Apologies, I guess the more accurate statement would have been:

The effectiveness is much more than the small percent difference in a single skill check that his math showed, and people are taking it the wrong way. Not necessarily that you particularly are taking it the wrong way.

Its a long thread and I got mixed up on who said what.

Either way, attributes at beginning are the more efficient spending of XP, but if you don't, its not GAME breaking. And quickly allowing someone to rebuild their character if they aren't having fun isn't a big deal.

Basically the game isn't broken, but it is the most efficient method of creating a versatile character who is more useful in more situations. (not necessarily more powerful in any one area, just capable of more, which I think is the point of the system.)

Apologies, I guess the more accurate statement would have been:

The effectiveness is much more than the small percent difference in a single skill check that his math showed, and people are taking it the wrong way. Not necessarily that you particularly are taking it the wrong way.

Its a long thread and I got mixed up on who said what.

Thank you. I sincerely appreciate and accept your apology, and in turn apologize the relatively terse response. It's rare someone actually bothers to do this sincerely on these boards. I'm just tired of having what I've stated or presented be misrepresented, and I went out of my way to allow everyone to come to their own conclusions in this specific case.

In [weak] defense of what I've shown, it's focused on a single skill, and in a few cases pulls in a second skill compare what could be done with the left over xp. It's also not a single check, it's the summarized results of 100,000 checks.

I agree with the sentiment: This isn't a major issue and doesn't require/necessitate the proposed house. I will point out that, when asked, the OP reported there was a problem noted at his table. I, like many others, agree the better solution is simply re-building the character.

I see little purpose in restating that, though, since that sentiment has been repeated so many times as to be absurd and this thread has long since devolved in dog-piling on the OP.

This thread made me play around with the character generator and I made a few characters without spending any xp on characteristics. I didn`t manage to make one that wasn`t capable.

Why are specialized characters supposed to be best or more "powerful" in a system that is so spread out, has a cap on skills and is narrative? This is by no means a system for power gamers.

Some characters are more rounded, some are more specialized and focused, is one better than the other?

And what does powerful even mean in a system that have so many areas of focus?

I`ve said it before, the character in my game with the most xp by far, has used no xp on combat skills or combat related talents. Does that make him less powerful?

The system does not give extra xp for combat, for solving encounters or killing monsters, the game gives bonus xp for roleplaying and following motivation. You are not meant to be able to fight everyone or everything,

you are not meant to win every encounter, or succeed every skill, failing is fun and interesting too.

Hypothetically it would be more of a late game issue. If you invest in making a character who starts with a balance of skills and characteristics, you're effectively capped in the late game in ways characters who invested more deeply in characteristics aren't. I think it's a negligible problem (if you could even consider it a problem), but I can see players who are obsessed with getting the most potential out of their stats and might sacrifice creating a nuanced, diverse character they might other build under the pressure of "If I don't invest in these characteristics now, they're not going to grow".

Again, it certainly wouldn't be a big deal in any game I ran or played in, but I find the viciousness of this thread pretty inexplicable. That dude thinks it's a flaw in the system and proposes house rules for anyone who feels the same way. Others think it's not a flaw in the system at all and is simply reflective of the setting. I couldn't be arsed one way or another, but can't we just live and let live?

as to the efficiency of stat dumping versus Skill dumping of points.

Though Putting points into a stat raises the over all effectiveness of multiple skills....

Putting those same points into skills allows you to raise More skills as they cost less.

Lets Just take it from the lowest stat value.

start with a stat of 1 It cost 20 points to raise it to a 2

Raising 1 skill to rank 2 from rank one costs 10 points for a career skill 15 for a non-career skill... In most cases, half the cost.

a stat cost 30 points to go to a 3.... 15 points for a skill to go to rank 3 for a career skill 20 for a non career skill... so again about half cost.

Now Since Most stats start at a 2....and Most skill start at a 0 rank

If I put my points into the stat to raise it 1, spend 30 points. Which might get me 1 more green Die in the end.

If I instead put those 30 points into skills, I can get 6 career skills to rank 1.... or 3 non career skills to rank 1... or 2 career skills to Rank 2.

Which nets me 1-2 yellow dice in the selected skills.

Now the skill raising might benefit you in the beginning over stat raising, being an over all more broad character...

How ever... In the Long Run, the Person that raise 2 or 3 stats to 3 or 4..will have an Expertise in his field, a pool of Yellow dice, That exceeds anything the the person who does not raise stats will likely be able to achieve given the same gains.

So it ends up being a choice of Expertise in ones field of choice or being a Broader skilled character but less of a specialist.

I see nothing wrong with either choice.

Having looked over some of the NPC stat blocks ...... all things considered not stat dumping doesn't really hinder you. The NPC's generally have average stats themselves with truly important ones having a 3 or 4 but overall they stay close to mostly 2's. So in the long run it's not like failure to buy up stats at char generation is going to put you at a huge disadvantage against NPC's.

Hey all, I just wrote up some MatLab code to calculate the probabilties of success for each combination of skills (GG to YYYY) given normal difficulty (PP). Below is an example of how the numbers were crunched, but here's the result:

P(Success > 0 | GGPP)
0.4351

P(Success > 0 | GYPP)
0.5020

P(Success > 0 | YYPP)
0.5651

P(Success > 0 | GGGPP)
0.5969

P(Success > 0 | GGYPP)
0.6508

P(Success > 0 | GYYPP)
0.7004

P(Success > 0 | GGGGPP)
0.7205

P(Success > 0 | GGGYPP)
0.7617

P(Success > 0 | YYYPP)
0.7681

P(Success > 0 | GGYYPP)
0.7987

P(Success > 0 | GYYYPP)
0.8472

P(Success > 0 | YYYYPP)
0.8812

------------

P(S>1) = probability of there being more than 1 successes

400xp in
Characteristic character:

You do know that no matter how solid your math is, it is still a matter of trying to fix a completely avoidable problem with an imbalanced houserule?

Hey all, I just wrote up some MatLab code to calculate the probabilties of success for each combination of skills (GG to YYYY) given normal difficulty (PP). Below is an example of how the numbers were crunched, but here's the result:

(subtitle: a head exploding from WAY TOO MUCH MATH!)

Who actually read throgh and really looked at JJRodney`s math just now? Be honest!...

You do know that no matter how solid your math is, it is still a matter of trying to fix a completely avoidable problem with an imbalanced houserule?

Also, letting everyone respec once is a fair houserule that I think most everyone on the forum has agreed with.

At character creation,

Skill human succeeds 40% of the time

Characteristic human succeeds 53% of the time

100xp in

Skill Human 49% success

Characteristic Human 55% success

400xp in

Skill Human 56% success

Characteristic Human 59% success

Um, big deal? We're talking a 3 to 6 percent difference of between one way or the other - that's a ridiculous tiny gap.

Also again, failure can be fun. Constant success is unnecessary.

At character creation,

Skill human succeeds 40% of the time

Characteristic human succeeds 53% of the time

100xp in

Skill Human 49% success

Characteristic Human 55% success

400xp in

Skill Human 56% success

Characteristic Human 59% success

Um, big deal? We're talking a 3 to 6 percent difference of between one way or the other - that's a ridiculous tiny gap.

Also again, failure can be fun. Constant success is unnecessary.

It's all a false argument because it doesn't take into account advantages and triumphs, both of which have significant impact on the overall effect of a given dice roll. A single triumph can activate a crit, which under certain conditions could disable a target and end a confrontation with a single success rolled.

He's fixing something that isn't broken and doing it without taking a look at the mechanical effects of a total dice pool.

It's all a false argument because it doesn't take into account advantages and triumphs, both of which have significant impact on the overall effect of a given dice roll. A single triumph can activate a crit, which under certain conditions could disable a target and end a confrontation with a single success rolled.

He's fixing something that isn't broken and doing it without taking a look at the mechanical effects of a total dice pool.

(1Y means 8% triumph, 2Y means 16% triumph, 3Y means 23% triumph)

EDIT: I take that back - the skill character has 12 more yellow dice at character creation and ~4 more yellow dice than the characteristic character at 100xp, but at 400xp the characteristic character has more yellow dice.

So maybe the lack of success rate is made up for in the larger probability of triumphs?

Still at 400xp and beyond, the characteristic character will have more triumphs and a higher success rate.

Add to that, the characteristic character has more dice in general (G and Y) and thus more advantages.

EDIT: These are really similar to LethalDose's math (Thanks LethalDose!), I only did this because there were a few dice combos that I couldnt find in LethalDose's posts (he's got mostly PPP and a few PP). I'm going to post a bit more using this info.

It's reassuring to see the stochastic results (my method; simulation) line up well with the deterministic results (JJR's method; calculation). At least I assume that his method is deterministic, and he simply found each combination of positive and negative dice results that provided the exact combinations of the S & F, which in turn lead to the total success space.

I'm going to make a subjective qualification below. Said qualification is in no way intended as a slight, or even criticism, in regards to what JJR has done above.

Overall, I dislike the deterministic method for estimating the dice pool results for .

Both methods have their merits, I simply prefer the stochastic method for the following reasons:

• Algorithmically simpler
• Easily evaluates multiple dimensions (success, adv, triumph, despair)
• Easily calculates probability ranges and variance
• Easily calculates correlations between dimensions
• Provides a holistic view of the dice pool results

The downsides of using the stochastic method are:

• Risks not getting a representative sample of the outcome space*
• Processor/time intensive*
• Changing the dice pool requires a complete re-run of the simulation
• Kind of a PITA to convince individuals ignorant** in regard to stats it's valid

*The former is offset by the latter, in that with a large enough sample, the results strongly tend to converge on the actual results (as demonstrated above)

** I use the term to mean uneducated in the subject matter. This doesn't mean stupid; it isn't an insult.

The deterministic method doesn't run the risk of getting a crappy sample at all, and doesn't require a ton of time to run the number of simulations needed to reduce that risk. However, programming the algorithms to enumerate all the possibilities and their respective frequencies requires more "biological processor" power (e.g. effort) than I'm willing to commit.

Beyond that, I deal with simulations all the time, so its more in my wheelhouse anyway.

tl;dr JJR's results appear to be legit, just gotten in a different method than I used. The remainder is an explanation of why I chose my method.

The mathematical analysis is spot on. Which is why the RULES suggest focusing on characteristics to begin with. Further, the racial differences exacerbate this point. Fundamentally, what it seems is that the OP has an issue in the way that the game has been designed.

Sorry, but I don't think that the core basis around which the game itself has been balanced can be easily tweaked such that an emphasis on skills over characteristics can be a more mathematically advantageous method of XP distribution (both at creation and over the course of campaigns).

You do know that no matter how solid your math is, it is still a matter of trying to fix a completely avoidable problem with an imbalanced houserule?

I'm sorry you feel negatively about it, but the proposed house rule has not been tested by the community yet so its imbalanced-ness can be argued either way.

Also, letting everyone respec once is a fair houserule that I think most everyone on the forum has agreed with.

I'm pretty sure though that FFG did playtest how it would go. I mean I assume they did their due diligence when designing the rules and thus weighed how the game would be affected by it, hence their eventual recommendation that you spend xp on traits during char gen. I mean I can't know for sure but it would be the logical thing to do when building your basic rules.

Well, here's the numbers for advanteges (just had to swap the probabilities on the dice in the code )

P(Advantage > 0 | GGPP)
0.2856

P(Advantage > 0 | GYPP)
0.3206

P(Advantage > 0 | YYPP)
0.3567

P(Advantage > 0 | GGGPP)
0.4573

P(Advantage > 0 | GGYPP)
0.4913

P(Advantage > 0 | GYYPP)
0.5248

P(Advantage > 0 | YYYPP)
0.5938

P(Advantage > 0 | GGGGPP)
0.6070

P(Advantage > 0 | GGGYPP)
0.6358

P(Advantage > 0 | GGYYPP)
0.6634

P(Advantage > 0 | GYYYPP)
0.7163

P(Advantage > 0 | YYYYPP)
0.7545

To calculate advantages:

120xp
Skill Character
6 skills at YY, = 6 x 0.3567
6 skills at GY, = 6 x 0.3205
21 skills at GG. = 21 x 0.2856
/ 33 = 0.3049 chance of at least one advantage

120xp Characteristic Character
6 skills at GGY, = 6 x 0.4913
12 skills at GGG, = 12 x 0.4573
15 skills at GG. = 15 x 0.2856
/ 33 = 0.3854 chance of at least one advantage

---

220xp
Skill Character

A)
6 skills at GYY, = 6 x 0.5248
1 skill at YY = 1 x 0.3567
5 skills at GY, = 5 x 0.3206
21 skills at GG. = 21 x 0.2856
/ 33 = 0.3365

B)
12 skills at YY, = 12 x 0.3567
4 skills at GY, = 4 x 0.3206
17 skills at GG. = 17 x 0.2856
/ 33 = 0.3157

C)
6 skills at YY, = 6 x 0.3567
16 skills at GY = 16 x 0.3206
11 skills at GG = 11 x 0.2856
/ 33 = 0.3155


220xp
Characteristic Character

A)
4 skills at GYY = 4 x 0.5248
7 skills at GGY = 7 x 0.4913
7 skills at GGG, = 7 x 0.4573
2 skills at YY = 2 x 0.3567
13 skills at GG.= 13 x 0.2856
/ 33 = 0.3990

B)
14 skills at GGY, = 14 x 0.4913
4 skills at GGG, = 4 x 0.4573
5 skills at GY = 5 x 0.3206
10 skills at GG.= 10 x 0.2856
/ 33 = 0.3990

C)
6 skills at YYY, = 6 x 0.5938
12 skills at GGG, = 12 x 0.4573
15 skills at GG. = 15 x 0.2856
/ 33 = 0.4041


---

400xp
Skill Character

A)
6 skills at GGYY, = 6 x 0.6634
6 skills at GYY, = 6 x 0.5248
4 skills at GY, = 4 x 0.3206
17 skills at GG. = 17 x 0.2856
/ 33 = 0.4020

B)
12 skills at GYY, = 12 x 0.5248
16 skills at GY, = 16 x 0.3206
5 skills at GG. = 5 x 0.2856
/ 33 = 0.3896

C)
15 skills at GYY, 15 x 0.5248
2 skills at GY, = 2 x 0.3206
16 skills at GG. = 16 x 0.2856
/ 33 = 0.3964


400xp
Characteristic Character

A)
12 skills at YYY, = 12 x 0.5938
6 skills at GGY, = 6 x 0.4913
1 skill at GY, = 1 x 0.3206
14 skills at GG. = 14 x 0.2856
/ 33 = 0.43613

B)
6 skills at GYYY, = 6 x 0.7163
4 skills at YYY, = 4 x 0.5938
2 skills at GGY, = 2 x 0.4913
6 skills at GGG, = 6 x 0.4573
15 skills at GG.= 15 x 0.2856
/ 33 = 0.4450



----

TL;DR:

At character creation,
Skill human gets advantages 30.5% of the time
Characteristic human gets advantages 38.5% of the time

100xp in
Skill Human 32% advantages
Characteristic Human 40% advantages

400xp in
Skill Human 39% advantages
Characteristic Human 44% advantages

Yeah, the problem is, at least in my view, that you can't report advantage results and success results independently, as you've done here. Go back to my data posts, and you'll see the correlation between the values are strongly, negatively correlated (r ~ -0.55).

The way you've presented the data above, it would infer, for example, a 100 XP Skill human gets, on average:

• > 0 Advantage, 32% of the time
• > 0 Success 49(ish)% of the time

And, while these numbers are fine separately , they are crap together . Since they are so strongly correlated, you get much fewer advantages on a success than you would on a failure.

This phenomenon is actually the reason I started simulating the dice results; Back in EotE Beta my players kept rolling huge #'s adv on failed rolls, and were like "WTF does this mean? YTF does this keep happening?".

To calculate triumphs:

P(Triumph > 0 | Y) =
0.0833

P(Triumph > 0 | YY) =
0.1597

P(Triumph > 0 | YYY) =
0.2297

P(Triumph > 0 | YYYY) =
0.2939

120xp
Skill Character
6 skills at YY, = 6 x 0.1597
6 skills at GY, = 6 x 0.0833
21 skills at GG. = 21 x 0.0
/ 33 = 0.0442 chance of at least one triumph

120xp Characteristic Character
6 skills at GGY, = 6 x 0.0833
12 skills at GGG, = 12 x 0.0
15 skills at GG. = 15 x 0.0
/ 33 = 0.0151 chance of at least one triumph

---

220xp
Skill Character

A)
6 skills at GYY, = 6 x 0.1597
1 skill at YY = 1 x 0.1597
5 skills at GY, = 5 x 0.0833
21 skills at GG. = 21 x 0.0
/ 33 = 0.0465

B)
12 skills at YY, = 12 x 0.1597
4 skills at GY, = 4 x 0.0833
17 skills at GG. = 17 x 0.0
/ 33 = 0.0682

C)
6 skills at YY, = 6 x 0.1597
16 skills at GY = 16 x 0.0833
11 skills at GG = 11 x 0.0
/ 33 = 0.0694


220xp
Characteristic Character

A)
4 skills at GYY = 4 x 0.1597
7 skills at GGY = 7 x 0.0833
7 skills at GGG, = 7 x 0.0
2 skills at YY = 2 x 0.1597
13 skills at GG.= 13 x 0.0
/ 33 = 0.0467

B)
14 skills at GGY, = 14 x 0.0833
4 skills at GGG, = 4 x 0.0
5 skills at GY = 5 x 0.0833
10 skills at GG.= 10 x 0.0
/ 33 = 0.0480

C)
6 skills at YYY, = 6 x 0.2297
12 skills at GGG, = 12 x 0.0
15 skills at GG. = 15 x 0.0
/ 33 = 0.0418


---

400xp
Skill Character

A)
6 skills at GGYY, = 6 x 0.1597
6 skills at GYY, = 6 x 0.1597
4 skills at GY, = 4 x 0.0833
17 skills at GG. = 17 x 0.0
/ 33 = 0.0682

B)
12 skills at GYY, = 12 x 0.1597
16 skills at GY, = 16 x 0.0833
5 skills at GG. = 5 x 0.0
/ 33 = 0.0985

C)
15 skills at GYY, 15 x 0.1597
2 skills at GY, = 2 x 0.0833
16 skills at GG. = 16 x 0.0
/ 33 = 0.0776


400xp
Characteristic Character

A)
12 skills at YYY, = 12 x 0.2297
6 skills at GGY, = 6 x 0.0833
1 skill at GY, = 1 x 0.0833
14 skills at GG. = 14 x 0.0
/ 33 = 0.1012

B)
6 skills at GYYY, = 6 x 0.2297
4 skills at YYY, = 4 x 0.2297
2 skills at GGY, = 2 x 0.0833
6 skills at GGG, = 6 x 0.0
15 skills at GG.= 15 x 0.0
/ 33 = 0.0747



----

TL;DR:

At character creation,
Skill human gets triumphs 4.4% of the time
Characteristic human gets triumphs 1.5% of the time

100xp in
Skill Human 6% triumphs
Characteristic Human 4.5% triumphs

400xp in
Skill Human 6.8-9.9% triumphs (0.08143 average over three builds)
Characteristic Human 7.5-10% triumphs (0.08795 average over two builds)
(basically the same)

I'm pretty sure though that FFG did playtest how it would go. I mean I assume they did their due diligence when designing the rules and thus weighed how the game would be affected by it, hence their eventual recommendation that you spend xp on traits during char gen. I mean I can't know for sure but it would be the logical thing to do when building your basic rules.

Or they just knew that it would be a pain in the ass to boost your attributes after the fact, having to buy your way slowly down to the bottom of a tree to get your +1.

Yeah, the problem is, at least in my view, that you can't report advantage results and success results independently, as you've done here. Go back to my data posts, and you'll see the correlation between the values are strongly, negatively correlated (r ~ -0.55).

The way you've presented the data above, it would infer, for example, a 100 XP Skill human gets, on average:

• > 0 Advantage, 32% of the time
• > 0 Success 49(ish)% of the time

And, while these numbers are fine separately , they are crap together . Since they are so strongly correlated, you get much fewer advantages on a success than you would on a failure.

This phenomenon is actually the reason I started simulating the dice results; Back in EotE Beta my players kept rolling huge #'s adv on failed rolls, and were like "WTF does this mean? YTF does this keep happening?".

Absolutely - great point they are absolutely dependent, that's probably the limit in my knowledge in probability though right now - I'm not entirely sure how to calculate those correlations.

So (ignoring the dependecies between advantages, successes, and triumphs that LethalDose pointed out),

A Characteristic-focused character succeeds 10 percentage points more often, gets advantages 8 percentage points more often, and gets triumphs 2 percentage points less often

Comparing the percentages,

Going Characteristics will increase your character success rate by up to 33% , down to 5% at 400xp
(53/40-1 = 0.325, 55/49-1 = 0.11, 59/56-1 = 0.05)

It will increase advantage rate by up to 25% , down to 11% at 400xp
(38.5/30.5-1 = 0.26, 40/32-1 = 0.25, 44/39-1 = 0.11)

It will decrease triumph rate by up to 65% at character creation, decrease it by 25% 100xp in, and they'll be the same at 400xp
(1.5/4.4-1 = -0.65, 4.5/6-1 = -0.25)

It's all a false argument because it doesn't take into account advantages and triumphs, both of which have significant impact on the overall effect of a given dice roll. A single triumph can activate a crit, which under certain conditions could disable a target and end a confrontation with a single success rolled.

He's fixing something that isn't broken and doing it without taking a look at the mechanical effects of a total dice pool.

For both advantages and triumphs, these increase with yellow dice used. And the characteristic character systematically has more yellow dice than the skill character so this means characteristic characters get more triumphs and advantages.

(1Y means 8% triumph, 2Y means 16% triumph, 3Y means 23% triumph)

EDIT: I take that back - the skill character has 12 more yellow dice at character creation and ~4 more yellow dice than the characteristic character at 100xp, but at 400xp the characteristic character has more yellow dice.

So maybe the lack of success rate is made up for in the larger probability of triumphs?

Still at 400xp and beyond, the characteristic character will have more triumphs and a higher success rate.

Add to that, the characteristic character has more dice in general (G and Y) and thus more advantages.

So the bigger stronger person, or smarter person, or more charming person with equivalent training is better? I am failing to see how this differs from real life.