Castle Clash: Guild Royale Today: 0    Total Posts: 12319

Total:#48

Create Thread

[Guides] Guild wars simulation – hero comparison

Hot Topics [Copy link] 47/18693

#31
Posted on 9/2/14 1:22:14 PM | Show thread starter's posts only

why does a 5/5 scatter ninja have a higher win rate than the 5/5 scatter Reaper i have? By my math Reaper will effectively stop any hero with an attackspeed of 1200 or higher from ever proccing, and will delay a non revit hero with an attackspeed of 1000 to almost 60 seconds per proc. If you for instance take the matchup against a revit 5 molt, ninja loses every time and reaper still has a 50% chance of winning, since both molt and reaper have the same range.

Signature
#32
Posted on 9/3/14 2:30:20 AM | Show thread starter's posts only

how do you do

Signature
KSA
#33
Posted on 9/3/14 5:36:51 AM | Show thread starter's posts only

Edited by jdone at 9-3-2014 05:40 AM

MoCJixx replied at 9-2-2014 01:22 PM
why does a 5/5 scatter ninja have a higher win rate than the 5/5 scatter Reaper i have? By my math R ...

Good question. Some thoughts:

1) There is a margin of error due to randomness. A difference of 1.5% is not particularly significant.

2) Ninja kills the enemy quicker than gr, allowing less time for Deadly Strike or Heavy Blow.

3) Sure, 5/5 Revite Moltanica may beat Ninja and lose to GR, but that particular matchup is only one out of thousands.



Signature
Viceleader of HaVoC, a friendly top 100 guild which often wins in guild war :-)

Main account: 197.7k might
LB points: 1100-1300
GW points: around 2500-2600
#34
Posted on 9/3/14 5:41:29 AM | Show thread starter's posts only

mtsubrent replied at 9-1-2014 09:51 AM
I wonder if flame guard will be effective in Torch battle.  It would be similar to wargod in terms o ...

Flame Guard is really good, approximately the equal of bulwark ...

Signature
Viceleader of HaVoC, a friendly top 100 guild which often wins in guild war :-)

Main account: 197.7k might
LB points: 1100-1300
GW points: around 2500-2600
#35
Posted on 9/3/14 4:57:00 PM | Show thread starter's posts only

jdone replied at 9-3-2014 05:41 AM
Flame Guard is really good, approximately the equal of bulwark ...

Thanks for the Flame Guard info. Just saw this thread, was surprised to not see FG listed among those top talents.
Also figured succ would be decent 1v1 and high scatter would be good, but not as good as you show, very interesting.
Also interesting that zerk5 heroes perform worse than a random talent on the same hero, for the heroes shown anyway. Of course more attacks just mean more energy for the opponent, didn't think it'd be that bad though.

Anyway, thanks for all this info, good stuff. A couple questions though, if you don't mind:

- In melee vs. ranged, assuming same attack speed, why does ranged's skill fire first? Getting hit by a hero gives the same energy as attacking (or do I have that wrong?) so won't they fire at the same time?

- I assumed the game steps through in 200 ms. increments, processing all attacks and skill hits "simultaneously". In other words, both heroes attack and skill fire at the same time, not 1 going a tiny amount of time before the other. No?

- In general, could you describe your simulator some? Just curious. I'm wondering how well it mimics the actual - not that this will tell me that, but may give a feel.

Signature
#36
Posted on 9/3/14 5:36:55 PM | Show thread starter's posts only

according to cc wiki hero skills and talents gives your enemy energy
so flameguard is actually pretty bad
http://castle-clash.wikia.com/wiki/Energy
but who knows if cc wiki is true they had a lot of mistakes in the past

Signature
#37
Posted on 9/4/14 3:14:53 AM | Show thread starter's posts only

JonnyBolt replied at 9-3-2014 04:57 PM
Thanks for the Flame Guard info. Just saw this thread, was surprised to not see FG listed among th ...

Thanks! Answers to your questions:

- The ranged hero gets to hit first while the melee hero is running towards it. After 4 ranged hits and 3 melee hits, both heroes have full energy, and the melee hero can proc on the fourth attack.

- I think when both heroes are scheduled to attack at 200 ms intervals, one attack still gets done a tiny bit before the other. In 1v1 one hero nearly always comes out on top rather than both dying frequently, which would often happen if both attacks were executed simultaneously. ForceClosed showed this in boss fights with cupid procs. In my simulator this is 50/50 chance.

- Good question. If you are interested I can send you the code. The basic method is:
- Choose both hero types, levels, talents and skills.
- Read stats from database
- Set up hero health
- Determine who gets to attack first, and by how much time. This is based on range. On each attack the attacked hero health is reduced by damage with a +- 10% random factor. Stone Skin reduces this damage, and Flame Guard, AC skill, and Blade Shell reflect damage.
- Perform attacks and procs, in time order, until one hero (at least) dies.
- Repeat many times to get statisically significant results.

Notes:
- Hero energy is tracked at all times, and is used to trigger procs just as in-game.
- Buffs and debuffs are fully supported, including the 200 ms attack speed rounding.

Signature
Viceleader of HaVoC, a friendly top 100 guild which often wins in guild war :-)

Main account: 197.7k might
LB points: 1100-1300
GW points: around 2500-2600
#38
Posted on 9/6/14 2:46:34 PM | Show thread starter's posts only

Any more simulations coming up? :)

Signature
#39
Posted on 9/18/14 10:19:09 AM | Show thread starter's posts only

Any more simulations comin up?

Signature
#40
Posted on 9/20/14 9:09:41 AM | Show thread starter's posts only

finally a good thing :)

ok moltanica is op but it's only one heroes.

other problem i found is how to stop a revive atlanticore? druid, immo, tg, executioner and champion fail

Signature