Shot Destroy Probability

In-depth tactical discussion on how to lose the least

Moderator: Defcon moderators

grzybiarz
level1
level1
Posts: 24
Joined: Fri Jun 21, 2013 10:06 am

Shot Destroy Probability

Postby grzybiarz » Mon Nov 18, 2013 8:21 pm

I read source code recently, found one interesting table - probability of shots destroying their targets. So:

Silos vs:
  • nuke - 25%
  • fighter/bomber - 10%

Sub vs:
  • sub - 10%
  • bs/vc - 25%

BS vs:
  • sub - 30%
  • bs - 10%
  • fighter - 30%
  • bomber - 20%
  • cv - 20%

Fighter vs:
  • sub - 3%
  • bs/cv - 10%
  • fighter - 10%
  • bomber - 30%

Bomber vs:
  • sub/bs/cv - 25%

CV vs:
  • sub - 30%

Enjoy.
User avatar
jon
level4
level4
Posts: 537
Joined: Sun Mar 06, 2011 10:15 pm
Location: London
Contact:

Re: Shot Destroy Probability

Postby jon » Thu Nov 21, 2013 1:15 am

Thank you sir/madam, very useful #bookmarked#
grzybiarz
level1
level1
Posts: 24
Joined: Fri Jun 21, 2013 10:06 am

Re: Shot Destroy Probability

Postby grzybiarz » Thu Nov 21, 2013 9:11 am

well.. I dug deeper into sources, it seems, that i lied.. for once - you have to take into account, that different vessels shoot with different probability. Also, it seems that there in fact are hitpoints in defcon, default to 1, but cvs having 3.

I will come back and post more detailed/accurate results when I'm done.
grzybiarz
level1
level1
Posts: 24
Joined: Fri Jun 21, 2013 10:06 am

Re: Shot Destroy Probability

Postby grzybiarz » Thu Nov 21, 2013 1:14 pm

(Since I'll be talking about ranges - CVs anti-sub range is 5).

Random fun fact: plane position in game world is where its tail ends.

Well, it turns out that CVs and BSs have 3 hitpoints, planes, nukes and subs have 1. Each gunfire has a probability of hit (see list below). If missile hits, it decreases HP by 1. HP do not regenerate over time.

Nuke splash-damage range is 2, destroyed plane/ship is 0.6. Planes running out of fuel and exploding gunfire do not give splash damage (although little explosions are drawn for gunfire). If installation is in splash-damage range, its counted as if was hit with nuke. Ships and planes are destroyed by splash damage regardless of remaining HP. Nukes are immune to splash damage.

When city is in splash-damage range, amount of people dead will scale literally with distance to explosion. For nukes it will be 50% with direct hit, 25% if distance is 1 and 0% when distance is 2. For shoot down planes its 15% when directly over a city, 7.5% at 0.3 and 0% at 0.6.

For depth charges (CVs vs subs), probability of destroying a sub also scales with distance. Countary to some post I've seen here, the closer the sub, the higher the chances that it will get destroyed:
  • directly over a sub - 60%
  • distance 1 - 59%
  • distance 2 - 55%
  • distance 2.5 (half range) - 53%
  • distance 3 - 50%
  • distance 4 - 41%
  • distance 5 (max range) - 30%

And now a ultimate table, of speed, radar and fuel range and other fun stuff. Speed is given in distance/s. So, for fighter it takes 50s to travel CVs anti-sub range. I've also run some experiments, so besides shot hit probability, I calculated average number of shots needed to destroy target, and average number of shots for those "just won't die" targets (top 10% most durable). Time it will take for fire all needed shots is also given in parentheses.

Fighter:
  • HP: 1
  • Range: 45
  • Radar range: 5
  • Speed: 0.1
  • Shoot cool-down: 20s
  • Combat range: 10
  • Hit probability:
    • sub - 3%, average: 32 shots (10:40), won't die: 104 (34:20)
    • bs/cv - 10%, average: 30 shots (10:00), won't die: 63 (21:00)
    • fighter - 10%, average: 10 shots (3:20), won't die: 33 (11:00)
    • bomber - 30%, average: 3 shots (1:00), won't die: 9 (3:00)

Bomber:
  • HP: 1
  • Range: 140
  • Radar range: 10 in naval mode, 5 in nuke mode
  • Speed: 0.05
  • Shoot cool-down: 30s
  • Combat range: 20 in naval mode, 25 in nuke mode
  • Hit probability:
    • sub - 25%, average: 4 shots (2:00), won't die: 12 (6:00)
    • cv/bs - 25%, average: 12 shots (6:00), won't die: 25 (12:30)

Sub:
  • HP: 1
  • Radar range: 3 when surfaced
  • Speed: 0.02
  • Shoot cool-down: 20s
  • Combat range: 5 vs naval, 45 in nuke mode
  • Hit probability:
    • sub - 10%, average: 10 shots (3:20), won't die: 34 (11:20)
    • cv/bs - 25%, average: 12 shots (4:00), won't die: 24 (8:00)

BS:
  • HP: 3
  • Radar range: 10
  • Speed: 0.03
  • Shoot cool-down: 20s
  • Combat range: 10
  • Hit probability:
    • sub,fighter - 30%, average: 3 shots (1:00), won't die: 9 (3:00)
    • bomber - 20%, average: 5 shots (1:40), won't die: 16 (5:20)
    • cv - 20%, average: 15 shots (5:00), won't die: 32 (10:40)
    • bs - 10%, average: 30 shots (10:00), won't die: 66 (22:00)

CV:
  • HP: 3
  • Radar range: 15
  • Speed: 0.03
  • Shoot cool-down: 60s
  • Combat range: 5
  • Hit probability:
    • sub at point blank - 60%, average: 1 (1:00), won't die: 4 (4:00)
    • sub at half range - 53%, average 1 (1:00), won't die: 4 (4:00)
    • sub at max range - 30%, average 3 (3:00), won't die: 9 (9:00)

Silo:
  • Radar range: 10
  • Shoot cool-down: 20s
  • Combat range: 30
  • Hit probability:
    • nuke - 25%, average: 4 shots (1:20), won't die: 12 (4:00)
    • fighter/bomber - 10%, average: 10 shots (3:20), won't die: 34 (11:20)

Radar:
  • Radar range: 20

Airbase:
  • Radar range: 10
User avatar
jon
level4
level4
Posts: 537
Joined: Sun Mar 06, 2011 10:15 pm
Location: London
Contact:

Re: Shot Destroy Probability

Postby jon » Thu Nov 21, 2013 1:47 pm

They have 3 HP? This is BIG NEWS
grzybiarz
level1
level1
Posts: 24
Joined: Fri Jun 21, 2013 10:06 am

Re: Shot Destroy Probability

Postby grzybiarz » Thu Nov 21, 2013 3:12 pm

I've tested it with 36 bombers vs 36 BS, launching single bomber vs single BS one at a time:
  • None of the BS died before 3rd shot (in fact the most "unlucky" one died with 4th shot)
  • I think BS died usually with 10-11th (didn't took precise notes), so it seems, that all above table is correct

So..
YES, BS/CV HAVE 3 HP!
User avatar
trickser
level5
level5
Posts: 1823
Joined: Thu Mar 06, 2008 2:15 pm
Location: The Senate ; GMT+1
Contact:

Re: Shot Destroy Probability

Postby trickser » Thu Nov 21, 2013 10:06 pm

Great table, bookmarked it :)

It's almost embarrassing i've never heared about/noticed the 3HP thingy.

And your average calculation is wrong. You cannot just lineary add the probability until you got 100. You have to account for the cases where target already died. I can not come up with a proper formular, but ihmo the averages need to be a bit higher.
grzybiarz
level1
level1
Posts: 24
Joined: Fri Jun 21, 2013 10:06 am

Re: Shot Destroy Probability

Postby grzybiarz » Fri Nov 22, 2013 8:30 am

Average shoot counts are right. For each scenario I've run 1000 experiment and counted shoots needed to kill target. In table I gave average for all 1000 experiments and for top 100 with most shoots needed (the "won't just die" part). So I'm pretty sure those are right.
User avatar
trickser
level5
level5
Posts: 1823
Joined: Thu Mar 06, 2008 2:15 pm
Location: The Senate ; GMT+1
Contact:

Re: Shot Destroy Probability

Postby trickser » Fri Nov 22, 2013 9:16 am

Oh you did it by experiment. Interesting. Guess then something with my thought must be wrong.
May I ask how you did the experiments? I assume you didn't issue 1000 attack orders by hand for every unit combination.
Mr.Cows
level2
level2
Posts: 211
Joined: Fri Nov 08, 2013 4:23 pm

Re: Shot Destroy Probability

Postby Mr.Cows » Fri Nov 22, 2013 11:49 am

What is the odd of 3 subs taking other 3 subs out all with first shots. That happened to me yesterday and my math sucks so can someone calculate me the odds of that happening?
grzybiarz
level1
level1
Posts: 24
Joined: Fri Jun 21, 2013 10:06 am

Re: Shot Destroy Probability

Postby grzybiarz » Fri Nov 22, 2013 12:21 pm

3 subs vs 3 subs, each sub firing at single enemy sub, destroying them in 1st shot: 0.1%

3 subs vs 3 subs, 3 subs firing at 1st enemy sub and destroying it, then at 2nd and destroying it, and then at 3rd and destroying it: 2%.
User avatar
jon
level4
level4
Posts: 537
Joined: Sun Mar 06, 2011 10:15 pm
Location: London
Contact:

Re: Shot Destroy Probability

Postby jon » Sat Nov 23, 2013 4:45 pm

Sir you win a medal for defcon contribution.
zoros dog
level2
level2
Posts: 146
Joined: Sun Sep 19, 2010 3:14 pm

Re: Shot Destroy Probability

Postby zoros dog » Tue Nov 26, 2013 1:44 am

Cool thread! Some of it I knew, some I didn't. Some I never cared to understand :P

Random fun fact: plane position in game world is where its tail ends.


I have been aware of this for a long time. How is it useful: the back of the plane is where you should put your mouse when using the insta-fighter trick. It allows you to use the trick very quickly and save time.

Well, it turns out that CVs and BSs have 3 hitpoints, planes, nukes and subs have 1. Each gunfire has a probability of hit (see list below). If missile hits, it decreases HP by 1. HP do not regenerate over time.


Every decent player knows this. They may not think they know it, but they can certainly feel it during play. Think about it... have you ever, ever seen a battleship die in 1 shot at the beginning of Defcon 3? Certainly in the hundreds of games you have played, you should at least have seen it once. But it has never happened! But we have all seen a surfaced sub get taken out in 1 shot by a stray fighter from time to time. Or a single fighter taking out 3 bombers in 3 shots.

Nuke splash-damage range is 2, destroyed plane/ship is 0.6. Planes running out of fuel and exploding gunfire do not give splash damage (although little explosions are drawn for gunfire). If installation is in splash-damage range, its counted as if was hit with nuke. Ships and planes are destroyed by splash damage regardless of remaining HP. Nukes are immune to splash damage.


Another fun fact. The nuke splash damage diameter is exactly the same size as the order circle (the target circle that appears in the world where you target a city, move a ship, move a plane, etc.).

When city is in splash-damage range, amount of people dead will scale literally with distance to explosion. For nukes it will be 50% with direct hit, 25% if distance is 1 and 0% when distance is 2. For shoot down planes its 15% when directly over a city, 7.5% at 0.3 and 0% at 0.6.


I noticed and tested this when I was bored once.

For depth charges (CVs vs subs), probability of destroying a sub also scales with distance. Countary to some post I've seen here, the closer the sub, the higher the chances that it will get destroyed:

directly over a sub - 60%
distance 1 - 59%
distance 2 - 55%
distance 2.5 (half range) - 53%
distance 3 - 50%
distance 4 - 41%
distance 5 (max range) - 30%


Definitely noticed this. We have all seen a sub that is running away from a CV, just on the edge of its attack circle, and seems to live forever. But a sub underneath the CV? Dead!

One time Weps told me that the way Defcon should have been made was with HP instead of probability. Each unit starts with a certain HP and each successful hit takes away a certain amount of HP... no chance involved. Well it seems Defcon has a little mix of both...

One more thing: one thing I was wondering about before I stopped playing Defcon was if the location a BB/CV gets hit affects its chance of dying? For example, I felt that attacking a BB/CV from the bottom seemed to be more effective than shooting it from the top. Does the source code show anything else interesting about damage?
User avatar
tllotpfkamvpe
level5
level5
Posts: 1768
Joined: Fri Aug 31, 2007 12:04 am
Contact:

Re: Shot Destroy Probability

Postby tllotpfkamvpe » Tue Nov 26, 2013 7:23 am

Random fun fact: plane position in game world is where its tail ends.


My experience says the hit detection zone is in the middle of the plane. Maybe the plane position and hit detection are 2 different things, ie a plane has landed on an airbase when it's tail has reached a certain point on an airbase or carrier. Also the plane orders line starts from the tail, and the object trails also start from the tail. Therefore if I was trying to nuke a plane I would aim to hit it's centre and not it's tail.

Well, it turns out that CVs and BSs have 3 hitpoints, planes, nukes and subs have 1. Each gunfire has a probability of hit (see list below). If missile hits, it decreases HP by 1. HP do not regenerate over time.


Did you actually test that Hp do not regenerate over time? The source code also has stuff about saucers, tornados, and fighters being able to kill nukes. If Hp regenerated it would explain why BOR tactics were so effective in the past.

One more thing: one thing I was wondering about before I stopped playing Defcon was if the location a BB/CV gets hit affects its chance of dying? For example, I felt that attacking a BB/CV from the bottom seemed to be more effective than shooting it from the top. Does the source code show anything else interesting about damage?

Image
Image
Image

You should aim to nuke near the top of units. I don't know about regular fire, maybe the effect you saw was psychological. If you fire from the top the explosion marks are actually outside the unit, but if you aim from the bottom the explosion marks happen inside the units.
Noryb
level2
level2
Posts: 133
Joined: Thu Apr 28, 2011 12:55 am
Location: (UTC -8)

Re: Shot Destroy Probability

Postby Noryb » Fri Nov 29, 2013 3:49 am

zoros dog wrote:
Well, it turns out that CVs and BSs have 3 hitpoints, planes, nukes and subs have 1. Each gunfire has a probability of hit (see list below). If missile hits, it decreases HP by 1. HP do not regenerate over time.


Every decent player knows this. They may not think they know it, but they can certainly feel it during play. Think about it... have you ever, ever seen a battleship die in 1 shot at the beginning of Defcon 3? Certainly in the hundreds of games you have played, you should at least have seen it once. But it has never happened! But we have all seen a surfaced sub get taken out in 1 shot by a stray fighter from time to time. Or a single fighter taking out 3 bombers in 3 shots.


I've seen it! Just depends on the type of shot.

Clarification: If they do have hitpoints (and I'm not saying that they don't), they're different from the hitpoints that silos and airbases have. At least with regards to airplane splash damage. Just did some testing to confirm, but a downed plane will definitely kill an undamaged BB or CV rather than take off a hitpoint.

zoro dog wrote:One more thing: one thing I was wondering about before I stopped playing Defcon was if the location a BB/CV gets hit affects its chance of dying? For example, I felt that attacking a BB/CV from the bottom seemed to be more effective than shooting it from the top. Does the source code show anything else interesting about damage?


I'm not sure about that, but I've heard people talking about bombers being more likely to die with direct hits to the nose. I'm not sure how valid that is, but it's interesting.

Return to “Strategic Air Command”

Who is online

Users browsing this forum: No registered users and 3 guests