I've always had really good results in optimal conditions with Ryu. Curious what your results say for it. If you can, can you also provide the inputs you're using?
I haven't been able to manually update my copy of the simulation to include ODD/ODT for Primes. I'm using Apex Bat as the target, along with Dia III / Haste II / SV +9 Songs (HM, Minuet x3, Aria) / +8 Chaos & SAM with 7 for rolls, Crooked on Chaos, and Bolstered GEO +10 with Indi Fury & GEO Frailty, with 25% potency. Didn't use Entrusts.
AM3 active
Angon & Hasso (/SAM)
Master level 50
WSing at 1k (holding TP seems to be DPS loss according to simulation)
I'm also using Ephramad's and R25 as my Ody rank.
No DT requirements.
I get Ryu just shy of 9% ahead of Gae Buide IV (again, no ODD/ODT taken into consideration)
All that being said. I know that's not always going to be a realistic scenario. I've always thought Ryu would be a beast when you can overcome the attack penalty. The TP gain with AM3 up, as well as jumps, is insane. Though I feel that also benefits Prime more in practice, as you can weave Jumps in. Prime WS benefits greatly from this due to TP scaling on Diarmuid, while Drakesbane only gains increase crit chance.
Sigurd's Descendants: The Art Of Dragon Slaying.
|
Sigurd's Descendants: The Art of Dragon Slaying.
Ryu is my absolute favorite weapon in the game but I am having a hard time seeing how it is ahead of Gae Buide in almost any scenario.
I’ll let the mathers math. I need to stress that in practice, I do not see this ever being a likely scenario. I have Apex Bat chosen as the target in this instance. It has a listed Defense of 1338 and 289 VIT. I am completely aware of Drakesbane's shortcomings due to not scaling with TP and the unfortunate attack penalty.
Most targets that are worth mentioning are going to have significantly higher Defense, which is going to absolutely change results. This is moreso an exercise for curiosities sake :) Edit - I actually played with the numbers a little and increased Defense to 5000 and adjusted buffs to be non-2hr (marcato instead of SV, BoG instead of Bolster), and still got similar results. Ryu edging out by a few % which I would anticipate being made up for if ODD/ODT were taken into consideration for Prime. But I think more importantly, Jumps heavily favor Prime due to TP scaling WS. Curious what you guys think / see in practice. Random DRG thought: you should gain your wyvern's wings as a visual effect on your back when you use Spirit Surge, similar to August's wings mode, or Tartarus Plaitmail wings, since you're merging with your pet for the duration.
No idea why that just came to my mind.
[+]
Veydal1 said: » I need to stress that in practice, I do not see this ever being a likely scenario. I have Apex Bat chosen as the target in this instance. It has a listed Defense of 1338 and 289 VIT. I am completely aware of Drakesbane's shortcomings due to not scaling with TP and the unfortunate attack penalty. Most targets that are worth mentioning are going to have significantly higher Defense, which is going to absolutely change results. This is moreso an exercise for curiosities sake :) Edit - I actually played with the numbers a little and increased Defense to 5000 and adjusted buffs to be non-2hr (marcato instead of SV, BoG instead of Bolster), and still got similar results. Ryu edging out by a few % which I would anticipate being made up for if ODD/ODT were taken into consideration for Prime. But I think more importantly, Jumps heavily favor Prime due to TP scaling WS. Curious what you guys think / see in practice. If you increased def to 5000 you pushed it to different extremum, where crit hits add massive damage over non crit hits, since crit hits are static +1 pDIF, it's extremely potent when you have for example 1.0 cRatio. I was using Apex Toad as target for high buff, II Samurai Roll, CC IV Chaos Roll, Light shot, Job Bonus, Angon, Hasso, Haste II, Dia III, Boost-STR, Song+7 Honor, Minuet V and IV, Aria, Marcato, None of GEO buffs/debuffs.
[+]
Yeah it would be weird if they pushed us back to abyssea era where crit WSs were massively ahead of almost everything else outside of brewing. Not the same math, but similar results.
That's a fair point regarding crits being favored in that scenario.
Thanks for clarifying what buffs you used, I think '2hr zerg' scenario when I see 'high buffs.' Using those same buffs (with the exception of +9 on songs instead of +7, because what BRD is going to have +7 songs when they have Aeonic and Prime...) I get Gae Buide edging out Ryu by just shy of 5%. This assumes AM3 as well. Other scenarios (in addition to using same buffs as above) - BRD + COR + Idris Indi Fury (no 2hr) - Gae Buide by 1% BRD + COR + Idris Indi Fury (2hrs used) - Gae Buide by 1% I will say to please take this with a grain of salt. I've said it a few times, but it's worth repeating that in practice, Gae Buide should do even better with Jumps taken into consideration and naturally overflowing TP if you're not hitting that WS macro perfectly every time. Additionally, from my understanding the simulation does not account for TP return from WS when calculating average TP when WSing. For example, the average WS return for Daimuid is 291 with a vanilla WS set (nothing special with STP). And average TP per attack round returns 624 TP. Meaning the average TP for WS should be ~1539 (291 + 624 + 624). However, instead the simulation shows average TP for WS as 1248 (624 + 624, no initial TP return from WS). This should favor Gae Buide even further. Veydal1 said: » Using those same buffs (with the exception of +9 on songs instead of +7, because what BRD is going to have +7 songs when they have Aeonic and Prime...) I use +7 because Aria isn't coded to stop at +7 and still adds more PDL at +8 and +9. I need to stop being lazy and fix it I guess.
[+]
Veydal1 said: » Additionally, from my understanding the simulation does not account for TP return from WS when calculating average TP when WSing. For example, the average WS return for Daimuid is 291 with a vanilla WS set (nothing special with STP). And average TP per attack round returns 624 TP. Meaning the average TP for WS should be ~1539 (291 + 624 + 624). However, instead the simulation shows average TP for WS as 1248 (624 + 624, no initial TP return from WS). This should favor Gae Buide even further. I dont think that's the case, at least not for simulation. Simulation calculates TP return on each simulated WS and it's the starting point for next round. Here the code returns TP returned from a WS Code if simulation:
# Cap damage at 99999 for simulations if enabled
if player.abilities.get("99999", False):
total_damage = 99999 if total_damage > 99999 else total_damage
return(total_damage, tp_return)Here it adds a TP threshold, at which you have made WS, to the list. Code avg_ws_tp.append(tp) Here it resets TP after WS to TP returned from a WS and then it is used at the beginning of a new round. Code tp = ws_sim[1] Here it writes down what you see as avg TP/WS, which is just an avg from the list of all TP thresholds you were WSing at. Code Average TP/WS: {np.average(avg_ws_tp):4.0f}Sim is set to simulate 24h of fighting I think, that's why your avg will very rarely be different, but you can see few TP difference between avg TP/WS. If you change the code to run the simulation for 1minute of fighting, you will get very different result for that Average TP/WS on each different simulation. Wouldn't the average TP/WS be above what's being reported due to the TP return from WS? Just doesn't seem to add up to me. A 624 average TP return from an attack round = 2 attack rounds to WS with 1k minimal threshold. If TP from WS was taken into consideration, the average would be higher than what is being reported. The fact that it runs for 24 hours only makes me even more skeptical that it's being taken into consideration properly due to the large sample size.
For DRG you also have to consider Conserve TP which will proc randomly.
I think a lot is being missed in the averages here... If your average TP per round is 624 then many rounds will be 375, some will be 750, some will be 1000+. You will not always WS at the same TP, sometimes you will proc a QA and end up at 1800 TP. Sometimes you can proc CTP and a TA on your WS and start at a higher TP after WS. When the sim says your average TP/WS, it's not saying you ALWAYS WS at 1248, it's saying that of the 100,000 WS it did, the AVERAGE TP while WSing was 1248. Sometimes it was 1050, sometimes it was 1470, sometimes it was 2057. On average, it was 1248.
[+]
Carbuncle.Maletaru said: » If your average TP per round is 624 then many rounds will be 375, some will be 750, some will be 1000+. You will not always WS at the same TP, sometimes you will proc a QA and end up at 1800 TP. Sometimes you can proc CTP and a TA on your WS and start at a higher TP after WS. When the sim says your average TP/WS, it's not saying you ALWAYS WS at 1248, it's saying that of the 100,000 WS it did, the AVERAGE TP while WSing was 1248. Sometimes it was 1050, sometimes it was 1470, sometimes it was 2057. On average, it was 1248. Exactly and if for example (simple numbers just to show it better) You have 300TP per hit and you have 50% DA and avg tp round is 450TP, you will NEVER get that 450TP, you will either get 300 or 600. So saying your actual WS TP threshold is for example 3x450 is impossible, because you will never get 1350TP, you will get 300 600 900 1200 1500. Maybe I'm being dense. If your average TP return from an attack round is 500, then on average you're going to WS with 1000 TP after two attack rounds. But if that's what's being reported (as is the case with the simulation) then that's not taking into consideration TP return from WS. If TP return from WS was also taken into consideration, then your average TP threshold for WS would be higher than the average TP to hit over 1000 TP from attack rounds alone.
Either way, I'll leave the topic at that. I like Dragoon. Mine's named Rygor. Veydal1 said: » Maybe I'm being dense. If your average TP return from an attack round is 500, then on average you're going to WS with 1000 TP after two attack rounds. But if that's what's being reported (as is the case with the simulation) then that's not taking into consideration TP return from WS. If TP return from WS was also taken into consideration, then your average TP threshold for WS would be higher than the average TP to hit over 1000 TP from attack rounds alone. Either way, I'll leave the topic at that. I like Dragoon. Mine's named Rygor. Not sure how I can explain this to you in english (im not native, so detailed posts are harder to write), but I will try once again. To make it easier, I added more statistics to simulation's report: Code Total time: 100.0 h
Time/Attack: 1.640 s
Average Dmg/Attack: 4474.8
Physical: 4474.8 (100.0%)
Magical: 0.0 ( 0.0%)
Average Dmg/WS: 60907.7
Average TP/WS: 1377 (+250 TP Bonus)
Median TP return on WS: 362.0
All the TP/WS thresholds: 1046,1066,1067,1086,1087,1088,1107,1108,1109,1110,1127,1128,1129,1148,1149,1150,1168,1169,1170,1188,1189,1190,1191,1209,1210,1211,1212,1229,1230,1231,1232,1250,1251,1252,1253,1270,1271,1272,1273,1274,1291,1292,1293,1448,1449,1468,1469,1470,1489,1490,1491,1492,1509,1510,1511,1512,1529,1530,1531,1532,1533,1550,1551,1552,1570,1571,1572,1573,1591,1592,1593,1594,1611,1612,1613,1614,1632,1633,1634,1635,1652,1653,1654,1655,1673,1674,1675,1830,1831,1850,1851,1852,1871,1872,1873,1874,1891,1892,1893,1912,1913,1914,1915,1932,1933,1934,1935,1953,1954,1955,1974,1975,1976,1994,1995,2015,2016,2017,2035,2036,2056,2057,2233,2254,2255,2274,2275,2295,2296,2315,2316,2336,2337,2356,2357,2376,2377,2397,2398,2417,2418,2438,2439
Median TP/WS: 1189
Total Damage: 5597.2M damage (Total DPS: 15547.6)
TP Damage: 528.1M damage (TP DPS: 1467.0; 9.4%)
WS Damage: 5069.0M damage (WS DPS: 14080.6; 90.6%)Now you can see "All the TP/WS thresholds", so list of all unique TP thresholds for all the simulated WSs. You can also see "Median TP/WS", so the most common TP threshold you would WS at. You can also see "Median TP return on WS", so most common TP value returned after WSing. You can now notice, that "Average TP/WS" is (like name/description suggests) just an average number of all TP thresholds you were WSing at, but it's not an actual TP threshold. So in other words if you make 5 WSs, four at 1100TP and one at 1500TP, then your average would be 1180TP, but it doesn't mean any of five WSs were made at 1180TP. Same applies for TP gain. If you make 5 TP rounds: 1: single hit 300TP 2: single hit 300TP 3: da 600TP 4: da 600TP 5: ta 900TP and your average TP per round is 540TP, it doesn't mean that you get 540TP in any of your rounds. I hope this will clear things up for you. I tried Ryu for lulz a few weeks back in dynamis (wave 3 boss is the only fight Ryu could hope to he a contender on imo). It failed miserably compared to Gae Buide.
With all the buffs / debuffs up, I cap damage with Diarmud at 1750tp in wave 3 (with tp earring. Didnt matter if prime am was up or not). Ryu at best was getting 50k (r25 gletis etc) at 1kish tp and that would be great if we're the rule and not the exception. It was the exception. Unfortunately the inconsistent nature of crits procing on drakes and the att penalty just hold it back too much. Jumping with ryu is a dps loss aside from getting am 3 up ( faster to to just swing with it) and the increased ws frequency just didn't matter. If drg didn't gain so much tp naturally, it may be a different story but jumps, hi DA and 4 hit build with Gae Buide wins the race. Ryu drakes spam vs trishula is closer and Ryu can pull ahead there but but it looses to the prime. Thanks for taking time to explain. May I ask what your simulation showed for average TP return per attack round? Is it ~688? If so, then I don't think the TP return from WS is being calculated for Average TP/WS.
The average TP return on WS is 362 according to your results. Your average TP/WS is 1377. What I would expect to see as your average TP per attack round would then be 507 (if you have minimum TP threshold set to 1000). Again, I understand these are AVERAGES and won't be the real values for each attack round. With those values, I expect (minimum TP threshold = 1000) - Average TP return on WS - 362 Average TP return per attack round = 507 Average attack rounds per WS = 2 Average TP/WS = 1377 (362 + 507 + 507) If TP return from WS wasn't taken into consideration, the above would result in an Average TP/WS = 1014 (507 + 507) The latter is what I'm seeing in the simulation. It's only taking into consideration the TP from attack rounds, NOT the TP return from WS. If it did, the average TP/WS reported would be higher or lower than displayed, based on target minimum threshold and average TP returned per attack round. Veydal1 said: » With those values, I expect (minimum TP threshold = 1000) - Average TP return on WS - 362 Average TP return per attack round = 507 Average attack rounds per WS = 2 Average TP/WS = 1377 (362 + 507 + 507) You're really not getting the average TP round thing. If average TP round is 507 (let's just say) that DOES NOT MEAN you will hit 507 per round. In fact, there's an extremely high chance you will NEVER hit 507 in an individual round. Sims aren't just taking an average of your TP round then applying that repeatedly until you're ready to WS. They're SIMulating the actual attack rounds. It will SIMulate the WS, check if conserve TP proc'd, if you got a multi-attack on the WS, missed any hits, crit, etc. Then it will calculate your resulting TP from the results to all that. After that, it will SIMulate your attack round (how many hits, what's the TP, did it crit), add that to your TP (what remained after your WS), then it will keep going until it reaches your target TP and fire off another WS. It will do this for 24h and then it can give you an average WSD, average TP/WS, average TP per round, etc. These are calculated AFTER all the simulations are done, not before. These averages are not used in any way to calculate any of the simulation data.
[+]
Veydal1 said: » Again, I understand these are AVERAGES and won't be the real values for each attack round. I don't think you understand at all, because I have just explained to you, that TP return from WS is a base TP for each new round after WS. It might be my fault to not being able to explain this to you. I'm lacking vocabulary and probably not being face tof ace is also the problem here, because I could read your facial expression and know on which details I should focus. Nevertheless I will try one more time before I will give up. Veydal1 said: » With those values, I expect (minimum TP threshold = 1000) - Average TP return on WS - 362 Average TP return per attack round = 507 Average attack rounds per WS = 2 Average TP/WS = 1377 (362 + 507 + 507) You need to completely erase this from your brain. You are trying to apply avg numbers to achieve simulation. It's simply wrong. To make it easier again. I have made very short simulation and I added even more statistic to report and I also replaced all the gear with Regain to make it simpler. Code Total time: 0.0 h
Time/Attack: 1.640 s
Average Dmg/Attack: 4526.1
Physical: 4526.1 (100.0%)
Magical: 0.0 ( 0.0%)
Average Dmg/WS: 56394.5
Average TP/WS: 1366 (+250 TP Bonus)
Median TP return on WS: 367.0
TP return on WS flow: [367, 348, 407, 387, 348, 348, 387, 407, 98, 367, 308, 387, 367, 387, 98, 348, 328, 367, 348, 387, 367, 387, 367]
All the TP/WS thresholds: 1066,1106,1126,1145,1165,1485,1504,1516,1524,1614,1923,2223
Median TP/WS: 1165
TP flow: ['tp from round: 758', 'total tp: 758', 'tp from round: 758', 'total tp: 1516', 'tp from round: 758', 'total tp: 1126', 'tp from round: 758', 'total tp: 1106', 'tp from round: 758', 'total tp: 1165', 'tp from round: 379', 'total tp: 766', 'tp from round: 758', 'total tp: 1524', 'tp from round: 1137', 'total tp: 1485', 'tp from round: 1137', 'total tp: 1485', 'tp from round: 758', 'total tp: 1145', 'tp from round: 1516', 'total tp: 1923', 'tp from round: 758', 'total tp: 856', 'tp from round: 0', 'total tp: 856', 'tp from round: 758', 'total tp: 1614', 'tp from round: 379', 'total tp: 746', 'tp from round: 758', 'total tp: 1504', 'tp from round: 758', 'total tp: 1066', 'tp from round: 758', 'total tp: 1145', 'tp from round: 379', 'total tp: 746', 'tp from round: 379', 'total tp: 1126', 'tp from round: 379', 'total tp: 766', 'tp from round: 379', 'total tp: 1145', 'tp from round: 758', 'total tp: 856', 'tp from round: 758', 'total tp: 1614', 'tp from round: 758', 'total tp: 1106', 'tp from round: 379', 'total tp: 707', 'tp from round: 1516', 'total tp: 2223', 'tp from round: 758', 'total tp: 1126', 'tp from round: 758', 'total tp: 1106', 'tp from round: 379', 'total tp: 766', 'tp from round: 758', 'total tp: 1524', 'tp from round: 758', 'total tp: 1126', 'tp from round: 1137', 'total tp: 1524']
Total Damage: 1.4M damage (Total DPS: 14447.0)
TP Damage: 0.1M damage (TP DPS: 1491.8; 10.3%)
WS Damage: 1.3M damage (WS DPS: 12955.2; 89.7%)Just a disclaimer, I used rounding to make it look cleaner, so sometimes some calculation could be 1TP off Ok so lets go step by step We start first round with 0 TP We got 758TP in first round (2hits) We have total of 758TP We got 758TP in second round (2hits) We have total of 1516TP We WS and we got 367TP return We got 758TP in third round (2hits) We have total of 1126TP (758+367 1TP off because of rounding) We WS and we got 348TP return We got 758TP in 4th round (2hits) We have total of 1106TP (758+348) We WS and we got 407TP return We got 758TP in 5th round (2hits) We have total of 1165TP (758+407) We WS and we got 387TP return We got 379TP in 6th round (1hit) We have total of 766TP (387+379) We got 758TP in 7th round (2hits) We have total of 1524TP We WS and we got 348TP return We got 1137TP in 8th round (3hits) We have total of 1485TP (348+1137) We WS and we got 348TP return We got 1137TP in 9th round (3hits) We have total of 1485TP (348+1137) We WS and we got 387TP return We got 758TP in 10th round (2hits) We have total of 1145TP (387+758) We WS and we got 407TP return We got 1516TP in 11th round (4hits, Niqmaddu proc) We have total of 1923TP (407+1516) and so on... I hope you can clearly see now, that TP return from WS is calculated into each new round. I hope you can also see, that "Average TP/WS: 1366" is only an average from all WSs, but it's not an actual WS threshold. Finally I hope you can see, that Avg TP/round, which for set I used is 710.9 has nothing to do with actual values per round I was getting (0,379,758,1137 or 1516). I forgot median is not always the most common value, so I used Counter from collections module and since I hope it's clear now, I cleaned whole report and added most common values for TP you WS at, TP return on WS and tp per round. Here is report from simulation of 100h fight.
Code Total time: 100.0 h
Time/Attack: 1.640 s
Average Dmg/Attack: 4368.7
Physical: 4368.7 (100.0%)
Magical: 0.0 ( 0.0%)
Average Dmg/WS: 53777.6
Average TP/WS: 1353 (+250 TP Bonus)
Most common TP return from WS: 367
Most common WS/TP throeshold: 1126
Most common TP per round: 758
Total Damage: 4992.3M damage (Total DPS: 13867.3)
TP Damage: 515.5M damage (TP DPS: 1431.9; 10.3%)
WS Damage: 4476.8M damage (WS DPS: 12435.4; 89.7%)Thanks everyone. I guess my mistake was assuming that you could just add the averages after the fact. It was just a coincidence the #'s worked out when I checked. Ran additional sims and am not seeing that be the case anymore.
Accepted bug on the Relic body being slow to update wyvern HP for the additional damage. Link
Not ground breaking by any stretch but it can be paired with vishap legs to do some extra damage on high jumps. I can't get good tp returns or normal damage on high jump, so I started using it to get extra damage. Slightly More Important: Do breaths, healing or otherwise, suffer from wyvern HP being slow to update? Could see about another bug report if so since this was accepted. Phoenix.Iocus said: » Slightly More Important: Do breaths, healing or otherwise, suffer from wyvern HP being slow to update? Could see about another bug report if so since this was accepted. I also have a vague recollection that flat WHP+ and percent WHP+ worked a bit differently in that regard... One of them might have been faster/more reliable in updating? It's been a loooong time since I messed with it any. For elemental breaths it's not typically relevant since those work on current wyvern HP rather than max. So to boost an Elemental Breath that way you'd have to increase wyvern max HP, and then restore that HP before the breath fired. Which I could maybe see happening sometimes from timely regen ticks, and very rarely otherwise.
[+]
|
All FFXIV and FFXI content and images © 2002-2024 SQUARE ENIX CO., LTD.
FINAL FANTASY is a registered trademark of Square Enix Co., Ltd.