Originally posted by Deskepticon
View Post
Announcement
Collapse
No announcement yet.
Ramblings from a Recent Savage Worlds Convert
Collapse
X

 1 like

Originally posted by Lightning91 View PostI would propose a simpler solution: In those instances were the target number equals the attribute/skill die, allow the player the option of rolling the next die down (and foregoing the odds of rolling one or more raises). The GM could do the same, but I think he/she shouldn't since the GM is tracking enough as it is (but the players would likely not notice either way).
In situations like this, player agency (choosing which die to roll in this instance) should result in player contentment. Additionally, I do not believe many players will select this option (see above re forgoing the odds on rolling raises). If after several sessions no one has availed themselves of this choice, suggest removing the option and continuing with RAW.
Since higher die types reflect additional skill and training, I visualize this odd but potential decreased chance of success (due to higher die) to be a result of the PC/NPC overthinking or overanalyzing the situation, when a lesser skilled character would have the benefits of decisive action (due to ignorance or lack of concern resulting from a lesser skill base). This of course only happens when the difficulty of the task is at the limit of their skill (i.e. when the TN and die type match).
 1 like
Comment

Riffing off of what ValhallaGH just said above, I'm an engineer and programmer by training, and deal with a lot of math. In my work, being very precise is extremely important. But in a game such as SW, that tiny little glitch doesn't bother me at all, I just ignore it. It seems like the OP and others are jumping thru a lot of hoops to fix something that doesn't really matter in gameplay. But I understand, to each his own.Savage SummariesRAW, with added info from Clint:Combat Actions,Cover,Healing,Using Powers,Grappling,Chases (all SWD)
Also:Persuasion (SWADE),Better Bosses (SWADE),Better Combat Rating (system independent)
And:historical tech levels,generic SW scifi tech levels (both system indepdent)
 1 like
Comment

Originally posted by Wzrd View PostFor some reason I love the idea of using fudge dice like this. Damn, that's clever. Not just for smoothing out the probability curve (especially with 2 fudge dice), but I'm thinking that if you roll ++ then the player, or GM, can choose the swap those ++ for a Bennie instead of adding them to the total. Need to ponder over that one for a bit.
Originally posted by ZenFox42 View PostRiffing off of what ValhallaGH just said above, I'm an engineer and programmer by training, and deal with a lot of math. In my work, being very precise is extremely important. But in a game such as SW, that tiny little glitch doesn't bother me at all, I just ignore it. It seems like the OP and others are jumping thru a lot of hoops to fix something that doesn't really matter in gameplay. But I understand, to each his own.My blog: Savage Stuff. I've also written some free tools and supplements.
 2 likes
Comment

Agree that it is a tiny glitch and the game plays fine without trying to fix it. If it was difficult to fix, I probably wouldn't bother. Adding 2 Fate dice to all trait rolls is fairly simple to implement so I'm going to go with that for a while.
The way Fate dice work is that for each + you roll, add one to the trait roll. For each  you roll, subtract one from the trait roll. These 2 Fate dice smooth the probability curve nicely. The beauty of this solution isn't just the smoothing that I need to satisfy my nerd brain, but I'm hoping that those Fate dice can also bring something more to the game, adding to the fun.
You have a 11% chance to roll ++ and 11% to roll  with 2 Fate dice. My initial thoughts are that ++ gives you a Bennie as long as you succeed and do something worthwhile. For example you roll Notice to check if anyone is hiding in the bushes. You succeed while rolling ++ and notice the bandit, so you get a Bennie. But if there was no bandit, you would not receive the Bennie since the roll did not achieve anything. I'm mostly concerned that player's don't try random rolls trying to get extra Bennies.
When  is rolled you are required to spend 2 Bennies, instead of 1, to reroll that trait roll.
I'm also pondering that maybe rolling a ++ might, instead of a Bennie, give you an extra action in that turn. This extra action uses the current multiaction penalty and does not make the penalty any worse. Need to think about this some more.
The 2 Fate dice idea wasn't mine, so make sure the check out http://www.godwars2.org/SavageWorlds...html#FudgeDice for the original.
For those curious, here are the probability curves for d4 to d12 with no wild dice but with 2 Fate dice:
Comment

First off, I want to say I'm just putting this out there for information sake. If the "statistical anomalies" bother someone, then it bothers them. End of story.
But I do want to be clear for anyone else reading the thread that taking the system of Trait resolution as a whole, a lower die type is not mathematically superior to a higher one as far as game play goes.
I understand looking at that one factor can be a bit like seeing 2+2=5, but 2+2=5 is just one part of the overall equation. Hopefully, I can make the larger equation a bit more clear.
Originally posted by gigacanuck View PostI have also noticed these statistical "hiccups". However, they're not really worth doing anything about. Not only are they small percentage differences, but: they only occur at very specific numbers (6, 8, 10 and 12)
 they only occur between adjacent dice
 the difference gets smaller as die size increases
 the lower chance to critically fail, as well as reach higher TNs in general, outweighs these hiccups
I think this covers most of the points, but I'll add a few things from the game design standpoint. Note that all of these points reference Wild Card rolls with a Wild Die since that's the factor that impact players.
The most "improvement" is between a d4 and d6 when specifically looking for a 6 (the d6 is superior for every other TN), and the difference is just 1.73% (32.29% vs 30.56%). The difference gets smaller from there (less than 1% for a d8 or d10, averaging 1.1%). Keep in mind that those slight differences only come up in game if rolling that specific die type looking to hit that specific number and only as it compares to the next higher die type. If we limited the potential highest TN to 12 (the highest one affected, though higher is possible), that's 4 out of 60 possibilities (5 die types x 12 TNs) which would average a 1.1% variance. In short, the effective impact on overall results would be roughly 0.000733%.
Another consideration is the higher die type in each of those 4 cases will actually never roll that specific number (without modifiers). Each of those numbers is the highest on that die type, which means if they roll that number, they Ace, and get to add to that result.
So if a d6 rolls a 6, at the very least it gets a 7 instead (adding a minimum of 1 from the reroll). So the odds of rolling a 6 on a d6 are in fact the same odds of rolling a 7 (30.56%). The odds of a rolling a 7 on a d4 however are 27.08%. So the overall chance of the d4 matching a d6's result is actually 3.48% lower. Not necessarily a huge factor, but one to consider when looking at overall results.
But really, there's one more major factor I haven't seen mentioned (though I admit I may have missed it).
Trait rolls (almost) always have 2 levels of success, not one.
There's the base effect of success, but also the improved result from rolling a raise, 4 points higher. And the higher die type will always be better at achieving the other level of success than the lower die provides for the other one.
Consider a d8 vs a d10 rolling against a 4 with a –2 penalty. To get the raise, the player needs to roll a 10 with the penalty. The d8 has a 18.36% chance and the d10 has a 17.5% chance. But the base success requires rolling a 6 (–2 for a 4). The odds for the d10 are 58.33% and 47.92% for the d8. So, in that case the player would "lose" a 0.86% better chance for a raise but gains a 10.41% better chance of a success. The d10 is definitely the better die to roll for that reason alone.
Anyway, all of these things are factored in to the system. The 2 levels of success, the 2050% greater chance of rolling a critical failure with the next lower die, the extremely small difference in the odds at those specific TNs between those two dice, the higher die type TN being 1 better, the rare instances when it would even come up, and the impact of trying to "fix" something that's imperceptible in actual play by adding more complexity and when taking the system as a whole, doesn't provide a true "mathematical" advantage.
Again, just looking to clarify in case someone were to think the "anomaly" breaks down game play in some way, that it doesn't, at all.
It does bother some folks, and it may keep some from trying the game, but for others, it's worth adding some complexity to not worry about those cases and keep playing Savage Worlds. It's really not much different from house rules anyone might have for their table, so let's respect it the same way we would any other.
Thanks!Clint Black
Forum Admin & Rules Answer Guy
Savage Worlds Brand Manager
 6 likes
Comment

When we first noticed this very small quirk in the system I decided to offer the players the option of always being allowed to roll a lower die type if they preferred.
They have never chosen that option, because the weighting still favors the higher die so much that that marginal percentage was not worth even remembering.
Comment

As I was looking over this thread, another item came to mind, which is the MultiAction situation. I previously did a bunch of calculations about when it's worth it to attack twice versus once, and the whole TN=trait die situation shows up there.
There are several interesting things that happen with TN 6 vs a d6 trait if you can try multiple times, and a lot depends on whether you're looking for at least one success or many successes, but here's some numbers anyways, for wildcards:
d6 trait, trying once TN 6: 30% chance of success, 16% chance of raise, 3% chance of crit fail
d6 trait, trying twice TN 6: 45% chance of success, 10% chance of raise, 5% chance of crit fail (all probabilities are for at least 1, though obviously there's some possibility of getting 2 of these results)
d4 trait, trying once TN 6: 32% chance of success, 13% chance of raise, 4% chance of crit fail
d4 trait trying twice TN 6: 35% chance of success, 8% chance of raise, 8% chance of crit fail. (again at least 1, but 2 is possible here)
Anyways, just some more numbers to think about, I might try to see how the fudge dice effect it...
Comment

I just watched a video where Shane and some others playtested some Savage Pathfinder. https://www.youtube.com/watch?v=pTaz...ature=emb_logo
It got me to remember some of the fun that I have had playing Pathfinder and the versions of D&D up to that point. For me, Savage Worlds better fits my play and GM style, so it will be great to get the feel that I miss but with the rules that work best for me. Frankly, I am excited. Full two boxedset Super Pledge all the way. I think that my son will run it, too, so that is a nice bonus to not have to be the GM. I can run other stuff for the group as I had planned.
 1 like
Comment

Originally posted by Theorel View PostThere are several interesting things that happen with TN 6 vs a d6 trait if you can try multiple times, and a lot depends on whether you're looking for at least one success or many successes, but here's some numbers anyways, for wildcards:
d6 trait, trying once TN 6: 30% chance of success, 16% chance of raise, 3% chance of crit fail
d6 trait, trying twice TN 6: 45% chance of success, 10% chance of raise, 5% chance of crit fail (all probabilities are for at least 1, though obviously there's some possibility of getting 2 of these results)
d4 trait, trying once TN 6: 32% chance of success, 13% chance of raise, 4% chance of crit fail
d4 trait trying twice TN 6: 35% chance of success, 8% chance of raise, 8% chance of crit fail. (again at least 1, but 2 is possible here)
The trying once is pretty straightforward as being (generally) the typical odds of success for a Wild Card with a d6/d4 Trait. It's the "trying twice" that's throwing me.
I took it to mean "If a character attempts the same action twice with that Trait die, the numbers represent the odds of a critical failure, a success or better, and a raise or better on at least one out of the two rolls." But the numbers clearly don't follow that premise (there's no way for the chance of rolling a raise or better to lower rolling a second time), so I'm not sure what they do represent.
Hoping for some clarification if possible.Clint Black
Forum Admin & Rules Answer Guy
Savage Worlds Brand Manager
 1 like
Comment

Trying twice in a single round, so MAP is making the probability for any single success go down.
There's some TNs where getting MAP is pretty clearly worse than getting multiple tries, i.e. if the TN is 2 then trying twice only makes things worse. (i.e. if you have a +2 to get TN 4 you only need to not crit fail, while trying twice means you negate that +2 bonus, and have to roll an actual 4 or better, actual probabilities later)
Actually, that works to show that this isn't exactly a TN=trait die problem. It's just a thing which happens when you're at particular points in the probability curve, which I assume has to do with the dice acing and changing the slope of the curve. In particular, if the TN is 9, 10 or 11 (or 5,6, or 7 and you want a raise), with a trait die of d6...you're more likely to get that value at least once by trying once, than by trying twice and applying a MAP (making the effective TN an 11, 12, or 13 respectively)
Of course, it is more pronounced when the TN equals a multiple of the die (i.e. going from 9 to 11 just lowers the probability by 0.5% while going from 10 to 12 lowers it by 5.4%) However, it's still present even with the fudge dice, even if again it's less pronounced. Anyways, my point I was going to make was that it's not just 'raises' that care about more than the base TN, but also multiple actions (with MAP), which then introduced a bunch more math, and things got complicated quickly, and so I just posted the numbers instead of saying all of that :P
By way of example, using d6 trait, and a TN 2:
Probability of success on a single die is 83%
If you do 2 actions, and suffer MAP, probability on any single die drops to 50%.
Putting that together, an extra has 83% chance of success if they try once in a round, and ~75% chance of success if they try twice.
The wild card starts with a 97.5% chance of success (probability of at least 1 die in 2 getting 2 or more), but trying twice means at least 1 die in 4 has to get 4 or more...that probability is 93.75%. If you just need 1 success, then in this case you should only try once.Last edited by Theorel; 01292021, 01:12 PM.
Comment