EPTIs the time listed real world seconds or in-game seconds (on the 'Faster' setting)?
SCFusion - WoL, HotS & LotV Build Order Optimizer - Page 3
| Forum Index > SC2 General |
|
Lunchtime
183 Posts
Is the time listed real world seconds or in-game seconds (on the 'Faster' setting)? | ||
|
DaBears57
United States300 Posts
+ Show Spoiler + 0.00: 50M 0G 0E 6/ 10S - Build Probe 17.00: 73M 0G 10E 7/ 10S - Build Probe 34.00: 107M 0G 19E 8/ 10S - Build Probe 44.44: 116M 0G 25E 9/ 10S - Chrono Nexus 48.81: 140M 0G 2E 9/ 10S - Build Probe 48.81: 90M 0G 2E 10/ 10S - Build Assimilator 63.15: 100M 0G 11E 10/ 10S - Build Pylon 78.81: 99M 0G 19E 10/ 10S - Move Probe To Gas 88.15: 159M 7G 25E 10/ 18S - Build Gateway 95.52: 50M 12G 29E 10/ 18S - Build Probe 95.52: 0M 12G 29E 11/ 18S - Chrono Nexus 95.52: 0M 12G 4E 11/ 18S - Move Probe To Gas 106.85: 56M 28G 10E 11/ 18S - Build Probe 119.52: 87M 45G 17E 12/ 18S - Build Probe 124.93: 75M 53G 20E 13/ 18S - Build Assimilator 136.52: 74M 69G 27E 13/ 18S - Build Probe 136.52: 24M 69G 27E 14/ 18S - Move Probe To Gas 154.25: 150M 103G 37E 14/ 18S - Build Cybernetics Core 154.25: 0M 103G 37E 14/ 18S - Move Probe To Gas 154.25: 0M 103G 37E 14/ 18S - Move Probe To Gas 154.25: 0M 103G 37E 14/ 18S - Move Probe To Gas 164.35: 50M 138G 42E 14/ 18S - Build Probe 164.35: 0M 138G 42E 15/ 18S - Chrono Nexus 175.68: 56M 178G 24E 15/ 18S - Build Probe 204.25: 199M 281G 40E 16/ 18S - Build Twilight Council 204.34: 50M 181G 40E 16/ 18S - Research Warpgate 204.34: 0M 131G 40E 16/ 18S - Move Probe To Gas 213.27: 50M 165G 45E 16/ 18S - Build Probe 213.27: 0M 165G 45E 17/ 18S - Chrono Nexus 222.22: 50M 199G 25E 17/ 18S - Chrono Cybernetics Core 254.25: 274M 321G 18E 17/ 18S - Build Dark Shrine 266.67: 203M 118G 25E 17/ 18S - Chrono Cybernetics Core 266.67: 203M 118G 0E 17/ 18S - Build Gateway 281.97: 150M 176G 9E 17/ 18S - Build Gateway 297.72: 100M 236G 17E 17/ 18S - Build Pylon 305.60: 50M 266G 22E 17/ 18S - Build Probe 322.72: 109M 331G 32E 18/ 26S - Build Probe 324.34: 72M 337G 32E 19/ 26S - Convert Gateway To Warpgate 334.34: 150M 375G 38E 19/ 26S - Convert Gateway To Warpgate 339.72: 192M 396G 41E 19/ 26S - Build Probe 346.97: 205M 423G 45E 20/ 26S - Convert Gateway To Warpgate 346.97: 205M 423G 45E 20/ 26S - Chrono Gateway 354.25: 267M 451G 24E 20/ 26S - Build Dark Templar 354.25: 217M 301G 24E 22/ 26S - Build Dark Templar 354.25: 167M 151G 24E 24/ 26S - Build Dark Templar 359.25: 162M 20G 27E 26/ 26S - Satisfied. | ||
|
Lunchtime
183 Posts
| ||
|
CarbonTwelve
Australia525 Posts
On November 12 2010 13:23 Lunchtime wrote: Is the time listed real world seconds or in-game seconds (on the 'Faster' setting)? It's in-game seconds (ie, should match the timer at bottom left). | ||
|
Chronopolis
Canada1484 Posts
| ||
|
DaBears57
United States300 Posts
- A quota system like others have previously mentioned where you can specify that you want 1 stalker and 1 zealot by 300 seconds while simultaneously achieving the other defined goal. - A total tally of units/structures/upgrades produced to allow for easy comparison between builds. For example, a table or something that lists how many probes were produced, how many gateways, if warp gate was acquired etc. - Incorporation of upgrades, like storm and forge. It would be nice to plan some fast +1 weapons upgrade for PvZ. - Incorporation of buildings/upgrades wanted. For example, it would be nice to use this program to find the absolute fasted time for a Korean 4 warp gate rush but in this current version if you want 4 zealots, the program will only use 1 or 2 gateways and not use warp gate tech. All in all, great job so far, I can't wait for future versions! | ||
|
Lunchtime
183 Posts
So there's likely going to be no Protoss equivalent of a 7RR being discovered with this thing. There's really nothing more efficient than the 4gate for gateway units. The program says ~6:00 for 3 DTs, add 20-25 seconds to that for practical results and you're not far off the 3 DT rush on Liquipedia which says 6:38, plus that build doesn't cut probes. | ||
|
CarbonTwelve
Australia525 Posts
On November 12 2010 15:04 Lunchtime wrote: There's at least a 20 second difference from the time the program gives you and the time you can actually achieve these things in-game. I guess it's because it takes time to get the probes to the minerals at the beginning, time for a probe to move to a spot (where it's not mining), move back, time for the warp-in animation to complete, etc, which all adds up. My program already handles the last two of those. It calculates a travel time of 10s for most buildings, and it allows 5s warp-in time for warpgates. But you're right, the mining rates probably are more of an 'optimal' calculation than a practical one. I should be able to adjust for that though (ie, add a 2s delay before the start of mining, and possibly adjust the income rate calculation so it's a bit more 'real world'). So there's likely going to be no Protoss equivalent of a 7RR being discovered with this thing. There's really nothing more efficient than the 4gate for gateway units. The program says ~6:00 for 3 DTs, add 20-25 seconds to that for practical results and you're not far off the 3 DT rush on Liquipedia which says 6:38, plus that build doesn't cut probes. It's not necessarily what you use to build things, more about the order in which you build them, and when cutting probes will give you a useful edge (you could compare this by asking the algorithm to get 22 probes as well as the units you want). I'm not expecting anything ground breaking to be found, but it might help people with sorting out what is the best way to implement a timing attack for instance. One feature I plan to add is for it to value income rate in the fitness function (and you could choose whether you want to include minerals or gas or both), which I think will be more useful than just saying to 'maximise probes' or something. | ||
|
Lunchtime
183 Posts
| ||
|
Gudeldar
United States1200 Posts
On November 12 2010 15:04 Lunchtime wrote: There's at least a 20 second difference from the time the program gives you and the time you can actually achieve these things in-game. I guess it's because it takes time to get the probes to the minerals at the beginning, time for a probe to move to a spot (where it's not mining), move back, time for the warp-in animation to complete, etc, which all adds up. So there's likely going to be no Protoss equivalent of a 7RR being discovered with this thing. There's really nothing more efficient than the 4gate for gateway units. The program says ~6:00 for 3 DTs, add 20-25 seconds to that for practical results and you're not far off the 3 DT rush on Liquipedia which says 6:38, plus that build doesn't cut probes. I wouldn't be so sure, I put 7 stalkers in and got a build that is only 34 seconds slower than 7RR. | ||
|
Evolve
Canada63 Posts
| ||
|
icezar
Germany240 Posts
| ||
|
divadretrac
United Kingdom36 Posts
| ||
|
ArcticVanguard
United States450 Posts
| ||
|
GWash
United States153 Posts
WARPGATES | ||
|
Lunchtime
183 Posts
On November 12 2010 16:58 Gudeldar wrote: Have you tried the build in-game?I wouldn't be so sure, I put 7 stalkers in and got a build that is only 34 seconds slower than 7RR. | ||
|
mechler
United States26 Posts
| ||
|
IamSoCrazy
France25 Posts
 | ||
|
Nolari
Netherlands51 Posts
A sorted linked list gives O(1)-time removal of the first event, but O(n)-time insertion of a new event and updating the time of an event. A heap would give O(log n) time for all three operations. Have you found that removals are so much more frequent and n so small that the linked list is superior? | ||
|
speedphlux
Bulgaria962 Posts
Thank you very much sir ! | ||
| ||
