# Portal parameters (hack speed, multihack, attack) #410

Closed
opened this Issue Jul 10, 2013 · 18 comments

Projects
None yet
6 participants
Contributor

### sorgo commented Jul 10, 2013

 I would like to add to portal detail data about these parametres: hack speed in seconds between two hacks total number of hacks possible until burnout total max time worth staying at portal (hack speed * nr of hacks) turret & force amp attack intensity (only RARE items are around, haven't seen any T or FA of C or VR rarity) UPDATE: The the HS and MH MODs are orderer by rarity DESC, so the most rare mod has always a full value and the rest 0.5, [1.0, 0.5, 0.5, 0.5]; Proven by field tests, the values we got are all here in "excel". https://docs.google.com/spreadsheet/ccc?key=0AtBPyzAvlFORdC1kemN0YU9MSS1UMmhDbWtBb090ZUE&usp=sharing
Contributor

### sorgo commented Jul 10, 2013

 These are official numbers: MH C 20% R 50% VR 70% HS C +4 R +8 VR +12 A quick field test at my work portal gives me these values: HeatSinks C+C 20+8% (80%*80% = 64%; difference +8%) C+R 20+35% (40%; diff +15%) R+C 50+5% (40%; diff +15%) R+R 50+12.5% (25%; diff +37.5%) MultiHacks C+C 4+2 (instead of 4+4; absolute difference -2) C+R 4+6 (4+8; -2) C+VR 4+10 (4+12; -2) R+C 8+2 (8+4; -2; ) R+R 8+4 (8+8; -4) R+VR 8+8 (8+12; -4)
Contributor

### sorgo commented Jul 10, 2013

 A deploy of second MH / HS diminishes the value of the less rare mod. MultiHacks Hypothesis C+C 4+2 = 6 = 4/2+4 C+R 4+6 = 10 = 4/2+8 C+VR 4+10 = 14 = 4/2+12 R+C 8+2 = 10 = 8+4/2 R+R 8+4 = 12 = 8+8/2 R+VR 8+8 = 16 = 8/2+12 HeatSinks C+C 20+8% = 28% = 1-(1-0.2)(1-0.2/2) C+R 20+35% = 55% = 1-(1-0.2/2)(1-0.5) R+C 50+5% = 55% = 1-(1-0.5)(1-0.2/2) R+R 50+12.5% = 62.5% = 1-(1-0.5)(1-0.5/2) Thanks for testing deploy of 3rd and 4th https://plus.google.com/102566933418982112950/posts/RGhptPerJzy It's 100%, 50%, 50% and 50% effective values of mods of one kind ordered by rarity.

### QVinto commented Jul 10, 2013

 HeatSinks C+VR 20+53% = 73% = 1-(1-0.2/2)*(1-0,7)
Collaborator

### jonatkins commented Jul 10, 2013

 Also, as a reminder to anyone who writes the code for this, the mod data from the servers include the stats. As mod stats can change (it happened for shield mitigation), use these values in calculations rather than assumptions based on rarity. Last time I looked the stats returned from the servers didn't include the reductions from multiple deployments of the same kind of mod - it would be nice if they did!
Contributor

### vita10gy commented Jul 10, 2013

 Yeah, or better yet it would be great if, in addition to that, the portal itself just said its mitigation and what not. Then we'd have to make even fewer guesses. Like there seems to be some debate if shields stack with addition or multiplication. (And then if links add) A max of 95% would imply addition, but then shields don't "fall off" the like everything else does. (Although if my own anecdotes count for anything, they sure do indeed be seeming to hold on to their almost total mitigation forever.)

### ghost commented Jul 11, 2013

 Taking into account, that link amps increase range by 100% for the first one, 50% for second, 25% for third, and 12.5% for the fourth, wouldn't it make sense to assume, these scale factors are effective for other mods, too? I remember having seen on reddit some data from the handshake protocol, showing an array with values [2000, 1500, 1250, 1125], which corresponds to my suggested values, with e.g. 1250 meaning 125% = 100%+25%. This would ultimately mean, the first mod is fully effective, the second one shows only half the effect of the first, the third only half the effect of the second, and finally the fourth once again half the effect of the third mod. Makes sense?
Collaborator

### jonatkins commented Jul 11, 2013

 @NThifrar link amps are actually 25% for both 3rd and 4th - the enlightened agents in Australia who created the continent-covering fields recently confirmed (and relied upon) this. Other reduction factors could be used for 3rd/4th slots for the other mods. Ingress, being beta, makes it a bit of a moving target. Niantic seem to be changing things server-side recently, they don't usually confirm how things are supposed to work, and I'm not sure the client app displays the correct values for the effect of mods during the deployment process in some cases at the moment (my suspicion is the current app calculates things based on the deployment slot order, but the server has recently been changed to apply rarest mods first)
Contributor

### sorgo commented Jul 11, 2013

 UPDATE: I'm sure it's definitely [1, 0.5, 0.5, 0.5] for heat sinks and multihack too Here are some more reports and it just proves my hypothesis from yesterday. https://docs.google.com/spreadsheet/ccc?key=0AtBPyzAvlFORdC1kemN0YU9MSS1UMmhDbWtBb090ZUE&usp=sharing And some more screens: https://plus.google.com/u/0/photos/107917688018975026970/albums/5899254812123440449 Thanks everybody for testing.
Contributor

### sorgo commented Jul 11, 2013

 I messed a bit with the code, but it's far from clean. If someone has time to work on it and make a correct commit, just say. I can send a working build.user.js file.

### ghost commented Jul 11, 2013

 The array [2000, 1500, 1250, 1125] was found on the verge of the release of version 1.28. Presumably those values are outdated. Could it be, there are some new scaling values like [2000, 1500, 1500, 1500] in the data stream? I must admit, I have not yet delved into the depths of capturing the data stream myself. I'm just interpreting data found by others and drawing conclusions from what I see and what I've seen in similar data way back in my time of active reverse engineering.

Closed

### QVinto commented Jul 11, 2013

 screens from testing RFA, RT and RLA https://plus.google.com/u/0/107917688018975026970/posts/eUVuYYEkeeq
Contributor

### sorgo commented Jul 11, 2013

 According to QVinto's screenshots: Link amps [1, 0.5, 0.25, 0.25] Force amps 2x [1, 0.25, 0.125, 0.125] Turret attack bonus 2x [1, 0.25, 0.125, 0.125] Turret hit bonus 0.2% [1, 1, 1, 1]
Collaborator

### jonatkins commented Sep 19, 2013

 link amps were added to the portal range calculation a while ago portal hack counts and shield/link mitigation details have just been added, in 54997ac
Collaborator

### jonatkins commented Sep 19, 2013

 if anyone can point me to details on portal attack details, including turret/force amp mods, and suggest what details would be useful to display in IITC for these...
Contributor

### nexushoratio commented Jan 19, 2015

 I think everything in this issue has been implemented.
Collaborator

### jonatkins commented Jan 19, 2015

 Force amps/turrets could have the totals displayed as well, so will leave open
Contributor

### nexushoratio commented Jan 19, 2015

 Ahh OK. You get the values when hovering. I wasn't sure there would be enough room to put them in that block. Pretty crowded as is.
Collaborator

### fkloft commented Jan 19, 2015

 Values for individual mods are included in their tool tip, the values for force amplifier, hit bonus and attack frequency are now included in the general info block below as well.