FYI: Incredible Facts!!! % Crit Chance vs. % Increased Crit damage

Hey,

BEWARE: some math down there but consider reading my other post containing a way larger data sample than the one beeing discussed here:

Lemme try show you what whiteknut and I are talking about:

I try to avoid math on this forum because it's hard for me to be as excact as one has to be when using nonbasic math because I don't know all the terms in english and because people get upset when i just assume for example that when i say avg, that i know it's for an infinite amount of times even though i don't have a sample that seize. But because you insinst, i'll give it a shot and just hope you know what I mean even if i use a wrong word:


Let's do a simpler version first: 90% Hit Rate (Normal + Crits) vs. 10% Evade, for 3200 trials:

The cumulative probability of a lower than 89% hit rate is: 3.2% (Your first Test 87.42+1.65=89.07%)

Addmittedly you are on the lower end of the hit rate scale, but nothing special there.

Since you are talking about halfs of percentage points here: the cumulative probability of a hit rate below 89.5% is 16.53%!!! It's just not feasible to talk about tenths of percentage points when having only this small of a sample seize.


Now:
Doing the same thing (90/10) with 70000 Trials and 89% hit rate results in a crazy low cumulative probability (1.635 x 10^-18)

For 70000 trials even a 89.8% hit rate yields only a cumulative probability of 3.865% of beeing lower.

That alone should show that 3200 is a very small sample seize!

Without going into why comparing 3 highly unreliable samples yields even lower probabilitys of beeing "correct", because the math behind that would be very unreadable without me beeing able to code the formulas in some way on this forum, let's look at the more complicated case with 3 variables: Hit, Crit, Evade:


So assuming you fired 3200 Shots total and have an avg hit rate of 87%, 10% misses/evades and 3% crits:

In your second test:

You had 3205 shots with 84 crits, assuming the real crit rate is 3% then the cumulative probability of you getting fewer than 84 crits is 9.3%!!! That also means that in 90% of the cases you will do more crits when you repeat this!

70000 shots fired with 3% assumed crit rate and a measured crit amount of 1834 (the crit rate you stated) would yield a cumultated probability of 8.24*10^-10 for beeing lower!



I could go on and i could also include the way the numbers influence each other and why but I guess since you are quite capable of understanding math and the reasoning behind my post I believe I don't have to.
I didn't mean to attack you when i said that the sample seize is too small and the conclusions unreliable because of that but it's just a fact!

Regards
Thark

I know what you say Adi but accuracy might mess up with Crit rate, evade rate aaaaand Hit Rate

I don't think I had that lower hit rate ( only in that hours) because of the game dynamics, it's because accuracy boot.

We don't know how MA made them

But what I am desperate for you guys to see is that in 4 hours with accuracy I had no difference in damage output vs normal damage output...and yet in almost 2 hours with Kill Strike I had ( I would say) quite a big difference

Second hour I only got around 4.1 because of the1 1 minute lag resulting in lesser shoots

Just give me cheaper Kill Strike nanobots and I'll be happy to test this over and over and over


Oh, wait, I know, after 31 when I'll get my Laser I'll reset chat log and do all the hunting first with Laser attached then without

Just let me know what data amount you want in order to believe me so I know when to switch from Laser to normal

You've calculated your hit % incorrectly by counting the critical hits separately. Your actual hit rate is what you've called "Hit Shoots" PLUS your "Critical Shoots".

Oleg said:

You've calculated your hit % incorrectly by counting the critical hits separately. Your actual hit rate is what you've called "Hit Shoots" PLUS your "Critical Shoots".

Click to expand...

I outlined them like that to make my point easier to understand but I see to no avail

bettlejus said:

I outlined them like that to make my point easier to understand but I see to no avail

Click to expand...

But it's misleading to claim that as your hit rate, especially when you then say that using the accuracy buff reduces your hit rate by 1.5%, which is not true.

If every shot was a crit, would you say your hit rate was 0%?

your example is not good, but what I am trying to say is that accuracy messes with the slider between crits/miss and maybe normal shoots

If I have not outlined them this way you would not see this

Overall the Hit Rate ( the one you are mentioning which is the real one) remains fairly the same, but you would not see the difference in Crits vs normal shoots

Also if this % Crit Change really adds 1% only to your Crit Rate it means all added up should do 101% but they only add to 100%

And leaving all the math behind, the only real fact is I had more damage output with Kill Strike in 2 hours vs. almost no damage output difference in 4 hours with accuracy, and leaving the fact that my data are way to small how do you explain this

bettlejus said:

Also if this % Crit Change really adds 1% only to your Crit Rate it means all added up should do 101% but they only add to 100%

Click to expand...

It adds 1% to the proportion of hits that are also critical hits (2% without buff, 3% with). It's not supposed to make you hit more often.

yeah, it adds 1% to crit hit by taking adding 1 percent to miss shoot or by taking from normal hits
But in the end it does not produce more damage per same ped/time/shoots spend nor it increases DPS

But...Kill Strike does

I have another question

Why are Kill Strike nanobots so expensive on AH and Accuracy so cheap? Because Accuracy are better then Kill Strike?

Wow this is a cacophony of terrible maths

bettlejus said:

Why are Kill Strike nanobots so expensive on AH and Accuracy so cheap? Because Accuracy are better then Kill Strike?

Click to expand...

I said this before but maybe you missed it. It's because of the candy cranes!


With my larger sample I now think it works like this:


Of 100 shots 10 shots miss. Leaving 90 to be split among crits and normal hits. 3% of those 90 hits are crits so 2.7 shots, so 87.3 of the shots are normal hits.

So:

10% Evade (90% hit rate)
87.3% normal Hits
2.7% critical Hits


At least thats what my log tells me. That +xx% crit dmg adds dmg is a nobrainer for me but +X%crit chance is stronger.

I don't like uploading an unedited chat.log but here is my log without any other information aside from hits/crits/evades etc.:

https://drive.google.com/file/d/0B4jlusaQ_exDREU0eG9nQnV1MGM/edit?usp=sharing

It seems you can only see the first 19 pages? if you download it, it'll show everything.

Regards
Thark

Well Thark, if this items don't bring extra damaga then where is there value?
I wouln't pay nothing for an item that only give me more crits whitout actually increasing DPS

I flipped a coin twice and get 1 heads, 1 tails.

Then, I put on my pink shirt, and got two heads.

Therefore, it is an undeniable fact that wearing a pink shirt increases the amount of heads you will get by 1. One is just way to many to be random.

Just a basic 2 tail test...

whiteknut said:

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
~!!!!!!!!!!ˇˇ

Following text is only about statistics and data, it has nothing to do with the theory. If you do not read about boring numbers, then skip this post, if you want to read about statistics it is your choice.

1h tests is far from valid data for statistics.
I can guarantee that if you did those runs all three runs same settings your results would be differing atleast as much.
Do something like 2 weeks tests (5 h a day, every day different hours), then you might get something.. But atm it is just 3 hunting runs with no statistical value at all.

It is like saying.... Today it took me 10minutes to get home from work, BUT yesterday i wore green hat and got home in 11 minutes.. So green hat reduces my average speed. And oh no!!! One day i lent my friends porsche and it took me 14minutes... that means faster car results in less average speed..

Get some data before creating statistics.

PS! If it was all sarcasm.. then i salute you, you got me. Yeah some ppl are really that crazy that they believe that and shout it in forums.

Click to expand...

...on first two sets of information suggests difference is by chance less than .71%(i.e. <1%) chance...so your spewing hogwash whitenut...you would presume that flipping a coin for an hour once a second would give you unreliable results as well? the main difference only being that likelihood of a desired result is ~2% instead of ~50%...it is your statistics that is silliness...my presumption based on a glance at the numbers was that there was enough data to have reasonable CI...I think the real issue remains the unknown...and that is...there is a hidden overlay(or underlay, however you want to think about it)...that influences why sometimes you have 3 crits in a row and why sometimes you may go an entire hour without a crit...same with skill gains and other things...if this timer or clock or 'function' works on some time frame significantly greater than one hour then the likelihood of the results being off is more likely to fall in that 0.71% chance that the numbers are wrong...and I think that is very likely...

Brick

First of all you have diffrent ammount of shots in all the diffrent rounds

And you made it unessecary hard for your self to do this test.
Infact you dont need liakon to make this test accurate.

Just use what ever amp you got and set it too 3000 shots before decaying.
And then you will have more accurate how the damages been divided.

So if your using a dante repair it 120 ped tt from broken and do 3 runs. Wont matter how much you get hit/lagg/faps or anything just that you break the dante within the hour.

Thanks Vodka, great idea

And I don't need to set it to 3000, will do full runs with a full dante or a full Adj A105 for let's say 1 day test in each situation, accuracy, kill strike ,normal

But after 31 Jan so I don't have to buy my kill strike nanobots

Scurvy Sityl said:

I flipped a coin twice and get 1 heads, 1 tails.

Then, I put on my pink shirt, and got two heads.

Therefore, it is an undeniable fact that wearing a pink shirt increases the amount of heads you will get by 1. One is just way to many to be random.

Click to expand...

Incredible Facts!!!!! Wearing pink shirt increases the number of heads you will get!!!!

Maybe that's why Kimmi likes pink?

jaydub said:

...on first two sets of information suggests difference is by chance less than .71%(i.e. <1%) chance...so your spewing hogwash whitenut...you would presume that flipping a coin for an hour once a second would give you unreliable results as well? the main difference only being that likelihood of a desired result is ~2% instead of ~50%...it is your statistics that is silliness...my presumption based on a glance at the numbers was that there was enough data to have reasonable CI...I think the real issue remains the unknown...and that is...there is a hidden overlay(or underlay, however you want to think about it)...that influences why sometimes you have 3 crits in a row and why sometimes you may go an entire hour without a crit...same with skill gains and other things...if this timer or clock or 'function' works on some time frame significantly greater than one hour then the likelihood of the results being off is more likely to fall in that 0.71% chance that the numbers are wrong...and I think that is very likely...

Brick

Click to expand...

95% confidence level (search it and you´ll understand)
And no flipping coins 3600 times(with two possible results each time with same chance of occurring) is not same as shooting 3000 times with hundreds of variables.

Try crafting 3000clicks of basic filters , same setup three times, i am sure the success rate will be off by more than 1% (no matter what boots your wear).

sitram said:

Incredible Facts!!!!! Wearing pink shirt increases the number of heads you will get!!!!

Maybe that's why Kimmi likes pink?

Click to expand...

Selling pink shirts tt+500. Selling pink paints 1500%.

I'm surprised it hasn't been mentioned yet (unless I missed it), but wouldn't this dataset be fine for something like an analysis of variance (ANOVA)? If you have a single column in say excel for each test, which each row having the amount of damage done in each shot, you'd be able to analysis the data pretty easy with any basic stats package. With that, you'd be able to get past any differences in the original test being just due to chance as that's the whole point of statistical tests. This shouldn't be as complicated as this thread has made it seem.

I'd still want this set up like any basic experiment though and repeat it once or twice more in case there is any effect of the timing of when you do to the test for some unforseen reason, but this data seems to be begging for a proper statistical analysis before a replication effect is even worried about. You'd basically be looking at average damage per shot and see if there is a statistical difference going on. It could very well be there isn't an actual difference in your data even at this sample size, but that's what the right statistical tests are designed to do.

Missed this discussion and the thread about the new scopes and sights that links to it.

Got some reinventing of the wheel going on with the discussion of crit rate. I posted much larger data sets investigating hit rate, accuracy enhancers, etc. a year ago that confirmed that maxed crit hit rate is 2% and that additional crit % (accuracy enhancers) are effectively additive. I would certainly expect the new attachments to work the same way, but the descriptions could have been more clear.

Thark said:

One more interesting thing:

when excluding the arso chip completely (as stated above - easy for my sample), and ingoring evades then:

Normal hits = 67465
Crits = 2090
Total Hits = 69555

Crits/Total Hits = 3% (3.00482)
Normal Hits/Total Hits = 97% (96.99518)

Which in theory means that the 1% increase is a perfect match! Sure there are evades in there as well BUT they shouldn't interfere with the ratio of hits and crits. It's just hard for me to exclude the arso evades from my data sample.

Click to expand...

Yeah my tests point to the CHA affecting the percent of hits that are crits, not the percent of shots. This makes sense as the algorithm checks for a hit, and if it's a hit it then rolls again to check for a crit. Consequences are that improving accuracy (i.e. accuracy enhancers) doesn't actually improve your percentage of hits, just your percentage crits. It also means entropedia's formula is a bit wrong, but not by much.

Doer said:

...Consequences are that improving accuracy (i.e. accuracy enhancers) doesn't actually improve your percentage of hits, just your percentage crits. It also means entropedia's formula is a bit wrong, but not by much.

Click to expand...

"Consequences" in this regard denotes negativity, to me at least. Are you questioning the title vs. effect (accuracy vs. crit chance increase) as incorrect, or are you questioning the effectiveness of the attachments themselves, as I've seen many other people here do?

I've done extensive testing with the new attachments. The 2% scope produces 4% crit rate vs the base of 2%, while still maintaining a 90% hit rate. Those extra critical hits equate to more damage produced for the same amount of ped spent, increasing eco as a result. I've done the math and the 2% scope saves me nearly 100k ped in 1 year of shooting, adding free dps along the way. What's not to love? These things are like an imk2 effect for any gun that will take a scope


2% Scope + 20% Laser Effects

• MM/H204: 3.078 - 3.16
• imk2/H204: 3.241 - 3.33
• Term/Mevil: 2.848 - 2.935

All data verified via ET client for hours and hours of shooting.

Stryker said:

2% Scope + 20% Laser Effects

• MM/H204: 3.078 - 3.16
• imk2/H204: 3.241 - 3.33
• Term/Mevil: 2.848 - 2.935

All data verified via ET client for hours and hours of shooting.

Click to expand...

Do 2x scopes give 40% critpower? And does it add up to the 60% critpower from earth shocker set?

Stryker said:

"Consequences" in this regard denotes negativity, to me at least. Are you questioning the title vs. effect (accuracy vs. crit chance increase) as incorrect, or are you questioning the effectiveness of the attachments themselves, as I've seen many other people here do?

Click to expand...

Consequently/as a result... I was trying to point out that the accuracy enhancers don't increase the proportion of hits, which is somewhat counterintuitive. I don't doubt their efficiency, and i believe i was the first to post a more or less comprehensive analysis of accuracy enhancers showing they can actually increase eco at a certain price threshold.

I've done extensive testing with the new attachments. The 2% scope produces 4% crit rate vs the base of 2%, while still maintaining a 90% hit rate. Those extra critical hits equate to more damage produced for the same amount of ped spent, increasing eco as a result. I've done the math and the 2% scope saves me nearly 100k ped in 1 year of shooting, adding free dps along the way. What's not to love? These things are like an imk2 effect for any gun that will take a scope


2% Scope + 20% Laser Effects

• MM/H204: 3.078 - 3.16
• imk2/H204: 3.241 - 3.33
• Term/Mevil: 2.848 - 2.935

All data verified via ET client for hours and hours of shooting.

Click to expand...

Hence my interest in knowing the mechanics of the bonuses on the new attachments.

To OP: the hit rate thread i referred to previously was mostly to verify hit ratio and also compared the affect of attachments and even with much larger data sets wasn't 100% conclusive about attachments. Your single, 1 hour tests are much less reliable still, but as always good to have people conducting tests and reporting.

Piotr said:

Do 2x scopes give 40% critpower? And does it add up to the 60% critpower from earth shocker set?

Click to expand...

Only one scope per weapon.

Would still like to know the answer to whether the damage increase is for the entire crit hit or just the crit damage. Also whether any of these effects are cumulative with enhancers, nanobots, etc. Previous reports are that of all objects giving the same effect, only the greatest is applied.

With a 20% laser the new max critical hit becomes 2.2*Max dmg. 99 dmg MM/H204 now caps at 217.8. I'll triple check that today.

The 60% crit armor buff (or 40% if you're using a different hat) does not stack, but rather the higher buff is used. The critical hit nanobot does stack with the 2% scope, yielding a new base 5% crit chance.

Bump because I want to follow up on this. Thanks for your reports, Stryker.

Stryker said:

With a 20% laser the new max critical hit becomes 2.2*Max dmg. 99 dmg MM/H204 now caps at 217.8. I'll triple check that today.

Click to expand...

I assume your triple check verified that? So it's just the critical damage itself that is increased, not the entire attack damage. That's as expected. Does that mean the minimum crit is now 1.2*Max dmg + min dmg (= 1.7* Max dmg for maxed)?

The 60% crit armor buff (or 40% if you're using a different hat) does not stack, but rather the higher buff is used. The critical hit nanobot does stack with the 2% scope, yielding a new base 5% crit chance.

Click to expand...

Does the crit nanobot and scope stack with accuracy enhancers? Do the laser sight crit buffs stack with one another?

Stryker said:

Those extra critical hits equate to more damage produced for the same amount of ped spent, increasing eco as a result. I've done the math and the 2% scope saves me nearly 100k ped in 1 year of shooting, adding free dps along the way.

Click to expand...

I assume it does save - I haven't done the maths - 100k of ped if you are calculating the cost of damage per year. However, in terms of loot, we don't know loot is generated purely in terms of damage inflicted, so it may be somewhat misleading to potential buyers to say it saves 100k per year.

I'd like to see some long term hunting data done with someone wearing a full set of Earth Shock Armor

Doer said:

Bump because I want to follow up on this. Thanks for your reports, Stryker.



I assume your triple check verified that? So it's just the critical damage itself that is increased, not the entire attack damage. That's as expected. Does that mean the minimum crit is now 1.2*Max dmg + min dmg (= 1.7* Max dmg for maxed)?



Does the crit nanobot and scope stack with accuracy enhancers? Do the laser sight crit buffs stack with one another?

Click to expand...

I've checked more times than I'd care to admit to be sure New maximum possible damage with a 20% scope is (max dmg+(max dmg*1.2)) The effects the attachments have on weapon economy is reflected here:

http://eldslott.org/entropia/

I checked shortly after receiving the scope and, unless things have changed, the crit nanobot does stack. Accuracy enhancers also stack. 2 lasers, however, do not. You seem to only receive the benefit from one source of any given buff (I guess nanobot and crit scope rating are considered different, despite sharing an identical icon)

sachi.mk2 said:

I assume it does save - I haven't done the maths - 100k of ped if you are calculating the cost of damage per year. However, in terms of loot, we don't know loot is generated purely in terms of damage inflicted, so it may be somewhat misleading to potential buyers to say it saves 100k per year.

Click to expand...

There are countless variables in Entropia, but one of the ONLY assumed facts VERIFIED by Mindark themselves: "This measures how efficiently, in terms of weapon decay and ammo consumption, a weapon produces each point of damage. An avatar will enjoy much better overall hunting results over time when using weapons with higher damage per pec." Eco matters in a huge way. Your cost of producing damage cheaper correlates directly to a percentage increase in returns you would otherwise receive 1) As far as we've believed for some time, 2) As far as we've been told directly by the game developers, and 3) By my and others tracking of % returns over time with varying setups (i.e. eco).

Eco boosts are reflected on the website above by utilizing the "free" damage given by the additional critical hits; an advantage users of these tools have above their competitors. Any financially savvy individual realizes that money saved is money earned, and from my personal napkin math, the "free damage" translates into a savings of ~20, 30, and 45 ped per hour* when using an unenhanced imk2, mod merc and Terminator+mEvil, respectively.

Believe and utilize this mentality for maximum profit, or be naive and wonder why you're always short on PED


* Assumes uninterrupted shooting for 1 hour; not feasible in most environments.

Stryker said:

There are countless variables in Entropia, but one of the ONLY assumed facts VERIFIED by Mindark themselves: "This measures how efficiently, in terms of weapon decay and ammo consumption, a weapon produces each point of damage. An avatar will enjoy much better overall hunting results over time when using weapons with higher damage per pec." Eco matters in a huge way. Your cost of producing damage cheaper correlates directly to a percentage increase in returns you would otherwise receive 1) As far as we've believed for some time, 2) As far as we've been told directly by the game developers, and 3) By my and others tracking of % returns over time with varying setups (i.e. eco).

Eco boosts are reflected on the website above by utilizing the "free" damage given by the additional critical hits; an advantage users of these tools have above their competitors. Any financially savvy individual realizes that money saved is money earned, and from my personal napkin math, the "free damage" translates into a savings of ~20, 30, and 45 ped per hour* when using an unenhanced imk2, mod merc and Terminator+mEvil, respectively.

Believe and utilize this mentality for maximum profit, or be naive and wonder why you're always short on PED

Click to expand...

Oh, I am sure economy, or efficiency as Mindark says, is very important and will produce better returns. I was just highlighting that loot may not be purely related to damage inflicted and therefore we cant base returns/savings solely on that basis. For example, I hunted with a ml35me+evil and adjm5+a106 at the same hit rate. One setup had a dmg/pec return of 2.45 and the other 2.80. However the hit rate was the same. The return rate in ped, over hundreds of thousands of ped for each, was 89% and 90% respectively. However the damage inflicted variance was between 15-20%. Likewise, I think you have posted you got 100% tt returns with a unenhanced mod merc and also with an imk2. But shouldn't you have got over 100% with the imk2, as its 5%+ more eco in terms of damage inflicted. Not that I doubt the effect of damage inflicted efficiency, just the level of its effect.

The thread that started it all!