1. Hello and welcome to the brand new home for PlayDota!
    Please read through our Welcome thread to see what's new!
    Dismiss Notice

Max attack speed?

Discussion in 'Advanced Mechanics' started by mapdesigner, Dec 13, 2016.

  1. mapdesigner

    mapdesigner Member

    2,040
    97
    48
    Aug 11, 2011
    Hello,

    What is max attack speed for drow?

    I made 4 moonshards + eat 1 + ulti should be 600 attack speed, with 0.283 seconds per attack, however this is the combat log I have

    [21:33.07] Drow Ranger hits Centaur Warrunner for 26 damage (6007->5981).
    [21:33.37] Drow Ranger hits Centaur Warrunner for 27 damage (5986->5959)
    [21:33.67] Drow Ranger hits Centaur Warrunner for 25 damage (5964->5939).
    [21:33.94] Drow Ranger hits Centaur Warrunner for 27 damage (5944->5917).
    [21:34.24] Drow Ranger hits Centaur Warrunner for 25 damage (5921->5896).
    [21:34.50] Drow Ranger hits Centaur Warrunner for 24 damage (5899->5875).
    [21:34.80] Drow Ranger hits Centaur Warrunner for 25 damage (5878->5853)
    [21:35.07] Drow Ranger hits Centaur Warrunner for 25 damage (5858->5833).

    if you notice, many times time between attacks is 0.3, but sometimes it is 0.27, and one time it was 0.26

    can someone explain? is this all just rounding or is it that front swing + back swing of drow > attack time for certain animations. is this even possible?


    EDIT: i even tested with 100 attack speed

    38:50.48] Drow Ranger hits Centaur Warrunner for 13 damage (6029->6016).
    [38:52.19]
    [38:53.90]
    [38:55.57]
    [38:57.27]
    [38:58.98]
    [39:00.65]
    [39:02.36]
    [39:04.03]

    1.71, 1.67, 1.70, 1.71, 1.67, 1.71, 1.67

    unliss interface is fucking up with me (it says attack time is 1.69 seconds which is the average between the two o_O

    EDIT: could it be that the projectile speed can differ from shot to shot?
     
    Last edited: Dec 13, 2016
  2. mapdesigner

    mapdesigner Member

    2,040
    97
    48
    Aug 11, 2011
  3. Jabutosama

    Jabutosama Member

    264
    98
    28
    Apr 15, 2015
    You can get from her page the following;
    Projectile Speed 1250
    Attack Animation 0.7+0.3 (attack time + backswing)​

    this makes up 1 attack per second.
    With maximum attack speed, it should be 3.53 attacks per second or 0.28 seconds per attack.

    if it differs, I don't know. At least base damage is randomly around 12% weaker or stronger.
     
  4. duble_dragon

    duble_dragon Member

    898
    48
    28
    Jul 10, 2009
    Projectile speed is factored in, but since it's a constant it shouldn't be the culprit changing your attack speed timings. Since your numbers are exactly .3, .27, and a single .26, I'd say it's definitely something else because projectile speed would only make a difference if the target is moving.
     
  5. mapdesigner

    mapdesigner Member

    2,040
    97
    48
    Aug 11, 2011
    Target was not moving.

    Further evidence proving that this has nothing to do with projectile speed is: that the between main attack hitting and splinter attack hitting the neighbouring target is always the same depending in the position of the target (so if the neiboughr was farthest away from main target, it will always receive the damage 0.34 seconds after main target;

    since distance between Neighbour and main target is constant, and time is constant, Speed is therefore constant
    + since distance from drow is constant, projectile speed is constant, but time is different, then there are 2 possibilites:

    either projectile path is different
    or
    time between attack is not constant
     
  6. EebstertheGreat

    EebstertheGreat Forum Manager Staff Member

    10,745
    79
    48
    Sep 3, 2009
    A third possibility is that the timers are inaccurate or at least imprecise, which I think is the most likely explanation.
     
  7. mapdesigner

    mapdesigner Member

    2,040
    97
    48
    Aug 11, 2011

    But how would this explain that The 2 times shown at the combat log when averaged will result in an average BAT of 1.69 seconds per attack (and it was shown 1.69 soconds per attack too)

    PS im not sure when I did the test, but I think it is after the patch
    EDIT nvm it was tuesday
     
  8. Oesile

    Oesile Member

    1,341
    279
    83
    Oct 8, 2016
    It could be because of rounding down numbers making Drow appear to have a minimally smaller BAT, perhaps. Considering it's so close to the correct value I think this is probably the case.
     
  9. roadto7kmmr

    roadto7kmmr New Member

    6
    2
    3
    Jan 21, 2017
    maybe ice frog did this on purpose. just like westworld. imperfection creates the greates XD
     
  10. EebstertheGreat

    EebstertheGreat Forum Manager Staff Member

    10,745
    79
    48
    Sep 3, 2009
    @Bu3ny
    I'm genuinely confused about what's going on here.
     
  11. mapdesigner

    mapdesigner Member

    2,040
    97
    48
    Aug 11, 2011
    I think I have the answer. It is that attack point is not always the same: one unit can have different attack animations each with different fron swings (I read this somewhere in wiki or the other forums)(the bug was talking about enchantress + alchema + PL)

    could this be it?

    but then the question rises: is the front swing randomly decided or is it just looping one by one?
     
  12. Blarrg

    Blarrg Member

    10,327
    330
    83
    Jun 12, 2009
    In Dota2 not every attack is performed the same. The best example is natures prophet, where his attack point I believe is constant, but his projectile will leave from a different point (sometimes about 100 range in front of the hero when he does the poking motion, sometimes from the model position when he does the swinging motion). This can have a small effect on attack timings.
     
  13. mapdesigner

    mapdesigner Member

    2,040
    97
    48
    Aug 11, 2011
    LOL I didnt know that xD

    ok then we have 2 reasons. Do you know if this happens to drow too?
     
  14. Abstract.Raze

    Abstract.Raze Member

    657
    84
    28
    Mar 14, 2012
    Looking how Icefrog created this game, he must be a genius, imperfection is simultaneously beyond and below perfection self, Icefrog is an artist.
     
  15. Bu3ny

    Bu3ny Moderator Staff Member

    8,217
    72
    48
    Dec 24, 2010
    You can ignore the +- 0.03/0.04 timings, they are caused by server ticks. If things don't perfectly align with the ticks, then you have these tiny timing variations.

    The game runs with 30 ticks per second, so everything can happen at a 0.03 seconds interval minimum. Attack speeds, when precisely calculated, have usually more than 2 decimals. Like in this example, 0.283 seconds per attack would translate to 0.27-0.3 seconds per attack.

    Another thing is (doesnt matter in this case), projectiles are launched from certain points on units. So if a unit has 2 different animations, it might be that the projectile travels a tiny bit longer with one animation than with the other. Good examples are Furion and Rubick, who have multiple animations, and with one of them, the projectile has to fly for longer.

    And yet another factor is target animations. Projectiles hit a unit when reaching a certain point of their model. Usually, it's their chest. So if a target moves his projectile target spot a lot, this may also cause a +-0.03 seconds.
     
    Last edited: Jun 18, 2017
    IFeelPrettyLow likes this.