Munin: That always seems extremely unfair, as whoever gets to roll the move and spend the hold has a massive advantage in terms of raw numbers.
Trasher has Hard+3, Hugh has Hx+2 with Trasher. Trasher goes seizing on Hugh trying to take his shit, Hugh doesn't want him to and resists.
AW1 using your example
Hugh has a 83.33% to hit and give Trasher a -2.
Trasher therefore has a probable 72.22% to hit anyway. That's a decrease from 91.67%.
This means Trasher can be pretty well expected to not only be successful in taking the shit, but also choose more hold to punish Hugh for resisting. He's got a 27.78% chance to choose two more punishments instead.
This roll doesn't account for Hugh's abilities at all (only his weapon and armor), and pretty much guarantee's he who rolls wins in PvP. This is definitely not an equitable exchange, and I'm actually very surprised your players let you get away with that. This way to handling the problem is even more problematic in AW2.
AW2 using your example
Hugh has a 83.33% to give -1 and a 41.67% to give -2.
Trasher probably still has a 72.22% but instead of being able to assume -2, now there is a large chance of only -1.
This means that interfering is actually less imposing then it was in AW1 by a small margin, and now with the roller able to pick one on a miss, the PvP is guaranteed to give Trasher what he really wants (at no other cost). A one sided seize wasn't at all fair before, it's definitely not fair now.
That's a problem. What works Best, since both players do have agency in that conflict, is that the instigator says what he wants to do and the other says how he plans to stop it. This isnt people talking over each other, this is people establishing the scene through conversation. If Hugh wanted to run away, then we're not at seize by force yet. We're cat and mousing. If He instead decides to stand and fight Trasher off, okay, both can roll a seize by force and both interfere with the other (there might be a better battle move now not sure, either way:)
AW using opposed rolls
Hugh has a +2 hard, +2Hx with Trasher.
Trasher +3 Hard, -1Hx with Hugh.
Now if they roll all the rolls (hardly takes a second to do, really!) They each get to first limit the other's successful hard depending on how well they know each other.
For Hugh it's: 83.33% of at least -1, 41.67% of -2.
For Trasher it's: 41.67% of at least -1, 8.33% of -2.
This now has a range of possibilities. I'm not doing the math for each, or collectively right now. (I will if you think you want that). But essentially it's +3 vs +2 hard rolls with this extra modifier. Much more likely to come out that Trasher doesnt absolutely get Hugh's shit, but they both are probably black and blue, and Hugh significantly less then previous examples. But it is equally clear that Trasher is going to have a much harder time predicting Hugh, and that the advantage of Hard is more likely to be equal or with Hugh advantaged instead. This makes Hx far more important then a mere bump of -few% to win.
This establishes the character's effective rolls. Now they each roll seize to find out how many holds they spend, and they prioritize what they want. It's clear they counter each other in options which is fine. Since this is mostly to see who has what advantages and then exchange harm and decide who has possession if either of them do. Which sets you up to make a move and describe how those rolls play out as an arbiter. The group should decide how the spend the hold: out loud first one then the other, taking turns hold by hold, the winner declares first, or if both declare secretly, etc.
EDIT::
As DeadmanwalkingXI says, this is a Single exchange of harm.