Does Triple Attack have a cap?

[Atomic Skull]

Was it ever determined if Triple attack % has a cap? You can get some pretty insane amounts of triple attack on THF now. I have 33% triple Attack on THF now from gear and job traits and with a few more difficult pieces and job points I could get it up to 45% or so. I'm wondering if that is a waste of time due to a cap.

Another question, is adding double attack into a build with high triple attack actually detrimental? If double attack procs, triple attack can't proc and if you have a high amount of triple attack would double attack actually end up slowing you down? I guess it would depend on how double triple etc. attack were calculated. If it calculates for all of them and just takes the highest one that proced then it wouldn't have a detrimental effect. If it does them in order and then stops checking when one procs then it would.

[Fwahm]

The game checks for TA before DA, so while adding DA will have a diminishing effect the higher your TA, it will never lower your damage.

If you're worried about a triple attack cap, go test in abyssea; atma can give you up to 32% bonus triple attack, which will easily hit said cap if it's relevant to you. With that much extra TA, it shouldn't take very long to see if there's a relevant cap at all.

[Byrthnoth]

Double Attack has no cap (100% DA is possible), and it's probably the same for Triple Attack.

[Atomic Skull]

So I guess triple attack is the new haste. Basically it allows you to break the 20% delay reduction cap by adding additional attack rounds. With 33% triple attack and max haste my THF attacks (much) faster than a MNK with HF and can actually self skillchain darkness, though not consistently. With a few more harder to get pieces and max job points I could have around 48% and I bet I could do consistent self darkness in that case.


Another question, If I have 35% TA and I have the choice between Raider's Earring (TA+1) or Brutal Earring (DA+5) is the TA+1 a better option than the DA+5?

[Fwahm]

You would need somewhere in the neighborhood of 60% TA for 1 TA fo beat 5 DA.

[Protey]

You would need somewhere in the neighborhood of 60% TA for 1 TA fo beat 5 DA.

That doesn't make sense. The higher TA you have, any additional will have diminishing returns. Did you mean 60% TA for 5 DA to beat 1 TA?

[Byrthnoth]

Nope~!

If you ignore QA:
1 + TA%*2 + (1-TA%)*DA% = Attacks per round

1 + (TA%+1%)*2 + (1-TA%-1%)*DA% = 1 + TA%*2 + (1-TA%)*(DA%+5%)

If DA%=0:
1 + TA%*2 + 2% = 1 + TA%*2 + (1-TA%)*5%
2%/5% = (1-TA%)
40% = (1-TA%)
TA% = 60%

That's why he said 60% was the tradeoff point between 1% TA and 5% DA.

Spoiler: show

For the full exchange rate between 1% TA and 5% DA:
1 + 2*TA% + 2% + DA% - DA%*TA% - 1%*DA% = 1 + 2*TA% + DA% - DA%*TA% - 5%*TA% + 5%
5%*TA% = 1%*DA% + 3%
multiply by 100 for simplicity
5*TA% = DA% + 3

5*60% = DA% + 3
3 - 3 = 0% DA (okay, this checks out with our earlier simplified example)
5*TA% = 20% + 3
TA% = 64% (so it turns out that having more DA actually slightly devalues TA because additional TA detracts from DA)

Regardless, not going to hit that 1% TA / 5% DA tradeoff point any time soon.
TAb+TA% = DAb*DA% + 100*(DAb - TAb*2)

[Protey]

Nope~!

If you ignore QA:
1 + TA%*2 + (1-TA%)*DA% = Attacks per round

1 + (TA%+1%)*2 + (1-TA%-1%)*DA% = 1 + TA%*2 + (1-TA%)*(DA%+5%)

If DA%=0:
1 + TA%*2 + 2% = 1 + TA%*2 + (1-TA%)*5%
2%/5% = (1-TA%)
40% = (1-TA%)
TA% = 60%

That's why he said 60% was the tradeoff point between 1% TA and 5% DA.

Spoiler: show

For the full exchange rate between 1% TA and 5% DA:
1 + 2*TA% + 2% + DA% - DA%*TA% - 1%*DA% = 1 + 2*TA% + DA% - DA%*TA% - 5%*TA% + 5%
5%*TA% = 1%*DA% + 3%
multiply by 100 for simplicity
5*TA% = DA% + 3

5*60% = DA% + 3
3 - 3 = 0% DA (okay, this checks out with our earlier simplified example)
5*TA% = 20% + 3
TA% = 64% (so it turns out that having more DA actually slightly devalues TA because additional TA detracts from DA)

Regardless, not going to hit that 1% TA / 5% DA tradeoff point any time soon.
TAb+TA% = DAb*DA% + 100*(DAb - TAb*2)

I didn't argue the 60% exchange. I'm saying that 1TA will have a lot more value at a low TA rate than at a higher TA rate; so at very high TA rate (60TA in this case) it should be 5DA beats 1TA, not 1TA beats 5DA.

[Fwahm]

No, that's not true. As you increase your triple attack, the static number of swings that each triple attack gives (+2 in an unmodified round of 100) stays constant, but the number of swings each double attack gives goes down because if you triple attack, you can't double attack.

1% TA always gives +2 swings per 100 unmodified swings (ignoring QA and Mythic AM3), but DA gives (1-TA) * DA extra swings per 100. If you have 50% TA, then each DA will only increase your number of swings by 0.5 per unmodified 100, and if you have 100% TA, then each DA will increase your swings by 0.

[Protey]

No, that's not true. As you increase your triple attack, the static number of swings that each triple attack gives (+2 in an unmodified round of 100) stays constant, but the number of swings each double attack gives goes down because if you triple attack, you can't double attack.

1% TA always gives +2 swings per 100 unmodified swings (ignoring QA and Mythic AM3), but DA gives (1-TA) * DA extra swings per 100. If you have 50% TA, then each DA will only increase your number of swings by 0.5 per unmodified 100, and if you have 100% TA, then each DA will increase your swings by 0.

Ah, I see. I was thinking about additional TA effectiveness % going down as TA rate goes up and not about TA reducing the probability of DA to proc.