Barspell Test
Barspell Test
jeez it's so crazy that it's getting the same % for set bonus @ 3/5 as it is @ 5/5... this REALLY is making me rethink my barspell set
Your samples are not statistically different. This would go a lot faster if you'd just use the spikes.
Lakshmi.Byrth said: »
Your samples are not statistically different. This would go a lot faster if you'd just use the spikes.
Suit yourself of course, as I'm not doing the tests, but you can Haste yourself (40% total), get 30% Dual Wield with /NIN, toss on Regen Atma, pull out Lady Bells (D1 clubs), and crank out a thousand trials in the time it takes you to get 100 worm nukes.
Lakshmi.Byrth said: »
Suit yourself of course, as I'm not doing the tests, but you can Haste yourself (40% total), get 30% Dual Wield with /NIN, toss on Regen Atma, pull out Lady Bells (D1 clubs), and crank out a thousand trials in the time it takes you to get 100 worm nukes.
I avoided using any atmas/buff/cruor buffs in order of not messing up the test.I was pretty sure that some ppl would haved called this test inaccurate if I did so.
Naw, if all you're looking for is the annul rate on AF3+2 you can use most Atma (no Annul's or Absorbs Atma) and Cruor buffs without fear. You can't resist spikes to 0 without some kind of annul proc. I'd recommend not getting STR/DEX Cruor buffs though, just because that'll let the Hecteyes last longer. xD
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
Lakshmi.Byrth said: »
Naw, if all you're looking for is the annul rate on AF3+2 you can use most Atma (no Annul's or Absorbs Atma) and Cruor buffs without fear. You can't resist spikes to 0 without some kind of annul proc. I'd recommend not getting STR/DEX Cruor buffs though, just because that'll let the Hecteyes last longer. xD
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
1/16 resist floors damage or something. D: Clickie
Lakshmi.Byrth said: »
Naw, if all you're looking for is the annul rate on AF3+2 you can use most Atma (no Annul's or Absorbs Atma) and Cruor buffs without fear. You can't resist spikes to 0 without some kind of annul proc. I'd recommend not getting STR/DEX Cruor buffs though, just because that'll let the Hecteyes last longer. xD
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
Cool testing, thanks Vaness! Good to see real numbers to prove that I can finally sell my Blessed Briault. :)
I'm definitely more interested in the difference between 3/5 and 4/5, than between 4/5 and 5/5. Relic legs seem obvious to me, but Augur's Gloves seem unlikely to be worth it.
As far as the spikes, I think it's a good idea for set bonus proc rate, since you'll always see that even for low damage. However, it's no good for testing things like MDB or benefit from resistance levels, because the damage is too low. You might add 5 more MDB and the spikes are still doing the same damage, so I'm glad you did the initial testing the way you did it.
I'd be willing to take out the Kraken for some blaze spikes testing if someone can point me in the direction of a good parser for capturing this kind of thing? I once wrote an extension to KParser to test some avatar stats, but it was a lot of hassle and I'd rather not go through it again if I don't have to...
I'm definitely more interested in the difference between 3/5 and 4/5, than between 4/5 and 5/5. Relic legs seem obvious to me, but Augur's Gloves seem unlikely to be worth it.
As far as the spikes, I think it's a good idea for set bonus proc rate, since you'll always see that even for low damage. However, it's no good for testing things like MDB or benefit from resistance levels, because the damage is too low. You might add 5 more MDB and the spikes are still doing the same damage, so I'm glad you did the initial testing the way you did it.
I'd be willing to take out the Kraken for some blaze spikes testing if someone can point me in the direction of a good parser for capturing this kind of thing? I once wrote an extension to KParser to test some avatar stats, but it was a lot of hassle and I'd rather not go through it again if I don't have to...
I gain an additional 1 bar effect using Augur's over the AF+2. I'll take any additional increase in % proc on AF+2 over this. I overcure half the time in Abyssea anyway.
If you choose to use Augur's, a Zenith Crown with 8 enhancing magic augmented on it would be useful in the head slot.
If you choose to use Augur's, a Zenith Crown with 8 enhancing magic augmented on it would be useful in the head slot.
I tested the 5/5 set bonus in Abys-Atto, still had my kparse file if you wanted to have a look at it (http://www.megaupload.com/?d=LUP2XBRZ)...
I'm just missing Augur body (for complete enhancing set), although I could test 3/5, 4/5 and 5/5 again next time I'm on.
5/5 set bonus procrate w/ aff:misery (so that I can confirm there was no cureskin on at any time). Spoilered below.
EDIT:// Derp derp, Anhur robe.. not augur
I'm just missing Augur body (for complete enhancing set), although I could test 3/5, 4/5 and 5/5 again next time I'm on.
5/5 set bonus procrate w/ aff:misery (so that I can confirm there was no cureskin on at any time). Spoilered below.
EDIT:// Derp derp, Anhur robe.. not augur
Fenrir.Aleste said: »
I tested the 5/5 set bonus in Abys-Atto, still had my kparse file if you wanted to have a look at it (http://www.megaupload.com/?d=LUP2XBRZ)...
I'm just missing Augur body (for complete enhancing set), although I could test 3/5, 4/5 and 5/5 again next time I'm on.
5/5 set bonus procrate w/ aff:misery (so that I can confirm there was no cureskin on at any time). Spoilered below.
I'm just missing Augur body (for complete enhancing set), although I could test 3/5, 4/5 and 5/5 again next time I'm on.
5/5 set bonus procrate w/ aff:misery (so that I can confirm there was no cureskin on at any time). Spoilered below.
So you have to wonder if you want to rely on the total randomness of the set procc rate or go for resist.I personally would go for the set #2 or #3.
Augur's hands only add 1 resist, but it looks like it bring it to Tier 2 or something. 1/2 resist are alot more frequent then others.
Sylph.Starstrukk said: »
Lakshmi.Byrth said: »
Naw, if all you're looking for is the annul rate on AF3+2 you can use most Atma (no Annul's or Absorbs Atma) and Cruor buffs without fear. You can't resist spikes to 0 without some kind of annul proc. I'd recommend not getting STR/DEX Cruor buffs though, just because that'll let the Hecteyes last longer. xD
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
1/16 resist floors damage or something. D: Clickie
Sorry, but did you have Phalanx up? I only ever got full and half damage spikes when I went out to test, I think.
oh and augur's body doesn't give you enhancing <.<
As far as I know? >.>
As far as I know? >.>
I've never put any faith in the set-bonus proc rate... it's kinda lame.
I normally barspell in 4/5 af3+2 w/ cleric pants (beneficus, coll torque, augmenting earring, merciful cape, cascade belt)... I don't honestly feel the (predicted) 2% procrate would be worth the loss of the extra potency...
EDIT:// Derp derp, Anhur robe.. not augur
I normally barspell in 4/5 af3+2 w/ cleric pants (beneficus, coll torque, augmenting earring, merciful cape, cascade belt)... I don't honestly feel the (predicted) 2% procrate would be worth the loss of the extra potency...
EDIT:// Derp derp, Anhur robe.. not augur
Vaness said:
Augur's hands only add 1 resist, but it looks like it bring it to Tier 2 or something. 1/2 resist are alot more frequent then others.
If this brought you up a tier, you could get a Merciful Cape to get you there and utilize the +2 gloves.
Phoenix.Bohgo said: »
Vaness said:
Augur's hands only add 1 resist, but it looks like it bring it to Tier 2 or something. 1/2 resist are alot more frequent then others.
If this brought you up a tier, you could get a Merciful Cape to get you there and utilize the +2 gloves.
and never any augmenting earring on AH lol
Otherwise I prolly would do the switch.
Fenrir.Aleste said: »
I've never put any faith in the set-bonus proc rate... it's kinda lame.
I normally barspell in 4/5 af3+2 w/ cleric pants (beneficus, coll torque, augmenting earring, merciful cape, cascade belt)... I don't honestly feel the (predicted) 2% procrate would be worth the loss of the extra potency...
EDIT:// Derp derp, Anhur robe.. not augur
I normally barspell in 4/5 af3+2 w/ cleric pants (beneficus, coll torque, augmenting earring, merciful cape, cascade belt)... I don't honestly feel the (predicted) 2% procrate would be worth the loss of the extra potency...
EDIT:// Derp derp, Anhur robe.. not augur
But that robe wouldn't be better the af3 body +2
Since af3 +2 gives + 10MDB on the barspell.
That robe would just had like 3-4 resist =(
Would be great for rdm phalanx though!
Lakshmi.Byrth said: »
Sylph.Starstrukk said: »
Lakshmi.Byrth said: »
Naw, if all you're looking for is the annul rate on AF3+2 you can use most Atma (no Annul's or Absorbs Atma) and Cruor buffs without fear. You can't resist spikes to 0 without some kind of annul proc. I'd recommend not getting STR/DEX Cruor buffs though, just because that'll let the Hecteyes last longer. xD
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
When I was testing Zanshin on RDM with a Lady Bell, I kept the same EM mandy alive for 2 hours <_<;
1/16 resist floors damage or something. D: Clickie
Sorry, but did you have Phalanx up? I only ever got full and half damage spikes when I went out to test, I think.
Nope, WHM/NIN.
Thats true, if anything I'd collect it for boost-spells ;D
Cerberus.Vaness said: »
I though so!
But that robe wouldn't be better the af3 body +2
Since af3 +2 gives + 10MDB on the barspell.
That robe would just had like 3-4 resist =(
Would be great for rdm phalanx though!
But that robe wouldn't be better the af3 body +2
Since af3 +2 gives + 10MDB on the barspell.
That robe would just had like 3-4 resist =(
Would be great for rdm phalanx though!
Quote:
this takes me to the point where I think the set bonus rate is totally random.Those are on big numbers of spell casted and isn't as effective when you fight a monster that will cast 4 to 6 spell max.
It's just statistically uncertain.
Aleste's test: 21 procs out of 203 samples = 10.3% proc rate with a CI range of 6.9% to 15.3%.
Vaness's test: 13 procs out of 104 samples = 12.5% proc rate with a CI range of 7.5% to 20.2%.
Combining the two tests: 34 procs out of 307 samples = 11.1% proc rate with a CI range of 8.0% to 15.1%.
You'd need about 3000 samples total to get down to about +/- 1% (need a relatively low sample size because the proc rate is so low).
If we assume that the actual proc rate is 10%, and that the proc rate scales at twice the rate as is assumed for other sets (ie: rather than 1%/2%/3%/5%, it's 2%/4%/6%/10%) then the gain of going form 4/5 to 5/5 is an increase in proc rate of 4%.
This in turn reduces damage from 94% D to 90% D, where D is the average damage taken after factoring in spell resists, MDB, etc. That works out to a 4.25% reduction in average damage taken.
Then you'd need to determine the effect of +20 resist on average D, and whether it's more than 4.25%.
If we assume that the +resist rate (m.eva applied to a single element) is used directly counter to the mobs +m.acc, and make use of the m.acc tests that used enspells to determine resist rates (not sure these tests were ever completed, so going off of vague semi-conclusions), we can evaluate the value of +20 resist.
A 'missed' spell does reduced damage rather than no damage. A 95% m.acc rate would mean 95% hit for full damage while 5% hit for half or quarter damage. +20 m.eva should reduce m.hit rate by 10%.
Worst case is that it does nothing. That is, the mobs have capped acc even with the better resist, or floored accuracy without the better resist.
In general, can expect it to reduce magic accuracy rate from 95% to 85% at the top end, down to 30% to 20% at the bottom end.
Overall damage at 95% should be (working only with full/half/quarter damage) 0.974375. At 85%, it should be 0.919375. Overall, 5.65% reduction in damage.
Overall damage at 30% should be 0.5275. At 20%, it should be 0.44. Overall, 16.6% reduction in damage.
So it would appear (assuming my understanding of resist effects and damage rates is correct) that as long as the additional resist from the relic legs has an effect, they're better than af3+2 pants. However for extremely accurate mobs, or mobs that have floored m.hit rates without the extra bar strength, the af3+2 legs would be better.
Other notes: Reduction from 95% to 20% m.hit rate (perfect application of 150 resist) would reduce damage taken by 55%. The model I used does not appear to adequately explain the damage from focused resist sets (eg: 300+ resist vs Tiamat, etc), though it might work out if I included 1/8 and 1/16 resists as well.
[+]
Fenrir.Motenten said: »
Quote:
this takes me to the point where I think the set bonus rate is totally random.Those are on big numbers of spell casted and isn't as effective when you fight a monster that will cast 4 to 6 spell max.
It's just statistically uncertain.
Aleste's test: 21 procs out of 203 samples = 10.3% proc rate with a CI range of 6.9% to 15.3%.
Vaness's test: 13 procs out of 104 samples = 12.5% proc rate with a CI range of 7.5% to 20.2%.
Combining the two tests: 34 procs out of 307 samples = 11.1% proc rate with a CI range of 8.0% to 15.1%.
You'd need about 3000 samples total to get down to about +/- 1% (need a relatively low sample size because the proc rate is so low).
If we assume that the actual proc rate is 10%, and that the proc rate scales at twice the rate as is assumed for other sets (ie: rather than 1%/2%/3%/5%, it's 2%/4%/6%/10%) then the gain of going form 4/5 to 5/5 is an increase in proc rate of 4%.
This in turn reduces damage from 94% D to 90% D, where D is the average damage taken after factoring in spell resists, MDB, etc. That works out to a 4.25% reduction in average damage taken.
Then you'd need to determine the effect of +20 resist on average D, and whether it's more than 4.25%.
If we assume that the +resist rate (m.eva applied to a single element) is used directly counter to the mobs +m.acc, and make use of the m.acc tests that used enspells to determine resist rates (not sure these tests were ever completed, so going off of vague semi-conclusions), we can evaluate the value of +20 resist.
A 'missed' spell does reduced damage rather than no damage. A 95% m.acc rate would mean 95% hit for full damage while 5% hit for half or quarter damage. +20 m.eva should reduce m.hit rate by 10%.
Worst case is that it does nothing. That is, the mobs have capped acc even with the better resist, or floored accuracy without the better resist.
In general, can expect it to reduce magic accuracy rate from 95% to 85% at the top end, down to 30% to 20% at the bottom end.
Overall damage at 95% should be (working only with full/half/quarter damage) 0.974375. At 85%, it should be 0.919375. Overall, 5.65% reduction in damage.
Overall damage at 30% should be 0.5275. At 20%, it should be 0.44. Overall, 16.6% reduction in damage.
So it would appear (assuming my understanding of resist effects and damage rates is correct) that as long as the additional resist from the relic legs has an effect, they're better than af3+2 pants. However for extremely accurate mobs, or mobs that have floored m.hit rates without the extra bar strength, the af3+2 legs would be better.
Other notes: Reduction from 95% to 20% m.hit rate (perfect application of 150 resist) would reduce damage taken by 55%. The model I used does not appear to adequately explain the damage from focused resist sets (eg: 300+ resist vs Tiamat, etc), though it might work out if I included 1/8 and 1/16 resists as well.
He's basically saying on higher level NMs, using the AF3+2 legs would be better as opposed to relic legs. Supranational, basically.
Austar
If I keep my relic legs in storage forever
Will the world end?
nvm, forgot my whole whm atm is in storage forever :(
If I keep my relic legs in storage forever
Will the world end?
nvm, forgot my whole whm atm is in storage forever :(
Cerberus.Wolfshadow said: »
Austar
If I keep my relic legs in storage forever
Will the world end?
nvm, forgot my whole whm atm is in storage forever :(
If I keep my relic legs in storage forever
Will the world end?
nvm, forgot my whole whm atm is in storage forever :(
Maybe my comon sense is ***.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
*** it, I'm just gonna use Cleric's pants.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
*** it, I'm just gonna use Cleric's pants.
Cerberus.Vaness said: »
Maybe my comon sense is ***.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
*** it, I'm just gonna use Cleric's pants.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
*** it, I'm just gonna use Cleric's pants.
keep in mind, watching your tank nomnom a Thundaga V for 0 is priceless.
Carbuncle.Tweeek said: »
Cerberus.Vaness said: »
Maybe my comon sense is ***.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
*** it, I'm just gonna use Cleric's pants.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
*** it, I'm just gonna use Cleric's pants.
keep in mind, watching your tank nomnom a Thundaga V for 0 is priceless.
Vaness said:
Maybe my comon sense is ***.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
5/5 +2 10% procc rate
on a mob that may cast 5 to 6 max heavy spell, that aprox give you a 0.8% chance that set procc in those 6 spell.
Incorrect.
Chance of at least one set proc with 4/5 AF3 vs 6 spellcasts: 31% (maybe one proc every 3 fights)
Chance of at least one set proc with 5/5 AF3 vs 6 spellcasts: 47% (maybe one proc every other fight)
Assuming 6% proc rate with 4/5 and 10% proc rate with 5/5.
All FFXI content and images © 2002-2024 SQUARE ENIX CO., LTD. FINAL
FANTASY is a registered trademark of Square Enix Co., Ltd.