Thursday, 9 December 2010

BattleTech 2.4: Purpose of House Rules

Why have house rules? Or a more helpful way of phrasing this question, what purpose do house rules serve?

From a purely pragmatic perspective house rules are the players of a game system using their power to audit what has been given them, by saying "yes but" to the rules of the game.  However, as soon as one enters house rule territory one has to get others to agree with you about your "yes but", which often leads to a reversion to a "yes but, let's just use the rules as written".

Now I've made it clear in past blogs that I'm an OSR kinda gal, and for me realism and BattleTech are not synonymous, which means I can accept I'm just playing a game.  However, the pedantic wargamer in me also wants the rules to touch some sort of base with reality.  Fickle female I suppose?

This also touches on some science fiction mantras that writers use when creating new universes for their stories.  Change one thing to the reality of your universe to allow your story to work, but follow the laws of physics etc with the rest.  Of course it's not that easy to do this, and writers end up using a combination of unobtainium and handwavium at times, but in principle less is more.  Stories where the writer moves further away from "keeping it real as much as possible" tend IMO to be writing science fantasy, rather than hard science fiction.

Like most things though, there is a continuum from hard SF to soft SF to pure science fantasy.  Hard SF tend to limit things like tractor beams and anti-gravity usually, because they indicate soft SF by having consequences to the universe that are conveniently ignored.  Scantily clad women and muscle rippling heroes with swords tend to fit better in a science fantasy setting.

Where does BattleTech fit then?

Well it's not ultra hard, despite of having no anti-grav and tractor beams, because too much unobtainium and handwavium exists around the size of the ships (mass to volume ratios too low), and having ludicrously high thrust drives with handwaved reaction mass.  For more on this I can highly recommend the Atomic Rocket's website.

Therefore I've decided I'm going to put BattleTech in the soft SF category, especially since the literary form of the universe is predicated on telling stories within a pulp fiction/juvenile rite of passage mix.  So it's a bit borderline to science fantasy, but on the whole BattleTech keeps itself in the middle-lane.

So what leads me down this tangential path from my starting position of house rules.  It's the "yes but" part.  The rules of the game do not simulate the novels particularly well.  To some extent it's good that they don't, because games follow rules, whereas stories are about emotional and physical responses by the characters to situation the plot has landed them in.  Both have logic driving them, but the purpose of games is to play, whereas stories tell a tale about people.  Of course games can also tell stories too, but that really would be going off on another tangent.

Either way both stories and games require people to spend time to enjoy the pleasures that each offer, and this is where my "yes but" comes from.  I no longer have the time to spend playing a game that will require more than a few hours of my time.  Long gone are the day where I would stay up until the early hours of the morning involved in a role playing game, or a weekend playing a wargame.

Books are easy enough to put down and pick up.  Films generally don't run more than three hours or so, but wargames can run for 10 to 15 hours, and what is worse not have been played to a satisfactory conclusion.  Large games of BattleTech fall into the latter category.  I'm not saying it can't be done, I'm just saying it is hard to pull off without a lot of preparatory work by everybody involved in playing the game, and the thing is that I don't have the time to do this anymore.  Again I'm long past the time in my life where I can rattle of chapter and verse on any set of rules, let alone BattleTech.

So what's my plan?  KISS; Keep it Simple Stupid.  Seriously, simplify the games and then simplify some more.  The problem is not only how, but by what has to be be simplified?

One evening this week I went to a lecture by Professor Jamie Angus on compressing sound recording through digital encoding.  She was a fascinating speaker, and even though I have little interest in audio engineering (I just like listening to music), when the presentation moved into system analysis and statistics I saw the same patterns I recognise in the probabilities from using dice to generate outcomes for games.  Fascinating stuff, and most of of it way above my head. Link. 

She described a three stage process for digital audio encoding, which is about reducing the number of bits of information it takes to record a piece of music.  The first stage is the use of lossless compression, which leads to the second stage of "quantitative entropy" that can allow music to be compressed further, but with various levels of information lost depending on format choice.  Running in parallel is an intermediary third stage, the "psychoacoustic components" to the digital encoding, which one can then use to massage the music.  Psychoacoustic modeling accounts for things like how loud music shadows underlying tones that precede it, which can then be removed with no audible loss.  Thereby saving a bit of information in the process.  This allows the encoder to take into account masking of sounds that occur due to the biological limitations of the ear's acoustic range, and the brain's processing of music and people's affective responses to sound combinations.

As I listened to her talk about this I had an eureka moment about applying this to games.  If you agree that games are based on information theory, then what we want is for the outcomes to be calculated with the minimum of loss: all be it of detail resolution, realism and time among many things, rather than soundwaves.  In this case I'm going to target one particular loss that BattleTech manages poorly, and that is the time spent by player's resolving each turn of the game, which increases as the number of battlemechs and players rise in a game for my examples.

From EastwoodDCs Giant Battling Robots I know that 2D6 encodes 3.27 bits of information, link.

I also glean from then same article that 1D6 rolled twice encodes 5.17 bits of info.  Both of these are relevant to BattleTech, because both mechanisms are used in the game to generate combat results.  The first 2D6 for one's to hit number, and 1D6 rolled twice when damage goes internal (Ed. correction, except that in BattleTech the first 1D6 roll is actually a 1D2 roll, so instead of this encoding 5.17 bits of information, it only has 3.58 bits, assuming I've got my maths right?).  These rule mechanisms tell us the maximum amount of information that can be encoded, or in game speak the granularity of the detail that can be resolved, or accounted for.  For instance as Steven Satak in his blog The BattleTech Reader made some relevant comments about the limit that the groundscale imposes on the game, link:

His comments were about the 30 meter hex size and how the game rules breakdown, and this is a good example of granularity in a games resolution of underlying detail, and illustrates what the game designers assumed could be ignored, discounted, or waved over.

Now if I've understood EastwoodDC & Professor Jamie Angus right, then approximately three bits of information is all that you can encode when rolling 2D6. Now I put my neck on the line and see how stupid I can be made to look on the internet. This means that rolling 2D6 encodes just a little bit more than rolling one D8, and I'm now going to round down and treat them as synonymous to each other, because it allows me to use heuristic analysis based on 8 bit computer systems.

Now when one applies modifiers to a 2D6 roll, if one knows that it only encodes 3.27 bits, then a modifier of one is equivalent to one bit.  If I'm correct then a modifier of three is equal to three bits of information, which has the effect of reducing surprise in the diced for result?  Such modification of the base 2D6 roll is therefore highly significant, which seems to me to support my proposition about the modifiers for targeting computers and pulse lasers in the BattleTech game being too coarse.

"Yes but" I hear, but what about the modifiers that you get from movement that make it harder to hit a battlemech, aren't they the same?  Yes they are, but they add to the granularity of the game, not take away from it.  By granularity I mean that more bits of information are added to finesse the combat results, whereas reducing the information in the 2D6 combat roll coarsens the granularity of the combat resolution results.

For instance, my hunch (WAG), is that if BattleTech had used 1D6 rolled twice to resolve hits, and 2D6 to resolve internal damage then the minus three modifiers of having both a targeting computer and pulse lasers would be far less unbalancing, because 1D6 rolled twice encodes 5.17 bits of information, rather than 3.27 that the current 2D6 system can encode.  Now if I've not just made a complete ass of myself, which will then require me to edit and delete said parts of this blog, where we go next is to look at "the what" consumes time when playing BattleTech?

Reducing the amount of time it takes to resolve the action each turn will streamline the BattleTech game resulting in either quicker games, or more turns in the same time period.  So what is the first step?

First we start by identifying where dice rolls, or modifiers are not needed, because if you can reduce the number of things one has to do in a turn, then the game will go faster.

Secondly, using the concept of digital encoding would allow the compression of game data, while still retaining the original granularity of the game play.  Where this choice comes into conflict with the need for speed, the encoding model would indicate the appropriate game mechanism (choice of dice to roll, or rule) to use so as to be able compensate for any loss granularity from the reduced number of dice rolls.  I imagine this will be equivalent to the concept of reverse shadowing in psychoacoustics; where a sound is revealed that was previously hidden by the change in timing of a louder passage in the score.

However, big problem number one.  Taking the standard BattleTech rules and rewriting them using this model of information coding goes way beyond any set of house rules that I've encountered in all the years I've been playing the game.

Quite frankly this really would be BattleTech 2.0, the next generation, or re-imagined game universe, and I can't see anyone from CGL reading this and saying to themselves "you know what, why don't we let her do this?"  I'm under no illusions here, as after all CGL have said that no outside contributors are being sought at this time, and as and when they do seek new freelancers, it will be to fulfill their long term business plan for the game.  Given that CGL have not yet finished their six volume core rule set, one which is rumoured may/has been cancelled, I'm not so much holding my breath, as waiting for Hell to freeze over.

So do I now go back to my metaphorical drawing board and rethink my approach to house rules, or do I run with the idea of compressing the core rules of BattleTech?  The former is quick and easy, the latter not so much.  Time is probably going to dictate what I do next.

Terms & Conditions, Errors & Omissions Excluded from this Blog

I unreservedly reserve the right to change my mind at any time about the subject that I've just written, and make changes to the spelling, grammar and contents as I see fit, up to and including the total removal of said blog posting if necessary.  Furthermore what I've written at this point may be just what I'm thinking at the moment I wrote it down.  Nothing should be inferred about any authority of the writer to write said thoughts down, as everything that has been written is meant to be entertaining for the reader, and it shouldn't be construed as anything more than the ramblings of yet another person on the world wide interweb.  However, no contract should be assumed between the writer and the reader about whether or not the article is entertaining or not.  Remember service brings citizenship, but reading a blog is purely voluntary, and you remember what your parents said about volunteering don't you?  If you don't remember, or know, then the writer of this blog cannot be held responsible, or liable for the outcomes that arise from volunteering for things that can have deleterious results to either your physical, mental, or emotional wellbeing.  All posts are condensed & abbreviated summaries of complex arguments posted for discussion on the internet, and are not meant to be authoritative in any shape, or form on said subject.


  1. Wow, Ashley! I wish I had written that, because you did a good job putting some difficult concepts into simple language. Your math about subtracting bits of information isn't quite right, but your intuition of how information decreases with increasing difficultly is spot-on. At least it's correct from a certain perspective, and I think you have helped me understand how to explain what that is.

    There is a lot more in your post that I should respond to, but I'm out of time for the present.

    I posted the follow up to my Dice and information post, and added a partial response for you. :-)

  2. Ha! You scooped me! I finished up my response to your last blog post at midnight, wanted to wait until I'd gotten the rest of the blog together for an update.

    Well, missy, it is coming. Meanwhile, I am going to print your latest out and read it over a cuppa and a cigar. Nice work! And the disclaimer! Who would have thought there was a lawyer hidden inside that nice English lassie? Go figure.

  3. Nice blog! Thanks for letting me know about it.

    Not sure I follow all the math, but are you saying that one way to speed up the game would be to combine to-hit and damage location rolls in to a single dice throw? That makes sense and you could easily duplicate the probabilities of two separate rolls, but would such a table look too complicated for ease of play?

    You might be interested in this thread on the Starship Combat News boards:

    It's by the author of Attack Vector: Tactical and concerns spaceship gaming, but the theories apply to all genres.

  4. Interesting stuff based on the EastwoodDC stuff. Good read. I am in the same boat as you in terms of battletech speed/playability, so I went and wrote a slew of pages about a streamlined but somewhat more realistic (in my mind) battletech tentatively called BattleCommand. Initial feed back was that it was not battletech anymore, but I would be interested in your opinion since you seem to be going down the same road.
    The core idea and subsequent extrapolations:

    First make the turns express 1 minute of time. This allows the hexes to be 6 times larger, and range 6 times greater, to better simulate a longer ranged ground conflict.

    Second, thanks to the 1 minute turn length, all weapons will fire 5 times in a round. Thus, armor/structure on a mech will be 1/5th. This allows the record sheet to be shrunk to the size of a playing card, which can be put in a plastic sleeve. Collectible card potential not withstanding, card sized record sheets have great appeal when running a company of mechs, as all 12 record sheets will be on 1 page and 12 decriptions will be on another page (or the rear). This is what war machine does, and they do just fine.

    Third, because damage is being inflicted 5 times as fast, game length will be reduced to 1/5th the duration. So a duel that lasts 20 rounds in battletech will be concluded in 4 rounds in BattleCommand... because the idea is that if you want intricate one on one action you have battletech as it is. But if you want to have lots of stuff battle it out at once you want to play faster.

    Forth, and this part really pertains to your post: Change the dice rolling conventions. 2d6 as a mechanic has the advantage of a bell curve, but in practice with modifiers and such requires numerous individual rolls and doesnt make use of the bell curve distribution well. The 2d6 probabilities can be reduced to a single d6 toss with limited statistical difference. Basically, compress the 2d6 toss to a single d6 toss, and then roll 30+ events at the same time if need be. No box of death, no 15 pairs of colored dice. 4 lasers, 4d6 droped, picking up successes. 20 hit locations need to be rolled thanks to that LB20x? Drop 20d6 at once, and done. Like you mention, compress the bit size of a roll.

    After that foundation, it was just a bunch of rules converting all of the Total Warfare 30 meter stuff into streamlined BattleCommand 180 meter stuff. Like I said, the initial reaction was that if I wanted to speed up battletech I should play Battleforce (despite my dislike of Battleforce for stripping too much personality of the mech away); also people were very adamant that their was nothing wrong with the 2d6 system, you just need something like a box of death to handle lots of 2d6 events.

  5. Desert Scribe said..."Nice blog! Thanks for letting me know about it.

    Not sure I follow all the math, but are you saying that one way to speed up the game would be to combine to-hit and damage location rolls in to a single dice throw? (snip)

    You might be interested in this thread on the Starship Combat News boards... It's by the author of Attack Vector: Tactical and concerns spaceship gaming, but the theories apply to all genres."

    I have a passing knowledge of Ken's Attack Vector game, and own a copy too. We've even exchanged the occasional email. So I found the link useful.

    In answer to your question though, I'm not yet able to say what my answer would be, as I really need to run some math now and examine the options through play testing, which is the only real way one can solidify stuff around look and feel.

    Remember what Sherlock Holmes nee Arthur Conan Doyle wrote, quote; "It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. There is nothing more deceptive than an obvious fact."

    Bradley said... "Interesting stuff based on the EastwoodDC stuff. Good read. I am in the same boat as you in terms of battletech speed/playability, so I went and wrote a slew of pages about a streamlined but somewhat more realistic (in my mind) battletech tentatively called BattleCommand. Initial feed back was that it was not battletech anymore, but I would be interested in your opinion since you seem to be going down the same road."

    I can see why people said what they said to you, as your house rules have changed some of the assumptions of the Battletech game.

    I personally wouldn't have a problem with going that route, but I think I may have found a another way between that stands dimensionally off from the original Battletech rules, house rules, or BattleForce rules axis.

  6. There are many ways to streamline the game, and they all boil down to personal preference.

    For my own, I i.e. cut down on irrelevant stuff (like FF and ES crits written on the record sheet when they're only relevant in the construction phase), took the good old medium lasers heat and damage as baseline for other weapons as well as armor and internal structure diagrams, and gave the players the option to switch to a 1d8 mechanic and a base THN of 0.

  7. Following up!
    I would put Battletech on the soft/pulp/fantasy side of fiction. I like it anyway!

    I once spoke with someone who was writing a "new" game, which was much like Battletech, except for hit locations he rolled a 1d6 or 1d8 if there were only 6 or 8 locations to be hit. This saves some time over the usual CBT method, but the probabilities of hitting each location are the same either way.

    You might streamline a game by combining rolls as suggested, but you might end up shifting the granularity to a different aspect of the game. For instance Bradley's example above - the scale in time and space is 6 times greater - and some Battletech games are nearly finished in 6 turns. In simplifying the game to do this you lose the variety of outcome - the information - that you would have by playing it the long way. That might be good, because too much detail in a large game is impossible to play.

    I feel I've caused some confusion with my discussion of information. The number of "bits" is associated with ONE random outcome. It could be a to-hit roll, or a hit-location roll, or a combined to-hit-and-location roll, and each would have a different total number of bits.

    Now here is a brain bender - Suppose you could roll one set of dice to resolve a whole turn of Battletech play, or a whole game - How much information would be in that? I don't know myself, but I'll think on it. My intuition is a simpler game should have less total information than a complex one, but I'm not sure yet what it would mean to compare them.

  8. I suppose, if an entire battletech game result came down to one die roll, it would contain perhaps a single 'bit' of information--win or lose. If it was a d4, with Massacred, Minor loss, Minor Win, and Major Victory, then it would contain 2 bits of info.

    EastwoodDC, we have info on bits of data in dice rolling thanks to you. Here is the real kicker though... call it my contribution to the discussion. Check out the amount of data in a movement phase!

    A 2/3 mech is the slowest standard design. Put it on an open field. He has the following options for a 3 MP run... Forward, Turn Left, Turn Right. If he spends 3 MP, there are 27 unique options available correct (though 3 turns left and 3 turns right are the same, and 2 turns right and 1 left is kinda useless)? Then in a walk for 2 MP, he has 4 options... Forward, backward, left and right, 2 times. Thus another 16 potentials (but some combos are silly, like backwards then forwards, or left then right). Plus he could only spend 1 MP, for 4 more potentials. Finally, he could stay still as a final potential. Thus, a 2/3 mech's data for moving in 1 phase is 48 items, or 5.58 bits. If you add 2 jump jets, like an urban mech for example, you add a huge amount more... 6 facings at the end of the movement and 19 possible hexes mean 114 combinations (here no move is technically a useless choice). This has 6.83 bits on its own, and 7.34 bits total with the walk/run/stand still options. Finally, add that torso twist in, which multiplies the 162 combos by three (turn left, turn right, and stay put) to 486--8.92 bits. Finally, if its an urbie it can flip its arms!

  9. @Bradley: Bear in mind this measure of information only applies to a set of random outcomes - a probability distribution. If you are choosing how you move randomly, and all choices are equally likely, then you can measure that information. However, that would be a really weird way to play.

    Really weird, but it might also be a really useful way to think about games. You are on track to describing "Mixed Strategies", concept from Game Theory on how to make optimal decisions - and a topic I hope to get back to one of these days. I can't go into this at length today, but here is a link to my GBR post on Mixed Strategies:
    Or search "Game Theory" on my blog for the whole series.

  10. I think if the time it took to make the move was reduced then this would increase the pace of the game, and result in more turns per hour. I'm aiming for four an hour with eight units on the table, with the goal of a linear, rather than logarithmic increase as more units are added to the game.

  11. For an interesting comparison, where attacks are resolved as a single roll of a handful of dice, you may want to examine the Leviathans game mechanic, PiP. The design criteria there was quite explicitly to speed up combat resolution, which I think has been achieved quite elegantly.


  12. That's the new game by CGL, I didn't even know it was out, or is it a PDF?

  13. I think by sticking to bits when looking at results you ignore a super super important aspect of the dice rolls- distribution. For example a 2d6 roll and a d12 roll have exactly the same number of result, but the distribution is entirely different. This distribution has a large effect on results, but is mostly dependent on the gunnery of the pilots you are using. Both movement mods and TC, pulse, etc all affect the same roll. They *have* to have the same effect on "granularity" However a TC is typically seen with clan pilots and super lower to hit numbers. This typically means each point of to hit number change is actually a pretty small change in percentage chance to hit. However movement mods affect all mechs. You're likely to see them against average to crappy gunners where it'll swing a 6 to a 9. Thats a huge change in chance to hit. Much more so than going from a 2 to a 5.

  14. Absolutely spot on analysis Brian.

    And that they say is the problem in a nutshell. How to maintain the feel of the original while making the game faster?

    I'm play-testing ideas in my current campaign. So I'll feedback progress in due course, but don't hold your breath.