Part of the issue remaining is that Proficiency Dice don't offer a significant upgrade to the percent of success over Ability Dice. In other words, a character with 5 Intellect and zero points in Medicine can roll about as well (minus Triumphs) ...

Key phrase: "minus Triumphs". Based on that comment I think you're still thinking about this along only the success/fail axis and discounting the real benefits of those yellow dice. It's true they don't increase chance of success by much, however, and more importantly, they push the average roll from "success with threat" to "success with advantage". Triumphs are just gravy.

Far more interesting narrative things can be done with Advantages and Triumphs than with simple success, and this is what the proficiency dice add. The most important part of this is it lets the player take a lot more control. A Doctor and a Slicer might have the same Intellect, but a Doctor player making a Medicine roll is going to have a lot more opportunity to customize their results than a Slicer player making a Medicine roll. If your players aren't making use of this, then they're missing out on the major differentiating factor.

Anyway, I'm sure a solution can be found for you that will help mitigate the issue without having to restructure the entire game. Some things I've seen on these boards:

- different skill rank costs (above)

- players can only take specs in their careers (i.e.: no Doctor/Assassin)

- make those Talents that reduce setback count by finding reasons to add setback on a regular basis

- upgrade difficulties for untrained use of a skill. I sometimes do this if it seems like failure could be more than unusually problematic, e.g.: somebody untrained in Medicine might cause more problems than they fix; or somebody with no Mechanic skill could break something critical

- tie a rank in a skill above 1 to a minimum row in a Talent tree.

It's also in how you narrate it.

"You manage keep yourself from panicking at the sight of the blood as you tear off bits of your shirt to tie around the wound and wipe the sweat from the injured man's head. This is definitely going to need more medical attention at some point but you've stabilized him and stopped the bleeding."

versus

"You whip out your medkit, give the injured man a shot of a sedative and then use a medicinal gel to stop the bleeding and seal up the wound. In a few minutes he'll be up and around when the sedative wears off."

In the first case, you could even have the wound reopen later or just deal with it narratively--the injured person who was treated by the unskilled practitioner could have their wound be irritated or reopen without it necessarily meaning anything in terms of Wound Points or whatever.

Or you describe the doctor as being careful and attentive, the slicer standing there with the "how to" manual open on his datapad and reading through the labels. I mean, the slicer is going to read and follow that 'Emergency First Aid" book that came in the medkit a lot more effectively than the gun bunny (higher intelligence), but isn't going to accomplish much more than patching up the direct wounds according to instructions (low advantage, no triumphs).

I'll grant you I don't really like the linear skill improvement scale, nor that the non-career skills are only 5XP per rank more expensive. Shifting that might resolve the issue for you while still making the game open for character development, e.g.:

Career skill XP requirements: 5, 10, 15, 25, 40

Non career skill XP requirements: 10, 20, 30, 40, 60

I'm not sure that solves the whole issue, but it isn't a bad start. Part of the issue remaining is that Proficiency Dice don't offer a significant upgrade to the percent of success over Ability Dice. In other words, a character with 5 Intellect and zero points in Medicine can roll about as well (minus Triumphs) as a character with 4 Intellect and 2 points in Medicine. And those two points in medicine? They're supposed to represent the equivalent of a college degree in the skill.

A Triumph on a medicine roll to heal wounds can heal a crit. That alone feels enough to justify the training. Sure, anyone can slap a bandaid on a cut, but someone trained in medicine can sew up an open wound so it heals better.

Also, look at the dice. (I'm away from my dice so I'm hoping the reference sheet I'm looking at is accurate.) A green die has successes on 4 out of 8 of the faces, or 50%. A yellow die has successes (or Triumph) on 8 of the 12 sides, or 75%. The odds of success are so much greater with yellow dice over green dice. That training means something because you're much more likely to succeed over just getting advantages. (Note that rolling advantages are the same for both, 4 and 6 sides for 50%. Blue dice are also weird since 3 sides have advantages for 50% but only 2 sides have successes for 33.3%.)

I once found an EotE dice statistics spreadsheet on here and just ran some numbers. 1 green vs 1 purple is almost two-thirds chance of failure while 1 yellow vs 1 purple is closer to a 50-50 coin flip.

Dice | Failure | 1+ Success | 2+ Advantage | 2+ Threat

1G vs 1P | 65.6% | 34.4% | 4.7% | 6.3%

1Y vs 1P | 54.2% | 45.8% | 6.3% | 6.3%

I'm too far away from my old statistics classes to know what constitutes as a significant difference, but let's look at the chance of failure of the quoted example above. I did find that more green dice is usually better than less dice with some yellows. Less dice didn't get better until it was 3 yellow and 1 green as compared to the 5 green, and that's only for success chance not advantages. (Since half of all facings on green, yellow, and blue dice are Advantages, the number number of dice rolled will always determine odds for advantages. 5 green will always have better odds at rolling more Advantages than 4 yellow.)

Dice | Failure | 1+ Success | 3+ Success | 2+ Advantage | 2+ Threat

5G vs 2P | 18.9% | 81.1% | 41.5% | 51.6% | 3.5%

4G vs 2P | 27.9% | 72.1% | 27.7% | 37.4% | 6.1%

3G1Y vs 2P | 23.8% | 76.2% | 32.1% | 38.4% | 6.0%

2G2Y vs 2P | 20.1% | 79.9% | 36.8% | 39.5% | 5.9%

1G3Y vs 2P | 16.9% | 83.1% | 41.7% | 40.5% | 5.8%

4Y vs 2P | 14.0% | 86.0% | 46.8% | 41.5% | 5.8%

How about if we look at the same number of dice? At what point is it a significant improvement? Just flipping 1 green to yellow gives 3 or more success a 4.6% chance improvement. I think that's pretty decent.

Dice | Failure | 1+ Success | 3+ Success | 2+ Advantage | 2+ Threat

5G vs 2P | 18.9% | 81.1% | 41.5% | 51.6% | 3.5%

4G1Y vs 2P | 15.9% | 84.1% | 46.1% | 52.5% | 3.5%

3G2Y vs 2P | 13.2% | 86.8% | 50.8% | 53.3% | 3.4%

2G3Y vs 2P | 10.9% | 89.1% | 55.5% | 54.1% | 3.4%

1G4Y vs 2P | 9.0% | 91.0% | 60.1% | 55.0% | 3.3%

5Y vs 2P | 7.3% | 92.7% | 64.6% | 55.7% | 3.3%

What I learned from crunching these numbers is that someone without training will be better than someone with training if they are rolling more dice unless there is a large number of yellow dice. 3 Int and 4 ranks medicine (or 4 Int and 3 ranks medicine) is better than 5 Int. 1 Int and 5 ranks medicine is better than 5 Int. And that's before you calculate in the perks that rolling Triumphs would give you. Triumphs are supposed to represent the advantage training gives you over raw talent.

I guess the question becomes, at what point do you feel that it's too hard to suspend your disbelief when a character with a high stat attempts to try something that they are not trained in. A lot of people only start with a 3 in their good stats, possibly a 4 if you want to maximize one thing, which is relatively easy to make up the difference with skill training. Without cybernetics, a 6 is the best. So, yeah, in Star Wars I'm ok with a Sherlock Holmes type character with a 5 Int being able to know how to patch up someone and Slice computers because he's super smart and read about it somewhere and has a near photographic memory and remembers how to do it. The best of the best of the best can do that sort of stuff. Would it still feel as weird if we were discussing a Doctor with 1 Int and 3 ranks medicine against a Slicer with a 4 Int?

Also, I don't believe that no ranks in something means that you know nothing about it. My ideas on skills is that everyone has had some basic education which brought them up to zero ranks with no black dice, which is what I'd consider high school and time living in this universe. That's why I'm ok with everyone being able to do anything, because basic public education is worth something.

*Edited to fix dice results to be readable.

- upgrade difficulties for untrained use of a skill. I sometimes do this if it seems like failure could be more than unusually problematic, e.g.: somebody untrained in Medicine might cause more problems than they fix; or somebody with no Mechanic skill could break something critical

This I believe is actually intended, or at least it's what I do. I also sometimes change the results of a Success by giving lesser quality information or not quite as good pay-off to unskilled PCs.

However I don't think this approach would apply to the OP's concern because once a PC buys into a Skill they are no longer unskilled. Plus you don't want Players to create one dimensional PCs so you shouldn't restrict Skills. So I think the best solutions if you are finding Players buying Skills up that are outside their Career Skills rather than Talents for the skills they do have is to either increase the cost of out of Career skills (+5 to +10 each or make it exponential - Level 1= +5, level 2= +10, 3= +15 etc.) and/or requiring a talent for each level of your highest Skill.

Upon reflection I think upping the cost is the easier way to go.

So I think the best solutions if you are finding Players buying Skills up that are outside their Career Skills rather than Talents for the skills they do have is to either increase the cost of out of Career skills (+5 to +10 each or make it exponential - Level 1= +5, level 2= +10, 3= +15 etc.) and/or requiring a talent for each level of your highest Skill.

...or (and I should have mentioned this above), run the game so that the Talents don't appear to be useless. If the GM never applies setback dice, all those Talents are wasted.

So I think the best solutions if you are finding Players buying Skills up that are outside their Career Skills rather than Talents for the skills they do have is to either increase the cost of out of Career skills (+5 to +10 each or make it exponential - Level 1= +5, level 2= +10, 3= +15 etc.) and/or requiring a talent for each level of your highest Skill.

...or (and I should have mentioned this above), run the game so that the Talents don't appear to be useless. If the GM never applies setback dice, all those Talents are wasted.

Yes, but adding Setback dice takes a while to get used to because not every scene feels like a Setback die is appropriate. So the trick is to always add an element of tension (however small) that you can tie that die to.

My answer to OP is "no".

I um... got nothing else.

I consider myself pretty broad for my PC. I have at least 1 rank in 11 of the 22 General skills and 3 of the 6 Knowledge skills but only 1 of the Combat skills. I enjoy environmental factors in the games, and think PCs should have a wide variety of skills to survive the various challenges they could face. A game that focuses on the core skills of a concept will quickly marginalize the importance of other skills.

Btw, I actually track the percentages of my XP expenditures (including initial XP from character creation):

• Characteristics: 17%
• Skills: 23%
• Talents: 48%
• Specializations: 12%

This is after several months of play, and these percentages have changed over time (characteristics were about 90% of xp during creation).

The skills are the core mechanic of the game. The Talents enhance their ability or totally change what a skill can do.

Both are totally viable expenditure of XP.

IMO, it's the Talents that truly cause "wheelhouse" specialty paths. Because with the right Talents, you ignore the setback that anyone else with the same stat/skill combo would be dealing with. Or you have a Talent that allows you to do something that can't be done otherwise. Period. The Signature Abilities from the Career books being the pinnacle of that.