On March 10 2010 00:34 micronesia wrote:
Can you explain how 50% miss means 4 hits to kill the marine?
Can you explain how 50% miss means 4 hits to kill the marine?
it says " on average " so really just in theory, in reality its about luck O.o
Forum Index > News |
UbiNax
Denmark381 Posts
On March 10 2010 00:34 micronesia wrote: Can you explain how 50% miss means 4 hits to kill the marine? it says " on average " so really just in theory, in reality its about luck O.o | ||
ckjy
United States12 Posts
Now suppose we upgrade the tank's weapons. If we assumed a level 3 upgrade of +6 to the tank's damage against marines, then the math changes substantially. A 50% miss rate will result in a dead marine with two or fewer volleys 75% of the time, with only one volley required 50% of the time. This is in contrast to a 50% damage reduction, which will always require two shots. Let's take it one step further, and upgrade the marine's armor to +1. Now we have a complicated situation involving damage reduction. Should the damage bonus from the upgrade be applied before, or after the fog-of-war damage reduction of 50%? If it is applied before (so damage taken is [(35+6)*50% - 1]), then three shots are required to kill the marine. If it is applied after (so damage taken is [(35 * 50%) + 6 - 1], then only two shots are required. If there is no damage reduction and only a 50% miss chance, the marine will still die with two or fewer volleys 75% of the time. The difference is drastic, when you consider going from 2 to 3 volleys equates to taking 50% longer to kill a unit. The difficulty of balancing a game goes far beyond deciding whether a number should be calculated one way or the other - it involves taking into account every possible way the battle can play out, and weighing every single variable against each other in a similar fashion. | ||
0neder
United States3733 Posts
| ||
Wintermute
United States427 Posts
On March 10 2010 00:34 micronesia wrote: Can you explain how 50% miss means 4 hits to kill the marine? It doesn't take 4 hits, it takes (on average) 4 shots. Sometimes the first two shots will hit, and sometimes it will take 8 shots, but on average, 4 shots, because you need to hit twice and you have a 50% chance each shot. | ||
theSAiNT
United States726 Posts
| ||
Crunchums
United States11143 Posts
your analysis of %miss increasing shots required to kill is oversimplified. hit miss hit miss is just as likely as miss hit miss hit but the first only requires 3 whereas the second needs 4 | ||
ckjy
United States12 Posts
On March 10 2010 00:41 UbiNax wrote: Show nested quote + On March 10 2010 00:34 micronesia wrote: Can you explain how 50% miss means 4 hits to kill the marine? it says " on average " so really just in theory, in reality its about luck O.o It takes two successful hits to kill the marine, so if it takes n volleys to kill the marine, the last volley must have been a hit, there must have been a hit in the first (n-1) volleys, and there were a total of (n-2) misses. Here's a table of all the possibilities, and the chance of it happening (death in n volleys). Shots Needed..Percent Chance 2......................25.0% 3......................25.0% 4......................18.8% 5......................12.5% 6......................7.8% 7......................4.7% 8......................2.7% 9......................1.6% 10 or more.......2.0% The average number of shots needed is equal to the probability of each possible outcome multiplied by the number of shots that were needed (2 * 25% + 3 * 25% + 4 * 18.8% + ... = 4). A shortcut to this calculation is (shots required) / (hit chance), which works out to be (2 / .5), or 4 shots. | ||
![]()
intrigue
![]()
Washington, D.C9933 Posts
| ||
Hrrrrm
United States2081 Posts
I think using the cliffs on LT and Kulas as an example of this mechanic needing a change is misplaced. These are more map decisions that can be changed. | ||
Maceifer
Germany11 Posts
| ||
Wintermute
United States427 Posts
On March 10 2010 00:44 ckjy wrote: The math for the average number of shots required to kill a marine is incorrect. The expected (average) number of shots required to kill the marine with no damage reduction is (2 / hit rate), so with a 33% hit rate, it will take an average of 6 shots. The math says 50% hit rate, not 33%, so it is correct. Now suppose we upgrade the tank's weapons. If we assumed a level 3 upgrade of +6 to the tank's damage against marines, then the math changes substantially. A 50% miss rate will result in a dead marine with two or fewer volleys 75% of the time, with only one volley required 50% of the time. This is in contrast to a 50% damage reduction, which will always require two shots. With upgrades, a marine will die every shot. 50% miss rate will result in an average of 2 shots per kill. Thus, 50% miss rate always results in causing units on high ground to have (on average) twice as much staying power. Let's take it one step further, and upgrade the marine's armor to +1. Now we have a complicated situation involving damage reduction. Should the damage bonus from the upgrade be applied before, or after the fog-of-war damage reduction of 50%? If it is applied before (so damage taken is [(35+6)*50% - 1]), then three shots are required to kill the marine. If it is applied after (so damage taken is [(35 * 50%) + 6 - 1], then only two shots are required. If there is no damage reduction and only a 50% miss chance, the marine will still die with two or fewer volleys 75% of the time. The difference is drastic, when you consider going from 2 to 3 volleys equates to taking 50% longer to kill a unit. This is why miss chance is suggested instead of damage reduction, because it always gives the same effective advantage, vs damage reduction, which will depend on how much overkill a unit normally would do to another. | ||
lossofmercy
United States29 Posts
Also, warping in through a pylon just seems like a late-game ability. And it is the beta after all. Balance can be fixed in patches as we find out stuff. Right now, Blizzard should be looking for bugs and finding the better build, which means they can change the tech tree. | ||
KOFgokuon
United States14892 Posts
| ||
3nickma
Denmark1510 Posts
On March 09 2010 23:28 damenmofa wrote: ive seen plenty of good players abusing the auto target worker mechanic in fights by bringing a cpl of scvs in big fights and abusing the fact that the ai targets workers with a-click. I think it not only takes away the defenders advantage, but it can also be used offensively to an extent. Then again, I agree with the cliff mechanics. What really is key imo though is bnet delay, muta stacking, right click functionality and a way to move your units in formation, forcing them to NOT clump like they always do. Lol I didn't know your offensive units would auto-target workers instead of the opposing forces when A-moving? That's acutually a pretty big deal but ofc your scenario also gives a new possibility. Bit at least nice to know. | ||
diehilde
Germany1596 Posts
On March 10 2010 01:13 3nickma wrote: Show nested quote + On March 09 2010 23:28 damenmofa wrote: ive seen plenty of good players abusing the auto target worker mechanic in fights by bringing a cpl of scvs in big fights and abusing the fact that the ai targets workers with a-click. I think it not only takes away the defenders advantage, but it can also be used offensively to an extent. Then again, I agree with the cliff mechanics. What really is key imo though is bnet delay, muta stacking, right click functionality and a way to move your units in formation, forcing them to NOT clump like they always do. Lol I didn't know your offensive units would auto-target workers instead of the opposing forces when A-moving? That's acutually a pretty big deal but ofc your scenario also gives a new possibility. Bit at least nice to know. yeah its quite frustrating because this is one of the few points in the game where ur units actually feel dumber in sc2 than in sc1. I wanna tear my hair out when I attack with massive roaches against his mmm ball and he has like 10 scvs in front and the roaches waste their attacks against the scvs while the marines/marauders happily shoot them down without being touched. | ||
onmach
United States1241 Posts
| ||
![]()
Mystlord
![]()
United States10264 Posts
The (non-existent) cliff mechanic is also troubling. It takes away an element of realism from the game. | ||
phyvo
United States5635 Posts
| ||
hugman
Sweden4644 Posts
On March 10 2010 01:35 Mystlord wrote:The (non-existent) cliff mechanic is also troubling. It takes away an element of realism from the game. Tell me you're joking! | ||
UnderWorld_Dream
Canada219 Posts
Cliffs should be just as it is in bw, you get sight on higher grounds only when they shot at you. After all, when a marine is gunning you, how can you not see him if he sees you, thats totally dump. You are still less likely to fight from the bottom if you have chances to miss and thus you will try to avoid these situations anyways. I say just put cliff rules like in was in bw. (i know its a different game, I can still point out it was better before) | ||
| ||
![]() StarCraft 2 StarCraft: Brood War Calm Stormgate![]() Rain ![]() Flash ![]() Stork ![]() actioN ![]() Mini ![]() firebathero ![]() PianO ![]() Hyun ![]() Rock ![]() [ Show more ] Counter-Strike Heroes of the Storm Other Games Organizations
StarCraft 2 StarCraft: Brood War Dota 2 League of Legends Other Games |
Replay Cast
Replay Cast
Afreeca Starleague
Snow vs Soulkey
WardiTV Invitational
PiGosaur Monday
GSL Code S
ByuN vs Rogue
herO vs Cure
Replay Cast
GSL Code S
Classic vs Reynor
GuMiho vs Maru
The PondCast
RSL Revival
[ Show More ] GSL Code S
OSC
Korean StarCraft League
RSL Revival
SOOP
Online Event
Clem vs ShoWTimE
herO vs MaxPax
Sparkling Tuna Cup
WardiTV Invitational
RSL Revival
Wardi Open
|
|