As the title says, is there a rules-lite game out there that has a formula for attackers?

What I'm looking for, in short is Barbarians of Lemuria with something to tell me how many mooks, or sergeants, or saber tooth cats, or villians three beginning PCs (for example) can handle.

Slightly longer: I'm mostly looking for a rules-lite game. And I don't generally like class systems. I love how BoL gives character 4 starting careers and does combat stats separately. I want to be able to run my friends through some existing adventures using such a system, but want something to tell me how much the characters can handle. I'm not looking for advice on the BoL issue exactly--I'm looking for a game that includes rues for building appropriate opponents.

As it is, I'm creating my own system, and getting my wife (who is working on her PhD in cancer genetics) to do a statistical analysis of all combinations of 1-on-1 and 1-on-2 and 1-on-3 attack combinations, to see if she can give me a ranking formula. I'm a software developer and I'm generating the data--the 1-on-1 file is 152,473,105 lines long and takes about 2 days to generate (each case is tested 1000 times).

Hard question, often rules-lite games don't seem to go in much for CR-estimation. The only thing I can really think of offhand is 3:16. It may be be exactly what you're looking for but it has some system for the GM to spread out alien "threat tokens" per mission.  

CR estimation seems popular in 3E-derivative games (3E itself, 4E, FantasyCraft) but they're generally not that light. JAGS' designer Marco was doing something similar to your software approach with their game on their blog too, but its not light either (more GURPSesque).

I also wouldn't be surprised if there's some boardgame or other that would have something like what you're after. Maybe.

Thanks for the reply.

3:16 sounded a bit interesting at first. I wish the previews would give something useful instead of 5 pages of useless fluff. But I read a few reviews and it sounds a bit on the narrative side (players get narrative control by calling for Flashbacks and narrative [that's a new RPG verb from me] their strengths/weaknesses). Combat is "balanced" by making the enemy statless--you attack by rolling against your skill and your weapon damage decides how many enemy there were based on how many your weapon kills.

I've looked at JAGS before and just looked at it again. The crunch--it hurts! :eek: *joke* I don't mind crunch during character creation, but if you're going to have 15+ pages of grappling moves and still not let me simulate Aikido... If PowerFrame, for example, had lighter in-game rules, I wouldn't mind that level of building complexity. (Especially since I'd just write a program to help if needed.)

My second test file of 1-on-2 and 1-on-3 combat finished last night.  Only 14,224,897 lines, since the opponents only have their modifiers go up to 3--but it looks like I'll have to redo it with the numbers going higher, since a maxed out attack easily wins against three opponents at half strength 100% of the time.

On 3:16 - Yes it does have flashbacks and such, which is not my bag either (I think I picked up the .pdf in one of the big disaster relief downloads).
rpg.net had a review of it with some more explanation of how the system worked
http://www.rpg.net/reviews/archive/14/14870.phtml

Oh also - there's, oddly, still a free older version sitting around on the author's webpage.
http://gregorhutton.com/roleplaying/


On JAGS - heh, yep I feel your pain. Just mentioning for interests' sake.
Good luck.

Quote from: Bloody Stupid Johnson;860246...CR estimation seems popular in 3E-derivative games (3E itself, 4E, FantasyCraft)...

Microlite20 (PDF here) is about as lite and stripped down as you can get, but it still allows you to draw upon the d20 SRD when you need to, which has the encounter calculator all ready to go.

Tunnels & Trolls defines monsters by a single characteristic "Monster Rating" or MR. MR determines how many dice and adds the monsters get in combat. This can be easily compared to the number of dice plus adds the PCs do. In combat the losing side suffers the difference in hits between the total dice score plus adds of the high scoring side and the low scoring side.

Not quite a Challenge Rating formula, and spells and special abilities will affect it, but it gives a good baseline.

BOL is straight 2d6 to hit, then some toss or other for damage, right?

Just figure the average damage per round, then you can work out how far down the winner is on average after felling the others.

If damage per hit is not much different, then eyeballing chances to hit should be close enough. Seriously, you should have a damn good ratio if you're thinking about taking on more than one guy in the first place.

Hit on 4+ vs. hit on 10+ is 33:6, or 5.5:1.

Is JAGS2 lighter enough in the relevant 'crunch' areas?

Averages don't necessarily tell enough of the story--some games are going to have a fairly high variance in damage per round, which would militate in favor of building a good margin of safety into your "CR".

That said, if a game's combat system doesn't offer much in terms of options, I have to wonder why have CR. If there are interesting options in combat, then you might want CR in order to keep combats balanced enough that the players can exercise those options in a meaningful way.

For a simple game, frankly, the bulk of interest with the combat is exploring and learning what you can handle and what you can't, and deciding what's worth risking. A CR system deprives players of that experience by guaranteeing that they can run into every combat without thinking.

Maybe replacing BoL's 2d6 with a single d10 or d12 roll would make the odds easier to figure out?

No, the difference is trivial mathematically and computationally.

The problem (and the reason you need a monte carlo simulation which I assume is what the OP is doing) is that there are so many intermediate results possible in the iterative process (i.e., round-by-round).

If instead you had a game where each attack would result only in no damage or defender incapacitated, it'd be a lot easier to write a complete flowchart of possible events and then calculate the probability of each final outcome.

Quote from: Bloody Stupid Johnson;860467Maybe replacing BoL's 2d6 with a single d10 or d12 roll would make the odds easier to figure out?

Sure, but it's only slightly obfuscated with straight 2d. BoL figures typically have separate attack and defense factors.

I typed in a table, but I guess it needed HTML to format it. Not too hard to figure if you know the 2d pyramid:

02 03 04 05 06 07 08 09 10 11 12
01 02 03 04 05 06 05 04 03 02 01
36 35 33 30 26 21 15 10 06 03 01

Quote from: Arminius;860471No, the difference is trivial mathematically and computationally.

The problem (and the reason you need a monte carlo simulation which I assume is what the OP is doing) is that there are so many intermediate results possible in the iterative process (i.e., round-by-round).

If instead you had a game where each attack would result only in no damage or defender incapacitated, it'd be a lot easier to write a complete flowchart of possible events and then calculate the probability of each final outcome.

Arminius, is very correct about average damage being meaningless. I did some testing on that early. If character A has 6 HP and does an average of 3 damage per round, and character B has 36 HP and does an average of 1 damage per round, what happens? Switch the damage, but add in armor with one having 1 and 3? Then reverse the armor. Also, average damage is going to depend on the opponent's defense (at least the way I do combat, and the way anyone else who uses the opponents skill or armor, etc as a target number or modifier). If you hit less often, you do less damage. But you don't want to have to calculate everything the way I did when designing monsters/encounters. Although, that's my fall-back position if necessary.

I actually did an exhaustive simulation--all possible combinations. According to my wife, this makes the ranking calculations easier. Hopefully, she'll have time to explain the analysis steps tonight or tomorrow. She is working on her PhD, so she's pretty busy.

I used something similar to but different from BoL combat. It "feels" better to me and has some features I like.

There are five traits--I'm uncertain whether these are stand alone or based on other traits as yet, but they are:

To Hit Modifier (0-6) modifies chance to hit
Damage Dice (1-6) number of dice of damage possible
Defense Modifier (0-6) modifies chance to be hit
Armor Dice (0-6) modifies the damage they receive
Hit Points (1-6x6) this is a number from 1-6 multiplied by 6 and is the amount of damage they can take

In tests, initiative made less than 10% (usually much less) difference in results for several different versions of combat, so this version got rid of it.

To determine if a hit is made you roll 2d6 and add the attacker's To Hit Modifier and subtract the defender's Defense Modifier. If the result is greater than or equal to 7, then you hit. (This gives a base chance of ~58%.)
        To Hit: 2d6 + Attacker Hit - Defender Defense >= 7

If a hit is made, damage is based on the number found by rolling the attacker's damage dice and subtracting the number found by rolling the defender's armor dice. Any value above 0 is subtracted from the defenders hit points.
        Damage: (Attacker Damage Dice)d6 - (Defender Armor Dice)d6

The damage method means, a tank with 6d6 armor is not going to be hurt by a 1d6 sword or pistol or even a 5d6 elephant gun. (Those numbers aren't meaningful, just illustrative of the conditions.) This is one of the things I like about this method. It gives some flavor of armor being proof against some weapons, but doesn't require a lot of paperwork keeping track of what is proof against what.

I ran with A attacking first. After A attacks, if B still has any hit points, they get to attack. That makes a turn.

Each set of characteristics was tested against each other 1000 times. (In testing this gave stable results.) Tests proceeded until either A or B's hit points were reduced to 0 or less or 21 rounds had been tried. At that point, whoever had taken less damage was declared "the winner" and if damage taken was equal, then a tie was declared. The average number of turns is recorded.

When designing battles, I'd like turns to stay under 5. So I'm hoping to be able to reflect that in the rating if possible.

​I have a second file, testing one against 2 or 3 opponents, with opponent characteristics going up to 3. (I'm going to need to redo that one and cover all the way up to 5 or 6, since a maxed out character A always beats 3 character B with all 3 characteristics.)

I went with d6 "just because" and because it limits the stat variances to a lower calculable number of permutations. Part of what I plan is, similar to Barebones Fantasy, "extra" skill levels allow you extra attacks at -2 (or something) per extra attack. So your first attack is as described earlier, your second attack at -2, third at -4 and so forth. I haven't checked the statistics on that yet. My earlier tests were one attack attempt each.

I also haven't decided how to do active defenses--probably some version of rolling to create a higher defense modifier.

You'll recognize stuff in here from a variety of games, particularly BoL, BareBones Fantasy, and EABA. I'm thinking of calculating the To Hit and Damage Dice similar to how Ryuutama does, off of a variety of characteristics. So, you can have a dexterous and intelligent character who does better with bows and firearms than a strong character, who is better with swords.

To add one, since I'm doing a simulation, the number and size of dice are also pretty irrelevant, except to max stat/skill levels and target numbers. Once I've figured out how or if we can get a useful ranking, I can share Java or C++ code for doing the calculations, or just run them for people. The stats part might be in R.

I'm worrying there is going to be a ranking calculation, but it is going to either be too complicated to be useful or end up dependent on the opponent.

Quote from: Phillip;860417Is JAGS2 lighter enough in the relevant 'crunch' areas?

Still too crunchy. In my case, "crunchy" often means "too many things to remember". Just figuring out the numbers to try for in JAGS2 requires a half page of explanation. I'm not going to remember that. That's my fault, and not the game's.