GU50, Tunso Testing, and Player Testing

Let me add another thing haha this thread lol

Ten second parses are to erratic for most powers so going by that isn't ideal imo. In one rotation i can get a 15k and in the next 12k all because of the part of the rotation that i"m in. Doing 30 seconds is the best way to get clean data.

5 minutes of data is the same thing no matter what way you count it lol.

The difference being that if you screw something up in a rotation you don't need to wait for the 30 second parse to stop, you can just continue on and pull your flubbed rotation out. It actually saves more time in the long run.

Nobody is taking just one sample. We're taking averages of many samples. Was that not clear?

If you are throwing away your flubbed data then you are losing the fact that some rotations are harder to pull off with consistency than others.

I get that but if you do 30's you can get basically the same number every time. And heres another example, I cant get a 15k in a 30 second rotation but i can't in a 10 cause of the part of the rotation it falls on. So the high from a 10 second is different from the high of a 30 which could also be misleading.

This testing stuff is getting so drawn out with these inaccurate methods. Lets just set a normal standard and be done. Like i said before this is all a waste of time i figured this out in a hour and now its going on like a week lol. This is such simple stuff. I'm not getting on you i just feel you could be spending your time doing something different. When i work i do things thoroughly and in a timely manor and it just kills me how this stuff goes.

The average is all that's really important. The high isn't all that important in either test. It's just nice to track since it's being collected. The real nice thing about 10s samples is that the first two represent your burst potential. 30s samples is a bit too long for that. Either way, let's just let this thread play out for now.

Not really your highs can be different in a 10s then in a 30s depending on the part of the rotation so that could skew your numbers at the end of the day.

Thats the only time i do 10s is when i'm testing ramp up times other then that 30s seems more reliable to me. And i agree about averages but its good to know the highs and lows of a power to. I like to take everything into consideration.

If your sample is big enough, it won't matter. Keep in mind that most rotations are 24s rotations so you'll have that same effect with 30s samples too. As long as you're focusing on the mean, it's not going to matter. It all washes out.

These guys just need to get with the program already, all that work you and Radium did just cause they don't know how to test properly smh

If you took 3 10 second parses, its still going to add up to exactly what the 30 second parse would be.

Same thing with 30 10 second parses. If you add them all up, its still going to show you the exact same thing as 10 30 second parses. We're just breaking it down further which gives us a gage on rampup time as well.

The only Tunso Test we're missing at the moment is Sorcery Ranged if anyone wants to tackle that.

This whole balance at 0m credo is bs. If you can successfully pull off some rotations at point blank you deserve to get a little boost in damage. Risk abd skill should equal a reward. Balance should be from a relatively safe difference

I know what your saying i just think 30s is easier and more consistent with providing the same numbers and i get ramp-up times i test them myself using the 10sec parse. Instead of seeing numbers that are up and down with 10s i like to just see the same numbers consistently, less work for me to figure stuff out that way.

It's seriously splitting hairs at that point. You can balance the powersets with 10s samples. The difference between seeing a high 10s sample and a high 30s sample is purely academic. It's only for a "hey, look at that" factor. It doesn't actually impact decision-making for anything.

It's less work to do all your tests at 10s in one run than it is to do some at 10s and others at 30s.

Im on it

You just missed what i said, I just said it was easier to use 30s (well for me anyway) cause you basically get the same numbers every time vs the 10s where its up and down. I like to see consistency but thats just me. Sure you can do some math and figure it out but why do all that?

And if your testing ramp up times you have to stop every 10s anyway and it only takes a sec to change the parser setting lol. We're never gonna get anywhere like this so im done lol. to much thought has gone into something so simple. We just need a normal testing standard thats all.

Pretty much. You're just taking the damage and dividing it by the time. With a 30 s parse, it's 30 seconds. With a 10 s, it's 10 seconds. The total damage recorded is the same. That's the real metric: damage per unit time.

But that's the point. This is the testing standard used by the Devs. We could come up with whatever we want. It won't be comparable. If we're trying to show the differences between the Tunso Tests and our tests, they should be as close as possible. A lot has already come to light from these comparisons.

But what i'm saying is why do all that math lol just do some rotations and look at the numbers..bam done