I had two odd ones for 1200 and 1300 dmg... crits I'm guessing
Crit rate doesn't vary with TP. It's not a crit-based weaponskill. Therefore they weren't crits. DA, maybe, though more likely standard pDif variance, but definitely not crits.

Stacking int brought it down? hmm
More controlled tests are needed before you can say for sure what was going on.

With 485 attack (what you had on the lower of the two), you will already be capping attack on those mobs, so the additional attack in the second test will not make any difference.

With only 10 samples you can expect a 10% margin of error on the average itself, for -each- test, assuming all other variables were controlled. So, the first test implies an average between 900 and 1100. The second test implies an average between 794 and 970.

Unfortunately there's one other significant variable that's not properly accounted for. This is a 4-hit weaponskill, and you can thus expect an average of 2 missed swings for every 10 weaponskills. Given the numbers of the weaponskills themselves, you can expect that for every missed swing in one set that didn't happen in the other, that the average would shift by about 25 points. It's entirely possible for there to be a 100 point swing in average due to this factor.

In addition, your str changed in the second test. Given that you were only barely above the limit for capped fStr on the lvl 0 mobs, you can say that your fStr is basically guaranteed to have dropped 2 points in the second test set, which will further affect the average.

Overall, you do not provide nearly enough data to make any reasonable conclusions about the weaponskill, except the vague general area of damage you're likely to achieve while using it.

The weaker mobs gave less damage overall.
It's a 4-hit weaponskill. On starter mobs, only the first hit (or maybe second if you whiff the first hit) is ever going to connect. Thus it will necessarily always have lower damage than on mobs where all hits connect.

However it also makes those tests more consistant (we can be sure that only one hit landed each time). You can say that adding 60 int increased your expected damage by somewhere between 0% and 39%.