EPTTvZ Winrates with Mass Widow Mine - Page 6
| Forum Index > SC2 General |
|
BearG
United States33 Posts
| ||
|
spalding
95 Posts
| ||
|
Talack
Canada2742 Posts
On April 08 2013 12:10 FallDownMarigold wrote: It's probably also true Terran has a higher win rate with >7 SCVs produced in a game vs. <7 SCVs. I chose 7 because I didn't want to choose 10. This doesn't really prove that SCVs are OP though... Ok now I want to see the win rates for when zerg makes > 6 drones vs <6 drones | ||
|
BronzeKnee
United States5217 Posts
On April 08 2013 12:23 spalding wrote: OP can you please check if zergs building more than 50 drones have a higher winrate than those who don't? I have a strong feeling that drones might be overpowered and that would prove it! lol some people are so dense, it's unbelievable. But won't the winrates for SCVs and Probes production mirror what Drone production does? In that case, there is no imbalance. All races benefit from producing more workers, and we know this. Anyway, comparing a combat unit that Terran can choose to make, and workers that are required in large numbers (especially for Zerg that loses a worker each time they create a building) is silly. | ||
|
Najda
United States3765 Posts
On April 08 2013 11:36 Emzeeshady wrote: except for the fact that Mutas cost more and are higher up the tech tree. It doesn't have to do with cost, 10 is just the number that would generally be made to determine whether or not you're committing to that playstyle. And just like with the widowmines, it doesn't show anything related to balance because there are too many variables for that. It just shows that these builds have a higher win chance on ladder, which could be attributed to a large number of things. I think it's important to do this with a wide variety of units to see if this is repeatable or if widow mines and mutas are an isolated case. Try it with units like collossus, medivacs, void rays, or infestors and see what information you get out of it. | ||
|
iky43210
United States2099 Posts
Also balance should not be made at the master level. I am a master player, and I'll tell you most master players are jokingly bad compare to even lower tier pros. | ||
teddyoojo
Germany22369 Posts
On April 08 2013 12:26 Talack wrote: Ok now I want to see the win rates for when zerg makes > 6 drones vs <6 drones does this mean drones might actually be underpowered? | ||
|
petered
United States1817 Posts
Before I continue, let me just say that I am an actuary and I do this sort of analysis for a living, so I am not talking out of my ass (not completely anyway). Off the top of my head, here are a few suggestions for controlling variables. I have no idea if this data is available on the site you are using, though. - whenever you are trying to measure effectiveness, you need to be very careful how you measure usage. For example, when determining how good of a scorer a basketball player is, a good analyst will look at shooting percentages rather than points scored. A person who scores 20 points on 15 shots is a heck of a lot better than a person who scores 25 on 30 shots. In our case, using 10 mines made isn't a good metric of how a player is using mines. A much better metric would be either percentage of army resources in mines or percentage of supply. Then we know the difference between a player who produces a couple mines at a time over a long game versus someone building a crapton of mines in a short game. - Control (as much as possible) for the style of gameplay. Lets say terrans lose 75% of the time they do a proxy rax cheese. Every one of those games will fall under a loss with the less than 10 mines, but it doesn't really tell us much about whether or not mines are OP in a standard macro match. While you can't perfectly control for this kind of thing with stats, you can do things like exclude all games under a certain game length or exclude games with less than a certain threshold of workers to get rid as much as possible the games where mines were never going to have an impact one way or the other. this is more of an open ended one that won't necessarily help but it is a thought. - As you have already started doing, use the established metric in a number of ways before commenting on what it might mean. If you get weird or unreliable results in other cases, then it will give you new insight into what might be driving the differences you see. I'm too tired to think of anything else, but I always enjoy more statistics! | ||
|
Empirimancer
Canada1024 Posts
I died. True story. | ||
|
DooMDash
United States1015 Posts
| ||
|
Talack
Canada2742 Posts
On April 08 2013 12:37 teddyoojo wrote: does this mean drones might actually be underpowered? I actually want to see what the win rates of 6-pooling are haha | ||
|
Aerisky
United States12129 Posts
| ||
lichter
1001 YEARS KESPAJAIL22272 Posts
| ||
|
Noocta
France12578 Posts
Zerg players need to shut up already about them and just live with it. It's NOWHERE near as good as the shit Terran had to deal with for the last months of WoL. Sorry to sound a bit harsh, but sometimes I feel players take more time to whine about something rather than finding solutions by playing. ( and not ranting with theorycrafting ) | ||
|
FlukyS
Ireland485 Posts
If it were me id either make the radius lower or lower the damage. Personally they are way more effective in TvZ than they are in any other match up (maybe TvT if the terran isn't scanning and focusing correctly) and its because they 1 shot mass lings no matter what upgrades you have and it requires no micro or babysitting at all. Like I understand that they cost minerals and gas but if you look at what they can kill they are incredibly cheap. I mentioned this in the beta but it still holds true, terran don't need 1 more answer to muta but id like to add that they don't need 1 more answer to lings. Lings are already one of the hardest units to make work in the game because hellions and bio being hugely cost effective vs them. Also the fact that widow mines hit air that makes it slightly annoying because I know they would be balanced otherwise if they changed them to only hit ground. It would make it so terrans would use them more defensively if they were picked off by a small pack of muta. Its all micro though but ive been trying to work ways that I can actually hold my third vs a bio widow mine push with like 5 or 6 mines and defending drops and ive found very few options that work for both. Most pros are going ling bling muta but ive been seeing on ladder them zoning with the widow mines while taking down a base and then falling back and in that case ling bling muta isn't good at all unless you use overlords to tank. Using overlords just is really annoying because then either you have to have more overlords you need or you get horribly supply blocked and it puts you all in pretty much on defending that 1 base. | ||
|
Killerkrack
664 Posts
| ||
|
Empirimancer
Canada1024 Posts
On April 08 2013 13:09 lichter wrote: The only thing that pisses me off about widow mines is the fact that I need hydra range to kill it without losing a unit =/ Or you can use one of your three units that generates free units. Or trigger it with a unit that has more than 125 life, such as a roach or an overlord. Or make a changeling with an Overseer. Or if there's only one or two, kill it with mutas before they blow up. | ||
|
plogamer
Canada3132 Posts
On April 08 2013 13:27 FlukyS wrote: The way it feels like in my games at master level is that I have to work a lot harder than the terran to win and that isn't a balance whine either I have a 60% win rate overall in ZvT. The thing is it either seems like im out playing them heavily or they are just making mines and bio and I get smashed, there is no middle ground or much micro that helps it. If it were me id either make the radius lower or lower the damage. Personally they are way more effective in TvZ than they are in any other match up (maybe TvT if the terran isn't scanning and focusing correctly) and its because they 1 shot mass lings no matter what upgrades you have and it requires no micro or babysitting at all. Like I understand that they cost minerals and gas but if you look at what they can kill they are incredibly cheap. I mentioned this in the beta but it still holds true, terran don't need 1 more answer to muta but id like to add that they don't need 1 more answer to lings. Lings are already one of the hardest units to make work in the game because hellions and bio being hugely cost effective vs them. Also the fact that widow mines hit air that makes it slightly annoying because I know they would be balanced otherwise if they changed them to only hit ground. It would make it so terrans would use them more defensively if they were picked off by a small pack of muta. Its all micro though but ive been trying to work ways that I can actually hold my third vs a bio widow mine push with like 5 or 6 mines and defending drops and ive found very few options that work for both. Most pros are going ling bling muta but ive been seeing on ladder them zoning with the widow mines while taking down a base and then falling back and in that case ling bling muta isn't good at all unless you use overlords to tank. Using overlords just is really annoying because then either you have to have more overlords you need or you get horribly supply blocked and it puts you all in pretty much on defending that 1 base. 60% win rate, still whines, Zerg. Guess what? Did you think that players you outplay heavily are losing because they are stuck in WoL mindset and not utilizing the HotS units? Cuz I'm one of them, lol. Ok, and about holding your third against bio widow mine push. Make static defenses, including the one that can detect. Then voila, widow mine is offensively neutered once detection + spines + units, especially in a counter-attack (controlled properly, let a couple lings run in vanguard to draw out widowmine shots). | ||
|
Ben...
Canada3485 Posts
On April 08 2013 12:54 Empirimancer wrote: What? How dare he use ingenuity and a new tactic to win a game! He should be on a forum whining about said strategy he doesn't want to adapt to being OP instead!I've played against a Zerg who knew I had lots of widow mines and he was obviously tired of sending small packs of lings ahead of his army so he just brought 20 overlords with the speed upgrade with his army, moved them a bit ahead of it, and a-moved his lings, zerglings and ultras. I died. True story. As a person who has taken a fair number of university level statistics classes, I also question the methods used here. As others have mentioned, the method used does have an inherent bias to it that basically makes it difficult not to be true. Several great examples why this doesn't work but here is another. It would be like complaining that ultralisks are imbalanced because zergs who make 20+ ultralisks win the game 70% of the time or something similar to that, when if you think about it if a zerg is able to produce that many ultralisks they probably have a lead of some type in most of the games samples. There are just too many gaps in the argument for it to hold any water. 10 widow mines is not exactly a ton in the first place, especially when many terrans use 2-3 per base for defence. Why isn't it more than a larger number, say 20 or 30 produced throughout the game. That would make more sense to me than 10. On April 08 2013 12:27 BronzeKnee wrote: But won't the winrates for SCVs and Probes production mirror what Drone production does? In that case, there is no imbalance. All races benefit from producing more workers, and we know this. Anyway, comparing a combat unit that Terran can choose to make, and workers that are required in large numbers (especially for Zerg that loses a worker each time they create a building) is silly. Err... I am pretty sure he is joking and making fun of the OP for using skewed statistics. | ||
|
Orek
1665 Posts
| ||
| ||
