KnightHawk270
Recruit
Posts: 4
Karma: 1
Platform: Android
League: Silver
Server Region: North America
Favorite robot: Galahad
|
Post by KnightHawk270 on May 21, 2017 12:42:44 GMT -5
Ballistic Spread, Missiles Flying Into Structures, Knife-fights with Stalkers and Cossacks
These and other things LOWER the damage inflicted and can make Cycle Damage, DPS, etc. statistics almost WORTHLESS!.I'm working on the latest and greatest (modest, I know) weapon spreadsheet. I've found with many weapon damage spreadsheets the data is taken from other sources AS IS, even if doesn't reflect real game play (e.g., DPS off for weapons that reload while firing).I'll provide the completed spreadsheet (and the current version, if interested) to those who help me with the EFFECTIVE DAMAGE metric. I need: 1) Which estimates I'm currently using are farthest off? OR 2) Your estimates of Hit % for weapons.
Effective Weapon Damage = ["Official" Damage for Weapon] x [Percentage of Projectiles Actually Hit]Here are my newbie estimates of % Projectiles Hit Enemy:
% of Rounds
Weapon Hit TargetECC Thunder 70% Trident 90% Zeus 95% Trebuchet 90% Nashorn 90% ETC Kang-Dae 90% Magnum 90% Punisher 85% AC Molot 85% EE Aphid 80% Pinata 75% CRV Pin 75% GEKKO XX 1 00% AT Spiral 85% Punisher Mk2 85% AC Molot Mk2 85% Taran 90% Orkan 80% Tulumbas 75% Effective Weapon Damage = ["Official" Damage for Weapon] x [Percentage of Projectiles Actually Hit] - "Official Damage for Weapon" comes from the official website for now; Will adjust weapon damage numbers for continuous reload, etc later - "Percentage of Projectiles Hit" is based on your personal experience with % of missles hit, % of Thunder damage inflicted, % of Trebuchet shots hit target overall, etc. These are my best guesses, but I'm POSITIVE others can blow holes through these estimates and have something way better. Again, I'll PM the spreadsheet [SEE BELOW] to those who help out first, then maybe at later date release publicly. Thanks! 
|
|
|
|
Post by Thunderkiss on May 21, 2017 13:27:37 GMT -5
%'s seem arbitrary. What are you using to determine this?
|
|
KnightHawk270
Recruit
Posts: 4
Karma: 1
Platform: Android
League: Silver
Server Region: North America
Favorite robot: Galahad
|
Post by KnightHawk270 on May 21, 2017 13:42:16 GMT -5
The %'s are based on personal experience and best guesses. The whole point is to get MORE EXPERIENCED PLAYERS estimates and base Effective Damage off their estimates instead of just mine.
For example, people post things like, "Yeah, Thunder's range says 500m, but it's more like 200m." I start firing around 350-400m and it does some damage, but nothing compared to when shooting at 100m distance. So overall I'm discounting the "Official Damage."
With Hydra's, as another example, seems like 4-5 rockets - on average - hit the target overall. Many times 100% hit, others maybe 1 of 6 hit. So when calculating DPS, Cycle Damage, etc. working off assumption 100% hit *every* time isn't realistic. An EDUCATED GUESS is much better, in my opinion.
Thanks for the question!
|
|
|
|
Post by shakingrabbit on May 21, 2017 17:22:27 GMT -5
Seems like a lot of your %'s are based on factors that may be out of your control. For example, moving vs stationary target. At 350 yds, Taran is 100% if the target is stationary, if the target is moving, it depends entirely on both pilots skills. The Gekko, is really good at long distances but get in close and the % drops off significantly if the target is running around you.
Shielded targets could get even trickier, for example Aphids vs a Gareth. Full front impact does not show any damage, but you can slip a couple past the shield depending on how you and the other pilot are moving.
I think, aside from the Thunder, Punisher and Molot, you should make the assumption that all projectiles are hitting 100% at range. Eliminates variables that you cannot control.
For splash weapons you could find estimations for damage away from the point of impact.
|
|
|
|
Post by moody on May 21, 2017 19:24:40 GMT -5
Seems very arbitary and dependent both on pilot skill and how tough the shot is.
If I'm in a trident bot for example, I am taking shots at things behind cover that are difficult to judge. If I were using kwk's I wouldn't attempt those shots.
Tarans at extreme range are less accurate than they once were - and again it depends on pilot skill. (also on torso traverse speed of the given bot).
|
|
KnightHawk270
Recruit
Posts: 4
Karma: 1
Platform: Android
League: Silver
Server Region: North America
Favorite robot: Galahad
|
Post by KnightHawk270 on May 21, 2017 19:28:37 GMT -5
shakingrabbit: You make a good point. This may be more effort than it's worth. I was working under the assumption for stationary v. moving targets it would be reflected in each person's personal experience with the weapon. For example, the Hydra, Spiral, etc. are very good at hitting moving targets on average and excellent at stationary ones. So you have that going for it. Then again, I've fired more than I care to admit (but getting better) into buildings, ceilings, etc which detracts from the hit %. However, overall, I'd think the % hit for those type of weapons are better than rockets (Pin, Pinata, etc) and their Effective Damage would be higher (assuming the official/stated damage is the same). They should be higher than the ballastic (Punisher, Molot) type weapon - on average - as well. I'm assuming whether the damage is directly to the target or the target's shield is the the same: Both count as a 'hit.' Appreciate the feedback!
|
|
|
|
Post by DogTrail on May 21, 2017 23:54:02 GMT -5
This is a great project, but % is pretty hard to pin down. First off, my hit rate goes down when I use iPhone vs. iPad. So that's entirely on me, not the weapon. Second, let's take Orkans for example. Moving target - is it moving sideways? backwards? up a ramp towards you? how fast is the target? All these contribute to hit rate, and on my best day I'm pretty sure less than 80% of it hits completely (not counting splash damage). I suppose some constants need to be assumed. Stationary = not moving, not behind a wall or building, unshielded by anything. Moving = Galahad, sideways, shield down, no elevation difference from shooter. Something like that.
PS: Note that DPS and DPM are two different statistics. DPS is what it is - how much damage does it deal in 1 second, regardless how long the clip empties. DPM is where reload time is considered.
|
|
buford
Destrier
Posts: 78
Karma: 62
|
Post by buford on May 22, 2017 8:30:40 GMT -5
This concept is fantastic and one I'm so surprised is not discussed more. Although, as others have said, getting numbers to support it is very very difficulty and subjective.
The idea that a weapon puts out X DPM and is better then a weapon that puts out half as much DPM is always the argument, but rarely is it discussed the difficulty of getting the higher DPM weapon to land the damage it's capable of, or how long the bot will last in range to deliver that damage. I have a feeling the treb (I know, gasp and cough) would end up being one of the highest damage weapons based on these statistics. A bot that can sit in the corner and charge up a weapon to deliver full damage from across the field without ever moving, although tremendously boring and annoying, has a huge advantage. Just one example I think of. It's also why tridents are so devastating. Little movement, safer range, splash, yada yada yada...
|
|
|
|
Post by K A Z on May 22, 2017 8:47:28 GMT -5
The only way to make it really representative would be to take actual stats from what game registered as a damage and the registered number of rounds shot (which again is prone to inconsistencies in the server-player link). Arbitrary numbers like above are no good if you plan to achieve truly representative results.
Many examples can be made, but suffice to say - chase gary or stalker and show me the 80% hit rate with Pinata/Orkan...no way. On top of that consider the non-registered damage issues (e.g. zeus).
Idea is nice but the statistical data you intend to produce requires...statistical data input.
|
|
|
|
Post by miatahead on May 22, 2017 12:29:53 GMT -5
Wow, that's a big spreadsheet to make just for pure speculation. You must have a lot of free time in your Mom's basement.
|
|
KnightHawk270
Recruit
Posts: 4
Karma: 1
Platform: Android
League: Silver
Server Region: North America
Favorite robot: Galahad
|
Post by KnightHawk270 on May 22, 2017 16:00:15 GMT -5
@mrpotatohead: LOL! You're hilarious. I value companies, buy/sell publicly traded stocks and manage investment portfolios for a living. This means I eat, sleep, breathe spreadsheets. I crank these out faster than it takes you to write idiotic comments. By the way, your Mom says, "Hi!"
|
|
|
|
Post by SGT D00M! on May 22, 2017 16:08:06 GMT -5
Wow, that's a big spreadsheet to make just for pure speculation. You must have a lot of free time in your Mom's basement. How the ?firetruck? was that called for 「bum-bum」-hat? Way to be a ?wee-wee?. |
|
|
|
Post by moody on May 22, 2017 17:20:21 GMT -5
Wow, that's a big spreadsheet to make just for pure speculation. You must have a lot of free time in your Mom's basement. How the ?firetruck? was that called for ?bum-bum?-hat? Way to be a ?wee-wee?. I think I'd agree if I knew what you were saying  |
|
|
|
Post by Tatamat on May 23, 2017 6:25:30 GMT -5
Would be nice, but I agree that the % is so variable that it won't really work.
|
|
|
|
Post by Thunderkiss on May 23, 2017 6:58:56 GMT -5
I think the only way to do this is assume 100% hit rate and have a caveat emptor, ymmv, whatever as a warning. For instance, what brought this to my attention was Zeus is 100% hit rate, barring cover. If your target uses cover you'll lose dpm waiting for a good shot. What that wait is is the caveat.
|
|
|
|
Post by 0ppressor on May 23, 2017 17:42:13 GMT -5
Too arbitrary for my taste, but good luck!
|
|
|
|
Post by silvernja on Dec 24, 2017 21:33:46 GMT -5
@mrpotatohead: LOL! You're hilarious. I value companies, buy/sell publicly traded stocks and manage investment portfolios for a living. This means I eat, sleep, breathe spreadsheets. I crank these out faster than it takes you to write idiotic comments. By the way, your Mom says, "Hi!" Spreadsheets and crunching numbers is also a hobby of mine. When I'm not able to actually play. I'll be researching something. When creating a dps kind of numbers. Perfect hit is sufficient for comparison.. inaccuracies for movement, barriers and such is out of control and unnecessary for weapon comparison. Given that, I plan on working on distance and bot size damage etc. How many bullets hit for example. Thunder. Barely hits at 500m. Hits great at 50m. How long will a ancile Shield last under fire by different weapons accounting for Regen.etc. I'm just getting started which brought me here looking for data. |
|
|
|
Post by optimusprime on Dec 25, 2017 8:16:48 GMT -5
Too many variables.Wont even start with this
|
|
|
|
Post by silvernja on Jan 2, 2018 20:53:14 GMT -5
Thunder damage. Not a varying percent but by the bullet. And the spread of each shot changes. Still running numbers. Obviously distance matters when it comes to damage using a thunder. But how much. As you can see in the image every fire of the thunder results in a different spread. (These are not all the spreads. I think at least 2-3 more) Which alters how many bullets hit. Especially in the 200-300 range. Tested [20 shots] perfect front on aim at each range 16 bullets hit consistently upto 70 meters. No less then 15 bullets up to 90 meters 100 meters 14-16 bullets hit. At 200 meters 4 to 9 bullets hit At 300 meters the thunder hits with 1-4 bullets never hit a Natasha with 5 bullets. Rarely 0 Testing every 10 meters. Hopefully will have time this weekend to finish. Questions?  |
|
|
|
Post by Pulse Hadron on Jan 2, 2018 20:58:58 GMT -5
So a Natasha is the target in all these tests? And what bot is carrying the Thunder?
Edit: oh, I just got the image now. So it’s also a Natasha that’s firing. Where did you fire to get those bullet pattern images? I tried to capture these too but didn’t find anywhere as clean.
|
|
|
|
Post by silvernja on Jan 2, 2018 21:06:56 GMT -5
Wow, that's a big spreadsheet to make just for pure speculation. You must have a lot of free time in your Mom's basement. For the strategic and analytical mindset this kind of speculation comes natural. |
|
|
|
Post by Pulse Hadron on Jan 2, 2018 21:08:26 GMT -5
Also, are you saying there’s only 8-9 specific bullet spreads emanating from the Thunder? I’d always assumed a random number generator jittered the bullets each time.
|
|
|
|
Post by moody on Jan 2, 2018 21:23:33 GMT -5
What is it with all the necromancy going on at the moment. Every second thread is 6 or 8 months old.
|
|
|
|
Post by sochilli (Saltesers) on Jan 2, 2018 22:07:19 GMT -5
What is it with all the necromancy going on at the moment. Every second thread is 6 or 8 months old. VNTF? |
|
|
|
Post by Strayed on Jan 3, 2018 3:22:50 GMT -5
|
|
|
|
Post by sochilli (Saltesers) on Jan 3, 2018 6:07:25 GMT -5
tbh what's even wrong with necroposting in the first place? |
|
|
|
Post by moody on Jan 3, 2018 7:40:42 GMT -5
tbh what's even wrong with necroposting in the first place? If it is a valid continuation of a subject - nothing. It is when the situation has changed yet the thread on the front page says buy a Galahad - you won't regret it. |
|
|
|
Post by ᎶƦ℮℮ƊᎽ ƤΛƝƊΛ on Jan 3, 2018 9:43:58 GMT -5
Have you built a factor for splash damage occasionally hitting multiple targets?
|
|
|
|
Post by Koalabear on Jan 3, 2018 9:52:24 GMT -5
Interesting project. I think it has merit. I've found that a lot of the posted damage stats to be way off so it's good to see someone taking the time to factor in misses and stuff - especially for weapons like Punihsers and Thunders. Maybe do a percentage range? Like Thunder from 500m to 50m (or whatever the range is where you get almost 100% damage). I think this would help people use their weapons better.
|
|
|
|
Post by krebby on Jan 3, 2018 12:00:21 GMT -5
Thunder damage. Not a varying percent but by the bullet. And the spread of each shot changes. Still running numbers. Obviously distance matters when it comes to damage using a thunder. But how much. As you can see in the image every fire of the thunder results in a different spread. (These are not all the spreads. I think at least 2-3 more) Which alters how many bullets hit. Especially in the 200-300 range. Tested [20 shots] perfect front on aim at each range 16 bullets hit consistently upto 70 meters. No less then 15 bullets up to 90 meters 100 meters 14-16 bullets hit. At 200 meters 4 to 9 bullets hit At 300 meters the thunder hits with 1-4 bullets never hit a Natasha with 5 bullets. Rarely 0 Testing every 10 meters. Hopefully will have time this weekend to finish. Questions? Very interesting. According to the official War Robots FAQ, and confirmed by your screen shots, a Thunder fires 16 pellets per shot. At level 12 (mk1) a full 16-pellet shot deals 14,720 damage, which means each individual pellet deal 920. Assuming your 20 test numbers are an accurate representation of expected hits, it means that: - At 100m or less, you are dealing between 12,880 (14 pellets) and 14,720 (16 pellets) per shot. Since you fire one shot per second, this DPS damage is comparable to a Level 12 Zeus (14,891).
- At 200m, you are dealing between 3,680 (4 pellets) and 8,280 (9 pellets) per shot. This DPS is comparable to a Magnum (4,080) on the low end, and a Taran (8,448) on the high end.
- At 300m, you are dealing between 920 (1 pellet) and 3,680 (4 pellets) per shot. Thus, you're dealing negligible (920) to Magnum (4,080) damage.
This makes me reconsider my Thunder play in general. The value of getting within 100m compared to 200 is enormous—an average of 13,800 DPS compared to 5,980. |
|
