|
After playing around with the different weightings for a little bit, I think orange, green and purple would give a pretty balanced set of keyscores. If one would weight CGs higher than abilities, red would be the way to go without getting too much out of balance. And if abilities are more important one should definitely look into teal.
weight for CGs: 1 - 1 - 1 - 0,9 - 0,8 - 0,5 - 0,2 - 0,2 - 0,1 - 0,1
two macro CGs and one army CG is a must have -> 1.0
second army CG is nearly mandatory -> 0,9
casters is also nice to have -> 0,8
I thinkt he harras CG will also get some use -> 0,5
I think the other CGs are not that much used, or at least not as frequently -> 0,2 to 0,1
weight for abilities:1 - 1 - 1 - 0,9 - 0,7 - 0,5 - 0,2 - 0,2 - 0,2 - 0,1 - 0,1 - 0,1
attack and build basic on the first 3 abilities -> 1,0
many unit abilities on the 4th key -> 0,9
a bit less ablities on the 5th key -> 0,7
from there on the ability count quickly decreases and is mainly used for alternatives and structures -> 0,5 to 0,1
But after all: with no real data to backing this up, it is hard to say, if this is a good approximation.
|
Can you explain you're reasoning there? Why do you think Green feels better than its opposite?
|
Yeah YoTca, I agree. It seems to me that ALL of these are good.
|
Another thing I noticed is, that you can lower the keyscore for CGs of orange and purple, by starting the CG 1 on the other finger. So for purple, put Control Group 1 on the middle finger instead of the pinky and for orange put CG 1 on the index instead of the pinky. All other keys change according to the rule to prevent finger repetition.
(This is valid for the weighting I proposed in my last post. And also if only the first 6 ability keys and 8 CGs are taken into account without any weighting, what is basicly your CG KS8 and Ablities KS6 score.)
|
I think that doesn't work because only one configuration perlayout satisfies the rules:
Macro 1 on different than macro 2
Macro 2 on third cg
Army 1 on same as Macro 2
I could be wrong though.
|
On May 16 2014 01:45 Antylamon wrote:
Wow, I have no idea what happened. Nothing is consistent.
For example, this is blue's full calculation for total ability KS:
KS3weight * (key1 + key 2 + key 4) + KS4weight * (key1 + key 2 + key 4 + key 6) + KS5weight * (key1 + key 2 + key 3 + key 4 + key 6) + KS6weight * (key1 + key 2 + key 3 + key 4 + key 5 + key 6)
(KS3weight and KS5weight don't even include KS3 and KS5 respectively)
And this is red's calculation:
KS3weight * (key1 + key 2 + key 6) + KS4weight * (key1 + key 2 + key 6 + key 7) + KS5weight * (key1 + key 2 + key 3 + key 6 + key 7) + KS6weight * (key1 + key 2 + key 3 + key 4 + key 6 + key 7)
(KS7 shouldn't even be included and only KS6weight includes its own KS)
Can someone please fix the spreadsheet in regards to this?
|
@morik
You can find the files on https://github.com/jonnyhweiss/TheCoreConverter
TheCore 1.0 and 2.0 are quite different. It´s hard to give any advice. I think TheCore 2.0 in it´s current state is superior to 1.0. So I would start with 2.0, even if there likely will be changes. 
@Ninjury-J
Yes. I forgot creep spread. I always do. ^^
|
@Ninjury_J:
I applied the rule 12211112 (1 = finger 1; 2 = finger 2) for Control Groups, as this is the distribution that is shown in the TheCore Master Spreadsheet on the sheet ‘R. Key-use Overview’.
With this only green showed a conflict, as there are not enough keys on the middle finger.
For the abilities I used the rule 1211 as can be seen on the ‘Ideal 1’ sheet.
The left over keys were used from top to bottom, to allow the best keyscore with the most used keys.
Or did I misinterpret something?
edt: and by switching the first CG to the other finger the rule became: 21122221 to also prevent finger repetition.
So indest of
CG1 - Finger 1
CG2 - Finger 2
CG3 - Finger 2
CG4 - Finger 1
CG5 - Finger 1
etc.
I got
CG1 - Finger 2
CG2 - Finger 1
CG3 - Finger 1
CG4 - Finger 2
CG5 - Finger 2
etc
|
P is attack, J is main army and I is UA1 on RRM2.0.
If you change RRM2.0 to green field, will J be attack, O be main army and P be UA1??
And do you have plan that you change RRM2.0 to green field??
Or other field??
|
On May 16 2014 07:22 YoTcA wrote:
@Ninjury_J:
I applied the rule 12211112 (1 = finger 1; 2 = finger 2) for Control Groups, as this is the distribution that is shown in the TheCore Master Spreadsheet on the sheet ‘R. Key-use Overview’.
With this only green showed a conflict, as there are not enough keys on the middle finger.
For the abilities I used the rule 1211 as can be seen on the ‘Ideal 1’ sheet.
The left over keys were used from top to bottom, to allow the best keyscore with the most used keys.
Or did I misinterpret something?
edt: and by switching the first CG to the other finger the rule became: 21122221 to also prevent finger repetition.
So indest of
CG1 - Finger 1
CG2 - Finger 2
CG3 - Finger 2
CG4 - Finger 1
CG5 - Finger 1
etc.
I got
CG1 - Finger 2
CG2 - Finger 1
CG3 - Finger 1
CG4 - Finger 2
CG5 - Finger 2
etc
The issue is that the 2nd macro is easily more important that the 2nd army, so it needs to be 3rd. If you switch CG 1 and 2 then to avoid finger repetition the 2nd macro will go down in priority, which is not ideal. It is not only the order of fingers but which specific function is on which CG.
EDIT:
I'm going to stream TheCore development process tomorrow morning at 9am PST until whenever. There will be a raidcall anyone can join and we can all discuss the best way to go about doing this, picking up from wherever we leave off tonight.
twitch.tv/thejakatak
|
Perhaps another way to approach this problem is to brute force it. I believe that certain issues will come to light once you start play testing it. For example, in Core 1.0, as a zerg player I had a problem with shift moving my overlord then holding position then creep spreading. The buttons at the time were very far apart from each other and it was difficult to reach. It was a technique I used very often for denying bases especially now because of cheaper and earlier overlord speed so of course I had to change it. Move shift click patrol also gave me issues. I use that technique to throw down buildings faster because a drone will need to accelerate first before putting down a building. I learned this from Stephano. This means that certain button combinations will be difficult to do even if you built the original Core on which hotkeys have the greatest keyscore.
I understand the problem you guys are facing and it's nice to chase after the super ideal optimally efficient absolutely best Core possible, but sometimes it comes down to personal preference. It's funny that you use colors to identify each Core set up. It would be like asking, "what's your favorite color?". From what I have seen, the colors are not much that different from each other. As zerg, I only use 5 control groups. 1 for creep queen, 1 macro, 3 for units. I do not use the creep camera because I find it's easier to track the camera on a queen and you would have to move your camera over a wide area anyways when you're creeping across a map. Sort of like how terrans rally units to a building.
Personally, I would prefer the hotkey setup with the highest control group rating because that's what you are actually controlling most of the time. Either to look at your units through cameras, to use them for whatever, to make units or research through buildings, practically everything. Often you would see pros simply just cycle through control groups and never actually do anything with them.
This may seem like I'm sidestepping the problem, but I wouldn't even have half of the success as I have in Starcraft 2 if I didn't try things out for myself and put things into practice. In my math class, there was a puzzle gizmo that I couldn't figure out. So I kept trying every day to figure it out by just attempting it. My body learned how to finish the puzzle and I had no clue what I was doing. It wasn't until much later that my mind figured out how the puzzle worked but only after I was able to finish the puzzle in under 30 seconds. When I first started learning the puzzle it took me 20 minutes.
|
I hear what you're saying Bulgogi, but we have to make a decision. And I'm not down for choosing randomly. Originally we chose Purple because it had the lowest overall score, but that wasn't taking distrubution into account and it was saying that the 13th abililty is just as important than the 1st, which it is not. We have the fortune of not selling this product and not having any bosses so we get to take as long as we need to find the solution, even if it makes a small difference, TheCore has been built on a series of small differences, and we'll continue looking for more small differences until we have to start again in Starcraft 3. Dispite this I don't think its reasonable to test out all 6 options. The time it would take would be tremendous and I'm not sure how we would go about collecting meaningful data from these tests.
Concerning combinations. In the past we have looked into specific combinations like the location key inject method and stim attack-move. What we need to make TheCore better is a more complete list of these combinations ranked by their importance. I'll make a tab on TheCore master spreadsheet and put the combos listed here on that tab and we can rank them to the best of our abilites. However I don't see this having a significant impact on choosing the split.
I'm going to look at the numbers with their keyscores individually to compare to what we have now (which is additive) and see if there is a significant difference. May as well stream it:
twitch.tv/thejakatak
come chat if you're interested
|
Hi, I just wanted to show my general support. I don't have anything much to offer you, cause key scores, weighting just screams 'MATTHHHHHSSS' down my ear-hole and my brain shrivels up.
I've been following The Core for ages now - in the end I never used it. I tried it, like 6 times and 6 times flipped my table - 'OMGGGGG THIS IS SO DUMB!!!!!!!' and went with something a lot closer to my setup in League of Legends. What I did was have control groups on 123456xcvb, which works for me, because I use the palm of my hand for ctrl. I mention this because it is a pretty smart setup - if you use your palm for modifiers you get all five fingers free. But I wanted to use The Core because xcv hit with your thumb isn't great for flicking through groups. As you have said somewhere, Jak, the thumb's great for just plonking down somewhere, but it's not great for, say, re-deploying tanks and microing your marines, cause you'd be using the same finger (thumb) I tried a few more setups and then I though, 'fuck it, let's just learn the hotkey scheme these guys have been working on for years.'
My thoughts on the 2.0 I've been using on github. I love the control groups mixed in the home row. I like how, now I've got used to it, it spaces out my fingers in a way that much closer to touch-typing - and that's kind of what you do with Starcraft, cause there's so many (too many) buttons to press. I love the trick with ctrl to add and shift to switch between cams. My problem with it is I feel like Y, H and N for control groups are pretty sucky - rather, it's the same as my xcv setup, except you're using your pinky instead of your thumb.
This probably isn't that helpful, since you're in the process of revising, but ... there it is.
Anyway, looking forward to today's stream. Watching the vod of yesterday's actually was pretty useful for me, I began to sort of understand what's what. I'm gonna spend the day thinking about all this stuff (cause I'm a nerd like that - it took me two years to figure out my LoL setup and I loved and hated every second of it).
I just took a walk to the shop to get some energy drinks to prepare for me day's Thinking and I had a little rant in my head, which I'll relay to you. Starcraft's a fucking terrible game. At least for a lazy fucker like me who probably has a touch of the OCD. I can't stand moving my hand, I can't stand things not feeling 'right', so I fiddle and fiddle.
If only Blizzard would let you bind abilites to key-combos. Like hold position on, say, e, and stop on alt. Or whatever. You CAN techically, I've found, bind a command card ability to, say, ctrl f in the hotkey doc but it breaks whatever you bound to f. And just generally borks in a way I found mind-boggling.
The Ultimate Core would be the Cheaty Core. If you used AHK or your keyboard software, you could cut the keys you need to press for abilities in half. You could have - random, not-perfect example - hold position on T and stop on, say, alt T, or even space T, if you want to get fancy with AHK or Synapse or similar. Or even double-tap T. You could use the same tricks for siege / unsiege, though it would get tricky when you're shift-queuing stuff.
Problem with all this is AHK is a bit tricky and it bugs out if you don't know what you're doing, which I don't.
Anyway, see you tonight (1 am for me, I think) . I'm invested now. After my 7th table-flip moment, I went back to my old method and it just felt cramped and wrong.
|
On May 16 2014 07:23 brook8128 wrote:
P is attack, J is main army and I is UA1 on RRM2.0.
If you change RRM2.0 to green field, will J be attack, O be main army and P be UA1??
And do you have plan that you change RRM2.0 to green field??
Or other field?? We don't know which colour will be preferred for RRM2.0 yet, because they are currently still very close. I can't guarantee what the homerow would look like for green, but I think P would be attack, and J UA1 instead.
***
@brothersadcase
The 'Cheaty' Core was actually a thing a while back. It was called TheCore+. It died off, because Blizzard patched the hotkey system in a way that made sketchy manual hotkey editing reset upon logging in to the game with that particular hotkey file selected. That's how I remember it anyway, I might be wrong. An exception to the resetting, is an indefinite number of alternatives to any command, which is why it is being used in 2.0.
If you have problems getting TheCore under your skin, I'd highly advocate you to use the Arcade map in this video. I find it extremely useful
***
On May 16 2014 07:00 Antylamon wrote:Show nested quote +On May 16 2014 01:45 Antylamon wrote:
Wow, I have no idea what happened. Nothing is consistent.
For example, this is blue's full calculation for total ability KS:
KS3weight * (key1 + key 2 + key 4) + KS4weight * (key1 + key 2 + key 4 + key 6) + KS5weight * (key1 + key 2 + key 3 + key 4 + key 6) + KS6weight * (key1 + key 2 + key 3 + key 4 + key 5 + key 6)
(KS3weight and KS5weight don't even include KS3 and KS5 respectively)
And this is red's calculation:
KS3weight * (key1 + key 2 + key 6) + KS4weight * (key1 + key 2 + key 6 + key 7) + KS5weight * (key1 + key 2 + key 3 + key 6 + key 7) + KS6weight * (key1 + key 2 + key 3 + key 4 + key 6 + key 7)
(KS7 shouldn't even be included and only KS6weight includes its own KS) Can someone please fix the spreadsheet in regards to this? Keys are counted in an unusual way, because we need certain control groups to be on different fingers. E.g. Spellcasters and Army 1. When we want to count the 6 most important CGs, they're counted as they can be placed through the suggested CGs. Same goes for abilities. In my opinion, the sequence for counting is still debatable, like YoTcA have shown with his own spreadsheet.
***
On May 16 2014 07:22 YoTcA wrote:
@Ninjury_J:
I applied the rule 12211112 (1 = finger 1; 2 = finger 2) for Control Groups, as this is the distribution that is shown in the TheCore Master Spreadsheet on the sheet ‘R. Key-use Overview’.
With this only green showed a conflict, as there are not enough keys on the middle finger.
For the abilities I used the rule 1211 as can be seen on the ‘Ideal 1’ sheet.
The left over keys were used from top to bottom, to allow the best keyscore with the most used keys.
Or did I misinterpret something?
Whether the last 1 or 2 (depending on your starting point) is the very last, second, or even third have much less of an impact than the first 2 of that finger are placed accordingly. So for example, in your spreadsheet of the finger distribution with multiple variations, which I'm looking to integrate into the main btw, you can safely swap 7 and 8 for variation 2 of CGs for blue for evaluation purposes.
|
Streaming LIVE Changes to 2.0
twitch.tv/thejakatak
|
Brother, remember keys like Control and Alt can be remapped externally, that is not cheating and that is not identifiable.
@Jak: Where are you getting the info that the pinky is the strongest finger?
Sure it can be trained to have high dexterity, sure it is connected to the muscles and tendons in your forearm, but so are all other fingers. The pinky is the weakest and less 'skilled' finger. Even if you train finger independence, and you make it super fast and strong it will never compare to the index or the middle finger.
I've had this kind of discussions in the past... i'm a guitarist and i like to shred  .
According to my teacher, who happens to know a shitload about the subject, the strongest is actually the third.
It is obvious there are differences between guitar playing technique and typing, but i feel you are making a huge mistake in giving the pinky so much credit, and this is coming from a guy who did trills for the pinky for about one year for at least one hour everyday untill i hurt my ligament a bit. I can now control it as any other finger when fretting, playing the piano, typing, i used to even force different fingering positions to get my pinky into super shape and it works, but it will never be the strongest.
|
Hey guys, a couple quick questions:
- Is there a visual layout guide for 2.0 like the ones in the OP for 1.0?
- How may I go about setting multiple alternate hotkeys? Not sure if this will be needed when I check out 2.0, but in 1.0 at least I am using AHK to rebind ' to i so that I can use rapidfire inject with 2 keys right next to each other (single quote & enter--had to rebind base cam off of shift-; because holding shift makes queens run all over if you have more hatches than queens).
|
On May 17 2014 02:08 Morik wrote:
Hey guys, a couple quick questions:
- Is there a visual layout guide for 2.0 like the ones in the OP for 1.0?
- How may I go about setting multiple alternate hotkeys? Not sure if this will be needed when I check out 2.0, but in 1.0 at least I am using AHK to rebind ' to i so that I can use rapidfire inject with 2 keys right next to each other (single quote & enter--had to rebind base cam off of shift-; because holding shift makes queens run all over if you have more hatches than queens).
To get multiple alternates, edit the hotkey file with a text editor. Find the entry for the command you want - some of them have weird names. TargetChoose is the command to add rapid fires. Separate the alternates with commas. No spaces.
To include punctuation marks you will need to use their names in the naming order, like Apostrophe, Minus, SemiColon. Some are non-intuitive like BracketOpen so scroll around to discover how the file names them.
EX:
TargetChoose=LeftMouseButton,K,P,Period,Equals
|
Its possible to add cgs with shift, but shift is bound to cameras
|
I'll be working on the Zerg layout now. If anyone is interested, please feel free to join.
|
|
|
|
|
|
EPT
