Teras Kasi Archive

---

Nerz wrote:

Just look at the health bar.. it does 3 times the damage shown in combat chat and floating numbers.

If you killed something faster using melee strike, its just because its considerable faster in your template.

---

Three times the damage?!? I'm lucky if I get a 200hp difference between melee strike and advanced combo! I've been watching mob health bars for the past week hoping to see a huge difference but I have not seen it at all. I'm barehanded. What weapon are you using?

Advanced Combo and other specials are not measured in DPS or damage per second. They are measured in damage, PERIOD. Advanced Combo does do roughly 3X the damage PER HIT. These parameters need to be understood beforeMEANINGFUL disagreement can occur. The Advanced Combo damage has been tested ad nauseum by many including yours truly. Watching the target's HAM bar, verbose combat spam and time to kill. DPS would be a totally separate conversation.

Damage of an attack can be measured three ways:

1. Damage per hit (DPH) -- how much damage you do each and every time you score a hit with the attack
2. Damage per second (DPS) -- the long-term average amount of damage you do per second with the attack
3. Damage per action (DPA) -- the amount of damage you do per action point used

These three measures are different for all your attacks. The weakest one (melee hit I think it's called, though I always mix up hit and strike) has the best DPA and the worst DPH, and probably median DPS (though I am not sure of the last one). That is, it does the least damage each time you hit with it, but uses the least action. The other basic hit (strike I think) has a higher DPH, but uses more action per point done, so a worse DPA... and probably has the best DPS in the skill tree. Combo has by far the highest DPH (each individual hit does the most damage), but it has an abysmal DPA (it just eats action) and since it is on a long timer and has a long "animation time" it is a low DPS (though it may not be the lowest of all the moves).

Which one you use depends on your situation. If you are low on action, you want your good DPA move (basic hit). If you are high on action, but the enemy is fresh and has lots of health, so you are going to be hitting a lot of times, you probably want strike, for the good long-term average DPS. If the target is low on health, enough that one combo will take it down, then you're better off doing combo and finishing it in one hit, rather than using several hits to finish it.

One of the fundamental misconceptions that people who think DPS is the "holy grail" of combat have is that your actual DPS in combat each and every second will equal the DPS -- that's simply false. DPS is the true or "parametric" mean (average) damage per second done, taken over many seconds and many attacks. As anyone with basic statistics knows, the average you observe, will depend on the sample size, and small sample sizes often deviate significantly from the average. For example... let's say Melee Strike does 100 DPS (that's just an easy round number, not the real number). Let's say combo does 150 DPS. That's what you'll get over the long run, if you use them for 1,000 seconds each.

However, in a single battle, say you fight one guy equal to your level, and he has 5,000 HAM. If you hit every single time with Strike, you will drop him in 50 seconds (5,0000 HAM / 100 DPS). However, you're going to miss sometimes, so it might take you 60 seconds instead of 50. And beyond this, your DPS will be off by an even greater amount in short periods. What if you had a run of 3 misses in a row over a 3-hit sequence? That's a ZERO DPS amount for that 3-hit sequence. Has that changed your attack's long-term DPS? No. But for those few seconds, your DPS was zilch, zip, nada.

This is the problem with doing things like DPS computations: people assume every second of combat they will exactly equal the average. It won't... it can't. Long-term DPS is just that -- a long-term average... not an instant per-hit automatic value.

C

---

Jack-Aubrey wrote:
>If melee strike is doing slightly less damage, but overall more dps because of its speed compared to combo, than combo really needs an increase in its speed, and maybe an increase in its damage. And MAYBE a fix for this stupid display bug?? wow, that sure would be nice..

---

Well now you are just being unreasonable. I mean... you expect our highest level attack to be our best attack, AND to report the damage it is doing correctly? Come now, might as well ask for the moon...

/sarcasm off

C

---

Chessack wrote:
Damage of an attack can be measured three ways:

1. Damage per hit (DPH) -- how much damage you do each and every time you score a hit with the attack
2. Damage per second (DPS) -- the long-term average amount of damage you do per second with the attack
3. Damage per action (DPA) -- the amount of damage you do per action point used

These three measures are different for all your attacks. The weakest one (melee hit I think it's called, though I always mix up hit and strike) has the best DPA and the worst DPH, and probably median DPS (though I am not sure of the last one). That is, it does the least damage each time you hit with it, but uses the least action. The other basic hit (strike I think) has a higher DPH, but uses more action per point done, so a worse DPA... and probably has the best DPS in the skill tree. Combo has by far the highest DPH (each individual hit does the most damage), but it has an abysmal DPA (it just eats action) and since it is on a long timer and has a long "animation time" it is a low DPS (though it may not be the lowest of all the moves).

Which one you use depends on your situation. If you are low on action, you want your good DPA move (basic hit). If you are high on action, but the enemy is fresh and has lots of health, so you are going to be hitting a lot of times, you probably want strike, for the good long-term average DPS. If the target is low on health, enough that one combo will take it down, then you're better off doing combo and finishing it in one hit, rather than using several hits to finish it.

One of the fundamental misconceptions that people who think DPS is the "holy grail" of combat have is that your actual DPS in combat each and every second will equal the DPS -- that's simply false. DPS is the true or "parametric" mean (average) damage per second done, taken over many seconds and many attacks. As anyone with basic statistics knows, the average you observe, will depend on the sample size, and small sample sizes often deviate significantly from the average. For example... let's say Melee Strike does 100 DPS (that's just an easy round number, not the real number). Let's say combo does 150 DPS. That's what you'll get over the long run, if you use them for 1,000 seconds each.

However, in a single battle, say you fight one guy equal to your level, and he has 5,000 HAM. If you hit every single time with Strike, you will drop him in 50 seconds (5,0000 HAM / 100 DPS). However, you're going to miss sometimes, so it might take you 60 seconds instead of 50. And beyond this, your DPS will be off by an even greater amount in short periods. What if you had a run of 3 misses in a row over a 3-hit sequence? That's a ZERO DPS amount for that 3-hit sequence. Has that changed your attack's long-term DPS? No. But for those few seconds, your DPS was zilch, zip, nada.

This is the problem with doing things like DPS computations: people assume every second of combat they will exactly equal the average. It won't... it can't. Long-term DPS is just that -- a long-term average... not an instant per-hit automatic value.

C

---

wonderfully explained, someone give this person a sticky!

---

Jack-Aubrey wrote:

If melee strike is doing slightly less damage, but overall more dps because of its speed compared to combo, than combo really needs an increase in its speed, and maybe an increase in its damage. And MAYBE a fix for this stupid display bug?? wow, that sure would be nice..

---

I'd be willing to bet that a cycle of Combo -> melee strike -> Combo -> melee strike has better overallDPS than melee strike alone...

The CU timer change (especially the addition of cooldown and warmup timers) is the key. Ignoring attacks with warmup timers (which TK alone doesn't have anyway), there's the "execution" timer (the "fast sweep" you see on attack icons other than the one you just did) and the cooldown timer (the extra time it takes for the last attack you executed to be able to go again). If all you do is Combo, roughly 1/3 to 1/2 of the time you're waiting you have other attacks ready and available to use.

It's the same thing doctors run into. Bacta jab (a novice doctor ability) actually heals less damage than Bacta spray (a medic ability). But they're on different timers (I also think jab is faster), so by using both, a doctor out-heals a medic.

I'm with you on a fix for the display bug, though.

Message Edited by Jack-Aubrey on 06-16-2005 07:01 PM

Some of you actually think this skill is useful, let me clue you in on a few things that may have slipped your notice.

1. It has very high action cost

2. Itdoes mediocre damage

3. It has slow recharge.

No matter what you say about healthbar this or reporting issue that the facts above remain the same.At best this ELITE MASTER skill can only be described asmediocre. Melee ASSAULT (mbrawler) does FAR more damage for equal or less action, and equal recharge time. It's so funny because so many people think it's actually good because they can't see the hard data. I will be laughing if the reporting bug is fixed and the damage issue is ignored. You'll all be back in here posting about how it sucks.

---

Brandon81 wrote:

Some of you actually think this skill is useful, let me clue you in on a few things that may have slipped your notice.

1. It has very high action cost

2. Itdoes mediocre damage

3. It has slow recharge.

No matter what you say about healthbar this or reporting issue that the facts above remain the same.At best this ELITE MASTER skill can only be described asmediocre. Melee ASSAULT (mbrawler) does FAR more damage for equal or less action, and equal recharge time. It's so funny because so many people think it's actually good because they can't see the hard data. I will be laughing if the reporting bug is fixed and the damage issue is ignored. You'll all be back in here posting about how it sucks.

---