Miska

Miska

Playtest report: Revving up the Core

The thing that excites me in game-design is seeing the thing actually running with all the bolts and nuts loosely attached. I guess this is what drives the people who hack Apocalypse World or create new Fate settings too. So you can guess the feelings running through me couple of days ago, when I started up explaining the basics of Ironcore engine to our company role-playing club members.

We had agreed earlier to play something light-hearted action, as the session was our first after summer break. The Ironcore’s design ideas fit perfectly for the task, so I prepared a small action packed one shot scenario for the group. With the added benefit of being able to  find out some Ironcore’s kinks while having some after-work fun with the colleagues.

For play-test purposes my goal was to double down on the Push-Strain-Flow system, which seems to work nearly perfectly on paper. To do this I needed the scenario to have scenes with real sense of danger and risk, and lots of action-hero action.

tl:dr: The Push-Strain-Flow system rocks. Roughly drafted hacks can have consistency issues.

The Push / Strain / Flow

Flowchart of the Push-Strain-Flow system.

Miska called the PSF an “iron-solid core” of the system, and based on the findings I have to agree. The system design is deviously clever – as its super simple to explain while still being able to create a surprising amount of excitement and suspension to the game.

The basic idea starts from the Push. A player character is able to exert their Abilities to enhance their performance when absolutely needed. The catch is that the Abilities effectively work as hit points, making Pushing a risky business.

I was originally a bit opposed to the idea of combining the Strain and damage as I thought players would feel adversely on using their primary damage reduction points for added performance. As Pushing increases the odds of gaining Flow for extra effects, damage or resistance – I did not think the issue was really pressing though.

The play-test session revealed that my worry was misplaced. The game group had no issue using the Abilities for pushing purposes, even when that made the characters temporarily more prone for danger. Quite the opposite: closer the characters got to being wounded, more Pushing occurred.

During a debrief a player made a remark on the issue: as pushing in a firearms fight is done with Focus, while damage strains Focus only if Body is Exhausted, it’s quite rational to use the Focus points on eliminating the enemy before the Body runs out. Choosing whether or not to Push is much harder if Pushing is done with Body, like in a mêlée combat.

Possibility of gaining additional Flow when Pushing had actually less effect to the player choices I had anticipated. I expected the players to game for Flow, but there was little need for actual Flow hunting within the current system. This was a highly positive surprise.

Another positive surprise was the player’s reaction to the main villain using his first Flow point during the boss fight. Before the scene using Flow had been the heroes domain, and giving this ability to a single game master character worked extremely well as the players had already seen the mini-bosses Pushing earlier.

Rough Edges

One of our goals with the Ironcore has been hackability. Simply put, we want the engine to be easy to bolt on to any action role-playing setting or scenario.  And even if the PSF system works well, there is work to be done on the hacking front.

One of the basic challenges of hacking a game with skill based resolution mechanic is to draft a simple and consistent Skill tree. A rough skill tree often causes both skill overlap and skill gaps in addition to super-skills like “Combat Sense” skill we added for survival and tactics in combat.

The issue lessens by each game published with the system. As it’s far more easier to tweak an almost working listing to a hacked one. Chthonian highways will set a good example for Post-Apo genre as will Hyperstorm for scifi. Modern action, horror and classic fantasy mods will need more work for now.

In our play-test Delta Green slash Mage Technocracy themed scenario – we used an experimental skill tree, and added a host of equipment, cybertech augmentations, and some supernatural effects to the base engine with surprisingly little problems. To me this underlines the design of Ironcore. Even with quite inconsistent hacks, the core engine keeps on running. Like a Mi-go powered superhauler on a worm ridden highway…

…which reminds me, I probably need to get back working on the accursed thing…

 

Facebook
Twitter
Reddit
WhatsApp
Email
Print

Myös nämä voivat kiinnostaa sinua...

Ironspinen uusi aika on koittanut

Uusi aika on koittanut Ironspinelle. Toiminimi jonka allekirjoittanut laittoi pystyyn vuonna 2006 on päättänyt toimintansa.  Mutta tämä ei ole Ironspinen loppu, vaan

Uutuuksia ja uutisia – Syksy 2019

Heinäkuussa julkaistiin Astraterran Kertojan käsikirja ja karttakokoelma Taivaltajan kartasto. Kertojan käsikirja viimeistelu Astraterran 2. laitoksen julkaisun siinä mielessä, että nyt säännöstö on

2 Responses

  1. Your results correspond exactly with what happened when we played Numenera. You have similar might, speed, and intellect pools and you spend from them to activate powers or to expend effort for a bonus. Outside of exceptional circumstances, all damage first hits might, then speed, then intellect. The result was that even though our party consisted of two glaives, two jacks, and a nano, none of us had might based powers. We just would not spend from our primary “hit points”. Similarly, we rarely used effort, because it was far too expensive for the momentary gain in most circumstances.

    1. I noted the same while playing Numenera couple of months ago. Design-wise these results imply, that any effects gained from spending “Might” -points should lessen the risk of losing “Might” -points or create an effect worth losing those points. In Ironcore this would mean melee, brawl, and other physical feats. Maybe even dodging attacks. Btw., there have yet been no mana/fate-style point use (to activate powers) in any of the test-scenarios I’ve created, and I am not sure if the push/strain mechanics can cover those needs at all. In fact as a personal preference I feel that powers often work better when they require no activation cost. If characters can breath, they can breath… and if they can breath flames, they can…

      On Flow/Effort: We are trying to balance the Flow generation to a level, where spending Flow would be both effective and reasonable, without being too cheap. And I do agree on Effort being too expensive in the Numenera’s system.