I have an RPG in development for Ten Worlds, the setting for the Attack Vector line of games.

In mechanics, it owes a lot to Jake Norwood's Riddle of Steel, and Jake has given permission for me to copy is Spiritual Attributes rules.

Basic mechanic is a roll-under dice pool system. Your TN is the attribute, your skill is rated in dice. If you have a Talent with a skill, you get double successes on die rolls of 1.

I developed a good chunk of my combat engine before seeing Riddle, and one of my alpha readers pointed me at Riddle, since they were similar.

Combat rounds have two Actions. Dice pools refresh at the beginning of the next Round.

I was hoping for a "Roll in the middle" combat system, and something that requires less real world knowledge of guns/weapons than Riddle does. What I got wasn't that...

You can do the following things in an Action:

1) Look Around. You can only keep track of a number of opponents equal to your REASON score. They're the only ones you get full pools against.

2) Move -- either 1 hex step (combinable with everything subsequent), or a number of hexes equal to AGILITY.

3) Yell -- Shout up to 5-6 words to another character.

4) Ready (some weapons take multiple actions to ready)

5) Aim (Aim a ranged weapon, benefits go away if you ever use dice defensively). Benefits let you use your full dice pool with a ranged weapon, otherwise, you use half your base firearms skill, and lose any specializations)

6) Fight (split your dice pools between offense and defense, a la riddle, everyone splits at the same time, attacks resolve in order of highest number of dice, with PCs winning any ties).

Take attacker's successes, subtract defender's successes, get a margin of success. If 0 or less, the attack misses. If positive, add MoS to the weapon damage, roll hot location, subtract armor, compare to lethality code and tell the defender what wound level they just took.

Immediately, we realized is that the game is quite munchkin-able in character generation. One character maanged to make a Combat Pool of 34D, at which point 5D Dramatic Hooks don't matter much at all.

First combat:

Man with sword charges man with gun. Man with gun takes an Action to aim. Man with sword ducks into a crouch, which doubles the range penalty, and crawls. We assumed both sides saw each other as a simplifying assumption (more on this later).

End result: Shooter rolls his 9D pool on the second action of the Round. With an attribute of 5, he gets 6 successes. Range penalties at 40 meters are 8 successes needed, doubled for the crouch to 16, the gun's sights drop that down to needing 8. Swordsman lives. Advances another hex.

Round two: First action: Swordsman charges, making a running roll and closes by 5 hexes. Rifleman aims. Second action: Swordsman is 3 hexes away, rifleman has a bead on his chest. Swordsman charges, and rifleman shoots; Rifleman gets 7 successes, the swordsman uses his default Agility check and gets 1 success. Damage for the Lee Enfield is 10, Margin of Success is 4, the Thuggee Cultist's armor is 4, which leaves 12...and the end result is a level 5 wound. Thuggee cultist dies at 3 hexes.

This was the "right result" for the tactics involved, however, the amount of handling was too much for a narrative game, and the amount of interesting decisions to make on a tactical level were too few for a simulationist game.

Second trial:

Two guys with swords and shields and armor duelling it out. One has 16 dice in the pool the other has 17. The guy with 17D had target numbers of 4, and the fellow with 16D had target numbers of 5.

Round 1, action 1: Guy with fewer dice splits them 2 offense, 6 defense. Ohter fellow had gone 6 offense and 3 defense. Successes match up on both exchanges. Action 2, it's 6 defense, 2 offense versus 4 offense, 4 defense. No hits scored.

Round 2, action 1: Pattern repeats. No wound scored. Action 2: Both go 4/4, no wound scored.

Round 3, we decide to not bother rolling buckets of dice, as the number thrown means that we're hitting the mean in most cases. We just allocate and add 10ths of a success, and roll a d10 apiece to add to the fractions to see if we go over the threshold. Action 1: 17D fighter tried a surprise with 10D in the attack and nothing in the defense, 16D fighter had put 8D in defense. Nominal successes were 5 versus 2.4, defender rolled a d10, got a 10, had 3.4 wich rounded to 4 successes. 17D fighter hits for 6+1-6 -- light wound, which the player described as a pump fake that grazed the thigh.

And we stopped it there, because it wasn't fun. It was too deterministic and too much work for the amount of enjoyment we were getting out of it, even if it was getting results that were correct for the genre we were looking for (gritty fantasy).

I tried for light, fast and lethal. Got ponderous, slow, and lethal. In trying to come halfway between simulationism and narrativism, I fell into the Marianas Trench and fell short of both. :)

The "combat action selection" (slide a penny between choices on the character sheet) went smoothly. Up until we threw dice we were fine. And the first two times we threw dice, it was "Well, getting right result..."

I suspect smaller die pool sizes would help considerably. My armor model is broken, and it might work better as a "saving throw" instead of a damage modifer.

And now the questions:

How do you streamline die pool games?

How do you keep large numbers of dice from being deterministic?

While this is arguably a good 'model' for realistic guns combat, it has the "player character assumptions" that PCs never suffer from buck fever or the shakes. Arguably with modern firearms, if you've had even minimal training, hitting a man sized target is pretty easy, and doing potentially lethal damage is in the same range of difficulty.

The interesting question isn't "can I hit him", it's "Is this something I'm willing to pull a trigger over". This game doesn't seem to hit that note well, at least as a combat test.

The tactical choices in this boil down to "don't get seen" and "hope your armor save works". It really doesn't give interesting choices in terms of mechanics yet.

Ahh well. Enough rambling.

AdAstraGames wrote: How do you streamline die pool games?

How do you keep large numbers of dice from being deterministic?

AdAstraGames wrote: Round 3, we decide to not bother rolling buckets of dice, as the number thrown means that we're hitting the mean in most cases. We just allocate and add 10ths of a success, and roll a d10 apiece to add to the fractions to see if we go over the threshold.

AdAstraGames wrote: The interesting question is ... "Is this something I'm willing to pull a trigger over?".

AdAstraGames wrote: The interesting question is ... "Is this something I'm willing to pull a trigger over?".