+ Reply to Thread
Results 1 to 3 of 3

Thread: 200% Threat Threshold Testing

  1. #1
    Join Date
    Dec 2006
    Location
    Tacoma, Wa
    Posts
    8,766

    200% Threat Threshold Testing

    Here's a write-up on our testing process:


    The first thing we do when testing is find repeatable patterns that explain behavior. Since we want our test subjects so stay alive, we used level 1 vendor-bought pistols while in melee range. We made sure we were dealing the same damage, and did a series of tests with this.

    1) Player 1 fired one shot. Player 2 fired three shots and pulled aggro.
    2) Player 1 fired two shots. Player 2 fired five shots and pulled aggro.
    3) Player 1 fired three shots. Player 2 fired 7 shots and pulled aggro.

    (On recounting this, my description that it's '200% or greater' is incorrect. It cannot be equal to 200%, it must be greater. I will amend this.)

    Next, we need to narrow down our numbers. We see there's a pattern, but we can't say anything definitive.

    To do this, we move one player to a substantial range, which will cause them to deal less damage with their pistol.


    Player 1 deals 82 damage; Player 2 deals 168 over the course of 10 shots, then pulls aggro.
    Player 1 deals 82 damage; Player 2 deals 170 over the course of 10 shots, then pulls aggro.
    Player 1 deals 82 damage; Player 2 deals 170 over the course of 10 shots, then pulls aggro.
    Player 1 deals 82 damage; Player 2 deals 173 over the course of 9 shots, then pulls aggro.
    Player 1 deals 82 damage; Player 2 deals 170 over the course of 10 shots, then pulls aggro.

    There is some variation in Player 2's damage due to the target shifting and changing range and due to the player critting.

    In each case, the final shot puts him over the 200% threshold, which establishes a regular pattern.


    At that point, we start expanding the scope.

    Do different weapon types bring different results? We switch to shotguns, then rifles, with no change.

    Do different weapon levels or damage types bring different results? We test using electric, sonic, and incendiary, the result stays the same.

    Do different enemy types bring diferent results? We test on fithiks, howlers, boargar, warnets, and kael, the result stays the same. We also test on Kenilaxx, a named boss with very fast respawn in a reliable and clear area. We originally tested on Treebacks and Magmonix, but due to the encircling/shifting that most creatures do, but takes 8-10 seconds for these things, we get mad and we give up on them.

    The next three subjects are a bit harder to parse since they could all produce differing results: level, class, passive abilities.

    In our initial tests, we used two Commandos of equal level with Firearms.
    We then test using one Commando and one Specialist with Firearms.
    We then test using on Guardian (level 46) and one Commando (level 27), where the Guardian does not have Firearms.

    The biggest issue came right there. Originally, we had casually checked the aggro threshold and it came out to 300%. The reason is that we had not let Threat fully disappear before starting our test, and instead of two test shots from the person with aggro, there were three still being counted. This is what took me 3x the damage. This is also why we had nothing conclusive until doing a more scientific test last night.

  2. #2
    Join Date
    Dec 2006
    Location
    Tacoma, Wa
    Posts
    8,766
    And, credit to the people who helped out, Velhym, Klor, Dietran, and Rikethe.

  3. #3
    Join Date
    Jan 2008
    Location
    San Diego, CA
    Posts
    866
    I'm useful.

+ Reply to Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts