On my Monk, I'm trying to figure out if I would rather do a Dodge build with Dodge-centered Passives and Mantras, or if I should go for more armor and healing. In my mind, it comes down to just how random Dodge is.

If I have a 50% dodge rate, which of these is true?

  1. Every time I am attacked, I have a 50% chance to Dodge the attack.
  2. Incoming attacks will be Dodged at a consistently 50% rate. Past results affect future results - eg, I have not Dodged in a while, I will be more likely to Dodge.

Essentially I am asking if the Gambler's Fallacy is true for Dodge chance.

Both case #1 and case #2 would average out to the same thing in a long fight, but in a short burst case #2 is preferable.

In case #1, a string of bad randomness will mean I could take several hits in a row without Dodging any of them (Dodge, Dodge, Dodge, Hit, Hit, Hit, Hit, Hit, Hit owww my spleen). If that happens, my Dodge wasn't really much use to me, as a bunch of unfortunate die rolling got me killed.

In case #2, having 50% Dodge chance is nearly equivalent to 50% damage mitigation, as attacks would essentially alternate (Dodge, Hit, Dodge, Hit, Dodge, Hit, Dodge, Hit). I can expect that I will consistently avoid incoming attacks, and don't have to curse Math.Rand() for getting me killed.

So, which is it? Can I expect consistent results from my Dodging, or is it all chance?

  • I know case 2 was used in Warcraft 3, another Blizzard game, for figuring out certain proc effects. It wouldn't surprise me if it was used again, but I can't say for sure. – Invader Skoodge May 23 '12 at 14:13
  • @StrixVaria in WoW case 1 is used. – Blem May 23 '12 at 14:20
  • @Blem since you mention WoW, I'll mention that at some patch case 2 was implemented for several aspects of the game (most prominently quest item drops)... just some trivia. Mag Roader: This is Diablo, so it's all about random. – Edurne Pascual May 23 '12 at 14:23
  • Case 2 is known as "pseudo-random". "The generation of random numbers is too important to leave to chance." – Raven Dreamer May 23 '12 at 14:25
  • 1
    @Raven: All randomness in a computer is technically pseudorandom, so the answer to the new question is necessarily "it's pseudorandom." This is not what OP was asking, though, so the original term "consistent" was a much better choice. – BlueRaja - Danny Pflughoeft May 23 '12 at 17:45

Since one of the Monk's achievements (Like Water) is to dodge 15 attacks in a row, I would say #1 is accurate: each time you are attacked you have a 50% change to dodge

This can further be backed up by many dodge-specced monks who can testify to dying many times due to a string of bad luck :)

  • I had this feeling yesterday fighting Belial as Monk. I died once after a really unlucky streak of dodge not proccing. – Drake May 23 '12 at 14:23
  • Other reasons to guess that case #1 is true is that from a programmer's perspective, case #1 is far easier to implement, and is far less likely to have math bugs that accidentally make players nigh-invulnerable, or walking pincushions! – Wikwocket May 23 '12 at 14:38
  • @Wikwocket, I doubt implementing either method would cause a first year college student any problems. – oscode May 23 '12 at 14:47
  • 1
    @oscode Have you met many first year college students? Besides, after seeing WoW Rogues achieve 100% dodge through creative use of game mechanics, I can imaging Blizzard being conservative here. :) – Wikwocket May 23 '12 at 15:29
  • Possible implementation: count # of incoming attacks and dodged attacks. If count is 0, randomly dodge, otherwise take #dodges/#attacks, and either dodge or not depending on whether the result would be closer to or farther away from your dodge%. If desired, reset the values occasionally, say after 5 seconds of no incoming attacks. – Mag Roader May 23 '12 at 17:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.