NichG
2013-01-30, 06:34 PM
The recent set of 'Tier 4 tests' made me think of the question: How does one go about thoroughly testing the difference between builds, characters, tiers, etc?
The main problem with play-based testing is that you have to keep switching up the scenario so that pre-knowledge isn't an issue. And really you probably won't get more than a few groups willing to actually play a test like this at a given time, so you only have one or two data points.
Instead, what about a testing protocol that each player can do 'at home' and report their results? Basically you'd write the scenario but more like a video game than a tabletop RPG would normally be. That is to say, enemies follow a script like:
1. Charge-attack the nearest detected valid target, else
2. Move towards/attack the nearest detected valid target, else
3. Attack the last known location of a target if within 1 round of vanishing, else
4. Spend move and standard actions to Spot/Listen every round
Furthermore, every character and enemy are supplied with a fixed, pre-rolled set of d20 rolls. That is to say, each character gets say 10 Skill checks, 10 Saves, 10 Attack Rolls, and takes average on all damage rolls.
Clearly the threat of said enemies will be lower than if run by an actual intelligent DM, and the system can be gamed a bit by writing characters to exploit the script. Furthermore, players can time actions to focus on their best rolls in the set. So anyone run through this will be more powerful than they'd normally be.
On the other hand, there could be some value in seeing what the actual optimum behaviors are in these situations. And its something that can be discussed and pored over on the forums without changing the situation of the test at all. The test would likely need to be highly constrained in certain ways to prevent the introduction of adjucation-required elements. Essentially I can see this breaking in all the usual ways when planar binding/etc becomes available. But for low level tests it may remain a useful protocol.
Summary:
- Fixed, fully exposed modular test with all rolls pre-rolled
- Expect the PCs to be much more powerful due to pre-knowledge
- Probably needs to have some extra constraints to keep things within the frame of the tests - avoid abilities that introduce adjucation.
- All of this allows each forum member to independently attempt a scenario with any combination of builds/characters/etc that they want to investigate.
The main problem with play-based testing is that you have to keep switching up the scenario so that pre-knowledge isn't an issue. And really you probably won't get more than a few groups willing to actually play a test like this at a given time, so you only have one or two data points.
Instead, what about a testing protocol that each player can do 'at home' and report their results? Basically you'd write the scenario but more like a video game than a tabletop RPG would normally be. That is to say, enemies follow a script like:
1. Charge-attack the nearest detected valid target, else
2. Move towards/attack the nearest detected valid target, else
3. Attack the last known location of a target if within 1 round of vanishing, else
4. Spend move and standard actions to Spot/Listen every round
Furthermore, every character and enemy are supplied with a fixed, pre-rolled set of d20 rolls. That is to say, each character gets say 10 Skill checks, 10 Saves, 10 Attack Rolls, and takes average on all damage rolls.
Clearly the threat of said enemies will be lower than if run by an actual intelligent DM, and the system can be gamed a bit by writing characters to exploit the script. Furthermore, players can time actions to focus on their best rolls in the set. So anyone run through this will be more powerful than they'd normally be.
On the other hand, there could be some value in seeing what the actual optimum behaviors are in these situations. And its something that can be discussed and pored over on the forums without changing the situation of the test at all. The test would likely need to be highly constrained in certain ways to prevent the introduction of adjucation-required elements. Essentially I can see this breaking in all the usual ways when planar binding/etc becomes available. But for low level tests it may remain a useful protocol.
Summary:
- Fixed, fully exposed modular test with all rolls pre-rolled
- Expect the PCs to be much more powerful due to pre-knowledge
- Probably needs to have some extra constraints to keep things within the frame of the tests - avoid abilities that introduce adjucation.
- All of this allows each forum member to independently attempt a scenario with any combination of builds/characters/etc that they want to investigate.