# What approaches are there to lessen or eliminate reliance upon dice in an RPG? [closed]

Is it necessary to use dice all the time in a dice-based RPG system? In other words, are there occasions or means by which you speed up gameplay by either reducing the number of rolls or removing the dice from the game altogether? How did that work out for you?

I ask this because Paranoia (yes, I have a theme running here) encourages the game master's use of dice behind a screen -- where no one can see the results. The idea is to use dice to intimidate the player, but to ignore the roll and do what's best for the story and the game as a whole most of the time.

The more I think about this tactic, and being a player in a D&D campaign, and having played and GMed D&D (2nd edition), Call of Cthulhu, Battle Tech and Tales from the Floating Vagabond, I realize that what's best for the game as a whole, the story and the narrative may not always be what the dice come up with. On the other hand, the dice can also be a great way of introducing entirely new situations which may not have come up otherwise -- I'll be point-blank, my critical failures have created some of the most entertaining aspects of the game I'm currently in!

It's given that some systems are diceless, but many are not; and some are designed to use few enough rolls that they go quickly, while others are designed to be more in-depth, randomized and cerebrally-paced.

What are the pros and cons of performing every roll? What are the pros and cons of removing or reducing dice rolls? In the event of the latter, should the dice still be rolled for dramatic effect, but ignored by the GM? Should dice be rolled behind a GM's screen to heighten the drama? (That answer is intrinsically yes in Paranoia, at least according to my read of Paranoia XP.) How did any of these approaches work out for you?

Have you tried reducing the number of rolls in a game, rather than removing them altogether? How did that work out for you?

-

## closed as too broad by SevenSidedDie♦, Tynam, Wibbs, doppelgreener, KRyanNov 1 '14 at 18:26

There are either too many possible answers, or good answers would be too long for this format. Please add details to narrow the answer set or to isolate an issue that can be answered in a few paragraphs.If this question can be reworded to fit the rules in the help center, please edit the question.

Y'all want to explain this sudden, tardy close? – mxyzplk Nov 1 '14 at 23:58
This question seems to underestimate the scope of the RPG hobby, and it's actually at least three different questions rolled into one (some of them are already answered elsewhere!). We've got answers about dice-reducing homebrew experience, system rules that limit dice rolling, and alternatives to dice like Jenga towers and negotiation, all of which address just part of this question's scope. – BESW Nov 2 '14 at 2:33
If the above is intended to explain the closure, then I'm good with it. (As for some being already answered elsewhere, please understand a few things: They weren't when this was asked in early 2010. In fact, this was asked before the RPG SE guidelines really came into form. Now that they are better-formed, I myself support the question's closure.) – John Rudy Nov 10 '14 at 22:24

There are plenty of alternatives to using dice. Gnome Stew has an excellent article on replacing a dice system in a survival-horror game called Dread (by Epidiah Ravachol) with a Jenga tower; every risky action requires a block pull, and the game continues until the tower -- representing the player's sanity -- falls.

Sometimes the more creative alternatives can be the most fun.

-
I don't understand why this answer has 20 votes. It's a good answer, but hardly the best answer to the question. – Adam Dray Sep 22 '10 at 14:32

Hmmm, the rewording of the question makes my and other's discussion about diceless systems mostly off topic... Rather than deleting that part I'll just expand on "Approaches/schools of thought on when to not roll or ignore dice in a dice-based system."

Techniques

Step one is don't require rolls when they shouldn't be necessary. I've had GMs that have you roll Driving whenever you just want to tool around town, it's obnoxious. The D&D 3e "take 10" mechanic is an attempt at removing nuisance rolls as well, rather than do without any mechanic it's "compare 10 + skill vs target" if it's not a risky play. Consider giving people info if they just have the relevant skill, especially in games with a lot of skills - reward that one player that bothered to take Underwater Basketweaving as opposed to having the one chance in the game to use that skill come up, be rolled against, and failed.

Let things get "mopped up" without rolls. 6 PCs on the last goblin? Just say "So you want him dead or captured or what?" Assume they can make it happen and save the rolling for the gripping fun parts and not the lame routine parts.

You might fudge rolls to expedite things. Though you might ask yourself, "If I plan to fudge this roll, why am I rolling for it?" Games like GUMSHOE take parts of the play experience, like finding crucial clues, and make them a non-roll event. You can roll to get additional info and whatnot, but the plot-advancing clues are just there for you.

As for ignoring dice in a diced game when people think you're rolling dice - I fudge from time to time when I think it's needed but generally I go with it; I've found that the dice frequently make more interesting decisions than a person trying to make interesting decisions does.

Also, using "hero points" or other options where the players get to cash something in to declare success rather than rolling helps reduce dice reliance at highly critical junctures, if your concern is "massive fail from bad roll" as opposed to "time wasted on fiddliness."

Existing Games

There are a fair number of diceless (more precisely, randomizer-less) games especially on the indie side of the fence (e.g. Nobilis). Amber is a venerable one. They're not for everyone, and are mainly best for "shared-narration" games, but it's definitely something everyone should try once to see that it's possible. You can then choose to use those sorts of resolution mechanics instead of dice. (Technically there's games that use cards or Jenga towers as opposed to dice; IMO that is not the point of this question.)

There's also "less-dice" games, like GUMSHOE or The Committee for the Exploration of Mysteries, which use dice some but generally construct their rules so a lot of play happens without rolling. In GUMSHOE it's a lot of "you find the clue just by going there" and "you get this information just from having the skill." In TCFTEOM it's more narration driven player consensus stuff; die rolls only happen at occasional critical junctures.

Personal Experience

Heck, my very first experience with D&D was when we'd play in cars on the way to Scout camp at night - so the DM didn't use any dice (or pretend to), he just made up results. Worked for us. We had all the fun and none of the math of normal D&D.

-
+1 for "the dice frequently make more interesting decisions than a person trying to make interesting decisions does." – yhw42 Aug 30 '10 at 18:23

Without any prior knowledge of any such system existing anywhere, I'll just share my thoughts on a system that may be good for getting the dice off the play. I have not tested this but it may be fun to try.

# Bargain for success

This system uses a currency of some sort(call it karma, chi, clout, mojo, energy, mana etc. Whatever fits your campaign world. I'll use the symbol @ for example's sake from now on.) that can be spent to buy successful actions from the GM.

### The basics

The player announces his intent for an action, and if the GM decides that the action is a non-trivial one that should have a cost, he tells the player so and asks if he would like to commit to the action. If the player says yes, then his character is committed to the action and is doing it, come success or failure. Then the bargaining part begins. The player or the GM may make offers and counteroffers that involve exchanging a specific result for a set amount of @. Once they agree, the player gives up the agreed amount of @ and the agreed result happens.

GM – Ok Frodo. You notice that the chain holding The Ring™ around Gollum's neck has broken and the ring is falling down as Gollum is shouting foamy threats at you. It's possible that the ring may bounce into the lava pit behind him. What do you do?

Frodo – I'll grab the ring before it bounces away.

– That won't be easy, will you commit?

Sure! I'll just dive towards the ring and grab it before it touches down.

– OK. Why not. You can do so, along with Gollum noticing your move and jumping in to stop you. You'd end up face down with the ring in your hands and a mad Gollum on your back. For @20 of course.

Mad Gollum on back doesn't sound good. What if I kick him out of my way while diving for the ring? I'd be willing to give @35 for that.

– Considering your respective agilities and the fact that you're also trying to grab the ring, that sounds more like @90 to me. Deal?

Nope. Way too much. I think I can handle one Gollum-on-back if you let me have it for @15

The result need not be a black-and-white success/failure affair. Any kind of outcome with its partial levels of success and related positive and negative side effects can be requested or offered. A magnificent result should cost much. A mediocre success should cost something. An acceptable failure should cost a bit. A catastrophic failure should cost nothing.

– No deal, that one stands at @20. For @15 let's say that Gollum notices and grabs The Ring™ but also you take the dive and grab his wrists.

– Good idea, but let's see… @25 and we have the action. Yes?

After the player commits to an action, aborting the action becomes a negotiable outcome. Its @ cost is up to the GM, who is advised to determine the cost according to how 'abortable' the said action is.

This isn't going to work… Maybe I shouldn't dive for The Ring™ at all.

– Well, why not. For @10 you suppress the urge to spring into action, and the ring bounces away while Gollum tries to figure out the hesitant look in your eyes.

And yes, the player can decide to pay nothing, accepting failure at its worst or the GM's mercy.

@10 for nothing? That's unacceptable! I'll just give up! Zero!

– Sure, but remember that you committed to diving after The Ring™. Zero @ would mean you take the dive but trip over, sliding past Gollum and finding yourself hanging by one arm from the edge of the chasm behind him.

Of course it is good practice to get every little bit of leverage you can think of.

Come on! You know The Ring™ is calling to me. I have every motivation to get hold of it now.

– Ah, good point. I'll give you a @3 slack as long as you end up touching The Ring™ [evil smile]

Fine, then I leap and grab it, and Gollum grabs my wrist. @22, right?

### How do skills work?

The skills of the character provide automatic discounts to the amount of @ agreed on. A skilled character gets to spend less @ on the same result than an unskilled character. Higher levels in a skill provide higher discounts.

– Yes, @22 normally, unless you have levels in Juggling.

Will acrobatics at level 1 help?

– Sure, L1 would give you a 10% discount. That makes this @20.
You spring towards Gollum, grabbing the falling ring. He tries to do the same but reacts a split-second too late, managing to get a strong grip on your left wrist instead.

### Opposed actions

When two characters oppose each other, it is an auction about whose result stands. Player A announces his intended result and amount of @ he's willing to pay for that. Player B can either fold and accept A's result, or she can raise the bid, and announce her own result. This goes on until somebody folds. The winner gets her way, but she has to cough up the @.

By the way, there is no reason for the auction to involve just two sides. Any number of actors may be allowed to bid on the outcome, pending the GM's approval or invitation.

### Combined actions

When two or more characters are working towards the same goal and able to help each other, they can share the @ cost of their intended result. It is up to the players to decide how they share the costs. The GM can also make NPC's offer their help and @ contribution.

### We the people…

It is not only characters that have and spend @. Other entities like teams, military units, guilds or even whole nations may have @ pools that represent their collective will to be applied towards their goals. Bigger entities will have bigger @ pools unless they depleted them. A nation may have millions to billions of @ at its disposal, to be spent when the action represents the interests of the whole nation(and not the state).

Players may have access to such pools when they are somehow in a position to influence the actions of the entity. A ship captain directly uses his ship's @ pool when running the ship, representing the collective will and skill of his crew. A rockstar will be able to tap into the @ pool of her audience, directing them towards a common goal, but that will be only when she is able to step into the spotlight.

This makes it easy to abstract any scale of conflict, the skirmish between border patrols and smugglers, the competition of two rival companies, or the global cold war between alliances of nations, into a single, even if elaborate narrative.

### Death, or maybe not…

There's an interesting consequence of this system. Since death of a character is either the ultimate defeat or extremely hard to recover from in most settings, PC or NPC's would be willing to bid all their @ to avoid death. This makes killing people an expensive affair at best, since you would have to spend more @ than that character can muster(maybe including the extra @ his allies may contribute to save him). It also means that going on a killing spree would deplete your @ reserves rather quickly, leaving you defenseless against all kinds of attack and exploitation. Most of the time, it will be more effective to get your target to agree on a lesser form of defeat. He will be willing to top your @10000 bid to behead him as long as he has it, but he may agree to be disarmed by you instead of having to top your @100 bid.

### Earning @

Just as awarding XP, the GM should award @ for advancing the story, solving problems, overcoming challenges, defeating foes, acting in-character and good roleplaying.

## Optional rules

These rules are not required for this system to work, but can make it more interesting.

Instead of accumulating XP, the players use @ for advancing their characters. Some of the accumulated @ can be used to buy new skill levels and perks. The players are advised to spend wisely though. A smart investment of the character's @ in a frequently used skill may mean a lot of @ saved during play. A bad investment is just precious @ wasted on a useless improvement, and could have been used to succeed in-game.

### Multiple currencies

In your game setting, there may be more than one type of @-like currencies. A different currency may be required for a different type of action. Using combinations of different amounts of different currencies for one action may be allowable, as long as they make sense in your world.

A few examples of multiple-currency setups:

• Life + Magic + Prestige
• Good karma + Bad karma
• Fire + Earth + Water + Air

### Accepting fate

Call it GM special; during bargaining the GM may offer a secret outcome or fate at a relatively low @ cost. It may be one of the outcomes discussed earlier or something new entirely. The GM writes this down on a card and offers it face down. It may be any kind of result, ranging from spectacular success to abysmal failure, and include side effects. The player has the option of accepting this offer and he doesn't find out about the outcome until he agrees. The GM is welcome to entice, trick or cajole the player into accepting fate, and can lie about what's written in there.

And the only way for the player to find out about fate is to accept it.

None of this has been playtested yet. If any of you are willing to give it a try, please let me know about how it holds up. Any comments and suggestions are more than welcome.

I also intend to add imaginary in-play examples about how this system works, but it's late now, hopefully I'll be able to edit this later to include some examples.

-
Sounds like the bartering will take up more time than the dice rolling and thus you might run into some of the same problems, except perhaps failure rolls... Sort of interesting system but it relates to the question enough, which seems to be more about a pro/con for/against rolling dice or experience with using such a system, while yours seems to be hypothetical so far, yes? Good luck with it anyways. – Julix Feb 20 '14 at 19:43
Thanks @Julix. I assumed that the need for eliminating dice came from the idea of reducing roll-playing in favor of role playing and storytelling. I had this system in mind for a while and I just used the question as an excuse to write it down. Wondering if anyone tried this or something similar after 3+ years :) – edgerunner Feb 20 '14 at 23:15
well you're replacing roll-playing but not with role playing but instead with out-of-character bartering with the DM. The prices for actions would be very subjective and depending on the people (and the stakes) involved that could lead to conflict when players disagree with the "price tags". Interesting idea though, but I doubt it would lead to more/better roleplay. Let us know if you try it :-P – Julix Feb 22 '14 at 0:52

Amber Roleplaying Game is the classic diceless RPG. You should try getting a copy and looking it over. At the bottom of the wikipedia article there are a lot of good links to follow.

-

I'd take a step back and ask why you're using dice in the first place. Dice can help determine random results quickly (most commonly in crunch-time, ie combats), but too many DMs rely on them to determine the course of the communal story. Very often a Fudge die (-1/0/+1) or equivalent (on d6) can determine reactions, choice, and the like; no need to 'go deep' by quantifying everything into percentages (for example). And forget not the "rule of cool"... if a player comes up with a good contribution to the group Fun, tilt the results to encourage same in the future!

-

Are dice necessary? It depends on what you're trying to achieve.

There seems to be an assumption by many that narrative and dice rolling exist on a sliding scale. The more dice are used, the less narrative the game becomes, and vice-versa. I don't think that is necessarily the case.

When deciding what kind of game to play, part of the equation to consider is what the players want. When a game starts, players are going to have expectations regarding how the game is executed. Some players are more concerned about the story and are comfortable not using dice. For them, minimizing dice rolls or using an alternative non-diced system is fine. Usually, however, it seems that a large portion of the fun for a player is the act of throwing the dice and discovering the results. There's a bit of an adrenaline rush there as they plunge into the uncontrolled unknown.

The d20 system has an alternative rule that allows for players to roll all of the dice. I've started using that with considerable success. Meanwhile, they don't know their target numbers for the rolls, which enables me to fudge a bit if necessary (though I recommend doing that sparingly).

Putting all the rolls in the players hands effectively does two things for me. First, it maximizes their fun with the dice. They love the clatter, and love it even more when they get to make it happen. Second, it frees up my time that I would normally have to spend on rolling and calculating, which I can spend on planning narration. As a result, the actual use of dice has enhanced the opportunity for narrative play, rather than lessening it. This has definitely turned out to be a win-win for my campaign.

-
I like the "players roll all" idea. It also gives the players a better (if only illusory) feeling of keeping the GM accountable, since they can see the actual roll results. The DM still retains some exclusive knowledge as you've mentioned, so he still has the ability to fudge the results a bit. However, this ability is somewhat more limited in cases of extreme roll results such as a natural 1 or critical hit. – Iszi Aug 24 '10 at 20:22

Think, don't roll (quote from http://arsludi.lamemage.com/)

Use dice for two things:

1. When the players have sufficiently explained their actions and intentions, and you (the GM) have determined an appropriate chance for success. Only do this when there is a consequence for failure.
2. To entertain yourself (the GM) and keep the game and story exciting (e.g. wandering monsters)
-

I'm currently involved in a Labyrinth Lord game that rarely involves dice. Sure, we have them if we need them, but we're more likely to play out situations, especially social ones, without turning to the dice. This works pretty well for us, though it can put a lot more on the shoulders of your GM to keep things consistent and fair.

Because this game is played online via text-chat software, and because text-chat is notoriously slow, I also set to streamlining the game in advance. This mostly meant simplifying combat, especially issues like initiative. Removing dice-rolls means less back-and-forth with me asking players for rolls, the players making the rolls, and me typing up the results. Again, we all seem pretty happy with the results.

In the end, RPGs are games, and games are about making choices. Dice inject a random element of chance that can be fun, but it's the choices the players make that are really central to play. Otherwise, we'd all just roll dice and consult random tables of results and events. ;p

-

Just hitting a few highlights of the various approaches I've seen CORPS - reduced dice; if difficulty less than skill, no roll needed. If difficulty greater than skill, roll vs 11-(2x difference) or less on 1D10.

FUDGE - autosuccess option: some implementations allow not rolling if skill exceeds difficulty.

Vampire The Masqerade 1E - autosuccess: No roll needed to take 1 success if pool exceeds TN. If 2+ successes needed, roll required. (Note: all later Storyteller games I've seen to use a fixed TN; 1E used a difficulty number as the TN for each die to count as a success, ranging from 2 (trivial) to 10 (insanely hard).

Theatrix - diceless: skill vs difficulty comparison, providing start point on decision tree.

Better Games' "Free Style Role Play" system - dice reduction: every character level, a number of "make-rolls" gained; spending a make roll grants a success result without rolling dice.

several diceless games: if skill greater than difficulty, succeed; otherwise fail. Most include a point pool for pushing skill up on specific actions.

Marvel Saga (TSR) - Cards: Cards in opposed action vs NPC, or vs fixed TN by difficulty. In-suit card also adds top card from deck; cards add to stat.

DL5A - Cards: Card + Stat vs NPC stat or vs fixed difficulty; in suit gets extra card from top of deck.

Castle Falkenstein - Cards: Card + stat vs TN or vs Card + Stat. Out of suit counts 1; in suit counts face value.

Marvel Universe RPG (Marvel) - point pools: spend points to succeed. Pools refresh at a known rate.

Unisystem Light (Eden Studios: BTVS, Army of Darkness, Angel, etc.) - slightly reduced dice: all "opposed rolls" are made with only PC rolls; opposition always gets a 6. Also adds reduced bookkeeping by allowing short form NPC's with 3 stats instead of 6, and no skills.

D&D 3.X and D20___ - Reduced Dice option: Take 10 and Take 20. Unstressed situations, may take a 10 result instead of a 1d20 roll. If 20x time available and no penalty for failure, may take 20. If expenditure for failure and 20x time available, may pay 20x and take 20.

Star Trek RPG (FASA) - standard success model precludes some rolls: certain types of action automatic if skill high enough; more difficult actions require normal rolls. (For example, hailing a ship requires a communications skill of 10+, or a roll by difficulty if skill 1-9).

Burning Wheel (and related) - autosuccess, reroll prevention: "Just Say Yes" - if failure not interesting, GM instructed to just say yes.
"Let it Ride" - no rerolls unless circumstances become notably harder. EG: I roll 5 success on stealth. Until I hit an area and/or situation where the difficulty would be 6+, or I get injured, or I quit being stealthy, that 5 still counts and neither the GM nor I may call for it to be rerolled.

For me, I don't like randomless games; Theatrix and Amber both left me cold, and Marvel Universe was just not interesting. Cards have worked well for me, but not for my players, and the requirements of sitting at table are antithetical to my normal gaming mode, but that isn't to say I object to doing so; I just tend to game in the living room on couches around the room's perimeter.

The reduced dice approaches have been well received by most of my players. A few, however, object to them.

-

It depends on the system you are playing. Diceless games are popular for sure, but if you run a game where the player ends up with more than just six stats, or where more than half of the system gives the player mechanical bonuses to increase the result of their dice rolls, the players may feel cheated in the end.

The mechanics would have to support no dice in order to run the game without them. If a game is dice heavy, then taking that away from the players is likely to be a disappointment.

I've played in a D&D game where the DM does not allow any dice to be rolled by the players, and I can see the experience as being deeply immersible, but it requires heavy DM-player trust. Personally it wasn't very fun... but then again I love rolling dice.

-
I agree with the love of dice-rolling, and need for DM/Player trust in a diceless or "DM rolls all" setting - especially the latter. I personally would feel very cheated, or at least cheat-vulnerable if the DM made all the rolls. As it is, I don't like it when a DM won't tell me the roll result of an attack against me - even if it's behind a screen and he could be fudging some. – Iszi Aug 24 '10 at 20:26

Generally speaking, I've drifted towards games that reduce die rolls to when the outcomes are truly important. As a GM I've taken to ask the question, is the result of this dice roll, whether it succeeds or fails, going to move the story in an interesting direction? If the answer is yes, I go to the dice. If the answer is no, I go with the interesting result.

-

I would say it depends what you want from the game. I like the "random factor" of dice. Sometimes dice come up with results I would have never picked (thought of). On the flip side, I think it's good for the GM to know when to roll the dice and when not to. If you really want someone to climb over the castle wall, because the adventure lies beyond the castle wall, having them make a roll that can result in them not getting to the adventure may not be a good idea.

-

In general, if a game is designed well, you should play it more or less as written. If the game relies on dice, roll dice.

Most games can be improved by rolling dice only when it matters. Really, if it doesn't really matter to anyone at the table, why are you leaving it to chance? Go with whatever the player wants. Paraphrasing Vincent Baker's Dogs in the Vineyard, "say yes, or roll the dice."

Another way to reduce unnecessary dice rolling is to roll only once for something until the circumstances change dramatically. The Burning Wheel RPG has a "let it ride" rule in this vein. Climbing a mountain? Roll your climbing skill once and apply that result to all checks that come up while climbing, even if the "difficulty" changes.

You should not remove die rolls and just replace them with GM fiat unless everyone is really on board with that. These games devolve into "the GM's story" and the players are just spectators instead of active, creative participants. If the player characters are to be the protagonists of this shared story, then the players need their choices to matter. Often, the dice rolls are the main point of contact with the system for players, and removing them removes a vital tool for player agency.

I've applied all of these techniques to my D&D 4E gaming with great results. Don't call for rolls when they don't matter; just let the players succeed until the situation is important enough to call for rolls. Don't make a player roll over and over for the same thing (especially not just to make sure they fail). Don't use dice to railroad the plot. Don't use the removal of dice to railroad the plot, either.

-

One of the fathers of diceless role-playing (along with Amber) is Theatrix. It, in particular, had a pretty good flow-chart system for determining success or failure (is it critical that the players succeed in the task, is the task well within the character's area of expertise, would failure increase the drama, etc.). While I've never actually run a Theatrix game, the techniques it suggested have informed every game I've run since then. It's well worth a read if you can find a copy...

-
Theatrix has an appendix on how to use dice with it... for those who want to let the dice decide. Uses d100. – aramis Oct 11 '10 at 23:06

I myself use dice as a responsibility relief only. I never interpret results to players nor tell them numbers. Dice only decide when I don't want to.

If a character does something risky or something that can fail - I throw 'em. But I've minimised my mechanisc so that I don't need no time to interpret the results.

During a fight I often roll 5 dice and just take into account if the scores were good or bad. That's enough to build a good and dynamic fight scene upon. The way we [me and my friends] played required no sitting and counting every swing of a sword.

-

The RPG Fiasco doesn't use dice in the usual way--that is, for resolving what happens in play. Instead, play moves forward in "scenes" for each character, and the player of that character chooses either to set how the scene starts or to determine how the scene turns out (but not both). The other players decide the things you choose not to control. This is a game every gamer should try once.

-

There are games that have many different conflict resolution systems--Rock Paper Scissors (LARPs), card based (Prime Time Adventures), and so on. The usual point is that you want some way to be able to resolve a conflict when it comes up--whether it be a battle, or something as simple as 'can the character do X'. If there's agreement over everything in the RPG, then there's no need for the dice, potentially.

I wouldn't count out random elements altogether, though. It's a lot more fun, potentially, to let random chance come in and shake things up once in a while. It's not always that exciting, even as the DM, to always know and control what happens next.

-

No way are dice necessary. It's possible to have a dramatic, fun system without them.

Amber is a fairly popular system, and I've heard good things about Nobilis and Active Exploits. Many people seem to prefer systems without the randomness, where more planned and intellectual play may be possible.

To answer your other question, no, dice should not be rolled behind the GM's screen. If you're rolling dice, your players need to know you're not fudging the results to "save" them. Only roll dice when there's an actual possibility of failure, or you're ready to accept the consequences. "Overriding" the dice shows your players that you're not playing by the rules of the game, and is no different than cheating, no matter your purpose.

-
That's a bit subjective for absolutes, isn't it? Many people find rolling behind a screen increases tension and drama. I've especially found that when running games for new players, nothing turns them off exploring gaming further quite like a TPK. The objective there is for them to have fun and continue playing. After that point, of course, it then is up to the DM's discretion, but it's not as simple as black and white. – Logan MacRae Aug 20 '10 at 15:10
I don't feel it's particularly absolute. If you're concerned about the random factor resulting in a TPK, then you shouldn't be rolling dice to begin with. If you're going to use the dice as a randomization factor, then either go by the result or don't use them at all. That's more honest than saving the party by cheating, after all. The only real reason to roll dice behind the screen is for things the PCs aren't meant to know- perhaps the result of some more sinister purpose such as traps or secrets. – LawfulIndifferent Aug 26 '10 at 22:59

No, dice (or other randomizers) are not necessary. Not even for systems built around them. But, having sources of half-way decent random numbers can be a help, for when the GM goes "Uh, OK... So, what would happen?", even if the game is not supposed to use dice (or cards or Jenga towers or, or).

-
It would be helpful if whoever downvoted this could add a comment as to why they did it, so I can provide better answers in the future. – Vatine Sep 26 '11 at 14:16
I'm not downvoting, but elaborating on why having a source of random numbers can be a help might garner an upvote. – DCShannon Nov 1 '14 at 0:36

One very interesting idea I've seen a few places, mainly Cinematic Unisystem and Legendary Lives, is semi-diceless gaming: only the players roll dice. Instead of monster attack rolls we have player defense rolls for example.

There is a document on using it with D&D3.x at the same place at the E6 document.

-

A novel system was described in Weather the Cuckoo Likes supplement for Other The Edge. It is based off the William Burroughs cut up idea: you take lots of words, cut ups from new papers. Your skill gives you time (slices of 30 seconds) and the difficulty is the number of words pulled. You then have to describe your action using as many of the words as possible. Success is based on whether the action makes sense (as voted by everyone), the number of words used, and of course, you then have a role played description for your action.

-

There are some interesting answers here, but apparently the question got a major re-write, because almost none of them are actually on topic now.

The Angry DM has a lot of great advice, and part of that advice covers adjudicating actions. Specifically, step 3 of this process has these steps (which are elaborated upon in the linked article):

3a. Determine Whether the Action is Possible
3b. Determine if the Outcome Needs to be Randomly Determined
3b1: Determine How to Randomly Determine the Outcome
3b2. Make a Die Roll and 3b3. Determine the Outcome of the Die Roll
3c. Determine the Outcome (and Consequences)
3d. Describe the Results

3b is what you're looking for. Cut out dice rolls if the outcome doesn't need to be randomly determined. The process of making this decision is covered in more depth in the article 5 Simple Rules for Dating My Teenaged Skill System. The short, sweet excerpt from step 3b is as follows:

If the action can’t actually fail or failure is extremely unlikely, THE ACTION SUCCEEDS. Huzzah! Skip to 3c.

If the action can fail, but failure carries no risk or cost and the player can freely try again and again, THE ACTION SUCCEEDS. Huzzah! Skip to 3c.

If the action can fail and failure somehow changes the characters’ situation, you need to use the dice baby. Keep going.

I use this method in my game, and it cuts out at least half of the dice rolls the rules would ask me to make. This speeds up the pace of the game, and reduces die rolls where we all know the outcome anyway.

-