I was really pleased with the simplicity of the original skill improvement section of the Version 0 rules. Specifically, skills could be improved in three disctinct fashions; a Flash of Brilliance, Trial and Error, or by Learning.
These three mechanisms have three seperate spheres of influence respectively; in game, end of story arc, and out of play context.
I’m really proud of the Flash of Brilliance idea. It gives the player a distinct advantage and allows for immediate improvement within the game. See my post on rewards as they apply to Dungeonsiege to understand why I like this idea so much. – KO
Here’s how the original improvement mechanics were put down…
Improving:
Flash of Brilliance
If a character succeeds remarkably in a skill by rolling a 1 in a non-opposed check or a 10 in an opposed check, that skill has the possibility of improving immediately due to a flash of brilliance. To see if such an event has occurred, the player rolls an additional d10. If the roll is above the current skill score, then 1 point is added immediately to the skill.
Example of Flash of Brilliance:
Odie is fleeing from Something Nasty that is pursuing him out of the cellar of the lighthouse. Odie’s player asks if a successful Occult Lore skill check would provide any help in determining how to defeat or subdue the Nasty thing behind him. The GM agrees and Odie rolls a d10 and gets a 1. Brilliance!
Odie rolls another d10 and gets a 6. Since 6 is greater than 4 (Odie’s current Occult Lore skill), Occult Lore is immediately increased to 5.
Odie recalls something he read in a dusty tome years ago and reaches into his pocket for the sprig of mistletoe he always carries. Praying that his gambit works, Odie turns and holds it directly in front of him…
Trial and Error
Characters get better at skills by using them.
If a character succeeds in a skill check, regardless of opposition, he or she should put a check beside that skill. Once the scenario is over, the player will roll d10 for each skill that is “checked.” If the d10 roll is above the current score, then 1 point is added to the skill. This simulates learning by success as well as making progress to high skills more and more difficult.
Example of Trial and Error:
Since Odie has survived the horror beneath the lighthouse and has returned home to rest and recuperate, he has a chance to learn from his mistakes and successes. During the investigation the following skills were “checked.”
- Club (a new skill that was previously unlisted and assumed to have a skill of 1)
- Occult Lore (currently at 5)
- Architecture (currently at 2)
- Investigate (currently at 4)
- Odie rolls d10 for Club and gets a 3. Since 3 is greater than 1, Club is increased to 2.
- Odie rolls d10 for Occult Lore and gets a 7. Since 7 is greater than 5, Occult Lore is increased to 6.
- Odie rolls d10 for Architecture and gets a 1… bummer. Architecture is left at 2.
- Odie rolls d10 for Investigate and gets a 4. Since 4 is not greater than 4, Investigate is left at 4.
Learning from a buddy
Any character with a skill greater than 5 can teach that skill to another. Learning takes 1 month for each point learned and can be raised in this fashion to a maximum of one-half the skill of the teacher (rounded down.) Learning and teaching are mutually exclusive events. One cannot teach and learn in the same month.
Example of Learning:
Thinking that perhaps there are easier ways to make a living, Odie calls his buddy Woz to teach him how to gamble effectively. Woz is a professional poker player with a gamble skill of 7. Since Woz’s skill is greater than 5, he can teach Odie how to gamble more effectively than a normal slot-jockey. While Woz is arguably an excellent player, the most Woz can teach Odie the gamble skill is: 7 / 2 = 3.5 = 3. Odie and Woz are taken out of commission for 3 months as Woz takes Odie under his wing. After the 3 months, Odie now has a gamble skill of 3.
The flash of brilliance mechanic caught one of my player’s eyes almost immediately. Here was his response:
The only thing I noticed right off is the Flash of Briliance. I really like the idea, but I think the number of times we roll a 1 or 10, then just have to beat our score, after a few session I think many of our skills would be maxed out. Might be better if you have to roll two 1′s or 10′s in a row, makes it more difficult and hopefully less common.
-L
Now, keep in mind that L is a pretty hardcore D&D player who likes crunchy rules and lots of dice to roll and chance to dictate the play. That’s not a bad thing, so don’t flame me for pointing it out. But I think he was working under the false impression that we would be rolling for a lot of checks. While I’ll admit I used to do so in the past, I really wasn’t envisioning this ruleset to be used to constantly roll lots of dice. Instead, I intended to make the checks only be rolled for if the outcome (success or failure) would be interesting. So with that in mind, I demonstrated the math in mock tabular format:
Intital Score: FOB Chance x Chance to beat current score = Chance to improve
Kevin’s Proposed Method
1: 10% x 90% = 9%
2: 10% x 80% = 8%
3: 10% x 70% = 7%
4: 10% x 60% = 6%
5: 10% x 50% = 5%
6: 10% x 40% = 4%
7: 10% x 30% = 3%
8: 10% x 20% = 2%
9: 10% x 10% = 1%
10: 10% x 0% = 0%
Results in a diminishing linear probability with faster skill increases due to FOB, but with a diminishing probability as skill competency increases.
L’s Proposed Method
1: 10% x 10% = 1%
2: 10% x 10% = 1%
3: 10% x 10% = 1%
4: 10% x 10% = 1%
5: 10% x 10% = 1%
6: 10% x 10% = 1%
7: 10% x 10% = 1%
8: 10% x 10% = 1%
9: 10% x 10% = 1%
10: 10% x 10% = 1% <– Ignored as 10 is considered the maximum score
Results in a flat linear skill increase probability.
-Kevin
I’ll admit I liked the “less likely” part of L’s proposal, but I didn’t like the part where mathematically there was an exception to the rule. Namely that at skill level 10, we had to handwave away the Flash of Brilliance improvements rather than let the math just eliminate the possibility. In an attempt to help the players understand that relentless dice rolling wasn’t going to be the modus operendi, I also shot this across the email system:
I guess I should probably add that I assumed this would help offset any “leveling up” desires in the deep seated D&D part of our collective brains. Since the proposed system doesn’t currently allow for additional points for buying (something we may also want to consider) this would allow you to improve a low score quickly to competency via use, while still holding off on us all ending up with 10′s across the board.
-Kevin
I also thought it would be appropriate to analyze the Trial and Error mechanic in the same mock tabular format, so I sent it out too:
I’ll also pass on the probability scores for improving a skill via my proposed trial and error improvement option
Score: Success Chance x Chance to beat current score = Chance to improve
1: 10% x 90% = 9%
2: 20% x 80% = 16%
3: 30% x 70% = 21%
4: 40% x 60% = 24%
5: 50% x 50% = 25%
6: 60% x 40% = 24%
7: 70% x 30% = 21%
8: 80% x 20% = 16%
9: 90% x 10% = 9%
10: 100% x 0% = 0%
Results in a pretty good bell curve.
-Kevin
I really liked the simple bell curve this improvement mechanic displayed, and while I thought the magnatude of the percentiles would also scare off the players, surprisingly they weren’t even mentioned intitially. Instead, another player chimed in and affirmed L’s fears.
I’d probably lean toward L’s method just to keep it from artificially throttling skills up right off the bat. The GM’s perspective on use of those skills is more important than anything. You can adjust it to fit as necessary based on good ideas which is less artificial.
-M
In hindsight I probably should have addressed the underlying assumption of their concerns, but I didn’t. And with that, a small bit of complexity crept into the system.
Then I too fell prey to the “we’re gonna throw a lot of dice” mind-set and wanted to revisit the Trial and Error improvement mechanics…
I’m a little concerned that the method I’ve proposed is developmentally too fast.
I’m concerned that a scenario may indeed utilize all a character’s skills and thereby provide massive advancement in a single setting.
Couple of ideas to limit this sort of thing:
- The maximum number of skill improvements can not exceed a character’s IKE score. (The player gets to decide which skill improvements will be tested for.)
- We kludge my initial FOB test and use it for Trial and Error like this:
Player 1 has “checked” 3 skills that he was successful in using during the scenario.
Player 1 now rolls a d10 for each of these 3 skills.
Results:
Skill 1: 1
Skill 2: 6
Skill 3: 3
Only skill 1 can be considered to improve.
Skill 1 has a current score of 3
Player 1 rolls another d10 and gets a 7.
This beats a 3, so Player 1′s Skill 1 is improved to a 4.
The resulting math states that the chance of improving a score due to trial and error would be:
Score: Success Chance x Improve Chance x Chance to beat current score = Total Chance to improve
1: 10% x 10% x 90% = 1%
2: 20% x 10% x 80% = 2%
3: 30% x 10% x 70% = 2%
4: 40% x 10% x 60% = 2%
5: 50% x 10% x 50% = 3%
6: 60% x 10% x 40% = 2%
7: 70% x 10% x 30% = 2%
8: 80% x 10% x 20% = 2%
9: 90% x 10% x 10% = 1%
10: 100% x 10% x 0% = 0%
Which I like better.
-Kevin
As you can see, I fell into the same trap of assuming a lot of rolling. M tried to provide me a way out, but apparently I failed to see the light:
Off the top of my head, the Top Secret system used “critical rolls” such as 11, 22, 33 … and 01-05 for something like Flash of Brilliance. The general odds were less than 10% to increase a skill/ability by 1%. Given the nature of the system, it was less than 8% most of the time since critical rolls had to be less than the percentile skill ability. If you scale that down to a single d10 system, L’s method is the simplest to utilize without having to keep track of what skills where successful — it is also in the 1% range, slightly above a system designed by so called “experts”. It also keeps the tie for a singular event tied to the skill that was used rather than picking and choosing. People are often skilled in areas they really had never considered.
Likewise, the most common abuse of such a system is rolling dice for a skill check just to be rolling. That happened entirely too often in the TS world. If the common approach is that skill checks are only necessary if the task is difficult in comparison to the current skill score, artificial inflation would be uncommon. There is the odd-chance that a character would spurt ahead for a session or two but that’d likely balance quickly.
What you are really looking for is an exponentially increasing difficulty for FOB checks that isn’t significantly easy to accomplish for the lower skill levels. To accomplish that, you need a 1 followed by an initial FOB check that must also be 1, followed by a roll that must be higher than the current skill level of the character. Best case is 10% x 10% x 90% = .9% (0.009). Eventually you get down to .1% chance to increase on a FOB check. While that seems low, it makes the chance of it happening very unlikely, which I find realistic. Most skills take a lot of practice to self-learn. I could get basic training in swinging a sword but might swing it a 1000 times before I understand how the technique I was taught is really effective.
On the other hand, if a person RP’s using a skill wisely and well within the scope of their ability, the GM might note a skill check of 1 even after a failed FoB check and award the point anyway. I would always vote to take the dice out of the game rather than luring characters to roll even if there is no purpose.
-M
At this point I allowed the complexity to enter the system, and that’s where I failed in one of my core tennents to the system design. Eventually I reworked the mechanics, eliminating the complexity and redundant dice rolling. I returned to the simple solution and allowed the GM to decide on how fast a player has the chance of improving by dictating how often skills should be tested.
Tags: D&D, Dungeons and Dragons, Dungeonsiege, Flash of Brilliance, Game Design, improvement, KORE rpg, Learning, mechanics, reward, rpg, rules, Skills, system, Top Secret, Trial and Error
