[Typhus]

One thing I am fuzzy on in 6E (probably due to the final text) is the intended way to do tests with thresholds vs when to apply a modifier.

I see the Threshold Guidelines, and that makes sense enough.  However, in days gone by complications would also warrant a modifier of their own.  In 6E there's no analogous process directly.  In combat it's an assessment of Edge.  Status effects apply modifiers, as does vehicle speed, etc.  However, there's no guidance for the GM on when to apply modifiers outside those cases in point.

Was there an intent to handle general difficulty through thresholds with modifiers **only** as indicated by those cases?  The examples also indicate my thresholds should be higher for shooting a target at a farther range, but of course that's not in the combat process at all.

Or is this another case of Editor Sweeny and his bloody pen?  Do we have any ability to know what the RAI is here?

For me, I assume that, outside of combat, a GM would set a threshold based on how difficult the task would be without miscellaneous complications, followed by an assessment of any outside complications that make it harder to do.  Things like Status effects, injury and speed mods make it seem like that's the right way to go.  Does that sound reasonable?

[Noble Drake]

Outside of the few explicitly mentioned added or subtracted dice from a pool, everything that used to be a dice pool modifier (or target number modifier in the editions long before) is meant to be boiled down into the Edge system: gain it if you've got what would otherwise provide significant bonuses, generally not gain it, and have your opposition gain it if you're in a situation that would otherwise apply significant penalties.

SR6 just doesn't do the granular pluses and minuses that Shadowrun used to be chock full of. So if you are GMing and you feel like a particular thing just really needs to have a modifier to it that isn't covered and doesn't mesh with the Edge system, adjust the Threshold.

[Stainless Steel Devil Rat]

Modifiers aren't completely replaced.  For example, Noise is still a -rating in dice penalty.  Wounds still apply a dice pool penalty.  In fact, you now suffer damage/wound penalty for matrix damage to devices now!  (i.e. if you data spike a gun, it's no longer a binary outcome.  Even if you fail to brick the device, you can still end up inflicting dice penalties when it's used)

[Leith]

The difference between threshold and dice mods is easy to explore. Subtracting 3 dice is equivalent to adding one to the threshold. Hence modifying the threshold is very similar modifying dicepools in terms of how it affects the probability of success, youre just doing it in 3 die chunks. However changing the dice pool changes the upper limits of the roll; rolling 5 dice means you can get a max of 5 hit while rolling 3 means a max of 3.
The dice pool modifier that remain are mostlikely just holdovers that didnt fit into edge but no one wanted to do it in chunky mods of 3 by adding a threshold or wished to avoid the complexity of having thresholds involved in an opposed roll.
The reason why these modifiers don't fit into edge is because they are, hypothetically, infinitely cumulative. Edge is capped in several different ways so you can't have a speed interval adding or subtracting or giving away edge with each iteration. But, for whatever reason the designers of these mechanics felt that limiting (or increasing) the potential hits, and the granularity of small modifiers was the way to go. Probably this is because they are just used to thinking this way even though much of SR6 does not follow this design philosophy.

Tl;dr: designers put in many ways to alter die rolls and adjust difficulty, but couldn't agree on what was best, nor bother to discuss which method suited which situation. Shadowrun books give you the rules, they don't tell you how to play.

[skalchemist]

For me, I assume that, outside of combat, a GM would set a threshold based on how difficult the task would be without miscellaneous complications, followed by an assessment of any outside complications that make it harder to do.  Things like Status effects, injury and speed mods make it seem like that's the right way to go.  Does that sound reasonable?

I think it is reasonable, but I think Leith above does a good job of explaining what the pros and cons might be.

For me, I think I would stick to adjusting the Threshold for anything that was not "hard coded" into the rulebook as a dice bonus/penalty or similar.  If whatever is happening doesn't feel like a big enough deal to adjust the threshold, then my instinct is that its not a big enough deal to matter.  This is one reason why the Threshold table on page 36 seems like the key mechanic to me.  It feels to me like any possible set of circumstances can be encoded into that table, and then you get on with the dice rolling.

[Xenon]

Modifying thresholds is not really an established mechanic currently being used for opposed tests. When you are opposed the idea is rather that you use the edge mechanic (and/or that you apply situational modifiers).

But modifying the threshold is something you are expected to do when it comes to simple tests (that are not opposed) as well as extended tests.