During my weekend gaming session we had an issue with failure and opposed tests.

A player had attempted a commerce test as part of an acquisition and failed with four degrees of failure, the GM also failed with four degrees of failure.

The rules state that when an opposed test is failed by both parties then the test is re-rolled or nothing happens.

However the rules also state that when a player fails a commerce, test during an acquisition, by at least three degrees they take a minus ten penalty on the acquisition test.

Which rule takes precedence in this situation?

Is the test count as a failure and the player takes a penalty?

or

Is the test considered a stalemate and the test re-rolled/nothing happens?

As I see it,

The first check/roll did not fail per se, as there was no determined winner.

However the "fail" by 3 DoS penalty kicks in.

Then,

The -10 penalty gets added to the reroll, as you still need to determine who won as per the rules for opposed tests.

I'm more inclined to issue a re-rolled test. If the penalty applies after a "test that resulted in failure", then a "test that resulted in a stalemate" shouldn't apply the penalty. As there is no clear winner, there can be no clear loser. As there is no clear loser, there can be no "failure". That is to say, the character failed his roll with 4 Degrees of Failure, not his test . By my reading, anyway.

The way we play it is that for opposed rolls we look at the difference between the two rolls as the final DoS rather than what the player rolled.

So in the above example it's a tie (we do a stat check in ties to pick the winner, who then gets 1 DoS), so the failure condition doesn't kick in.

In another example, The player gets 1 DoS but the NPC is fantastic and gets 4 DoS. That's a comparative 3 DoF, so the failure condition kicks in.

I agree with Naviward. Opposed tests follow a different mechanic than regular tests. Any opposed test you compare your rolls and your success or failure is determined by the comparison: tally up the DOS and DOF, this becomes your end result, not the test itself . Not the action roll itself. So in your listed example Where both parties had four degrees of failure that would be a tie. Neither party got and success over the other, which means they did not fail the Opposed test by three or more degrees. However, the tie should be settled by whoever has the higher commerce skill stat: example whatever the fellowship or intelligence (I forget what its based off of) is higher would win. Even if its only by a single point. If they are happen-chance the same, either reroll the test or have them do the DH 1 system and just roll a d10, highest wins.