Monday 28 January 2019

Aerial Rules Revisited (Part 4) - Activation & Playtesting

As I have a bad habit of starting things I don't fini

...I thought I'd kinda "tie off"my thoughts on air combat with playtest results and comments on the movement/activation system

Adding on the bare minimum:
I played a 1v1/2v2 with two planes with 6"/12" subsonic/sprint stats. Each plane had only a 20mm gun firing 4"(90d arc) which rolled 3 dice (+1 dice if in the rear 180d arc of the opponent). Both planes were identical.

Planes could attempt to "dodge save" using their pilot skill (4+ for an average pilot). Planes were identical on both sides so no agility modifiers.

Each unsaved hit had a 1-2 no effect, 3-4 damage, 5-6 kill. A damaged plane was -2"subsonic/-4"sprint speed, and -1 to all rolls.  A second damage destroyed the plane.
A non-damaging hit "stuns" a pilot temporarily next turn but has no effect on the aircraft.

Basically a cheap and nasty 40K-ish hit-save-damage system. I wanted to concentrate on movement and activation and not the dice mechanics. Neither did I use any of my detection rules (as I wanted to concentrate on what I was testing).

Playtest verdict:
It was a bit too easy to transition between subsonic and sprint speeds. Perhaps a rule where you have to move at top subsonic speed (i.e. no turns or climbing) before transitioning to sprint speed next turn. Planes just seemed to spurt forward and brake too easily and dramatically, like something from a Robotech cartoon.

I wanted to restrict supersonic maneuver partly realism/G-forces and party to make it a trade off i.e. supersonic-limited maneuver, subsonic-very agile. However there's no reason a supersonic fighter can't climb OR dive OR turn - just not both climb+turn at once.  So I'd ease up on my original restrictions - especially if I had a "transition" phase like mentioned in the paragraph directly above.

The subsonic turns seem stilted... ...but perhaps I am just too wedded to oldschool/ hex based plane games

The subsonic turn just seemed weird; to pivot and move directly forward seemed.... unplane-y. However I may be merely carrying over my prejudices from old aerial games.  I could do a 90d-ttun-at-halfweay then 90d-turn-at-end  like the supersonic move. However I'l need to look at the pros and cons of changing this.

Weirdly, the mechanic I REALLY liked was my activation system, a kind of Bolt Action bastardization. I am not a fan of the Bolt Action activation normally, but my system worked well here. What I WANTED was a semi-chaotic card draw, with better pilots having more cards (thus getting more opportunities to act). However I decided stuffing around with cards was too slow (a key design goal was to speed up the painfully slow aerial rules genre).

So I decided to draw dice from a pot instead, a la Bolt Action.

This is what I did:


DICE VERSION - Modified Bolt Action
A dice is placed in a pot for each plane,  + 1 “high” wild dice + 1 ace dice per ace (each faction uses their own dice colours)

When your colour dice is pulled, pick a plane and roll a Crew Check.

(+) If the Crew Check succeeds, you can move and lock+shoot at a target (in any order)
(-) If the Crew Check fails, you may either (a) make any move, but forgo locking/firing; or (b) lock and fire but move directly straight ahead (at max subsonic speed).

Wingman: you may interrupt the turn sequence and move directly after your wingman if you are not damaged or stunned. Simply remove a dice from the pot and roll it like normal.

High & Fast: if you are the highest (or if tied, the fastest) when the high & fast wild dice is rolled, you can act on it. Also remove one of your dice from the pot and roll both it AND your wild dice – and use the best roll.

Stunned/Damaged: If you are stunned or damaged, -1 from your roll.

TL:DR Basically, you pull one of your dice out of the pot, you roll it. If it's not yours, give it to your opponent and he rolls it. If it beats your crew skill level (4+ for an average pilot) you get to maneuver (turn/climb/dive) AND shoot.

If it fails you only get to choose between moving OR shooting. I.e. move straight ahead, but shoot...   ...or make any maneuver, but miss out on shooting.

This worked really well. It added uncertainty (whose dice would be pulled out, and would you get to make 1 or 2 actions) and also choice (which plane to move, and choosing move vs shoot if you rolled badly).

For a quickly cobbled together mechanic, I really liked it and I can seem myself using it again, perhaps in my helicopter or fighter sub homebrew games.

Anyway, aircraft will be left sitting on the bench for a bit as my dieselpunk tanks have arrived... but I'll get back to this topic, I swear!

Wednesday 9 January 2019

Game Design #77: The Dice Mechanics Aren't Important

I'm more and more convinced that the dice rolling mechanics are relatively unimportant as to whether a game is good or not.  Whether you use d6 or d10, whether you add or subtract modifiers, whether you use contested rolls or a fixed target number....     ...it's actually not vital to make your game good. There are other (more important) elements to consider.

The mechanics merely need to be simple, consistent, and (if possible) familiar.

I'm not trying to controversial, just trying to be helpful. I think many game designers overly focus on kewl tricksy dice mechanics where their time could be more usefully spent elsewhere. In fact, when making home rules, the dice rolling mechanics are the last things I consider; in quite a few, I've swapped dice types and mechanics after playtesting, usually to "speed things up" when I realise I could do it simpler, with zero end effect on gameplay.

Why Dice Mechanics Aren't Important
Dice mechanics do not define the style of game. They do not help make tactics more historical.  They do not make players play a particular way, or define the meta.  It's the percentages that matter (lethality) and the activation (who goes first); whether you are using a single d20 or buckets of d6, you can get similar end effects. Dice mechanics don't necessarily make players play a particular way.

Move:Shoot Ratios >>> Dice
Changing the ratio of movement distance to shoot distance can change your game vastly. The "normal" wargame has units move 4-6"and shoot 24-36". The ratio of move:shoot is usually 1:4 or so, favouring shooting. This is mostly (I believe) due to tradition and commonsense impacts of a normal 4x6 game table and small amounts of terrain. However, imagine a game where units moved 1" and fired 20"(sounds like a modern naval wargame).  Now imagine a game where units moved 20"and fired 1"(ancient skirmish/melee?).  The two games would play vastly differently.

Modifiers >>> Dice
The modifiers for your dice rolls are more influential than the dice mechanics and dice types used.
Let's say a game has 3+ (67%) to hit enemies. But if they are in -1 if in cover: they are only hit on a 4+ (50%).  But what if we changed the modifier to -3? They are only hit on a '6' (17%) which means that cover is so massively beneficial that I predict units would seldom move.

Table Setup >>> Dice
Even something as simple as table setup - making your game table devoid of cover vs buildings every 4"with no long sight lines will impact your game experience more than whether you are using a d6 or a d8. A game dev who agonises over which dice size to use but does not consider table setup or deployment rules has made poor use of his time. Even victory conditions (increasingly wargames have ways to win without "kill em all" or "scrum in the middle" can have a bigger impact.

Activation/Initiative >>> Dice
Longtime readers would know how much importance I place on activation and initiative; I was hating on IGOUGO long before it was fashionable. Activation determines the "flow" of the game; the "when" of your movement is just as important as the "where." Simply changing from IGOUGO to alternate activation will make vast changes to your gameplay flow, let alone reaction mechanics, action points/pools. I spend a lot of time on these in other game design posts so I will not rehash their importance here, though I recommend #68 and #69 on momentum and breaking up the turn.

It's the final result that matters: Lethality
At the core, it is the end percentage of success created by the die/dice rolled, rather than how you got there. I tend to look at lethality in combination with modifiers. If your percentages are simple it's actually quite easy to predict how your game will "pan out" before you even playtest.

 Now I'm not saying that the topics above are the only ones to consider; nor am saying what dice resolution you use is completely irrelevant. I'm just saying it should be a long way down your list of priorities.  

Best Practice: Lowering the Barrier to EntryBasically, dice mechanics should keep the skill floor low (i.e. the knowledge you need to be able to play) with very little knowledge needed. You should be able to pick up dice and chug them without much thought. Simplicity, consistency, familiarity are all good.

Simple (KISS)
Basically, as this means rolls are uncomplicated as you can get.  After all, dice rolling detracts from the actual "meat" of gameplay - the decision making. Unless you are using a dice pool or some sort of system where you "game" the dice, every minute spent on dice is a minute not spent on decision making or tactics. Computers can do this instantly, under the hood so to speak. But wargamers manually rolling dice take up a lot of game time.  If you are spending more time rolling dice than moving minis, then something has gone awry.

Ideally, I should be able to absent-mindedly chug the dice while thinking about my next move, just noting the results at the end. 

As an example of what not to do: I remember Silent Death had a dice system using d4s, d6s, d8s and d10s (even d12s and d20s). Each weapon had a different rule and even different combinations of dice.  I.e. "for a blaster, roll 2d6 + 1d8 and use the highest two dice for the "to hit"; then use the middle dice for damage." But a phaser might roll 2d10 + a d12, use the lowest two dice "to hit"and the highest two dice for damage.  They were so proud how they managed to combine the "to hit"and "to damage"into a single roll...  ...but didn't notice they'd actually made it more frickin complicated!
This is a classic example of how trying to be overly clever with dice mechanics actually made the game worse.


Consistency
I've used the example of Bag the Hun (and almost any TFL ruleset) vs say Warmachine. In BtH, the game uses seven completely different mechanics to resolve actions. That's incredibly messy and you need to remember both how (and when) you need to use a particular method. In Warmachine, you pretty much use 2d6 vs a target number in every situation.  Consistency means you only need to learn one set of dice mechanics.

Familiarity
There's a reason games like Bolt Action and Flames of War merely uses a thinly disguised version of WH40K's dice rolling mechanics. Or why 40K hasn't changed much over decades. Familiarity with a system lowers the barrier for entry - a player instantly can grasp the "feel" of the game and there is little new knowledge needed. You don't need to reinvent the wheel, and in most cases it doesn't benefit your players anyway.

 But.... what about Dice Pools and Probability Curves?  : Interesting but Messy
Obviously a single d10 and 2d6 are not the same. Adding dice together (or using buckets of dice) can "smooth" the rolls, making them more predictable.  The buckets of dice method (throwing handfuls of dice with say a 4+ or 5+ as a success) does something similar.  But for a budding game designer, they can be a pain to balance.  A +1 modifier on a single d10 is 10%. On a d6 is is a flat 17%.  But modifiers on 2d6, for example, are not equal depending on your target number. Changing a +1 from a 7+ to an 8+ on 2d6 changes it from 58% to 42%.  A +1 changing 11+ to 12 on a 2d6 is 8% to 3%.
The +1 modifier does not have the same "value."   If this is confusing, then it's a good reason why these methods are not ideal for the amateur game designer.  Whilst I like managed probability, unless you have a compelling reason to use these methods (or love math), they make game balancing/tweaking far more difficult. I do like dice pools (which can add depth to gameplay through resource management aspects) but they do add to the game's complexity/play time and "mental cost."

Anyway, I hope I've shown how agonizing over which dice to use is fairly unimportant in the big scheme of things.  Focus on other stuff - starting with a mission statement/rationale aka success criteria, consider how all the elements (move/shoot, activation/initiative, table setup, deployment, lethality, etc) will combine to make players play the game using tactics you envision.  Ruthlessly keep to your original goals - e.g. if your aim is a fast play game, consider hard before you add ANY complication that does not promote your core philosophy.  Sometimes a cool mechanic is actually not the best for the specific game. This is especially true of dice mechanics.  Think about the big picture, and keep the dice rolling simple, consistent, and "under the hood."

Aerial Rules Revisited - Part 3 Detection & Distance

I've covered maneuver in the last post - it is a very simple system, far more "gamey" than realistic. The aim was to keep the "feel" of air combat (tradeoffs between speed/position/height) without strict scale besides "it looks OK on the table."

I toyed with "limited fuel" i.e. after say 3 game turns all planes using supersonic must make a crew check or be bingo fuel." However this would mean tracking another token on the table, so instead I made supersonic less attractive by making supersonic planes unable to climb. Technically this is wrong, but it forces the tradeoff between speed/height, and makes supersonic less attractive, so players don't use it so often.  The end effect is the same. (i.e. supersonic not overused)

The player is a flight leader/squadron leader controlling 4-12 aircraft, and the aim is to make the game play no slower than a skirmish game like Mordheim or Infinity.  As a result, we don't need to micromanage each plane, measuring precise airspeed, doing 5 degree turns etc. Instead we are giving pilots general instructions.

I do have a more traditional style of movement (has similar effect to CY6 but without the recording and charts, uses 3 colours of d6 to track 3 height and 6 speed bands) but this slows the game.

Some FA/18s and Tornados. I enjoy painting 1:600s - they are simple, satisfying, and "pop"well at tabletop ranges. 
Detection
Besides maneuver trinity of positioning/speed/altitude, detection is the vital element of air combat. Most planes are shot down without seeing their attacker, and nowdays planes carry complex electronics for detection, stealth, countermeasures etc.

I'm going to use the same 180d hemispheres as the movement rules; enemies are either "in front" or "behind" and can be "above" "below" or "same altitude." I like 180d as you can simply eyeball things without measuring - or if needed, hold your ruler in front of the model's base to check. No fancy charts or measurement aids needed.

Radar Detection
To limit the amount of die rolls, players may attempt to spot using radar OR Mk1 eyeball, not both (unless a 2-seater like F14).

I may allow hidden movement - i.e. counters which are only revealed when enemies detect them, with dummy counters etc - but at the moment I'll focus on the ability to engage enemy targets.

Radar range will vary depending on the plane
Radar arcs - the front 180d arc, + targets above or below must be at 4" range or longer
Radar lock is a roll based on the quality of the radar, say 3+ for a good one

There will be penalties if targets are lower, and/or on the deck, and if a target is stealth (F-22, F-117).

A single roll is made and ALL planes in the correct arcs/range who pass the roll will be spotted/locked.

As an example, a plane might need a 3+ on a d6 to lock.  Target A is at the same altitude so needs the usual 3+, but Target B is below, so needs a 4+.  A single roll is made for all targets, so a 4,5 or 6 would mean both planes are detected but a 3 would mean only Target A was detected.

Visual Detection
Visual range - enemies are hard to see, if they are: above and sunward, behind or below or in clouds. Each condition halves the visual range. E.g. if behind AND below, 1/4 usual visual range.
Visual "lock" - the lock/detect number is based on crew skill i.e. say 2+ elite, 3+ veteran, 4+ rookie?  Perhaps add a -1 penalty if to the rear with a bad-visibility cockpit like a MiG-25.

Sunward - an edge of the board is designated as sunward.  Clouds can be seen into/out of (at half range) but not through.

Again, like radar a single roll is made and all targets that roll equal or above are spotted.

Well, this may seem quite detailed compared to my casual approach to maneuver rules, but that is because I think detection is so important to simulate air combat.

These F-14s remind me of Macross fighters. I may do an anime space spin off (eschewing height and adding zany heroic feats and swarms of missiles).

Move/Shoot Ratios
As I've said, scales are pretty loose - it's more about how it looks on the table and the 'feel' rather than precise measurements like 1" = 500m, 1 turn = 5 seconds etc like some rules specify. 

What I DO want is to keep firing ranges relatively low and movement high. A normal skirmish game has a move/shoot ratio of 1:4 (i.e. 6" move, 24" shoot); I'm aiming for 1:1 or better in favour of movement.  I want to encourage planes to move and maneuver, not snipe from long range.

Let's do some examples:
F-14 subsonic move 6"; sprint 16"
= vs =
Cannon range = 4"
AIM-9 HS range = 2"min to 8"
AIM-7 RH range = 4"min to 16"

These are just ballpark figures, but you can see the firing ranges seldom exceed the movement ranges. In practice, this emphasizes maneuver.

Game Mechanics? Bah Humbug!
Well, it's getting late so I'll save game mechanics for another day. Spoiler: I'm actually going to use the Warhammer 40K roll to hit-save-roll to damage using simple d6s.

The older I get the less I care about mechanics. Move shoot ratios, activation, resource management, modifiers, maneuver, etc - all these things matter. The mechanics themselves don't so much (I'll do a game design post on this sometime soon) as long as they are simple and consistent. A bonus if they are familiar, to decrease the "buy in" for new players. I personally prefer d10s (10% increments make balancing the game easy) and d6s (everyone has them, familiar).

Tuesday 8 January 2019

Aerial Rules Revisited - Part 2 - Abstracting Movement

Last post I discussed how aerial rules (simulating combat that takes place over minutes or seconds) tends to be bogged down with written orders, rivet counting and an innate complexity. Most aerial wargames have more in common with doing a tax return than a swirling dogfight.

I identified four factors (speed, height, positioning vs foe, detection) as key elements to track. Now, I'm considering how to simplify them.  Note these rules aren't good, but more an example of my thought process and what I am striving for.

I've decided there will be 3 heights - high, medium, and "on the deck." As I am trying to do away with recording (and also I'm not using fancy bases) I'm simply going to record it thus: high = blue token, low = green token, medium = no token.  I considered using a d6 and having 6 heights (using the dice to record it) but sometimes I bump the dice over, and the token can be fit neatly under the model's plastic base. Besides, I'm trying to abstract here.

OK, now for the speed/height/positioning interactions.  I'm avoiding plotting moves or recording anything (including speed) so I have decided to divide aircraft speeds into supersonic and subsonic.

Supersonic is a sprint move. It's used to transit the battlespace quickly.  A supersonic plane can dive one level and/or turn up to 90d after the midpoint of the move. A supersonic plane has high speed, but poor positioning (maneuver angles) and whilst it can dive it has no ability to gain height. You can see the trade-offs.

Here is an example of a F-14 making a supersonic sprint. It moves quite far (up to 16") and may make a turn after the midpoint.  Interceptors with a high Mach number topspeed (like the Mig-25) have a great sprint ability, whereas light dogfighters (like the F-16) have a lower top speed.
I also call this supersonic sprint extending or pursuit as it is primarily to chase down or flee enemies.

Subsonic speed is where you perform extreme dogfight maneuvers i.e. it is strong in positioning and changing heights but much slower than supersonic sprint . A jet making a subsonic move can move into the front 180d arc and move up to it's subsonic speed (6"in the case of this F-14). 
The jet may either turn OR climb - if it does both its move is halved and it can only move 3". It may dive without penalty.  As you can see, attempting to climb (height) and turn (position) has a penalty on speed. I'm trying to main the trade-off.


The F-14 can make a tight turn (high or low yoyo) combined with dives or climbs (chandelle). The F-14 can even stay still (simulated by a loop). As you can see, I am lumping lots of maneuvers into the one rule, to keep things simple. Whilst I could have had separate rules for all these maneuvers, I've kinda made a single catch-all rule, so you never need to consult the rules or a maneuver chart.  (Savage Worlds does this well). 

I'd like to add an "evade" subsonic maneuver where the plane gains saving rolls against hits, perhaps. This evading would encompass scissors and barrel rolls and probably restrict speed to half. I'll think more about it when I add a combat mechanic.

Reversals are basically half-loop maneuvers into a plane's rear 180d arc.  The main difference from a frontal tight turn is that the plane MUST either climb (Immelman) or dive (Split S) - it cannot remain at the same altitude. The plane's height will impose some restrictions; you cannot Split S on the deck; or Immelman when you are high.


Now last post I mentioned plane performance and pilot skill as being important modifiers.  You might include pilot skill as a "crew check" on a d6, perhaps 2+ for ace, 3+ for veteran, 4+ for rookie - in order to pull off such a violent subsonic turn or reversal.  A failure would see them move straight ahead or 180 behind, and get some sort of penalty (perhaps a "stun" token that gives them -1 on all rolls until removed, or being unable to fire or evade this turn).

Plane performance could be simply movement based - i.e. supersonic movement is based on pure top Mach speed and thrust, whereas subsonic movement is more based on power/weight + wing loading + agility benefits like fly-by-wire). So a F-14 in the example would have 6" subsonic, 16" sprint - while a more nimble but slower F-16 might have better 8" subsonic, but only a 12" supersonic sprint.

I haven't mentioned detection at all, but this gives you an idea of how I am thinking about movement.  I'm not claiming these are even good. Just showing how I am striving to show the interactions between height, speed and positioning in as simple a way as I can. 

+ There's no charts A-F, or complex maneuvers or speed to track. There's no "accounting."
+ There's no fancy bases needed or complex turn templates; the only "clutter" is a token that fits under the aircraft's base (so it's not really clutter) showing if it is on the deck or high. 
+ There's no need to consult rulebooks - you can memorise the rules and move a plane as quickly as a 28mm infantry figure in a skirmish game; so a player can handle 8-12 planes (like a Mordheim warband) rather than 1-2 planes each in traditional aerial wargames, making it a wargame rather than a laboriously slow duel.

Obviously I'm not saying this whole idea works well - or even works at all - (I need to bolt on detection as well as shooting/damage mechanics). Obviously activation is important; I'm thinking something somewhat random (like a card draw?) to simulate the chaos of a dogfight.

But what I hope I've shown is the intent to keep key concepts (height vs speed vs positioning) while abstracting where possible and removing complexity, charts, fancy gear and accounting. I'd like to see aerial wargames evolve similar to skirmish wargames instead of remaining a laborious chore based on 70s mechanics, while retaining the "essence" of air combat.

Wednesday 2 January 2019

Aerial Rules Revisted (2019) Part 1

Longtime readers will know that aircraft wargame rules are an area of wargaming which I generally dislike. Or rather: I love air combat, and 1/600 models are both affordable and fun to paint, but the rules suck.  My run down of the various rules (in 2015) was negative.  I don't think the rules have evolved since the 1970s, and most seem fundamentally flawed. I have lots of lovely plane models but no desire to use them on the table.

I created a manifesto of key things my ideal aerial rules would contain, and looked at activation mechanics. I then kinda left it there, as (like usual) I dallied with other projects.  But I'd still love to get my aircraft minis on the table....

I really like my 1:600 jets. Easy to paint, and they look good at tabletop distances. I admit I secretly like the Su-27/MiG-29 over Western designs.

I revisited the topic last year, making a few points:
*rules are too complex, more accounting than playing. Why am I writing down "orders" for a fast paced dogfights with actions that last seconds?
*the 3D environment seems to necessitate complexity/aka energy management etc
*reusing tired old mechanics from 1970s where other genres have innovated/aka clunky mechanics
*tendency to rivet counting and excessive detail
*rules tend to be more suited to 1v1 duels than a wargame, aka one player flies one plane.
*wrong "command level" i.e. should you be controlling the exact throttle settings and maneuvers for each plane, if you command more than one..

The article wondered what we could abstract, yet maintain the essence of dogfight combat. (With some pretty abstract ideas)

Now I've been playing a lot of War Thunder (a free PC games) and I've identified a few key elements in air combat:

1. Speed (which can be converted into height/favourable position) aka kinetic energy
2. Height (which can be converted to speed/favourable position) aka potential energy
3. Position (relative to enemy aircraft; can be gained by trading in speed or height) attack delivery

A fourth factor which is not as directly related is:
4. Detection (allows planes to position themselves without opposition)

The more boring "good guys". As I was painting them, I was struck with the fact the majority of modern jets have their origins in the 1970s...
Flying my P-51 Mustang on my PC, I'm struck by how efficiently trading speed/height/positioning seems to be so important. Without the fear of real combat, it can be quite clinical in a PC game "I can make a tight turn and get a shot off (position), but I'd bleed off too much energy (speed), so..."  or  "If I dive now, I can convert height to speed, and run down that bomber."  Or choosing between a high or low yo-yo.  I spend a lot of time doing mental calculations of time/speed/distance/angle. Further, detection is vital. I'd say the vast majority of aerial kills occur without the attacker being seen.

Then there are two "modifiers"which strongly impact the above elements:

(+) Pilot Skill (a good pilot finds it easier to convert speed/height/position for attacks/avoidance) and
(+) Plane Performance (a good plane has better potential speed/height/agility to position and even may have better detection).  I.e. things like Thrust/weight and wing loading, the firepower to execute attacks, robustness to resist them, visibility from the cockpit)

While very important, I wouldn't say pilot skill and plane performance are key elements in themselves, but rather impact the elements.

I've always had a soft spot for the much-maligned MiG-23. I like their practical, businesslike look and although they don't have notable agility they have good thrust/weight and speed.

In my April post, I wondered whether height could be removed and simplified into "potential energy" but I think height is integral to air combat and abstracting it could lose to much of what makes air combat what it is. Differing altitudes impact positioning strongly in a way abstract "potential energy"does not. (There are a few rules, notably C21 Air War, which do away with height; incidentally it is a set of rules where you could attempt to fight a decent sized dogfight in a reasonable amount of time).

By the way, I'm not claiming my analysis of air combat (based on a PC game!) is exhaustive or accurate - this post is a kind of "train of thought" exercise showing my evolving thoughts on the topic (and hopefully inspiring someone more talented than me to write a set of relevant, modern rules instead of copying Blue Max for the tenth time!)

Anyway, I'm looking at the "four factors"and thinking "how simple could you make these, while retaining the essentials of a dogfight?" I'll probably start with a trimmed down set of conventional(ish) rules, and contrast it with more abstract ideas.