46 replies to this topic

[Karl Streiger]

Hey there,

i need your help. I'm working on a weapon table from time to time. This weapon table should consider each and every weapon with all of their stats and grant a point value.

While I have some accuracy calculations that seems to be ok, i think it would be much better to have "real" accuracy stats.

So, ifi you want to help me, please go:
Profile -> Stats -> Weapons
and copy & past the table and send it to me as a PM.

You don't have to bother about formatting - i have a program that does the computing and formatting.
Plain text and we are good to go

I will not forward this information to anybody. Thanks in advance.

Why to bother?

---

The goal:
a summary point value for all weapons.
this should include:

• average accuracy (based on player stats)
  
• effective damage (based on player stats)
  
• load value (weight, ammunition, heat, crits)

as an option i want to elaborate if it is possible to see the impact of burn duration, velocity, volley delay, spread and volley size in the absolute numbers - to see when changes are made how the summary value will change.

How?
I just extract 3 values for each of the player weapons:

• Fired
  
• Hit
  
• Damage

All the other values are unimportant

In the first step I will check the plausibility of those stats and provide some charts so you could check the process:

Spoiler

Edited by Karl Streiger, 20 June 2016 - 02:53 AM.

[El Bandito]

Sent you a PM.

[DarthHias]

Sie haben Post.

[Karl Streiger]

Thanks, I would prefer to have plain text rather than images
(don't bother El' my secretary Karline did type everything)
Don't bother about formats - I will make a program to extract the data.

Thanks for the first answers

[Wintersdark]

While I think your goals are a fool's errand that'll never work, I'm totally behind helping you try. PMing stats, and will share anything else you need.

[Wintersdark]

Keep in mind:

Lasers count as a hit with just one tick doing damage.

Clan AC's and missiles count each shell/missiles as separate shots.

Etc. There's some other fiddly bits that escape me at the moment, as it's 3:30am and I'm having a ****** night at work. But just keep such stuff in mind.

Calculated damage per shot, for example, shows true laser accuracy&range usage, but is complicated for CUAC's.

[Karl Streiger]

Wintersdark, on 07 June 2016 - 01:29 AM, said:

Keep in mind:

Lasers count as a hit with just one tick doing damage.

Clan AC's and missiles count each shell/missiles as separate shots.

Etc. There's some other fiddly bits that escape me at the moment, as it's 3:30am and I'm having a ****** night at work. But just keep such stuff in mind.

Calculated damage per shot, for example, shows true laser accuracy&range usage, but is complicated for CUAC's.

I know:
there is also the "background noise" of difference patches:

• speaking of pellet damage for bullets and missiles,
  
• beam durations and velocities
  
• critical damages
  
• not to forget about quirks that fiddle with those stats

so I know that those values will not be 100% true. But they don't have to.

[Karl Streiger]

Data from 27 players while it must be little bit inaccurate because of multiple patches in the time of those stats.
But I think the direction is obviously.
The weapon most fired was the IS-Machine Gun ~3,600,000 shots. but merely 167,000 dmg.

Best accuracy is achieved with the C-ER-M-Laser with 81% hits but of those hits only 45% deal damage. Or in other words of each shot with your ER-M-Laser only 3dmg on the average will connect.

Damage per shot is interesting. C-ER-PPC is on par with the IS-Gauss Rifle.... interesting the C-Gauss Rifle deals 8.5dmg per shot - 1 point more than the C-ER-PPC / IS-Gauss Rifle.

The ASRM6 deals only 0.04dmg over the SRM4 without Artemis for each missile fired.
I will now go in detail and throw all the other stats into the huddle and see where i will end.

[Karl Streiger]

Hello a minor update.

While I'm still working on a "quality" ranking I want to share the intermediate state this is in a simple json format including most relevant data (or at least I think it is relevant) about weapons.
I also included the summary of your (players) weapon stats.

BTW:
I may need more date (is there anybody who was so mad to drive a ClanSRM2 w Artemis?)
Why do Clan LRMs w Artemis deal less damage - in comparison to Clan LRMs without Artemis (really it look like Artemis on Clan Missiles is general the worse choice, not only does this combination weight more but the average damage done is also lower (so the average damage for a volley of C-ALRM20 is 4.2 while the average damage without artemis is 5.8!!)

Considering weight and heat of a weapon system + the accuracy of those 27 players the Clan Gauss Rifle is the best weapon in game - followed by the IS Small Pulse Laser
(both got ~14 points)

[Widowmaker1981]

Karl Streiger, on 06 June 2016 - 11:28 PM, said:

Hey there,

i need your help. I'm working on a weapon table from time to time. This weapon table should consider each and every weapon with all of their stats and grant a point value.

While I have some accuracy calculations that seems to be ok, i think it would be much better to have "real" accuracy stats.

So i you want to help me, please go;
Profile -> Stats -> Weapons
and copy & past the table and send it to me as a PM.

I will not forward this information to anybody. Thanks in advance

I dont know how many times ive told people on these forums to stop looking at weapon stats. They are utterly garbage, completely incorrect. For gods sake, mine show a TOTAL number of games where i have used the IS LPL of 194. My mech stats show 1478 matches in the BNC-3M, and i have NEVER run that mech without LPLs installed. ever.. and i also use LPLs on pretty much every single IS mech that runs lasers (because IS-LPLs are a bit OP).

I.e. my weapon stats are missing at least 90% of games, likely more. They literally mean nothing.

[Karl Streiger]

Widowmaker1981, on 16 June 2016 - 06:23 AM, said:

I.e. my weapon stats are missing at least 90% of games, likely more. They literally mean nothing.

it doesn't matter if games are missing or not.
I just look at the numbers: fired, hit and damage - don't care about the rest

So even when there are games missing or even registered incorrectly the same mistake is applied to the shots you fired - and those other two values.

It is unlikely that the stats will only register "shots" but not "hits" - also its unlikely that they will register damage but not the other values.

Of course, I know that I won't get 100% accurate values. It is not longer possible with the current stats (to much changes, quirks and other stuff)

But I would be surprised if there wouldn't be a tendency

Edited by Karl Streiger, 16 June 2016 - 07:08 AM.

[Afuldan McKronik]

Sent. My terrible accuracy with some weapons is getting better though.

[Yellonet]

Widowmaker1981, on 16 June 2016 - 06:23 AM, said:

I dont know how many times ive told people on these forums to stop looking at weapon stats. They are utterly garbage, completely incorrect. For gods sake, mine show a TOTAL number of games where i have used the IS LPL of 194. My mech stats show 1478 matches in the BNC-3M, and i have NEVER run that mech without LPLs installed. ever.. and i also use LPLs on pretty much every single IS mech that runs lasers (because IS-LPLs are a bit OP).

I.e. my weapon stats are missing at least 90% of games, likely more. They literally mean nothing.

Yeah I've run a ERLL raven plenty of times, but still I've only 50 matches with ERLL... hmm..

[Karl Streiger]

Thanks for your participation:

I have a minor update: it is primary for those that say: that those stats doesn't work. Stop arguing:

As you can see out of just 32 participants i have very few aberrations, but non representative absolute numbers. But for the most the accuracy is obviously almost the same.
So when i will add all those shots and hits together into a single value they will align around a single peak. This means those statistics could be used as valid foundation for additional research

[Afuldan McKronik]

My brain is trying to start. So the IS LL is a little more accurate for all the players combined than the IS LPL?

[Whaler]

PM sent.

[Bishop Steiner]

Karl Streiger, on 07 June 2016 - 12:58 AM, said:

Thanks, I would prefer to have plain text rather than images
(don't bother El' my secretary Karline did type everything)
Don't bother about formats - I will make a program to extract the data.

Thanks for the first answers

add this post to the OP, since a lot of people will be skipping anything below it.

Will mad stats mess up your extraction?

[Karl Streiger]

Afuldan McKronik, on 17 June 2016 - 06:31 AM, said:

My brain is trying to start. So the IS LL is a little more accurate for all the players combined than the IS LPL?

Yes only because of burn duration - its not necessary that the full 'burn' hit the target - things would look different when I will provide the second graph showing the 'average' damage.

For example the CERlarge got the best toHit rating but almost the worst to damage rating

Bishop Steiner, on 17 June 2016 - 07:32 AM, said:

add this post to the OP, since a lot of people will be skipping anything below it.

Will mad stats mess up your extraction?

Don't know maybe mad stats will mess up nut don't bother - can fix it.

[Afuldan McKronik]

Karl Streiger, on 17 June 2016 - 07:44 AM, said:

Yes only because of burn duration - its not necessary that the full 'burn' hit the target - things would look different when I will provide the second graph showing the 'average' damage.

For example the CERlarge got the best toHit rating but almost the worst to damage rating


Don't know maybe mad stats will mess up nut don't bother - can fix it.

Oh I see now. Figured it out a little late. Seen "new graph", forgot "rest of thread".

[Oberost]

Sent.