 |
 |
 |
Asked the Experts, For archive purposes only. Please use Ask the Expert! for questions |
|
Oct 16 2015, 21:11
|
Superlatanium
Group: Gold Star Club
Posts: 7,643
Joined: 27-November 13

|
QUOTE(Frederiksc @ Oct 16 2015, 12:14)  Ok, let's do math.
Dmg = base * type * speed * (1 + crit * ( 1 - crit_mod))
With your IW favoring speed, you should keep in mind, that speed in battle is multiplicative to any buff you have. The first buff as DW you always go for is comming from AGI, capped at AGI = 2 level = +10% speed. The second buff you should always have is haste +25% speed. The third is a light armor ability called "Light Speed" with max points @ 250 invested is 0.025 * LA_proficiency So with these estimated active buffs in battle, you already have speed = 1 + (1 - (1 - AGI_buff) * (1 - haste) * (1 - LightSpeedBuff)) speed = 1 + (1 - 0.9 * 0.75 * (1 - 0.025 * 380 / 100)) speed = 1 + (1 - 0.6109) ~ +38.91% With your IW this is now in battle: speed = 1 + (1 - 0.6109 * 0.9) ~ 0.45%
You may be disappointed now, that you just have a gain of about +6% in battle, but for a DW you can not have enough speed to shorten your "turn duration". To see something like a "clock" in your battle log, take fixed time ticks, like regen ticks occuring, because the battle log always counts your turns, however "slow" or "fast" they are. Also I recommend to forge your weapon to all stats +5 (without the use of bindings) to maximize the utility gained. To close this review on DW I state, that for potencies on weapons of balance: - Swift strike > fatality > butcher. (If you want to do the math yourself, use the Cobb-Douglas ulitity function) - Do not go for butcher, if you do not plan to forge that weapon further than 5. - Always go for potencies, that can not be forged, because you can always re-IW, if you meet the utility border when forging / leveling-up. This shows what SS does to improve attack speed stats, but isn't the real issue how such improved attack speed compares to the effect of other potencies? There are two main factors potencies can improve: turns required to kill monsters, and cure/rebuff frequency. Except for those using Niten and Skyward, SS doesn't improve killing speed, right? But, it directly reduces rebuff frequency, and slightly reduces Full-cure and rebuff frequency. On the other hand, other weapon potencies improve killing speed, and also slightly reduce Full-cure frequency. First let's assume everyone who might consider SS is not using heavy armor - makes the math easier. SS casting improvement: Depends on IA. Eg., with IA 4, the player is only casting Heartseeker, Regen, and maybe Protection (if player is 1h, they won't use Shadow Veil, so Protection will be on IA as well) - 1 every ~400 * 1.0 action time, plus 1 every ~110 * 1.0 action time, combine to get one turn consumed rebuffing every ~86 * 1.0 action time, which might be ~one turn consumed every 140 turns. (average frequency could be less if player gets Channeling.) (overall time multiplier of 140/139) With SS 5 and your numbers, it's around one turn spent every 156 turns. (overall time multiplier of 156/155) So, eg., 10,000 turn arena without SS -> = 9929 turn arena with SS 5. Then there's full-cure frequency, which is reduced with SS due to monsters dying after less action time. (can't say "fewer turns" (IMG:[ invalid] style_emoticons/default/wacko.gif)) I think for most, Regen provides the vast majority of healing, and full-cure is not common, though this depends on fighting style. Unfortunately, although monsters die after less action time with SS, Regen also provides proportionally less healing during that time, for the same reason - less action time has elapsed. Wish I had hard numbers on Full-cure frequency for other fighting styles - for 1h it's pretty much nonexistent. Then there are non-SS weapon potencies. I'll just look at Fatality because this is long enough already. Say crit damage is 160% of normal damage. Fatality 5 increases that to 170%. If crit chance is 37%, overall damage dealt increases by around (.63 + .37 * 1.7) / (.63 + .37 * 1.6) = +3%; perhaps turns required to kill monsters is reduced by 2.6% (extra damage on monsters who would die anyway doesn't increase speed). Considering only increased monster killing speed, a 10,000 turn arena without Fatality -> 9750 turn arena without Fatality. If equipment is forged, CC is higher, and CC synergizes with Fatality's CD, so the improvement might be 3.3% or so instead of 2.6%. Then there's reduced Full-cure frequency due to monsters dying quicker - somewhat similar to the Full-cure frequency effect of SS. Again, with harder numbers on that for non-1h styles, conclusions can be drawn, though SS's current -0.7% seems underwhelming compared to Fatality's -2.5% and Butcher's ~-1.3%. Though if a player is very low level and poor and doesn't have IA and also doesn't have much crit chance, SS will be comparatively better against Fatality - maybe even absolutely better (though they probably aren't IWing and reforging in the first place). QUOTE(izpekopon @ Oct 16 2015, 19:03)  Do 1h+Shield user need Ability Points in Better Imperil? We don't benefit very much from Imperil do we? You're right, it's useless. Well, you might be able to clear rounds in slightly fewer turns, but the concentration and time required to select and cast would vastly outweigh the benefit.
|
|
|
|
 |
|
Oct 16 2015, 21:20
|
jacquelope
Group: Members
Posts: 10,459
Joined: 28-July 15

|
I've done experiments with Imperil, it's true, the time required to cast the spell negates the (often unrealized) savings in reduced rounds. Perhaps mass imperil could make it worth it? I haven't put points into that yet so IDK.
|
|
|
|
 |
|
Oct 16 2015, 21:36
|
Frederiksc
Group: Gold Star Club
Posts: 694
Joined: 12-November 08

|
QUOTE(Scremaz @ Oct 16 2015, 18:12)  shouldn't it be:
Dmg = base * type * speed * ((1-crit) + crit*(1 + crit_mod))
?
Dmg = base * type * speed * ((1-crit) + crit*(1 + crit_mod)) Dmg = ... * (( 1-crit) + crit*(1 + crit_mod)) = ... * ( 1 - crit + crit + crit * crit_mod) Dmg = ... * (1 + crit * crit_mod) = wait.. let's do this again... Dmg = base * type * speed * (1 - crit) + base * type * speed * crit * crit_mod Dmg = base * type * speed * (1 - crit + crit * crit_mod) Dmg = base * type * speed * (1 + crit * crit_mod - crit) Dmg = base * type * speed * (1 + crit * (crit_mod - 1)) I fumbled at writing down the last term *sigh* QUOTE(danixxx @ Oct 16 2015, 18:57)  Why is there speed in the formula? (IMG:[ invalid] style_emoticons/default/unsure.gif) Because speed actually exists in HV. If you look for a "clock" for measuring your speed, take fixed occuring ticks like a tick of the regen spell. The system just counts in "your turns", but actually you have mild information about how "fast" or "slow" you are. This post has been edited by Frederiksc: Oct 16 2015, 21:40
|
|
|
|
 |
|
Oct 16 2015, 21:40
|
Superlatanium
Group: Gold Star Club
Posts: 7,643
Joined: 27-November 13

|
QUOTE(Frederiksc @ Oct 16 2015, 19:36)  Because speed actually exists in HV. If you look for a "clock" for measuring your speed, take fixed occuring ticks like a tick of the regen spell. The system just counts in "your turns", but actually you have mild information about how "fast" or "slow" you are. But what really matters is damage per turn (clear speed), rather than damage per action time - which is what you're doing - right?
|
|
|
|
 |
|
Oct 16 2015, 22:41
|
Frederiksc
Group: Gold Star Club
Posts: 694
Joined: 12-November 08

|
QUOTE(Superlatanium @ Oct 16 2015, 21:40)  But what really matters is damage per turn (clear speed), rather than damage per action time - which is what you're doing - right?
Well, that depends on your play style. If you stick to a certain play style and do not use haste to speed you up, the clear speed, as you define it, is exactly the same or some turns more. If you clear a given round in 20 turns, you will still need 20 turns without haste, but this time at least 20 turns. You "loose" turns, because this time monsters are allowed to attack you more frequently and you have to react on that. Buffs like PA or Imperil run out faster. There are play styles, who take advantage of high speed. Builds with light armor should always go for speed, because they have less armor and therefore rely on less damage incomming per turn. Also you can use any skill on CoolDown, because you have enough hits in between. Taking a look on Niten, you have a limited time window of the skill Skyward Sword to boost damage. When you "slice time" by a time flow with 40% speed, you have some valuable hits more while being in this window, and you clear time decreases. When I look on my mace, it states I stun for 4 turns. Actually its more, because I have the speed to be granted more hits in the "stunned time window". My buffs run longer, so my mana consumption is lower. Heavy Armor slows you down. The according playstyle is said be 1h, to have some extra damage while you wait until your next turn and parry some attacks. I do not know if there are Rapiers of Swiftness and according shields, but even as 1h speed should be not underestimated. The extra damage is limited to 3 parries per turn, so you need only 3-5 monster attacks per turn and not all monsters attacking to have some considerable damage boost. Since we do not have a solid model on this topic, that said is just my two cents.
|
|
|
|
 |
|
Oct 16 2015, 23:48
|
Superlatanium
Group: Gold Star Club
Posts: 7,643
Joined: 27-November 13

|
QUOTE(Frederiksc @ Oct 16 2015, 20:41)  Well, that depends on your play style. If you stick to a certain play style and do not use haste to speed you up, the clear speed, as you define it, is exactly the same or some turns more. If you clear a given round in 20 turns, you will still need 20 turns without haste, but this time at least 20 turns. You "loose" turns, because this time monsters are allowed to attack you more frequently and you have to react on that. Buffs like PA or Imperil run out faster. There are play styles, who take advantage of high speed. Builds with light armor should always go for speed, because they have less armor and therefore rely on less damage incomming per turn. Also you can use any skill on CoolDown, because you have enough hits in between.
Taking a look on Niten, you have a limited time window of the skill Skyward Sword to boost damage. When you "slice time" by a time flow with 40% speed, you have some valuable hits more while being in this window, and you clear time decreases.
When I look on my mace, it states I stun for 4 turns. Actually its more, because I have the speed to be granted more hits in the "stunned time window". My buffs run longer, so my mana consumption is lower.
Heavy Armor slows you down. The according playstyle is said be 1h, to have some extra damage while you wait until your next turn and parry some attacks. I do not know if there are Rapiers of Swiftness and according shields, but even as 1h speed should be not underestimated. The extra damage is limited to 3 parries per turn, so you need only 3-5 monster attacks per turn and not all monsters attacking to have some considerable damage boost.
Since we do not have a solid model on this topic, that said is just my two cents. Yeah, I know about the action-time vs player-turn interaction, that was the main point of this post (IMG:[ invalid] style_emoticons/default/tongue.gif) But the problem still exists - increased action speed, by itself, isn't bad (except maybe when grinding proficiency). But its effect on clear speed is minimal to unnoticeable, especially if the player is using Haste (which they certainly should be if they have any sense), especially if the player is 1h (next to no Cures). This post has been edited by Superlatanium: Oct 16 2015, 23:48
|
|
|
|
 |
|
Oct 17 2015, 01:51
|
Frederiksc
Group: Gold Star Club
Posts: 694
Joined: 12-November 08

|
QUOTE(Superlatanium @ Oct 16 2015, 21:11)  First let's assume everyone who might consider SS is not using heavy armor - makes the math easier. QUOTE(Superlatanium @ Oct 16 2015, 23:48)  Yeah, I know about the action-time vs player-turn interaction, that was the main point of this post (IMG:[ invalid] style_emoticons/default/tongue.gif) But the problem still exists - increased action speed, by itself, isn't bad (except maybe when grinding proficiency). But its effect on clear speed is minimal to unnoticeable, especially if the player is using Haste (which they certainly should be if they have any sense), especially if the player is 1h (next to no Cures). I never argued for Clear Speed, since this is not something I consider a goal after those Hoverplay / Spellspam scripts arrived. I argue for utility - something on which everyone has a personal view. I may have some "special" views on some topics, but that's me ^^ Showing a different way than the "most favored", thus mostly meaning "most costly", is my thing. I do not consider everyone reading and posting here to share my views or accepting any advice I give without questioning. That makes this thread - seen as idea exchange - highly appreciated and treasured, well, at least for me. ^^ The effect on clear speed, I think we can cut this now, and take a look on survivability in terms of speed. A speed build, so to say. No, we are not 1h and not heavy armor, nor mage. To throw in a sweet spot for speed outside battle, I'd go for about 33.33%, meaning in battle a solid 50% with haste, slicing time to 2 "player turns" a "tick". I think we can use this wording, since we have read each others posts. I expect light armor users to start with Imperil and then skills. Since the early RPGs we know, that you first apply a debuff and than attack accordingly to your build. So the debuff phase - even being the 1st turn of a round - is the most critical. Light armor has no cast speed, so we have a solid tick-long turn. At PFUDOR we can expect all monsters testing our defences. Then we enter the flow and slice the turns. At this point, slow attacking monsters like giants become just an amout of HP waiting to be sweeped away. Monster attacks now come in not that massive, since the turn is not long enough to have most of them attacking. Also reaction time shortens, since you can actually use pots/cure before Spark triggers, which saves you some spirit points. So I think we can say, you increase survivability. To avoid free turns of monsters, I think we should go for a rapier mainhand and Wakizashi offhand. Both ethereal and swiftness, IW'ed to Swift Strike 5, But/Fat 4 and any elemental strike plus ADB forged to 5 (no bindings, low cost ^^) With this build we cut incomming monster damage per turn in half, but we prolong "clear speed" by +33%. Maybe not really a build for high levels, but maybe for 1-250...without the etheral prefix for starters
|
|
|
|
 |
|
Oct 17 2015, 01:56
|
danixxx
Group: Catgirl Camarilla
Posts: 7,340
Joined: 3-September 10

|
QUOTE(Frederiksc @ Oct 16 2015, 22:41)  Well, that depends on your play style. If you stick to a certain play style and do not use haste to speed you up, the clear speed, as you define it, is exactly the same or some turns more. If you clear a given round in 20 turns, you will still need 20 turns without haste, but this time at least 20 turns. You "loose" turns, because this time monsters are allowed to attack you more frequently and you have to react on that. Buffs like PA or Imperil run out faster. There are play styles, who take advantage of high speed. Builds with light armor should always go for speed, because they have less armor and therefore rely on less damage incomming per turn. Also you can use any skill on CoolDown, because you have enough hits in between.
Taking a look on Niten, you have a limited time window of the skill Skyward Sword to boost damage. When you "slice time" by a time flow with 40% speed, you have some valuable hits more while being in this window, and you clear time decreases.
When I look on my mace, it states I stun for 4 turns. Actually its more, because I have the speed to be granted more hits in the "stunned time window". My buffs run longer, so my mana consumption is lower.
Heavy Armor slows you down. The according playstyle is said be 1h, to have some extra damage while you wait until your next turn and parry some attacks. I do not know if there are Rapiers of Swiftness and according shields, but even as 1h speed should be not underestimated. The extra damage is limited to 3 parries per turn, so you need only 3-5 monster attacks per turn and not all monsters attacking to have some considerable damage boost.
Since we do not have a solid model on this topic, that said is just my two cents.
Attack speed is a mostly defensive stat, he doesn't really count as more damage so i was surprised you put it in that formula.
|
|
|
|
 |
|
Oct 17 2015, 02:07
|
doom9ra
Group: Members
Posts: 789
Joined: 11-December 10

|
Del,sry
This post has been edited by doom9ra: Oct 17 2015, 02:36
|
|
|
Oct 17 2015, 03:18
|
karyl123
Group: Gold Star Club
Posts: 1,659
Joined: 9-January 11

|
well, I choose shortsword over axe, because it has parry for survivability.
|
|
|
Oct 17 2015, 06:16
|
Void Domain
Group: Catgirl Camarilla
Posts: 2,131
Joined: 30-May 10

|
QUOTE(Frederiksc @ Oct 17 2015, 07:51)  I never argued for Clear Speed, since this is not something I consider a goal after those Hoverplay / Spellspam scripts arrived.
You must have 4t/s (IMG:[ invalid] style_emoticons/default/anime_cry.gif)
|
|
|
Oct 17 2015, 09:06
|
Falbala456
Group: Members
Posts: 1,176
Joined: 21-April 09

|
Even the ping is poor.
|
|
|
|
 |
|
Oct 17 2015, 10:59
|
Frederiksc
Group: Gold Star Club
Posts: 694
Joined: 12-November 08

|
QUOTE(Void Domain @ Oct 17 2015, 06:16)  QUOTE(nobody_xxx @ Oct 17 2015, 06:37)  I think I have a quite moderate ping... henativerse.org [94.100.28.52] 13 ms average egth.org [81.21.76.62] 27 ms average Ping states just a "Hello, you there?" answer by a server or cluster, not the actual page load / service delivery time. The most time you loose by your own set up. Browser: Two major points are page refresh and garbage. You need to costimize the Garbace Collection (just Firefox offers this, Garbage Collection per window) and to insert functions per script to support Garbage Collection, to minimize garbage, like stetting all image objects to NULL after you click something in the monster pane. That AJAX script is a perfect example of how to minimize the use of a browser's GC, because you have no "old" Hentaiverse pages to collect and dump, because you deny a full page refresh. Page Refresh: At the page refresh all your addons, plugins, scripts and even your virus scanner (if it has web capabilities) take a look at the page. So the more you have (active), the more time it takes. Scripts can be optimized by the page_load_event, if you want to activate the functions after a full load (mostly not necessary) or if you jump in at the point, when the Document Object Model finished loading. For most scripts I use, loading the DOM is enough. Here again AJAX shines. Since DOM is similar on old page and new page, replacing / altering it by a query is bypassing all that. Well, it still has some issues with the monster pane, but that is ok in an alpha stage. I hope AJAX for HV is developed further, like a AJAX HV Spellspam. Then it should be achieveable for everyone to go top speed without touching browser core settings or doctoring on page load events.
|
|
|
|
 |
|
Oct 17 2015, 11:24
|
mozilla browser
Group: Gold Star Club
Posts: 2,131
Joined: 22-December 11

|
All the browser and page refresh optimisation will not help significantly when the network latency is of this order of magnitude: CODE 20 273 ms 274 ms 273 ms 94.100.28.52 Contrapositively, when the network latency is around 15ms, I would think that there is little need for setup optimisation.
|
|
|
|
 |
|
Oct 17 2015, 11:25
|
izpekopon
Group: Members
Posts: 1,498
Joined: 27-August 15

|
Currently with around 47% Block, i find it hard to gain enough OC to fire off the OFC while staying in Spirit Stance. So i need to off Spirit Stance for a few turns every time i want to use the OFC.
Around how much Block % will i actively gain enough OC to use the OFC without having to trigger on and off of the Spirit Stance?
That said, would forging Block on my shield greatly improve my OC gain rate? Or would forging parry chance on rapier be more cost efficient (Don't need all that defense matrix mats)?
|
|
|
|
 |
|
|
 |
|
Oct 17 2015, 11:43
|
karyl123
Group: Gold Star Club
Posts: 1,659
Joined: 9-January 11

|
QUOTE(Frederiksc @ Oct 17 2015, 15:59)  I think I have a quite moderate ping... henativerse.org [94.100.28.52] 13 ms average egth.org [81.21.76.62] 27 ms average Ping states just a "Hello, you there?" answer by a server or cluster, not the actual page load / service delivery time.
The most time you loose by your own set up.
Browser: Two major points are page refresh and garbage. You need to costimize the Garbace Collection (just Firefox offers this, Garbage Collection per window) and to insert functions per script to support Garbage Collection, to minimize garbage, like stetting all image objects to NULL after you click something in the monster pane. That AJAX script is a perfect example of how to minimize the use of a browser's GC, because you have no "old" Hentaiverse pages to collect and dump, because you deny a full page refresh.
Page Refresh: At the page refresh all your addons, plugins, scripts and even your virus scanner (if it has web capabilities) take a look at the page. So the more you have (active), the more time it takes. Scripts can be optimized by the page_load_event, if you want to activate the functions after a full load (mostly not necessary) or if you jump in at the point, when the Document Object Model finished loading. For most scripts I use, loading the DOM is enough. Here again AJAX shines. Since DOM is similar on old page and new page, replacing / altering it by a query is bypassing all that. Well, it still has some issues with the monster pane, but that is ok in an alpha stage.
I hope AJAX for HV is developed further, like a AJAX HV Spellspam. Then it should be achieveable for everyone to go top speed without touching browser core settings or doctoring on page load events.
it just your ping is superior. 13 ms. many ppl here got 200-300 ping
|
|
|
|
 |
|
Oct 17 2015, 11:44
|
karyl123
Group: Gold Star Club
Posts: 1,659
Joined: 9-January 11

|
QUOTE(izpekopon @ Oct 17 2015, 16:25)  Currently with around 47% Block, i find it hard to gain enough OC to fire off the OFC while staying in Spirit Stance. So i need to off Spirit Stance for a few turns every time i want to use the OFC.
Around how much Block % will i actively gain enough OC to use the OFC without having to trigger on and off of the Spirit Stance?
That said, would forging Block on my shield greatly improve my OC gain rate? Or would forging parry chance on rapier be more cost efficient (Don't need all that defense matrix mats)?
to counter. dot forget the parry. more parry, better
|
|
|
|
 |
|
Oct 17 2015, 11:45
|
Frederiksc
Group: Gold Star Club
Posts: 694
Joined: 12-November 08

|
QUOTE(mozilla browser @ Oct 17 2015, 11:24)  All the browser and page refresh optimisation will not help significantly when the network latency is of this order of magnitude: CODE 20 273 ms 274 ms 273 ms 94.100.28.52 Contrapositively, when the network latency is around 15ms, I would think that there is little need for setup optimisation. Hmm... You ever considered asking your ISP to set your line to "fast pass"? That costs around 10 bucks as one-time payment and varies from ISP to ISP. Some even do not offer it, because they do not have access to that feature (rented lines). Mainly it removes the checksum test on your line. My ping was about yours before, but then I played some games online (in some clans) and concluded sadly, that ping is life for such hobbies (IMG:[ invalid] style_emoticons/default/sad.gif)
|
|
|
|
 |
|
2 User(s) are reading this topic (2 Guests and 0 Anonymous Users)
0 Members:
|
 |
 |
 |
|