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

New armor mechanics

Discussion in 'Advanced Mechanics' started by HHHNNNGGG, Oct 31, 2017.

  1. HHHNNNGGG

    HHHNNNGGG Member

    17,814
    438
    83
    Sep 21, 2010
    This thread is to discuss about new armor formula.
    Previous formula: (0.06*armor)/(1+0.06*armor)
    New formula: (0.05*armor)/(1+0.05*armor)

    There is also changes with Agility:
    Previous Agility: 7 Agi = 1 armor
    New Agility: 6 Agi = 1 armor

    With this new formula armor stacking will be less effective (each armor increases physical EHP by 5% raw HP instead of 6% as before) while Agility can stack armor up faster (but the same Agility yields slightly less armor reduction than previously).

    Damage increase from armor reduction also appears weaker than before, due to the fact that each value of armor is worth less. What gets hit the most is negative armor, since the damage amp becomes weaker.

    This is quite apparent on towers, which just got buffed a lot. Their armor, in particular, gets increased to match the same damage reduction as before (claimed by Volvo) but Desolator reduces even less armor (while each armor is already worth less).

    Are there any other consequences?
     
  2. vvjacobo

    vvjacobo Member

    2,096
    21
    38
    Jun 10, 2009
    Well, -armor + armor skills slightly weaker. Which indirectly is buff to raw fisical damage items as its a bit harder to counter it.
    Armor suffered the least of changes imo. Magic resistance and debuff resistance are the big ones now imo.
     
  3. flamegod0

    flamegod0 Member

    656
    145
    43
    Sep 26, 2009
    maybe one of the extra consequences is the extra armour for each people nearby the tower (+2/heroes iirc). so even deso won't change so much with 5 people near the tower.
    kind of nice ninja buff for rat, i guess.
     
  4. HHHNNNGGG

    HHHNNNGGG Member

    17,814
    438
    83
    Sep 21, 2010
    Ah yes that too, I kind of forgot it.

    Hmm... So early physical damage gets buffed in fighting situation. Armor will kick in much later in the game, and at that time carries can have a lot of damage.
     
  5. kamukag3e

    kamukag3e Member

    3,528
    384
    83
    Jun 11, 2009
    While Armor is not actually changed that much, Strength cores are significantly buffed. With the new Strength mechanics, they receive another perk which significantly favors them.

    Sorry for derailing the thread, but can someone check a Morph with 400 agility's Movement speed now ?
     
  6. Oesile

    Oesile Member

    1,690
    349
    83
    Oct 8, 2016
    It's also worthwhile to note that while armor in general is nerfed, armor gained from stats isn't nerfed as much, due to armor per agi being buffed.

    So if you had say 42 agility before this patch, you'd have 6 armor, which is (0.06*6)/(1+0.06*6) = 0.264 or 26.4% physical damage resistance.
    Now however, 42 agility equals 7 armor, which is (0.05*7)/(1+0.05*7) = 0.259 or 25.9% physical damage resistance.
    25.9/26.4 = ~0.98 So it's nearly the same.

    If you had obtained the 6 armor due to raw armor rather than stats, you'll still have 6 armor in 7.07, and with the new formula about (0.05*6)/(1+0.05*6) = 0.230 or 23% physical damage resistance.
    23/26.4 = ~0.87 so it's quite a bit worse.

    Many lategame armor talents were removed as well, so raw armor will be not only worse but also much harder to get, which is technically a buff to physical damage.

    On the contrary, I'd say that raw armor is now a lot worse.

    I made a quick spreadsheet and a graph to show how much worse raw armor is:
    upload_2017-11-4_1-13-0.png

    I don't know how to make my shitty graph smaller, but basically I saw that raw armor is 15.87% worse when comparing 1 armor in 7.06 to 7.07, and that it's just 4.76% worse when looking at values of 50 armor (which is a pretty absurd amount of armor).
    You can replicate this by calculating the amount of physical resistance per point of armor in a spreadsheet [so (0.06*1)/(1+0.06*1), then (0.06*2)/(1+0.06*2), etc] for both the new and old armor formula. Then, you calculate (1 - New Physical resistance / Old Physical resistance). I hope someone will check this because I've always been awful with percentages.

    Anywho, if the math is correct, then raw armor is much worse than before in the early game, and your first armor item or armor talent will have essentially much lower effectiveness than before (relevant for items like Basilius/Aquila for example), while later on bigger armor items like AC will still leave most heroes at about ~20-25 armor-ish, making the armor 8.3% to 7.41% worse.

    Well, these values are a bit larger than in reality because you'll also have armor through agility, which hasn't lost efficiency.

    Lastly, I haven't made calculations with -armor because those are a pain.

    Morphling with 6 Ethereal blades and maxed agi (not counting talent as you can also select movement speed) has 433 agility.
    Sadly this means that I can't test for movement speed with boots and such, so I'll test with 397 Agi instead.

    When at 397 agi, Morph has 345 ms without boots. With boots, that goes up to 395. With boots of travel, Morphling instead reaches 469 movement speed. With the 10 agility talent, that becomes 471 movement speed.

    Given that Morph is a hero with 280 base movement speed, and that Bots gives 100 ms, you stant to gain 89 movement speed if you have 387 agility.
     
    mapdesigner, flamegod0 and kamukag3e like this.
  7. flamegod0

    flamegod0 Member

    656
    145
    43
    Sep 26, 2009
    Thank you very much for the data senpai. will check this up later. interesting graph you got there.
     
  8. Oesile

    Oesile Member

    1,690
    349
    83
    Oct 8, 2016
    Keep in mind that I'm a turd in calculating % increases or decreases :x
     
  9. mapdesigner

    mapdesigner Member

    2,795
    139
    63
    Aug 11, 2011
    Where are the numbers?

    For 42 agility:

    1) in previous patch 42/7 = it gave 6 Armor = 36% HP Increase

    2) in current 42 / 6 = 7 Armor = 35% hp increase.
     
  10. Oesile

    Oesile Member

    1,690
    349
    83
    Oct 8, 2016
    That's another way to do it.

    Here I was just using the formula and found it to be 0.98% as effective as before (so basically 35/36 = 0.97222, though I did round some numbers a bit which is why my number is a bit different).

    If you meant where are my numbers in my graph, the answer is that they are not because I only looked at raw armor;

    if we did the same thing for 84 agility:

    In previous patch this was 84/7 = 12 armor. With the old formula we'd have (0.06*12)/(1+0.06*12) = 0.41860465116 or 41.8 % physical damage resistance, or simply 12*6 = 72%

    Now it's 84/6 = 14 armor. Meaning it's like (0.05*14)/(1+0.05*14) = 0.41176470588 or 41.1% physical damage resistance, or 14*5 = 70%.

    Both numbers barely differ from each other at all, so I consider the change negligible for heroes who gain armor through stats.

    Edit: accidentally posted before finishing oops
     
    mapdesigner likes this.
  11. mapdesigner

    mapdesigner Member

    2,795
    139
    63
    Aug 11, 2011
    Mr @Oesile

    Why would we want to calculate new resustance / old resistance. Isnt just looking at efffective ehp enough?

    But you know what, since I can make vague sense from your calculation, you can make it clearer a little bit in 2 ways. You can plot current EHP & old EHP linearlry vs the same armor in X-axis. And then you can analyze it in 2 ways:
    1. Ratio of new/old EHP
    2. Difference of old minus new EHP
    Thats the only way I can think of it to make sense and understanding of these numbers. You can easily see that graphs have same starting points but the newer slope is less.

    For example we can take 1 point here to analyze point x = 50 armor =>
    EHPold = 400%
    EHPnew = 350%

    The ratio would be 87.5% and the difference would be 50%

    This can probably be all done on the same graph.

    Now I would like it if you would kindly redraw your graph either this way, or maybe some way you think gives more meaning to our case here :)

    If you aint wanna do it or a little busy maybe I'll do it when I'm back home.

    Im planning to recreate my deso thread anyway with the update