Question: Hit ability

[WoenK]

Yes, ofc, though nobody said anything about "feelings." I said "thoughts" and Blackhawk said "convinced." I've noticed Blackhawk takes a very thoughtfull approach to EU, and obviously has a massive well of experience.

If nothing else, it at leasts suggests routes for useful testing.

Real testing could only be archieved if all informations are relyable and I am not very sure of the stuff the client shows me. There was a post a few days ago about first shot never hitting and when I watched it more closely, I noticed that it can happen that the damage is not displayed in chat, but health bar is reduced.
Does that mean I really hit it, is it a display bug or lag and the damage is displayed later on ?
For me decay is also not reliable, because it seems that the decay reduction can also lag (or at least I had quite a few times that the weapon was usable until I killed a mob and not after that).
Maybe the only half way reliable thing would be to notice both if ammo was spent per shot and if damage was done.
I think using only chat messages,decay and ammo spent over a complete hunt does not give a reliable outcome at the moment.

 

[Xen]

Real testing could only be archieved if all informations are relyable and I am not very sure of the stuff the client shows me. There was a post a few days ago about first shot never hitting and when I watched it more closely, I noticed that it can happen that the damage is not displayed in chat, but health bar is reduced.
Does that mean I really hit it, is it a display bug or lag and the damage is displayed later on ?
For me decay is also not reliable, because it seems that the decay reduction can also lag (or at least I had quite a few times that the weapon was usable until I killed a mob and not after that).
Maybe the only half way reliable thing would be to notice both if ammo was spent per shot and if damage was done.
I think using only chat messages,decay and ammo spent over a complete hunt does not give a reliable outcome at the moment.

I think shots not showing up in chat would be a very small percentage, if it exists at all. I have watched for this happening and have not seen it (I know, not scientific). Anyway, it would still be very useful to see testing results relative to others, even if we can't control every possible minute imperfection.

 

[WoenK]

I think shots not showing up in chat would be a very small percentage, if it exists at all. I have watched for this happening and have not seen it (I know, not scientific). Anyway, it would still be very useful to see testing results relative to others, even if we can't control every possible minute imperfection.

If 1% does not show in chat that would already mean that the outcome is 10% screwed according to the tests that were done pre VU9.
Testing relative to others is surely useful and surely does give some hints in the right direction, but my gut feeling is, that those numbers will not be comparable to the ones that were done in previous tests (because of bugs still floating in the client).

 

[mastermesh]

do the test again...

Do the test a few more times. With statistics, you have to increase your population size or decrease your population size to change what the math is telling you... In other words EU IS DYNAMIC... i.e. random... random number generators are usually pretty random, but not always. Skills seem to be add ons to random number algorithm in eu, but how they play in to it all, only Lootius knows...

If you do 2 shots your chances are 50/50... just the same a n00b taking his first steps... ultimately that's the reality of it. Skills have some role to play, but in reality, percentages have no meaning because you either are successful or not successful... if it's a success x happens, if it's a non-success y happens... With crafting non-successes can give partial loot back... I think it's the same with hunting... non-success means you did not global but get something back... figuring out that something, and how much it's worth vs. how much was paid in is, I think where the skills play in to it somewhere.

Mobs are basically slot machine handles, and there's multipliers on them like slots... skills seem to help you either get more success or get bigger successes, but it's arguable which way it really goes.

 

[Xen]

If 1% does not show in chat that would already mean that the outcome is 10% screwed according to the tests that were done pre VU9.
Testing relative to others is surely useful and surely does give some hints in the right direction, but my gut feeling is, that those numbers will not be comparable to the ones that were done in previous tests (because of bugs still floating in the client).

True, if 1% does not show up, that would skew results. That's why I would like to see test results of others. If many people are getting 90-92%, and I continued to get 87%, that would mean I have a problem.

 

[Immortal]

True, if 1% does not show up, that would skew results. That's why I would like to see test results of others. If many people are getting 90-92%, and I continued to get 87%, that would mean I have a problem.

The test I did a couple of years back to track hit % I manually counted the number of shots it took to kill a mob. I think that's about as reliable as you can get. Just make sure it's not regen mob, and all should be well. I tested on formics when I did this(almost zero regen). It's actually easier postvu10 as mobs don't regen if there is no gd with it on their radar, so even if you are killed by it, your data should still be reliable.

 

[Xen]

The test I did a couple of years back to track hit % I manually counted the number of shots it took to kill a mob. I think that's about as reliable as you can get. Just make sure it's not regen mob, and all should be well. I tested on formics when I did this(almost zero regen). It's actually easier postvu10 as mobs don't regen if there is no gd with it on their radar, so even if you are killed by it, your data should still be reliable.

So, you are suggesting figuring out the average number of shots to kill a mob where you are hunting (there can be different maturities, and usually are), then count mobs killed? I would think that would be less accurate, no?

 

[Coop_]

True, if 1% does not show up, that would skew results. That's why I would like to see test results of others. If many people are getting 90-92%, and I continued to get 87%, that would mean I have a problem.

So, you are suggesting figuring out the average number of shots to kill a mob where you are hunting (there can be different maturities, and usually are), then count mobs killed? I would think that would be less accurate, no?

I happened to be in the middle of collecting data on the argos I've been hunting. The downtime today gave me a chance to input the numbers and play with them a bit.

As part of my data gathering, I record the maturity of each argo I kill during a hunt, and at the end of each hunt figure the "killing cost" (ammo+amp+gun w/ mu). I use several weapons to minimize overkill.

So I added the total hp of the argos I killed, and divided by the killing cost of the hunt, to calculate my "actual damage/pec" for the hunt. I expected it to be below what entropiatools.com calculates for my weapons because it does "seem" to me that I have extra misses due to lag in vu10.

I was surprised to find that my calculated dmg/pec of 2.93 is very close to the calculated eco of lr32+a106 (2.882) if you figure in a crit rate of 2% (increases dmg/pec by ~1% to 2.91).

Guess for now I can't complain about hitrate.

Anyway, recording the maturity of each of your kills might be more work than checking the log for damage msgs, but could provide an alternate way of estimating your true hitrate.

 

[Immortal]

So, you are suggesting figuring out the average number of shots to kill a mob where you are hunting (there can be different maturities, and usually are), then count mobs killed? I would think that would be less accurate, no?

If you are hunting different maturities, split the data.

If you don't want to manually count, hunt non-aggressive mobs. LA's set to lowest possible maturity have a distribution like 80%-17.5%-2%-0.5% (young-mature-old-provider)

An alternative to not manually counting shots, without the hassle of worrying about maturity is this:
Count number of each maturity killed
Sum their HP
Divide by number of shots you took for the entire session.
Multiply by hp of lowest maturity to give average number of shots for that maturity.

Doing the above method(s) for only a few runs is statistically sound, and isn't nearly as time consuming as wading through reams of data and worrying whether all of it was captured.

 

[rhogenbe]

How can you tell, that this is the right number of hits?!
Even with the log it does not really tell you if due to lag you did not hit and got the ammo back afterwards.
Can remember a few updates ago the lag was so horrible and the ammo count went up and down all the time, one could not really tell if it was a real hit or not. Happens way less often now and I rarely notice it, but it still occurs.
Also not sure if that ammo gets added to the main stack or the loot stack.

If you get ammo back because the shot did not reach the server due to lag you also do not get decay on your weapon for that shot.

Cheers
Siam

 

[Xen]

I tried this method and on the test run (5709 shots) with a maxed weapon had a 87.4% hit rate and 1.77% crit rate. Hit rate with a maxed weapon is supposed to be 92.5%, so I'm losing 5% of what I spend into never never land.

I'll do more tests. I'd be interested in what hit rates others are getting (and what country they live in).

2nd test, same as first:

88.2% hit rate (regular hits)
1.72% crits

 

[BlackHawk]

I tried this method and on the test run (5709 shots) with a maxed weapon had a 87.4% hit rate and 1.77% crit rate. Hit rate with a maxed weapon is supposed to be 92.5%, so I'm losing 5% of what I spend into never never land.

I'll do more tests. I'd be interested in what hit rates others are getting (and what country they live in).

No one ever said that hit rate is supposed to be that. Some people got some results and said it is so for them. No one bothered to prove otherwise. That does not make it true.

 

[Xen]

No one ever said that hit rate is supposed to be that. Some people got some results and said it is so for them. No one bothered to prove otherwise. That does not make it true.

Well, there's this, which is from this, which followed from this. Quite a bit of research was done by intelligent people. I'm not saying it's guaranteed to be true, but...

And part of the reason I wanted others to post was to see what others are getting for hit rates, and see if others get the theoretical max hit rate with a maxed weapon.

 

[Xen]

3rd test, maxed weapon, this time small furor.

86.98% hit rate, normal hits
1.87% crits

 

[BlackHawk]

Well, there's this, which is from this, which followed from this. Quite a bit of research was done by intelligent people. I'm not saying it's guaranteed to be true, but...

And part of the reason I wanted others to post was to see what others are getting for hit rates, and see if others get the theoretical max hit rate with a maxed weapon.

I never said that those people are not inteligent. I just said that that was true at that time, and in the conditions of those experiments. Since we don't actually have anything to compare with, we can't say it should be like this or like that. I think the number of misses is as random as loot is, and even tho it was made a lot of research on the community, no one actually knows what will be the loot on the next mob or the average on your next hunting run. We all know what it "should be" but we don't know what it will be. And if is not what it should be, we can't complain, since no official information is available about normal behavior of the system.

 

[Xen]

I never said that those people are not inteligent. I just said that that was true at that time, and in the conditions of those experiments. Since we don't actually have anything to compare with, we can't say it should be like this or like that. I think the number of misses is as random as loot is, and even tho it was made a lot of research on the community, no one actually knows what will be the loot on the next mob or the average on your next hunting run. We all know what it "should be" but we don't know what it will be. And if is not what it should be, we can't complain, since no official information is available about normal behavior of the system.

I know you weren't saying that they aren't intelligent.

And having other numbers to compare with is what I am looking for here. It's amazing how many people complain about loot, but they won't run a simple test to see if they are missing possibly more than normal. If you hunt 2000 ped a day, and you are missing 5% more than most other people, that's a loss of 100 ped per day, or 3000 ped per month, or 36,000 ped per year.

 

[BlackHawk]

I know you weren't saying that they aren't intelligent.

And having other numbers to compare with is what I am looking for here. It's amazing how many people complain about loot, but they won't run a simple test to see if they are missing possibly more than normal. If you hunt 2000 ped a day, and you are missing 5% more than most other people, that's a loss of 100 ped per day, or 3000 ped per month, or 36,000 ped per year.

All that, assuming loot system counts the "good hits". There is no prove or reason to think that it is so.

 

[GeorgeSkywalker]

All that, assuming loot system counts the "good hits". There is no prove or reason to think that it is so.

I think you have hit it on the head there Blackhawk. I suspect the system looks at total shots fired and works with that i.e. the actual misses do not really figure into the equation but the hit ability based on your gun and skills would be somehow used. I have no proof of this just a "feeling"

 

[Immortal]

3rd test, maxed weapon, this time small furor.

86.98% hit rate, normal hits
1.87% crits

1. What is your damage distribution curve on all those shots
2. How does this data compare to the method I suggested previously to calculate hit rate?

 

[Lugzan]

Fjolla, are you getting an idea that connection might be the reason of your somewhat over average losses in the past? dang, if that is the case...

damage distribution might compensate for the hits not made

this is something I've never heard of before...

it sounds more like a belief of perception lovers =) remember, MA compensates ess-eitch-ai-tii.

now, are you sure MA did not confirm miss rate 8% for maxed weapons at some point? my mind might be cheating but I kind of recall it being confirmed either via forum or support... anyways, it has been tested more than once and proved. no correlation to dmg dealt, skills gained or any other "mights".

P.S. what do you guys think, should MA implement a connection test possibility (inloader/ingame)? I mean, how the heck should I know if my connection is good or bad enough not to lose 5% of hitrate which is like A LOT!


J.

 

[xaph]

Fjolla, are you getting an idea that connection might be the reason of your somewhat over average losses in the past? dang, if that is the case...




this is something I've never heard of before...

it sounds more like a belief of perception lovers =) remember, MA compensates ess-eitch-ai-tii.

now, are you sure MA did not confirm miss rate 8% for maxed weapons at some point? my mind might be cheating but I kind of recall it being confirmed either via forum or support... anyways, it has been tested more than once and proved. no correlation to dmg dealt, skills gained or any other "mights".

P.S. what do you guys think, should MA implement a connection test possibility (inloader/ingame)? I mean, how the heck should I know if my connection is good or bad enough not to lose 5% of hitrate which is like A LOT!


J.

Interesting !

 

[Shady]

atm I have a very small connection 512Kbit/s I was wondering if I miss shot because of this.

Maybe we should make the test using the right mouse button (it only fire if there a target)

 

[Tim McDohl]

Speaking of evade and accuracy... Is the evade/dodge level compared against the hit profession, or do the defensive professions have the same effect regardless of the skill of the user?

 

[Witte]

In my opinion MA should implement "you missed the target" and "no target detected" messages. As for the tests, I suggest to calculate shots from decay instead of ammo, and use the right mouse instead of left mouse.