vice versa) Some say critical hits and misses are the answer. But this only answers a narrow cut-off of possibilities. what if I missed the Ciritical Hit roll by one point, but still had a very good margin of success? Damage should be intimately related to how well you hit. If you barely hit the other guy, that means you grazed him with the edge of your weapon....the damage should be minimal. conversely, if you hit really well, you hit him in a vital spot you landed a really solid blow, or you hit a weak chink in the defenses. Solution?: Have a system that has Margins of Success (see below) and influences the damage done. 5. No direct relation between how well you did your task, and the outcome of the resolution (Margins of Success) A game system which seperates the success of a task from its effect is what some call a "binary" system...meaning you pass or fail. The system doesn''t check how greatly you made your skill resolution roll by...it just wants to know whether you passed or not. This has all sorts of ramifications (like the one in #4) on how the game is played. Solution?: The system must have some sort of margin of success system built into the task resolution. 6. Rounding off Many systems have derived attributes or other functions that are rounded off (usually in the character''s favor) to produce another characteristic or game attribute. Unfortunately, this leads to mini-maxing effects so that Players only buy enough to get the round off in their favor. Solution?: The task resolution should be able to handle attributes without dividing them. Have them in their unadulterated form if possible so that all the "data" of that attribute is retained. 7. Damage systems that do not factor in what kind of damage they do Damage comes in many different forms. A bullet does its damage differently from an arrow which does it''s damage differently from a laser. They are stopped by different kinds of armor, they have different penetration, and they have different effects on the human body. Solution?: The damage system must be designed in conjunction with the health track system to account for the different kinds of trauma that can be inflicted as well as how armor affects damage types. 8. Probability systems Statistically speaking, one die systems (like d20) have an equal chance to roll any number. Therefore it is just as likely to roll a 1 as a 20...or a 5 and a 20. This makes critical hit or miss rules (and Margins of Success in general) more difficult to do. Solution? Multiple die systems on the hand are bell curve systems which make individual numbers unequal in terms of how hard it is to roll them. With some ingenious tinkering it is also possible to create non-bell curve systems with multiple dice (as in EABA "best of three" system, or systems which let you re-roll 6''s and add the result). Non linear probabilities allow for more interesting skill resolution systems and help avoid the "damn, kung fu charlie came out swinging with his sword and cut his own damn fool head off" syndrome by making fumbles and critical hits harder. 9. Little attention to personality traits Most games only cover the physical and mental aspects of a character, leaving psychology for the roleplaying. But roleplaying is more than just the physical and mental capabilities of a character, it is also his passions, drives, flaws and weaknesses and how they affect his behavior. Too often games assume that a player will responsibly act in character according to the character''s personality, quirks, history, faith, etc etc. But often this does not happen. More often the GM is forced to use negative reinforcement to make the player act in character. Few systems however reward with tangible rules-backed in-game benefits for acting according to character. Pendragon and The Riddle of Steel are the only two games I know of that do this. Now let me make a few comments about the computer medium in particular. One of the main design concerns in CRPG''s is to decide whether combat is handled by the character (most turn-based systems) or by the player with some modifications based on character ability (most real time systems). In the former, the player generally selects tactical information and then the computer does the number crunching and determines the result based on the enemy capabilities. Unfortunately, many find this rather boring since it creates too much of a detachment between the player and the character. In the real time system though, we have a possible dilemma in which good reflex players will have an unfair advantage over not-so-good reflex players even if the latter is supposed to be playing a better fighter character-wise. In other words, we have a dichotomy between player and character. In turn based systems the danger lay in creating a non-relatable protagnoist. In real time systems, we have the danger of putting too much emphasis on player ability rather than character ability. And roleplaying is supposed to be about assuming another character identity and capability is it not? Choosing which style of play is perhaps the first thing a CRPG designer should worry about. I tend to prefer that the player himself have little physical input into combat process and instead solely make tactical choices. For those who''ve been weaned on CRPG''s this may not make as much sense but for those who grew up on PPRPG''s, this was the only way to do combat. Moreover, it encouraged imagination and creativity. So I''m not sure it''s really a pet peeve, but I hope more CRPG designers find ways to make turn-based combat more intriguing as well as better able to connect the player with the character on a personal level.

Hm. This is a nice thread.
Basically, most of my problems in RPG design lay in gameplay. However, since I play exclusively turn-based console CRPGs, I can''t say anything about PPRPGs.
Main problem I have with CRPGs is quite obvious to anyone who''s ever played one: your characters are apparently incapable of jumpgin or climbing except in certain instances. Take this example: you are walking along a mountain path. There''s a rock in the way. Normally, one would be able to climb over a single rock. I do it all the time on hikes. But no, in the game, you can''t. So, these games really should implement climbing/jumping. Otherwise, the degree of unrealism gets on my nerves, not to mention it destroys the fantasy world created by a game and reminds me that it is, after all, just a game.
Now, with battles, which seem to be the main focus of your point, I agree with you on the player control vs. character control thing. There should be a balance - obviously, as nothing good is entirely one thing or another. I myself prefer a turn-based system over a real-time, as it allows for more thought. Realism in this case is moot, as the turn-based system by default is quite unrealistic. However, one thing that I''ve seen is a nice thing to imlement. If you''ve ever played Super Mario RPG for the SNES, it has a timed-hits sytem. Basically, you hit a button at the right time during the attack, and you do more damage, or get an extra attack, or some other nice thing happens. This inserts an extra degree of player control into an otherwise almost all character-controlled system. It''s nice.

I really like your ideas of the using multiple different stats to level up talents. That seems like a really under-used tactic. One thing I was thinking about in the implementation of something like that is to never list the correlation between the stats and abilities, but to sit down and make it so that every single stat one can have has a realistically proportional influence on just about every ability.

In this way, you could have something like damage. This could be composed of the accuracy, power, skill with a certain weapon, intelligence (to some extent, knowing not to aim for heavily armored spots), agility, etc. and each would have a different level of influence on the total damage done. What do you think?

In each of us there is a force that compels us to act selfishly, no matter the consequences. It is the force that will eventually cause our downfall. It is the ID - Thanatos.

I agree and disagree on different points, but I will play the devil's advocate here.


As for points 1, 6 and 9 you are assuming that all people who want to play RPGs are actually in it for realism. Personally I am willing to bet that 90% of RPG gamers are not in it for the sheer realism of it. In every one of these points you proved my own.

#1 "...they concentrate all their attributes in whatever skills they think they'll need".
#6 "...Players only buy enough to get the round off in their favor".
#9 "More often the GM is forced to use negative reinforcement to make the player act in character".

Take a look at #9 again... Why would anyone want to play a game that tries to force you to play it the way the designer thinks it should be played? This doesnt contribute much to the players experience of the game if it doesnt just frustrate him alltogether. So for all of these points, I say let the player play the game in whatever way he thinks is fun. If he wants to try to play a computer game by maxing his stats for efficiency, let him. If he pays money for the game then goddamn let him! Also its EXTREMELY difficult to make a CRPG that actually inspires a player to roleplay. Putting believable emotion into the NPCs of any CRPG is the most difficult thing I could imagine, but if the player did want to take a shot at playing it with some true roleplaying then I'm sure he will do it just the way you are trying to force him to anyway.

As for point 2, I should say that initiative is stupid! If your combat lasts 20 turns, everyone is going to get to act 20 times, plain and simple. If its one on one combat and one person is much quicker in battle by whatever factors you put in, then guess what? Unless the other guy gets a lucky roll once in a while, They'll take turns!!! If their initiative factors give neither an advantage, then your worst case scenario is that they'll switch of taking 2 turns each. P1 then P2, P2 then P1, P1 then P2, etc... Waste of dice rolling/cycles if you ask me. Although oddly enough you even stated this in your point, but what I dont get is how it could unbalance the game if the rule is nearly inconsequential.

Point 3 could be argued that yet again realism is not always best and your obsession with the realism of combat is creepy, but that doesnt nessicarily mean that all players would share my opinion on this matter. Like I said, just trying to play the devil's advocate.

I will agree though completely with the remaining points, numbers 5, 7 and 8, because these things things do seem as though they would contribute to the interactive experience of every player whether or not they to play for points or play for the roleplaying. I dont see much a way to play devil's advocate on these.


The thing is that as a developer you shouldnt care how a player plays your game. If you want the player to truely roleplay, you should make a game with such depth and emotion coming out of it, that it inspires your player to roleplay. Can you imagine trying to actually roleplay along with one of those older DOS D&D games? They lacked dramatic feedback so back it was damn near impossible to roleplay with. Best way to inspire roleplaying is to make your NPC's as believable as possible, or even the whole game world for that matter. I personally think that the GTA games accomplished this better than any other game I've ever played. Vice City was just very believable place if not a bit twisted. Although sadly it wasnt even an RPG...

~Vendayan

[edited by - Vendayan on December 20, 2003 2:24:47 AM]

I agree that I''m primarily interested in more realistic design considerations. I was weaned on old school "simulationist" style roleplaying games like Phoenix Command, Runequest, Champions, and Gurps. Games which ventured into "narrative" style came much later and in my opinion do a disservice to players who are not good actors for they stress too much in-game modifiers depending on how well a player acts things out. There is nothing stopping a rules-heavy game from having strong narrative flow as long as you remember one critical rule: "the rules are not your God".

In computer terms this is a bit more difficult to answer though since they don''t have the judgement capabilities of a human. Many were a time as a GM where combat took place and I simply fudged a roll in order to save the PC''s lives. While not entirely "realistic", it preserved the dramatic flow of the game. One option that is used in some games are "fate points" which allow the players to modify ex post facto certain die resluts, in effect making their own luck.

However, I feel that if one wants the true PPRPG experience (to which IMHO CRPG''s simply can not hold a candle to), then they must have some form of moderation ala a GM akin to what Neverwinter Nights does.

As for point #9, it''s not me the game designer that enforces personality traits on a character, because you the player get to chose what your traits are and their rank. For example, in Pendragon you had a list of traits with a scale. at 10, the trait was deemed average. Above 10 and you were seen as more and more an exemplar of that trait...less than 10 and you were seen as contemptuous of that trait. Traits were things like Honor, Piety, Loyalty and other "knightly" traits as well as a few others. The player himself got to chose what rank he had in each trait. Therefore if the player committed an action which seemed to go against a certain personality trait, the trait was qualitatively defined giving a better barometer for what the GM could do in such a circumstance. In The Riddle of Steel, a player could actually gain in-game bonuses if an action coincided with a personality trait. For example, a player could have a spiritual trait defined as Love for a certain person. If this person were threatened then the player could declare that he was using his Love trait as a bonus dice pool to his actions. Again, such a system in a non-moderated game would be virtually impossible.

Given that this forum is about the design of computer games, most if not all of your complaints can be solved. The things you're complaining about are PNPRPG rules that were created for the benefit of the GM/DM who has to put them into effect. As a GM, would you want to spend 30 minutes trying to add, subtract, multiply, and divide numbers in order to figure out exactly which limb you tore off the monster before it died? To fix the problem, simply stop using DND rulesets, or stop buying CRPGs that use DND rulesets.

And then rolling die. In the computer world, the random function is just a complex algorythm thats guarenteed to produce a regular series of numbers given an input. If you make the input some arbitrary number like the number of seconds passed since January 1st, 1970, then you get a series thats somewhat unpredictable. Conditional to the formula, its like having a die with 4.2 billion sides, and with the amazing speed of most computers when it comes to simple calculation, you could easily produce numbers with any degree of probability.

[edited by - inmate2993 on December 20, 2003 12:22:38 PM]

I think you should look at the Hero and GUPS P&P RPG systems. While they aren't perfect, they adress a lot of the issues you seem to have.
[1] In Hero, Skills start out as 9+(stat/5) {have to roll under that to succeed} or base 11 {for skills not based on stats}, which makes stats effect skills not at all. Of course, Hero is designed for heroic games, so you might want to make stats cost more, give more points and make skills start at 2+(stats/5) so you can realistically represent all skill levels (since realism seems to be your goal).

[2] Hero also has a nice initiave system: there are 12 'segments' in each turn, and which segment you act on is determined by your speed. Speed can be from 1 to 12, and you act that many times per turn. 1 speed means you act on segment 12, with 2 you get to act on segment 6 and 12, 3 means 4 8 12, etc up to 12 where you act every segment. This makes slow people act less often and fast people act more often. You could extend it to 100 phases and make default speed 20 instead of 2 (or something like that), so that speed changes have less effect but still do. You could also borrow from GURPS, which has an advantage you can buy that says basically 'When the GM says you have to decide NOW, you can take your time'. It is meant to allow for characters that think very quickly.

[3] Hero has 2 types of damage, stun and body. Almost every kind of attack does both, but they do more stun than body damage. This means characters get knocked out long before they get killed. You don't die for sure until you're at -HT{essentially hp stat} damage, but when you reach 0 HP, you start losing 1 every turn. It also has optional rules for wounding, which make you roll vs a stat with a penalty based on the amount of damage, and if you faily you can't take offensive actions next move {but you can still dive for cover, cower in a corner, etc}. In a realistic game, you might want to adjust some attacks to do more body than stun, but for heroic games what is written works well.

[4] Hitting well and damage are two seperate things. I can hit you REALLY well and do no damage, or I can barely hit you but rupture a vein or something like that. Basing damage on how well the hit succeeded instead of a seperate role is definitely not realistic, but it could make combat more fluid and thus more fun. It would depend on the dice you use. On a D20 I definitely wouldn't want it, but with 3d6 it might be okay.

[5] In Hero, skill rolls are only made for exceptional situations (ie dangerous or unusually difficult). The section on skills talks about how bad failures are based on the roll, but doesn't mention that a better roll should have better outcomes if it succeeded. IME, any competent GM would make the amount of success depend on the roll if they also make the amount of failure depend on the roll as the book suggests.

[6] Hero rounds off normall, and it is a problem with it, but since some things divide by 5, others 3, etc, it means that increasing it only enough to round off might round up for one thing and down for another. If you added in more divisors(maybe just the # 2), it would pretty much entirely fix the 'problem'.

[7] Hero only has body and hits damage, but you could borrow from GURPs here and have crushing, impaling, cutting, and special (for weapons that do extra things, like ensnaring the target). Each one has different multipliers for different situations.

[8] The 3d6 of Hero and Gurps is very nice. As you can see on this 3d6 probability table, the 18 and 3 for criticals are fairly rare at less than half a percent chance.

[9] With GURPS Advantages/Disadvantages and Hero Character Disadvantages, you can pretty much describe any type of personality. GURPS has the better coverage of the issue since Hero only covers general categories, but it is also less restrictive since it just gives point ranges and examples instead of a list of what you can pick (though you can make your own in GURPS as well of course).
I find that a lot of what you say in #9 is more based on the group and GM than the system. With a good group, people play their characters and the GM gives the characters interesting choices. In a bad group, either the players make the choices(without concern for the characters) or the GM doesn't give any, and then you have to start coaxing players to play. Also, negative reinforcement is the least suggested method for getting people to play in character. If you read any RPG newsgroups, you'll find that almost everybody recomends positive reinforcement as well, but that means you need at least one good player in the group to reward for playing in character and other 'good things'.
If you're going to talk about computer 'RPGs', you should remeber that it is really a misnomer, and they are just action games in a fantasy setting, no more an RPG than Half-Life {which was a great game and had a good story IMO, but was in no way an RPG}. Also, I'm not sure what you're saying about P&P RPG players knowing about only making tactical decisions. When I played, I could decide exactly what actions my character takes and then have a % chance of my character succeeding at the action, which is just like the realtime CRPG games that I've played. In wargames, the choices seem to be generally limited to tactical decisions, but I haven't played much of them so I can't really comment.

[edited by - extrarius on December 20, 2003 3:13:27 PM]

quote: Original post by Dauntless
Now let me make a few comments about the computer medium in particular. One of the main design concerns in CRPG''s is to decide whether combat is handled by the character (most turn-based systems) or by the player with some modifications based on character ability (most real time systems). In the former, the player generally selects tactical information and then the computer does the number crunching and determines the result based on the enemy capabilities. Unfortunately, many find this rather boring since it creates too much of a detachment between the player and the character. In the real time system though, we have a possible dilemma in which good reflex players will have an unfair advantage over not-so-good reflex players even if the latter is supposed to be playing a better fighter character-wise.

In other words, we have a dichotomy between player and character. In turn based systems the danger lay in creating a non-relatable protagnoist. In real time systems, we have the danger of putting too much emphasis on player ability rather than character ability. And roleplaying is supposed to be about assuming another character identity and capability is it not? Choosing which style of play is perhaps the first thing a CRPG designer should worry about. I tend to prefer that the player himself have little physical input into combat process and instead solely make tactical choices. For those who''ve been weaned on CRPG''s this may not make as much sense but for those who grew up on PPRPG''s, this was the only way to do combat. Moreover, it encouraged imagination and creativity. So I''m not sure it''s really a pet peeve, but I hope more CRPG designers find ways to make turn-based combat more intriguing as well as better able to connect the player with the character on a personal level.

I saw this post on RPG.net and thought both it and several of the replies were very interesting. I know the quoted text isn''t the central topic of the thread, but it''s one of my own areas of interest so I thought I''d chime in . I''m not opposing your viewpoint btw, just suggesting an alternative way of looking at it.

Instead of thinking about turn based or real time as the split, how about considering which character attributes are to be provided by the system, and which are provided by the player. Doing this may also help determine where the challenge of the game lies, i.e. is it the player''s skill at selecting enemies with the mouse quickly or is it his ability to think tactically. This isn''t just based on TB or RT, for instance take X-COM: Apocalypse and Quake, both RT games. In Quake it''s the player''s intelligence, dexterity, perception and speed that are tested. In Apocalypse''s RT mode tactical missions it is mostly the player''s intelligence only. The dexterity, perception and speed stats of the units he controls are far more important than his own.

It made me consider whether in a C/RPG the character should have an intelligence stat, if intelligence is provided by the player. An example with dexterity is the shaking sniper view in Deus Ex. The game tests dexterity twice, once with the character''s sniping skill and again with the player''s mouse ability. In this case players who had greater mouse ability could substitute that for the character''s skill, which provided an in-game benefit of freeing up skill points. Is this the way games should be, or should each test be dependent on the character or the player, but not both?

Apologies for the somewhat meandering post . What thoughts have others had about these subjects, and what did you conclude?

Fulby

What good is to have more realistic Combat when the environments and tools are so unrealisticly non-interactive?

The "climb the damn rock" example is one. Also the outdoors are usually just big rooms with definite entrance and exit points and boundaries.

In P&P games, with the right tools you can climb any tree, wall, etc. You can dig a hole anywhere that has dirt. Bury items under autumn leaves. Gather water from any river, well, bucket, etc. Go anywhere and overcome any barrier with the right tools.

What happens to the bottle after you quaff a potion?

Why can''t you spill the oil from your lantern on furnature and light it? Why can''t you pile furniture or brush to create a barricade?

How about hiding under a pile of bodies to evade a terrible foe?

How about writing a message with your blood?

When you walk in sand, there should be footprints.

So basically, I think CRPG combat is already in much better shape than the environment and objects. I would put priority on making them more interactive.