← Forum archive

PVP Dodge/Hitrate/Damage/HP

An archived conversation. New replies are closed.

Eld
I might wait until MR is changed on this though, based on values the higher HP races do make it out by a slight amount, which is technically right. Each point is 'technically' 4% in fights, 2% attack and 2% dodge. I don't think that's too terrible atm but since MR doesn't really work its not working so well. I.E. Supposed to be like rock paper scissors, (though obv you can still win as a high agi/lower hp char) vs high hp/low agi chars, but should be able to resist magic byfar more.

Through simulations right now tis 18.5 attacks to kill a Ling with an HO to 20.1 for a HO to kill a Ling. Obviously thats using the formulas and sample, and different things can come up. I would expect slayers to be better when HO's generally.

Obviously for HO's the variation was far greater. they could kill in 14 attacks, or sometimes 24 attacks. While a ling was more balanced, around 17.5 to 22.5. The randomness was greater for HO's that is. It probably sometimes feel like HOs or dwarfs can do more when you get those lucky kills, but ultimately i don't think it's too far off where it should be. aka yeah you can kill faster, but sometimes it can also take longer too. I can see raising this slightly, but I don't think by much. maybe 2.2% per agi instead of 2%.



Are we considering how much hp is regained during battles for these high end characters (im not sure of the formula but i believe higher end regain more hp per round?) Dwarfs can have up to 7 more end than halfing so I assume they are gaining maybe double (not sure if its much or if the formulas are posted).

Something else that doesnt really come up in 1v1 simulations is hp pool and how valuable that is in not only team events but pve as well. I think things are pretty close but I agree agi could use a boost.

I see nothing but dwarf being created and I dont blame them haha!
-casters cant touch them because they have 20wis and a 10 spell damage reduction. Even if their wis was around 16 they still have 23 end to take all that damage)
-high agi races dont really stand a chance since its a bit skewed atm.
-really the only thing that can touch them is another dwarf or half orc (they have the edge)



My dwarf slayer has 19 wis and 10 int, if you +1 a stat you cant get 20 wis or any int.
His MR is almost non existent tbh, the only time any spells fail on me are from 15 int druids with their faeriefire.

If i'm facing a Chanter/DM then I have terrible MR and get ruined.

TBH I wish i rolled orc and got the +3 str, i dont know why people keep rolling dwarves but they're gonna be disappointed like i was
daedroth

TBH I wish i rolled orc and got the +3 str, i dont know why people keep rolling dwarves but they're gonna be disappointed like i was

Because dwarves have beards and sound Scottish when they speak.
NiteHawk
Well I mean, wisdom is going to be changed so that would be the reason to use a dwarf, it will be a overall better tanker.

For the tests I mean, I did include HP pools and even char bonuses at an average per race.

If 2% is too low per point, maybe 2.5% would be better.
JadeFalcon
Well I mean, wisdom is going to be changed so that would be the reason to use a dwarf, it will be a overall better tanker.

For the tests I mean, I did include HP pools and even char bonuses at an average per race.

If 2% is too low per point, maybe 2.5% would be better.


Out of sheer curiosity - how do you test this out?
Do you have a sort of number generator that sims combats, or you actually get people fighting?
daedroth
You know usually when I make a character I am making it to a theme. I go with what race/class suits the theme/name I have in mind. I always find it amusing when people make what is considered the best combination (for that time). Then the mechanics change slightly and "OH NOES!" they have to make a new all powerful alt based on the changes <img src="{SMILIES_PATH}/icon_rolleyes.gif" alt=":roll:" title="Rolling Eyes" />
I have an elf barbarian based on a character from a book (was 50/50 with knight, but barbarian due to flurry seemed most fitting) will he be OP do you think ? <img src="{SMILIES_PATH}/icon_e_biggrin.gif" alt=":D" title="Very Happy" />
NiteHawk
Well I mean, wisdom is going to be changed so that would be the reason to use a dwarf, it will be a overall better tanker.

For the tests I mean, I did include HP pools and even char bonuses at an average per race.

If 2% is too low per point, maybe 2.5% would be better.


Out of sheer curiosity - how do you test this out?
Do you have a sort of number generator that sims combats, or you actually get people fighting?


I take all my programming data and simulate how many rounds it takes to kill based on HP, AGI, END, Armor, etc and simulate about a million fights to get the average amount of hits it takes to kill. It simulates everything though, misses, deflects, absorbs, etc.

I don't exactly line up two races in a combat but I could do that too probably. The whole 'how many rounds' is an average. It's mostly how many attacks would it take a race to kill another race in that regard.

It obviously won't take into effect skill vs stupidity, though. <img src="{SMILIES_PATH}/icon_e_ugeek.gif" alt=":ugeek:" title="Uber Geek" /> And obviously it's still paper info vs reality.
NiteHawk
Did an actual attack thing. Two knights going at it:

Attacker is HO with 19 agi, 22end, 12char. 150-193 damage, no div. Around 1213 hp.
Defender is Ling with 23 agi, 17 end, 16 char. 123-158 damage, no div. Around 952 hp.

Attacker and Defender swapped who attacked first. Each attack would be one at at time. So attacker attacks once, then defender attacks once. This would be straight up cavaliers smacking each other.

Current Formulas.
Attacker Wins: 1337061 (53.5%)
Defender Wins: 1162939 (46.5%)

2.2% agi
Attacker Wins: 1230130 (49.2%)
Defender Wins: 1269870 (50.8%)

2.5% agi
Attacker Wins: 1066778 (42.7%)
Defender Wins: 1433222 (57.3%)


Even though you can't factor in RH etc right now (i could simulate it i guess) It still gives a perspective on how different a couple decimals do. (Aka 2.5% is probably too much).

Maybe it's just in peoples heads cause the HP was raised a good deal but AGI was not really changed so instead of the 'feel' that you dodge more, you simply just soak up more damage as a ling.
NiteHawk
For example with 2.5% and different HP:

Attacker is HO with 19 agi, 22end, 12char. 150-193 damage, no div. Around 1215 hp.
Defender is Ling with 23 agi, 17 end, 16 char. 123-158 damage, no div. Around 850 hp.

Attacker and Defender swapped who attacked first. Each attack would be one at at time. So attacker attacks once, then defender attacks once. This would be straight up cavaliers smacking each other.

2.5% agi
Attacker Wins: 540331 (54%)
Defender Wins: 459669 (46%)

What happens though is the attacks to kill change alot more. But the risk is then HO's will feel like they aren't hitting enough, which isn't the case. Maybe it's the balance of dodge vs HP, though the final values are actually the same, in your 'head' it needs to 'feel right too'

So suddenly 2.5% agi isnt' so bad. It's about the same, but HO's would have a higher random variance while lings would probably be even more set. Thus the illusion that dodging works better as an average. Though in rare cases lings might have a chance to get fucked a little faster, but ultimately the amount of hits would generally average out to be higher.



It would be like having a 1HP monster that dodges 99% of the time vs a 9999 HP monster that dodges 1% of the time. In reality, the average amount of hits to kill it would be the same, around 100 hits, but you would probably feel a whole lot better hitting the 9999 HP monster because you know attacks are landing. I think same might be the case here in terms of how it works and how players are feeling. Just an example.
Folder
That is good math, thanks for running the sims.

I think why I feel AGI is not as good as END is mostly based on group PvP settings. Assuming you have a healer it is, typically, easier to keep someone with more END alive over someone with more AGI. My dwarf cav -feels- very durable in combat compared to my ling guard (which has more AC because of aegis). This is of course assuming there is a priest in the group healing me.

I'm not sure how you could sim that one out. All I'm getting at here is bigger HP pool + a healer = far more durable character than something with more AGI and less HP.

Hard to balance though because if you make lings just as good at soaking damage in a group setting they will be superior 1v1 and that's not good either. Bleh!
Lateralus
Did we realize hp regain during a battle isn't much then? I didn't see it mentioned above. Do we have a formula for that?
Folder
Pretty sure it's negligible. Unless it's been changed passive HP will tick at:

17 end:
every 5.7 seconds

23 end:
every 5.1 seconds

HP ticks are miniscule if you aren't in a temple. They are 1 or 2HP a tick I think?
NiteHawk
Pretty sure it's negligible. Unless it's been changed passive HP will tick at:

17 end:
every 5.7 seconds

23 end:
every 5.1 seconds

HP ticks are miniscule if you aren't in a temple. They are 1 or 2HP a tick I think?


Yeah. 1 or 2 per, passive HP really isn't meant to be a massive thing though either atm.
NiteHawk
That is good math, thanks for running the sims.

I think why I feel AGI is not as good as END is mostly based on group PvP settings. Assuming you have a healer it is, typically, easier to keep someone with more END alive over someone with more AGI. My dwarf cav -feels- very durable in combat compared to my ling guard (which has more AC because of aegis). This is of course assuming there is a priest in the group healing me.

I'm not sure how you could sim that one out. All I'm getting at here is bigger HP pool + a healer = far more durable character than something with more AGI and less HP.

Hard to balance though because if you make lings just as good at soaking damage in a group setting they will be superior 1v1 and that's not good either. Bleh!


Well I think lowering how much END gives per point a little bit (more so on lings but all END should take a hit), and boosting per AGI to 2.5% would be a start to making END vs AGI more balanced. Here is something I tested now:


Attacker Wins: 535067 (53.5%)
Defender Wins: 464933 (46.5%)
HO 'three in a row successful hits': 416778
LING 'three in a row successful hits': 1251935


What's interesting is this. This is how many attacks it took to win, removing some of the lower points:

HO killing Ling:
7 attacks to death: 64
8 attacks to death: 383
9 attacks to death: 1336
10 attacks to death: 3023
11 attacks to death: 5603
12 attacks to death: 9483
13 attacks to death: 14235
14 attacks to death: 19599
15 attacks to death: 25670
16 attacks to death: 31204
17 attacks to death: 35940
18 attacks to death: 39741
19 attacks to death: 42421
20 attacks to death: 42720
21 attacks to death: 41828
22 attacks to death: 39203
23 attacks to death: 35302
24 attacks to death: 31515
25 attacks to death: 26465
26 attacks to death: 21874
27 attacks to death: 17204
28 attacks to death: 13793
29 attacks to death: 10496
30 attacks to death: 7879




Ling killing HO:

7 attacks to death: 0
8 attacks to death: 0
9 attacks to death: 0
10 attacks to death: 7
11 attacks to death: 102
12 attacks to death: 495
13 attacks to death: 1598
14 attacks to death: 3890
15 attacks to death: 7802
16 attacks to death: 13370
17 attacks to death: 19841
18 attacks to death: 26564
19 attacks to death: 33154
20 attacks to death: 38170
21 attacks to death: 41320
22 attacks to death: 42192
23 attacks to death: 40706
24 attacks to death: 37950
25 attacks to death: 33530
26 attacks to death: 28700
27 attacks to death: 23882
28 attacks to death: 18976
29 attacks to death: 14850
30 attacks to death: 11258



Obvisouly here again it's the feel. Highest peak for lings is 22 attacks, for HOs its 20.
If you take the top 8, lings killed 63.5% in this bracket while half orc only killed 57.6% in their bracket. Meaning Lings have a more stability to them, though lings are +1 attack in terms of the top 8 too.

I could see people feeling like lings were weaker by 'alot' because of the first few rows. Lings will never kill a HO in 7,8, or 9 hits while a HO can do it, though lower chance.


Edit: I just included the 3 attacks to see the difference. You can see that lings really shine when hitting 3x attacks. I will probably make a simulation for ninjas in a sec.
Eld
Not sure how i feel about the prospect of everyone's HP being lower, it would have a lot of consequences.

1. PKing would be a lot easier, especially with the potion timer, at the moment most people have enough HP to absorb a round and then run.

2. General PVP at Moshes/OADs/Normal group fights would also change entirely due to lower hp reservoir.

3. One hitting people would potentially return, so this would be a big boost for spike DPS classes

4. All OADs would need to be rebalanced to take into account lower HP

There's probably more things i havent considered
NiteHawk
Guess no difference much, so simulations are actually alright:

HO Wins: 543929 (54.4%)
Ling Wins: 456071 (45.6%)
HO RHs: 413430
Ling RHs: 1242978

HO kills
7 attacks to death: 58
8 attacks to death: 351
9 attacks to death: 1134
10 attacks to death: 2737
11 attacks to death: 5206
12 attacks to death: 9117
13 attacks to death: 14046
14 attacks to death: 19751
15 attacks to death: 26098
16 attacks to death: 31848
17 attacks to death: 37449
18 attacks to death: 42087
19 attacks to death: 44313
20 attacks to death: 44697
21 attacks to death: 43565
22 attacks to death: 40819
23 attacks to death: 36763
24 attacks to death: 31954
25 attacks to death: 26834
26 attacks to death: 21987
27 attacks to death: 17096
28 attacks to death: 13441
29 attacks to death: 9965
30 attacks to death: 7195
31 attacks to death: 5138
32 attacks to death: 3553
33 attacks to death: 2306
34 attacks to death: 1597
35 attacks to death: 1030


Ling kills
7 attacks to death: 0
8 attacks to death: 0
9 attacks to death: 0
10 attacks to death: 0
11 attacks to death: 43
12 attacks to death: 285
13 attacks to death: 1158
14 attacks to death: 2973
15 attacks to death: 6659
16 attacks to death: 12012
17 attacks to death: 18681
18 attacks to death: 26033
19 attacks to death: 33383
20 attacks to death: 38439
21 attacks to death: 42025
22 attacks to death: 43309
23 attacks to death: 41678
24 attacks to death: 38589
25 attacks to death: 33982
26 attacks to death: 28506
27 attacks to death: 22991
28 attacks to death: 18271
29 attacks to death: 14023
30 attacks to death: 10054
31 attacks to death: 7500
32 attacks to death: 5165
33 attacks to death: 3665
34 attacks to death: 2424
35 attacks to death: 1562
NiteHawk
Not sure how i feel about the prospect of everyone's HP being lower, it would have a lot of consequences.

1. PKing would be a lot easier, especially with the potion timer, at the moment most people have enough HP to absorb a round and then run.

2. General PVP at Moshes/OADs/Normal group fights would also change entirely due to lower hp reservoir.

3. One hitting people would potentially return, so this would be a big boost for spike DPS classes

4. All OADs would need to be rebalanced to take into account lower HP

There's probably more things i havent considered


Well, I'm talking bout small amounts, I'm not sure what you're talking about here. 25-60 HP is enough normally to make a slight help to making AGI more useful. (One hitting I don't think is anywhere near even now. Takes at least two backstabs for sure.)

But if we don't change things around, AGI classes are never going to work in the game. You can't just up AGI here because even giving them .5% makes them dominate.

I don't think AGI or END are too far off, we dont need major changes here. MR is another story though.
NiteHawk
With 2.5% and nothing else changed.
HO Wins: 427321 (42.7%)
Ling Wins: 572679 (57.3%)

Ho wins:
7 attacks to death: 70
8 attacks to death: 277
9 attacks to death: 867
10 attacks to death: 1987
11 attacks to death: 3866
12 attacks to death: 6617
13 attacks to death: 10302
14 attacks to death: 14579
15 attacks to death: 19182
16 attacks to death: 23526
17 attacks to death: 27988
18 attacks to death: 31749
19 attacks to death: 34172
20 attacks to death: 34679
21 attacks to death: 34140
22 attacks to death: 32058
23 attacks to death: 29190
24 attacks to death: 25802
25 attacks to death: 22176
26 attacks to death: 18115
27 attacks to death: 14532
28 attacks to death: 11421
29 attacks to death: 8738
30 attacks to death: 6439
31 attacks to death: 4700
32 attacks to death: 3372
33 attacks to death: 2259
34 attacks to death: 1547
35 attacks to death: 1079



10 attacks to death: 18
11 attacks to death: 185
12 attacks to death: 742
13 attacks to death: 2301
14 attacks to death: 5626
15 attacks to death: 10954
16 attacks to death: 18176
17 attacks to death: 26790
18 attacks to death: 35317
19 attacks to death: 43565
20 attacks to death: 49224
21 attacks to death: 51927
22 attacks to death: 52383
23 attacks to death: 50005
24 attacks to death: 45438
25 attacks to death: 40110
26 attacks to death: 33523
27 attacks to death: 27211
28 attacks to death: 21392
29 attacks to death: 16593
30 attacks to death: 12391
31 attacks to death: 8985
32 attacks to death: 6521
33 attacks to death: 4511
34 attacks to death: 3096
35 attacks to death: 2041



I might need to do percent per attack amount to get proper. Obv you see that lings hit max kills at 22 HR, but they have a shit ton more kills in most.
Lateralus
I mean it makes sense to me that hit based classes like monk and barb would be better with high agi (you need those attacks in a row) while less dependent ones like all the other classes might be better with high str. I do think we are somewhat close. Maybe your suggested .5 would work out.
NiteHawk
Anywho, the real thing is what do we want to do.

1. If END per point matters more then per AGI then you can do one of the following:
1a. Make HP per END point more but start with a higher base, keeping the maximum HP the same. This means that you'd balance the AGI via HP being less of a variance.
1b. Increase AGI .25% per point so that HO's and Lings are about 50%/50% in fights. HOs' would still have the benifit of randomness and lower amount of attacks needed, but also has the same effect in terms of requiring more attacks. So HO's would be a random chancer while Lings would be more averaged out. AKA HO's might take anywhere from 8 to 18 hits to kill while lings are more stable and normally need around 12-14 hits to kill.

2. If the values are fine above are fine (technically HO's should be better then lings in physical combat by a little) but it's just the 'feeling', aka what I've talked about, then decrease END per point by around 60-80HP at lowest END (so around half of what it was before the changes) and increase AGI to 2.5% per agi point. (HOs would probably only lose like 10-15HP) This the above is around the same as it is now in terms of average attacks to kill, however, it would 'feel' like lings dodge more, because you're getting more dodges in replacement of less HP.

3. A different direction would be to make AGI per point only in PVE better. Which would mean that high AGI races would generally be better at PVE in some cases, not all.


Edit: I mean it's really up in the air, it's tough what to do atm. The problem as I see it is that you can't have lings feeling like AGi is 'great' without making lings have a higher win ratio overall, but having HO's still have those random blows. It's a tough balance.