New ATK and Accuracy value analysis

[Jibbatwinkers]

I put this together to figure out how the new values (atk and accuracy) work.

You can get the file here it's an excel file so hopefully most of you can get that open. Sorry that my setup isn't exactly awesome in terms of format.
http://zebserver.com/download/atkacctests.xls

All of the NPCs had a min damage of 100, max damage of 1000. Attackspeed value was -75, STR and AGI were set to 75 (default) and the level of the NPCs was 70. All NPCs were engaged by the same warrior with approximately 10,000hp and 2000AC. The actual values here aren't really as important as the general trend of the data.

I chose to measure the time it took this specific warrior to die at each different value. I feel like this is more important and less ambiguous that figuring out the average damage and hit rate. Also, the trend is more important here than any variable because any given tank has very different stats/AA.

First I tested ATK. I set all Accuracy values to 500. The general trend was hyperbolic. You can see from the graph that any ATK value past 1000/2000 has very little effect on the time to death where as very low 1/10 attack values cause a very large difference. Adjust your attack ratings as you see fit. My graph isn't especially accurate, but it gives you the right idea. Also note that it doesn't show up in the graph but with ATK set to 0 the time to death was 27 seconds, at 1 it was 51. I'm not sure if the code calculates an ATK value if there isn't one set.

After that I tests Accuracy with all ATK values set to 500. Again this is rough data. In the log it mentioned 10accuracy = 1%. The results seem a bit more linear. Again 0 Accuracy behaves much different than 1 Accuracy. 1000 and 2000 were basically the same, probably because it maxes it out at 100%. Also the NPC wasn't hitting 100% of the time (probably not even 50%) at 1000+ accuracy, just more often. The important relation here is the 1000 Accuracy is about 2x more DPS (at least in my tests) than around 10 Accuracy.

I hope this helps other server developers. Thanks to the EQEmu team and submitters for these great new variables I was very excited when I saw todays patch!

Zebuxoruk Server
Special thanks to Tiny the Troll Warrior for dying for me, a bunch.

[trevius]

Nice, that should save me some time with testing the new system when I get off work!

It sounds like there are defaults for these new settings depending on mob level as long as the setting is left to 0. I was hoping that was the case since it means you shouldn't have to redo stats for every mob in the database, just the ones you want to adjust, and mainly higher level encounters.

I wonder if KLS's spreadsheet factors this new stuff in properly. Here is a link to that:

openoffice
http://spreadsheets.google.com/pub?k...IWQ&output=ods
or excel
http://spreadsheets.google.com/pub?k...IWQ&output=xls

The combat system has made huge leaps and bounds lately and is really starting to seem finalized. Huge thanks to KLS for getting this stuff in the code. I know me and Jibba have been drooling for this stuff for quite some time now

[KLS]

Yeah I made it as a bonus, if it's set to 0 combat should be pretty much the same as it is now.

Accuracy isn't exactly right in that description. 10% = 1% but that 1% is then applied to the already existing accuracy. So you have say a 50% chance to hit and a 250 accuracy rating it would be 50% + (50% * 250 / 1000) or 62.5%

[ChaosSlayer]

Originally Posted by KLS

Yeah I made it as a bonus, if it's set to 0 combat should be pretty much the same as it is now.

Accuracy isn't exactly right in that description. 10% = 1% but that 1% is then applied to the already existing accuracy. So you have say a 50% chance to hit and a 250 accuracy rating it would be 50% + (50% * 250 / 1000) or 62.5%

Great work KLS!
I don't wont to start another thread, so my quetsion is regarding:

"KLS: Agility will now affect a defenders chance to be missed slightly."

what are the exact numbers?


Also - how about making DEX improve To Hit chance (for players mainly, and on related note- does DEX curently totaly useless for mobs?)? I am not looking for damage, just for ability to hit , cuase curently (for exmaple) caster/priest classes ability to score a hit is prety much non existant at higher levels

I mean, even if my wizard/druid will hit 2-3 times as often in melee, for a meager 10-20 dmg at lev 70 - that can't posibly break the game- can it? =)
But this will give people a purpose to build a DEX for something else other than procs/parry (specialy for classes who do not get such skills)


PERHAPS consider this - add a RULE which when turned ON, will make DEX to act as a BONUS to Hit chance. (so 0 is dex turned off, greater than 0 - how extra +1% to hit is gained per X dex specified as value)
and perhaps even same for agilty?

Oh another thing- a RULE for To Hit chance cap and to Dodge chance cap for ALL classes

[ChaosSlayer]

something else come to mind. In eq2 players stats keep rising continiusly every level. Essentialy there are no hard caps, only a cap in comparecen betwen player and a mob. So if hit your cap for say Dodge for lev 50 - this is only vs a lev 50 mob. As you go to lev 51, you will need MORE AGI, to maintain same dodge cap vs a lev 51 mob (cuase lev 51 mob automaticly ingnores soem of your Dodge). essentialy the system is based on comparecen betwen your level and stats and mobs level and stats. This way there is prety much NEVER a cap on anything - you will become inviciable to lower level mobs, and higher level mobs are invicable to you

[trevius]

If you want your druid or cleric to land hits more often, try adjusting your skill_caps table.