Future Imperfect: Appendix - 1, Action Cards

The action cards contained in Future Imperfect look very different from cards used in other games. To many, their layout might not seem intuitive. Information is densely packed, yet it is also distributed consistently to make referencing them simple. Introductory instructions on their use can be found in Future Imperfect - Action Cards. This appendix covers the minute details of the layout, as well the how and why of the entries on the cards themselves.

Card Map

How do I know what the numbers mean? What is the difference between the two grids? Why is there a clock face? The definitions that follow map the entries and can be used as a reference when other rulebook sections refer to results.

The Four Corners

Every action card has a number in 3 corners, and either yes or no in the other. These entries can be used for various situations, but their meanings describe the origin of the action cards themselves. They are not part of standard task resolution. None of the corner results are randomly distributed. In other words, as will be come clear later, knowing the number in the upper left will also allow the other corner results to be derived, while knowing the upper right and lower left will allow the upper left to be derived.

Upper Left

In the upper left corner is a number from 0 to 53, referred to as impulse. The action card deck is 54 cards, based on a standard poker deck with jokers included. Each card has a "value" (more on this later), and that is what determines the number in this location. The higher the number, the higher the value. Numbers in this location are unique.

Upper Right

The number in the upper right is the first clue to how the value of a card is determined. These numbers range from 0-14 and represent the rank of the card, with the jokers being 0 and 14. The rank is in order based on poker rules, with the deuce being 1 (the lowest) and ace being 13 (the highest). There are four each of the 1-13, and one each of 0 and 14 (the jokers). These numbers are called steps.

Lower Left

The lower left represents the suit of the card, in alphabetical order: clubs (1), diamonds (2), hearts (3), spades (4). The jokers are assigned based on color, with the red joker being zero, and the black being 5. These numbers are called fragments. There are 13 of each fragment from 1-4, and one each of 0 and 5 (again, the jokers).

Lower right

The lower right result is a yes or no value representing color. If no, then red, if yes, then black. These results are toggle. There are 27 of each toggle.

Origin

The value ranking for cards was initially developed as an initiative system. It still can be used as such, and it also works well for determining random timings (discussed in a later section). The names given to the corner results hint at the timing system that was originally developed. Eventually, the current initiative system evolved, and this usage was abandoned.

The unique number in the upper left provides a simple way to reference cards. Both the step and fragment are useful for non-homogeneous distributions that are mostly normalized but can also produce extraordinary results. These usages are detailed in various sections of the rules.

Random Event Timing

The original initiative system was developed using poker style card ranks to order actions. Because of this, the ranks are abstracted onto the card corners in a way that makes them easy to read for this purpose. The current initiative system is not related to the poker ranks in any way, however, it fortunately does have four action speeds, just like cards have four possible suits. Therefore it is simple to use entries already on the card to determine randomly when an event will occur.

To determine a random timing, cut the deck and consult the step and fragment. The fragment is the phase of the turn where the event occurs. The step, rounded to the nearest die type, is the reflex value. Fragments are rounded down, so 1-5 corresponds to d4, 6-7 to d6, etc. If a 0 is drawn, the event happens just before the round, if it is a 5/14, it happens just after.

For example, the Master determines that an unstable power cell will explode in approximately 10 seconds. He determines that in this case, each action round is about five seconds, so the explosion will occur sometime during the third round. He cuts the deck to secretly determine when that will be (to preserve tension to the Crew).

The card chosen is 20. Consulting the lower left hand corner (fragment), the explosion will occur in the 4th phase of the turn. The listed step is 5, so it occurs on d4. The event will occur near the end of round 3.

Random event timing is optional. It is provided here as an illustration of how the given tools can be used in creative ways.

The Cause Section

An action card is divided into three sections, the outer edge, where the 4 corner results (discussed above) are located, the cause section (the upper half of the box that separates the outer edge from the rest of the card) and the effect section (the lower half of the box). An example of a cause section is located below.

Note: Currently the hit locations and burst results are in the "cause" section. These are more correctly defined as effects, and in a future card design iteration will be located with the effects, while the armor penetration, equipment failure and scatter will take their place here. For simplicity I am describing the cards as they are currently arranged.

The Upper Grid

Most of the cause section is a 5x5 grid (called the cause grid). This grid is used for all skill checks, to determine the pass/fail result. Because these results determine whether the action is successful, they are referred to as causes.

The grid is arranged with column headers shaped like polyhedral dice (along with a number representing the number of sides on the given die type), and row labels from 1-5. The row labels are in a distinct font to differentiate them from the grid results. Each cell in the grid is one die of the given type. When speaking of a single given cell, the nomenclature is row-column, such as 3-8, for the d8 cell that has an 8 result (third row, above). When speaking of a result set, where multiple dice of a given type are rolled, the standard RPG convention of [row]d[column] is used, so for 3d8, the results of 5, 3 and 8 would be included (again, using the above image).

Cell Values

The values in each cell are results typical of that given die type. Therefore, the results in the d6 column tend to fall between 1 and 6, just like a standard six sided die.

Hey, wait...how come there is a 7 in 2-6, and an 8 in 4-4?

The die results are open ended, sometimes called "exploding". Many games use the concept of exploding dice, and there are multiple ways to implement them. The way the action card set implements open ended results combines simplicity and mathematical accuracy.

The concept of exploding dice is based on the idea that when the highest possible result on a given die type is generated, more dice can be thrown to increase the result. This makes it possible for dice of all sizes to be used, and for a d4 to achieve a success result even on a TN (target number) of 5 or more. On its face, it is a straight-forward idea.

However, when using dice for open ended results, the designer needs to choose between incomplete data sets and functional simplicity. Because the die is re-rolled when the highest possible result is generated, it becomes impossible to generate a value equal to any multiple of the value of the die face (6 for d6, for example) unless a guard condition determines whether or not the die is re-rolled. Both choices are sub-optimal outcomes.

This does not mean it is impossible or unplayable. For example, in Warhammer Fantasy Roleplay 1st edition, damage was rolled on a d6. If a 6 was generated, the player rolled their attack again, and if it hits, the d6 may be rolled again and added to the previous result. If another six is rolled, it is rolled again and added recursively (no second instance of the guard condition is necessary). This makes it possible to generate a 6 in damage (but not a 12 or 18, etc), at the cost of some convoluted rules and an extraneous die roll. This is an example of both a guard condition and an incomplete data set. In this case, the data set is nearly always complete (assuming a 40% skill level the set is complete almost 99% of the time). Warhammer Fantasy Roleplay 2nd edition gave this operation a name, Ulrics Fury, and changed the die to a d10, making the data set complete even more often.

But still, more time is spent rolling, adding, and generally doing things that should not be necessary. Surely there is a better choice.

Another possibility is that when a die explodes, the new die is added, minus 1. This seems pretty simple at first. Now, any number can be generated and other than a simple subtraction, no more operations are necessary. Except that when that die explodes again, now 2 must be subtracted. This is just as bad as the previous possibility.

How do the action cards handle this? First of all, I calculated the formula to simulate the average values generated on open ended dice, in this case n/2 + n/(n-1), where n is the number of sides on the given die. Then I use a Monte Carlo approximation with the characteristic given average for each die type and 54 members in the set (one per each card). So that makes the card deck perform exactly like exploding dice, in a single step.

However, you may notice that some cells in the card section pictured above are bold, italicized and highlighted yellow. What is up with that?

With a set of 54, a good approximation of dice can be made, but the upper bound is not nearly as high as might be seen over years of play with dice. So each cell has 5 results (about 9%) that are marked as previously mentioned. These are called exceptional results. When an exceptional result is generated, another card is drawn and added. This continues recursively until a normal result is generated. All results are added.

This creates a substantially higher upper bound, much closer to what gamers might see in the most extraordinary circumstances. It required the numerical characteristics to be adjusted a bit to preserve the average (so the cells still perform like the given die type), but it all happens seamlessly from the perspective of a player.

Look again at the image above. In the 2-6 cell, there is a 7 but it is not an exceptional result. This is because in 54 rolls, with only 9% being exceptional results, some "explosions" occur that are not in the top 9% of results. The set takes this into account and provides the necessary result.

Notes Regarding Exceptional Results

The reason exceptional results are bold, italicized and highlighted is to make them simple to identify, even if the user has vision impairment or difficulty with color. Since no other kinds of results need to be notated, it is feasible to use all three signifiers together for increased ease of use.

Exceptional results are optional. They can be used when the Crew deems necessary.

Critical Failures

The action cards also support a critical failure system. Critical failures are denoted by a bold CF in the given cell (as in 5-10 above). Each cell has one CF in the deck. Rules for critical failures are found in the using action cards chapter, as well as the tasks chapter.

Single Cell Versus Multiple Cells

You may be asking yourself why a multiple die result, such as 3d8, requires checking three separate cells rather than just displaying the result in a single cell. Should it not be possible to do the sampling automatically, so that the player need only to reference one cell and take the value displayed?

Yes!

In fact, the initial iterations of the action card deck did exactly this. The reason why that design was abandoned is that while elegant, it was thoroughly disliked by players. Understanding why requires some background regarding the math that powers the cards.

Each cell represents one entry in a 54 member set that approximates the given type of die. In the current iteration, all of the sets in a column are identical across the deck. In other words, if you took all 54 entries from 1-6 across the deck, and compared them to all 54 entries from 2-6, you would see the exact same distribution. This makes it simple to randomize the cells, because they are completely discrete. It doesnt matter at all if the 2-6 cell reads 7, and the 3-6 cell reads 3, as in the diagram above. The player can choose either.

However, if the player must use the designated cell only, then if a higher value exists in a cell represented by fewer dice, then players feel like they have been cheated. It is not true. Why? Unlike the previous example, if you took the 54 entries from 1-6 and compared them to the 54 entries of 2-6, the values are not equally distributed. The 2-6 results are higher on average (as they should be). But due to the nature of randomness, sometimes the result from more dice is lower (such as when one die rolls 6, while two dice roll 3 and 4), and though playtesters could understand this intellectually, it still led to a diminished gaming experience. This causes some design challenges.

In order to use the single cell model, one of the following choices must be made: either players must accept that sometimes a higher number will appear in a set of fewer dice, or the results must be engineered across the cards to ensure that this never occurs, rather than randomly distributing them. As a designer, I felt like the first choice was perfectly acceptable, but other players disagreed. The second choice is completely unacceptable. The purpose is to achieve randomness, engineering the results is the antithesis of that.

By using a single cell equals single die model, the issue is resolved in a simple way, but it does require a player choice instead of being completely automated.

Alright, I get it. Each cell is a single die. But if players are always choosing the die they want, the highest, then why can the results not be randomized, and then have the cells "corrected" to show the higher of either the number in the given cell, or any cell below? So, for example, in the image above, the 2-6 result remains 7, while the 3-6, 4-6, and 5-6 results are updated to reflect 7 as well. Problem solved!

On the surface this seems like a workable solution. In effect you are randomizing the sample and then updating the value so that the player need not consult other cells. This presents numerous issues. The first of which is that by manipulating the values in the cells, the average is no longer preserved across the set, and the cards no longer perform like the dice they are meant to emulate. If the cards are only used one way, and there is only one possible card set, this might be acceptable, because, as illustrated, the given number will be chosen anyway.

But the cards are used in multiple ways. The simplest example is exceptional results. These are based on a single cell. When drawing a second card for an exceptional result, the value added is the one that corresponds exactly with the cell that produced the draw. Lower results are not allowed to be chosen (see the section on exceptional results in the using action cards chapter for a full explanation).

Another issue is that of perception. Take the example of the d6 column referenced in the conjecture. What was once a set of varied numbers becomes a block of 7s. This makes the card look superior to other cards. In one sense, it is (for skill checks). But it seems to be a lot less random than it actually is, and the entire reason all of this extra work was being done was to quell the perception of players.

The cards are displayed as they are to preserve the random characteristics they emulate. Additionally, they contribute to player confidence in them by virtue of the design choices that created them.

Hit Locations

To the left of the cause grid are four rectangles, each with a hit location written inside. The second one from the bottom uses a red font, and the corners are rounded. The other three rectangles use a black font with sharp corners. When performing an action that requires a hit location, these rectangles are used. The standard hit location is indicated in red font.

Why are there four locations, if only one is used?

The other three rectangles have two uses. One is with multiple hits due to automatic fire (or other situations as determined by the Master). The next section discusses burst fire, consult the combat section of the rules for more information about determining hit locations for bursts. The other usage involves optional advanced rules.

If your Crew is utilizing the optional bump rules, the other three locations may be chosen. Bumps allow highly skilled characters to use their skill to change hit locations (among other things). See the section on bumps for more information.

As with the exceptional results, the rounded corners and font color combine with a static arrangement to make it easy to see which result is the standard hit location.

Burst Fire

Beneath the hit locations are three paired entries. The left number, 3, 5, or 10, is the size of the burst. The right number, which is inside a burst icon, is how many of the shots hit the target. The three burst values are grouped. In other words, it is not possible to have a lower number of hits in the 5 (or 10) result than in the 3.

The difference between the size of the burst and the number of hits is called stray shots. In the case of a single shot, if it misses that is also considered a stray shot. Stray shots use the unintended targets section, on the right outer edge. See the combat section for rules regarding unintended targets.

Each shot that hits will have a separate hit location. See the combat section of the rules for more info on arranging multiple hits due to burst fire.

The Effect Section

The effect section is in the lower portion of the inner box of an action card. There are four possible results to be gleaned from the effect section: damage/victories, armor penetration, equipment failure and scatter. An example effect section is located below.

The Lower Grid

The lower (effect) grid is on the left of the effect section. The grid is 5x7, and resembles the upper grid in that the columns are headed by polyhedral dice symbols and the appropriate number. The rows are numbered, similar to the upper grid, except these numbers are surrounded by parentheses. This allows the effect grid to be differentiated from the cause grid via discrete nomenclature. The third cell in the d8 column of the effect grid, which has a 20 (in the card section pictured above), is referred to as (3)d8 (as opposed to 3-8 for the cause grid 8 result in the d8 column, or 3d8 for the 5, 3, and 8 results in the d8 column, both from the diagram in the cause section).

The effect grid is a sum of the given number and type of dice. Unlike the cause grid, the results in the effect grid are arranged such that a greater number of the same die type always leads to a greater result on the same card. This is much simpler to engineer than in the cause grid, because the results are summed rather than sampled. Also, it allows for game effects to increase the size of an effect pool without the possibility of reducing the effect magnitude.

Generating Victories and Wounds

The effect grid is used to generate victories (also damage). The standard magnitude necessary for a wound or victory is 6.

If victories and damage both increment on 6, why not just list total victories/wounds instead of requiring division?

There are multiple reasons behind this decision. The first is that when damage is dealt, any leftover magnitude is applied as concussion (see the combat section). There are also some weapon types that deal damage entirely in concussion.

Finally, however, by having the full result the size/victory threshold can be modified as necessary. For example, an extremely large adversary may have a size (the number required to achieve a wound, see the combat section) greater than 6. One character may have an advantage over another in a contest, which allows them to have a lower victory threshold. The possibilities are endless, and this design choice allows the flexibility for the Master to vary challenge as the story dictates.

Exceptional Results

Like the cause grid, exceptional results can be generated in the effect grid. Unlike the cause grid, however, exceptional results are not differentiated by cell. Because the members of each column (in other words, an entire die type) are grouped into a single set (which ensures the increasing magnitude as the number of dice increase), exceptional results apply to entire columns. The column is shaded red, and the font is bold and italicized.

Note: In the current card iteration, all results in the effect grid are either exceptional or standard on a card. In future card iterations exceptional results will be grouped by column instead of entire grid.

Note 2: Currently, there are 5 instances of exceptional results for each column in the effect section set, however this is significantly higher than necessary given the way the set is generated. Because the "die" results are summed, and have the appropriate characteristics, the small die explosions are automatically accounted for in a way that is more accurate here than with the sampled results in the cause grid. Future card iterations will reduce the number of exceptional results in the effect grid, likely to either one or two column instances per deck.

Armor Penetration

In the upper right of the lower portion of the inner rectangle of the action card is a shield with a numeral inside. This represents armor penetration. See the combat section of the rules for an explanation of how to use these results.

The range of values is +2 to -2. The results are arranged in a bell curve, with the majority of values being 0. Negative values are in red font, positive values (and zero) are in black.

Note: Currently, only the armor penetration/ablation rules utilize these results. As the game system is codified other systems may also use this section, and therefore the symbology and nomenclature may evolve.

Equipment Failure

Below the shield is a wrench icon. A solid wrench means no result. If a broken wrench appears, then a number will also be present. This number is compared to the reliability of the equipment being used (if any) to determine if breakage occurs. See the gear section for more information.

There are a total of 13 equipment failure results (besides no result). This means the least reliable type of gear can fail approximately 24% of the time. Only improvised and previously damaged gear falls into this category. The highest reliability gear fails significantly less than 1% of the time (see the gear chapter for a full explanation, this is one of the very rare cases where a second card draw is drawn).

Equipment failure rules are optional.

Scatter

The final result in the effect section is a clock face and number. The value is 1-12, with 4 of each odd result and 5 of each even result. See the combat section of the rules for how to use the scatter diagram.

Initiative

The initiative section is on the left outer edge of the card. It consists of a key and 4 squares, each subdivided into 4 squares. The key provides the location of each of the given action speeds.

The initiative bar provides the TN required to have an action in the given phase. See the initiative chapter of the rules for how to use the results.

Note: In the current iteration of the action cards, the initiative bar and key is identical on every card. In future versions, once the initiative system is well-tested, the bar itself may vary as well.

Unintended Targets

On the right outer edge of the cards is a set of figures representing potential unintended targets. Each card has 10 potential unintended targets, and each target has a 1 in 6 chance of being hit. See the unintended targets section of the combat chapter for rules on how to interpret the results.

Why Action Cards?

Most tabletop RPGs use dice. Often the mechanics around the dice may be novel, but in general, dice are well understood by the gaming public. Many other games use cards for their resolution mechanics, but few if any use cards as seemingly complex as those provided with Future Imperfect. From the perspective of a new gamer, considering Future Imperfect for his table, why do I want to learn to use these action cards? What value do they bring to the table that is not provided by more familiar mechanics and tools?

The origin of action cards starts with the origin of Future Imperfect itself. When I initially contacted the friends who were to become my co-designers of this game, I knew only a few things about what I wanted to achieve. One of those things was that I wanted to use "big handfuls of dice". Something about that appealed to me on an intuitive level, and it drove the design decisions.

My gaming group has had many players come and go over the years. The skill levels of these players varied widely, as did the individual desires (in a gaming sense) of the players at the table. Over time, it became clear that it is often difficult to satisfy players who are strongly simulationist while also catering to those who wish to focus almost exclusively on narrative. The narrative players are frustrated by the time spent on minutiae and gritty detail in the mechanics, while without it many simulationist players tend to lose focus.

The reason I liked big handfuls of dice is that I craved the flexibility that so many possibilities offer. Moreover, I wanted to "cure" some ills I perceived in the games available on the market. First of all, I find it frustrating and immersion breaking to see seven foot tall giant brutes hefting heavy machine guns jumping around, going first, and maneuvering perfectly in a room scarcely larger than an elevator. While this is an extreme example (a clear reductio ad absurdum), the point is that many types of players "game" the system by finding the best piece of equipment of a particular type, and never vary from it. Why should they? The mechanics punish them for doing so.

These sorts of things can be handled by a crafty GM, but why should it need to be? Why not develop mechanics that reward solutions based on the problem at hand, rather than a broad, general case.

Therein lies the problem. Solving "gaming" problems via mechanics seems counter-intuitive. The players who are most able to get around the problem are the ones the issue itself is attempting to address. Not only that, but narrative players, as previously mentioned, abhor extra mechanics...or do they?

The more I thought about it, the more I realized that the narrative focused players did not eschew mechanics, or even more mechanics, they disliked when mechanics forced them to break their immersion or took more time than they are willing to spend on the given detail. My task became clear: deliver diverse mechanical depth and detail without sacrificing speed and immersion.

What breaks immersion? Arguing about rules. Looking things up in books and tables. Rolling unnecessary dice, or spending more than a few seconds fiddling with rules that do not make the game, and story, move forward.

Back to the action cards. How do they prevent the immersion breaking faux pas?

1. The action cards "know" the rules. Once you understand how to use the cards, you almost never have to refer to a rulebook for anything. All that is needed are the cards and your character sheet.
2. Because of how the cards are organized, the order of operations is always clear, and rules, even rarely used ones, are never forgotten. You know that guy who always seems to forget to check if he is out of ammo? Action cards have you covered.
3. All of the results that are necessary for any action, no matter the complexity, are complete in a single card flip.

The goal of action cards, as initially envisioned, was: resolve anything in 30 seconds or less. This doesnt sound like such a big deal, most things are easily done in that time. In fact, most resolutions in most games are 5 seconds or less.

Same with cards! The value gained via action cards increases as the complexity of the action increases. Performing a simple skill check with dice or cards takes 3 seconds or so. However, performing an attack action involving to hit, location, armor penetration, potential equipment failure, damage, unintended targets...this can be a long series of rolls and multiple book/table/chart references. With action cards its still a single flip, and even with all that detail it is done in 30 seconds or less every time, with no errors.

Future Imperfect action cards do not make the game faster than most other games. They allow a mechanically deep game to be delivered at a comparable speed to other games, but with significantly increased detail and reduced extraneous activity.

The vision of action cards came to me, and the implementation consumed a good portion of time. Intuitively I knew that it could be done, but developing the design and card mechanics was an arduous process. However, that work only needed to be done once. There is a learning curve to Future Imperfect action cards, but it is not as steep as it seems at first glance.

Spend 15 minutes reading the action card introduction chapter, and try a few sample actions. The time you invest in learning to use this tool will be paid back with increased fluidity, detail and immersion at the table.