|
Notes |
|
|
|
Neither fencing, tactics, archery, Macefighting, any skill (neither anatomy, begging... etc) ... Something weird's happening.
(and doesn't works very good, because if you have 0 you can't hit just like ANYTHING)
Last build. |
|
|
|
(0002922)
|
|
Rizz
|
|
13-07-15 23:27
|
|
|
@gain is triggering but none can really improve |
|
|
|
(0002924)
|
|
Rizz
|
13-07-15 23:51
(edited on: 13-07-15 23:52) |
|
I think there is a problemi with the default argn value, in this way is working
on=@gain
local.Swordsmanship=<src.Swordsmanship>
if <local.Swordsmanship>==0
local.Swordsmanship=1
endif
ARGN2=<eval 220000/<local.Swordsmanship>>
|
|
|
|
|
Revision 2382, non-combat skill increase for me, combat skills doesn't increase because of the high chance of miss.
I think the new hit chance formula have a problem.
In the revision is written:
New Hit Chance combat formula (more info: http://uo.stratics.com/content/arms-armor/combat.php#hitchance [^])
According to Stratics formula, i should have a 100% chance of hitting when i have 100 in a combat skill and the defender 0, instead the actual hit chance is around 50% miss / 50% hit. Interesting to note that if i have 25.0 in the combat skill and the defender has 0, i still should have a 100% hit chance, but actually the hit chance is much much lower.
This also happens when both values are equal, you should have a 50% chance of hitting the target, this seems true when both attacker and defender combat skills are around 50.0% but at low low value of skills the hit chance seems a lot lot lower. |
|
|
|
(0002926)
|
|
darksun84
|
14-07-15 16:57
(edited on: 14-07-15 17:24) |
|
I managed to made a "small benchmark", the results probably aren't 100% accurate but they should give an idea about the high miss chances of the new formula.
First test:
Attacker skill: wrestling 1.2
Defender skill: macefighting 0.0
I should get around 52% hit chance with that values
Results
Total Hits: 25/1280 (1.953125) - Total Misses: 1254/1280 (97.968750)
Second test:
Attacker skill: wrestling 100.0
Defender Skill: macefighting 0.0
My chance should be around 100%
Results
Total Hits: 665/1280 (51.953125) - Total Misses: 616/1280 (48.125000)
Third test(weirdest result) :
Attacker skill: wrestling 0.0
Defender skill: wrestling 100.0
I should have an 8% chance of hitting target
Results
16:52:(sphere_skills.scp,1431)Total Hits: 412/1280 (32.187500) - Total Misses: 868/1280 (67.812500)
|
|
|
|
|
it's insane. The miss chances and the skillgain chances need to be fixed ='( hope Coruja comes and save us hahahaa.
Thanks for the work darksun, like ever <3 |
|
|
|
(0002930)
|
|
Coruja
|
|
16-07-15 22:54
|
|
theres a difference between miss and block the attack, you're testing only the hit chance but not the block chance
the combat behavior works like this:
CheckAttackerHitChance
IF (hit)
CheckDefenderBlockChance
IF (blocked)
-HitBlock-
ELSE
-Hit-
ENDIF
ELSE
-HitMiss-
ENDIF
but anyway I found a weird issue calling twice the hit chance formula, making all chars have a double chance to miss attacks. Next nightly will be released in some minutes with this fix, maybe it will fix this combat problem |
|
|
|
|
Still have so much missing. Can you implement a system to get back to last formulas?
And the wrestling, sword, mace, fencing skills doesn't raise..
[hope this can be resolved :P]
|
|
|
|
(0002932)
|
|
Rizz
|
|
19-07-15 11:58
|
|
|
|
|
(0002933)
|
|
Shinaa
|
|
19-07-15 14:33
|
|
I think that something is wrong with the gainchance for combat skills overall, not related to the miss problem (which of course increases the problem).
I had the following happen:
Tactics has ADV_RATE=1.0,10.0,60.0,150.0,450.0,950.0
Wrestling has ADV_RATE=1.0,10.0,60.0,115.0,325.0,700.0
I set ACTDIFF in @HitTry to value of Wrestling, all skills were 50.0:
At first both skills increased like crazy, gainchance (argn2 at @SKILLGAIN) was something like 2000+, Tactics having a higher chance then Wrestling, which already made no sense to me, but after a few percent Tactics gain dropt to zero.
Then I tried not setting ACTDIFF to a set value, which of course increased the misses, but even when I actually hit, gainchance was often 0 at diffrent skilllevels.
I think the new formula sets ACTDIFF to values that cause the gaincalculations to go wrong, often resulting in zero chance to gain.
Skillgain function always had some problems with this, having ACTDIFF values higher than your skill for example, even if the skill would not fail because of it, would often result in zero gainchance. |
|
|
|
|
Tried the old formulas of actdiff that XuN shared in forums...
That formulas are wrong. Something weird is happening there. (no errors but bad calculations about actdiff)
hope this get fixed soon :'( |
|
|
|
(0002990)
|
|
XuN
|
|
30-10-15 18:01
|
|
|
This has been fixed in last rev, gain checks will be done with a minimum of 5.0 skill to prevent it to fail. |
|