Doing the math on science transmissions

[Jarin]

Edit: Take with a grain of salt. I may be misunderstanding the mechanics. See this post below.

So, I decided to do some science on SCIENCE, and find out what's worthwhile to transmit and what's not. Now, here's my setup; I have a vehicle in free solar orbit with plenty of solar panels, so data transmission is effectively free and limitless. This is our "research probe" setup. Granted, mine is manned, but Hermore Kerman volunteered for a one-way mission permanent deployment, so it amounts to the same thing.

Here's the data, transmitted up to the "break even" point where the transmissions equal the value of a single return.

(sorry if the formatting screws up for anyone, it's being a bit inconsistent.)

		Materials Bay			Mystery Goo			EVA Report
Total		1444	|	288.7		313.9	|	125.5		202.9	|	95.7 (86.2 @3)

1		275	|	55		110	|	44		88	|	44
2		235.7	|	47.1		85.6	|	34.2		52.8	|	26.4
3		202	|	40.4		66.5	|	26.6		31.6	|	15.8
4		173.2	|	34.6		51.8	|	20.7		19	|	9.5
5		148.4	|	29.7						
6		127.2	|	25.4		
7		109.1	|	21.8		
8		93.5	|	18.7		
9		80.1	|	16		

Now, the left-column totals are a bit misleading, because you'd have to do a full round trip for each and every one of those values, which could get exceedingly time-consuming. On the other hand, by running repeated transmissions, you are effectively "drilling the science-reservoir empty" for that particular area, so it could decidedly be seen as wasteful, so the return totals are a comparison for the potential that's lost.

Crew reports, of course, are at 100% return value, so you can and should run transmissions until you've mined every area empty for those.

Conclusions:

With the materials bay, you should *really* consider multiple return missions if you can manage it, as just two returns is far more than you will *ever* get with transmissions. Run-dry transmissions for the materials bay ONLY for locations you don't feel like you can reasonably get return missions from.

Mystery Goo is a bit better. Its 40% return rate means that it breaks even at only 4 transmissions, and not nearly as much is lost. A good candidate for any far-flung science probes.

EVA Reports are even better, very nearly breaking even at 3 transmissions. Send an outbound-flight flyby of a bunch of planets and transmit back your sacrificial kerbalnaut's observation's to your heart's delight.

Edited November 21, 2013 by Jarin

[Cmmndrme]

There is also value associated with the achievements of the capsule too which this doesn't account for, however I'm not sure on how significant they are

[allmhuran]

So.... with the materials bay and goo container, can you only go out and do one experiment in one place before the part is "full" and needs to be returned? Or can you carry one of each to multiple places, run all the experiments, return to kerbin and get all the results? (Would try myself but not at home)

[FITorion]

So.... with the materials bay and goo container, can you only go out and do one experiment in one place before the part is "full" and needs to be returned? Or can you carry one of each to multiple places, run all the experiments, return to kerbin and get all the results? (Would try myself but not at home)

The mat bay and the goo get used and can't be used again if you intend on returning them. You would need to take with you a seperate unit for each place you wanted to sample and return data from.

[allmhuran]

Thanks for that FITorion. Now I can spend my day at work trying to imagine how many such parts I can send to how many places in career mission 2!

[yongedevil]

I think the values for the totals are misleading because it seams they were obtained by added the total worth of each subsequent experiment when only transmuting the data back. However, actually recovering experiments lowers the value of future ones faster (Best guess is there's a set amount of science available at each location that your points are subtracted from), so if you actually go and recover the experiments from solar orbit you get the following values:

experiment       Value on retrieval
number	      Material bay     Mystery Goo
1               275	         110
2               78.6             48.9
3               22.4             21.7
4               6.4              9.7
5               1.8              4.3
6               0.5              1.9

total           384.7            196.5

EDIT: For completeness I went and did a complete test of both transmitting and recovering data for the SC-9001 Material Bay and Mystery Goo from solar orbit until achieving a return of 0.0 science. Long story short, the only penalty to transmission is the electric charge needed and the time spent clicking on the experiments.

                        Science Points Recived
test number          SC-9001               Mystery Goo
                recover   transmit      recover   transmit
totals          384.8     384.6         198       197.9

1               275       55            110       44
2               78.6      47.1          48.9      34.2
3               22.4      40.4          21.7      26.6
4               6.4       34.6          9.7       20.7
5               1.8       29.7          4.3       16.1
6               0.5       25.4          1.9       12.5
7               0.1	  21.8          0.8       9.7
8               0	  18.7          0.4       7.6
9                         16            0.2       5.9
10                        13.7          0.1       4.6
11                        11.8          0         3.6
12                        10.1                    2.8
13                        8.6                     2.2
14                        7.4                     1.7
15                        6.4                     1.3
16                        5.4                     1
17                        4.7                     0.8
18                        4                       0.6
19                        3.4                     0.5
20                        2.9                     0.4
21                        2.5                     0.3
22                        2.2                     0.2
23                        1.9                     0.2
24                        1.6                     0.1
25                        1.4                     0.1
26                        1.2                     0.1
27                        1                       0.1
28                        0.9                     0
29                        0.7
30                        0.6
31                        0.5
32                        0.5
33                        0.4
34                        0.3
35                        0.3
36                        0.2
37                        0.2
38                        0.2
39                        0.2
40                        0.1
41                        0.1
42                        0.1
43                        0.1
44                        0.1
45                        0.1
46                        0.1
47                        0

Edited October 18, 2013 by yongedevil
added data

[Jarin]

Oh, so repeated transmissions are a lot better than I thought, then. Nice.

[lammatt]

there's nothing oher than the surface samples are worth returning indeed, transmissions are bascally free after u get your first solar panel, the diminishing return and the transmission loss are underpowered.

[SkyRender]

The diminishing returns on transmissions CAN be a problem for some things, however. For example, unless you plan to circularize, most fly-by missions of the outer moons in the Jool system have you in the "low orbit" region for less than a minute and the "high orbit" region for perhaps 20 minutes. Nowhere near long enough to gather and transmit data 47 times. It actually does balance out nicely in that regard.

[Apophis775]

This actually makes me think of something I had been debating doing originally. What I wanted to do, was create a "superprobe" that performed the experiments, then detached a "return" module after doing one of each, and the module returned to Kerbin. Maybe, something like that would be effective? If your going to, say, Jool, maybe launch a few "lab modules" containing Goo cannisters and Mat Bays, then as they are used up, set them to return to Kerbin?

Maybe a bit more work to get setup, but less return trips, especially for long-range missions.

[AlternNocturn]

So basically transmitting gets you almost exactly the same amount of science as a return mission, which is just like real life. In that case I've got a probe sitting on Minmus that hasn't been fully used up.

[Pbhead]

well. i dont like that. bringing back rocks from space should be worth a whole lot more than not bring back rocks from space.

[Jarin]

EDIT: For completeness I went and did a complete test of both transmitting and recovering data for the SC-9001 Material Bay and Mystery Goo from solar orbit until achieving a return of 0.0 science. Long story short, the only penalty to transmission is the electric charge needed and the time spent clicking on the experiments.

This is everything I wanted to see. I can have a long-flight probe around the system, and only worry about the sample returns for manned missions. And that only because it would take forever to transmit and re-collect that many samples.

Edit: Honestly though, it does feel a bit off. This might fall into the "consider rebalancing" category?

[SkyRender]

well. i dont like that. bringing back rocks from space should be worth a whole lot more than not bring back rocks from space.

If we ever get tiers of quality to science experiment equipment, this probably will get fixed. The idea being that tier-1 versions of equipment will have an upper limit to the quality of transmitted data no matter how many times you do the experiment with them, but will still return the same level of quality results as a top-tier model if returned to Kerbin instead.

[magnemoe]

there's nothing oher than the surface samples are worth returning indeed, transmissions are bascally free after u get your first solar panel, the diminishing return and the transmission loss are underpowered.

You can also transmit surface samples, say you do three samples you transmit and bring the fourth back it should work out as if you did two missions.

[boolybooly]

I just take 3 materials bays along for a return mission with a couple of goo and seem to get big numbers. Personally I like doing a separate mission for each location. I am thinking it would be amusing if there was a way of enhancing science return by specifying the destination for the mission in the VAB, so that if you hit the target you get a boost on data recovery.

[OrtwinS]

-snip-

EDIT: For completeness I went and did a complete test of both transmitting and recovering data for the SC-9001 Material Bay and Mystery Goo from solar orbit until achieving a return of 0.0 science. Long story short, the only penalty to transmission is the electric charge needed and the time spent clicking on the experiments.

Science Points Recived
test number SC-9001 Mystery Goo
recover transmit recover transmit
totals 384.8 384.6 198 197.9

1 275 55 110 44
2 78.6 47.1 48.9 34.2
3 22.4 40.4 21.7 26.6
4 6.4 34.6 9.7 20.7
5 1.8 29.7 4.3 16.1
6 0.5 25.4 1.9 12.5
7 0.1 21.8 0.8 9.7
8 0 18.7 0.4 7.6
9 16 0.2 5.9
10 13.7 0.1 4.6
11 11.8 0 3.6
12 10.1 2.8
13 8.6 2.2
14 7.4 1.7
15 6.4 1.3
16 5.4 1
17 4.7 0.8
18 4 0.6
19 3.4 0.5
20 2.9 0.4
21 2.5 0.3
22 2.2 0.2
23 1.9 0.2
24 1.6 0.1
25 1.4 0.1
26 1.2 0.1
27 1 0.1
28 0.9 0
29 0.7
30 0.6
31 0.5
32 0.5
33 0.4
34 0.3
35 0.3
36 0.2
37 0.2
38 0.2
39 0.2
40 0.1
41 0.1
42 0.1
43 0.1
44 0.1
45 0.1
46 0.1
47 0

Thank you very much! I was thinking about doing exactly this today, but you saved me the efford.

It seems science is currently a 'pot of points' and no matter what way you drain it, you get it all.

I was already planning a Multi-lander mega-mission with full return capability to drain the Mun of every science point availeble. I already designed the thing in my head, backbone with 7 (for every Mun-biome) landers, each with 6 bays and pods, all that docking together again to recover in one big swoop .

However the current system actually promotes sending a small one-way mission and simply transmitting everything since you get the same science in the end.

Now a more pressing issue: does this also count for soil samples and EVA's...?

I haven't touched down on Minmus yet, I think I"ll use it to test this.

Edited October 18, 2013 by OrtwinS
cleaning up post, adding some details.

[magnemoe]

Yes, however science should not be lost, that is you lose points permanent then doing it the "wrong" way.

Outside of other reasons this tend to promote stupid meta gaming methods. However I have no issues to having an cap on how much science you can get from transmitting home sample returns.

Say Minmus has 500 points for sample retuns, you can only get 300 by transmitting so you need to do 2-3 returns too, if you do the returns first you don't get much more science from on site analyze.

[JoCRaM]

Say Minmus has 500 points for sample retuns, you can only get 300 by transmitting so you need to do 2-3 returns too, if you do the returns first you don't get much more science from on site analyze.

Except an "engineering bonus" that makes future on site analysis more effective (please?)

[Tommygun]

There also seems to be a lot more science points out there than you will ever need, at least right now.

I have gone through most of the tech tree and I have only landed on the Mun and Ike, and sent orbital craft to Minmus, Duna and the Sun.

[KerbMav]

There also seems to be a lot more science points out there than you will ever need, at least right now.

I have gone through most of the tech tree and I have only landed on the Mun and Ike, and sent orbital craft to Minmus, Duna and the Sun.

As I wrote in another thread:

If the nearer space already offers enough science to unlock the whole tree, than less experienced/adapt players can simply fly around Kerbin and its two moons and land on them in different locations to do so, and more experienced/adapt players can take the new parts and fly new missions to other planets to do the same, as there is no need to "farm dry" everything.

Yes, however science should not be lost, that is you lose points permanent then doing it the "wrong" way.

Outside of other reasons this tend to promote stupid meta gaming methods. However I have no issues to having an cap on how much science you can get from transmitting home sample returns.

Say Minmus has 500 points for sample retuns, you can only get 300 by transmitting so you need to do 2-3 returns too, if you do the returns first you don't get much more science from on site analyze.

I would even make it a bit easier - as Squad doesnt seem to want it to be that difficult to gain science at a steady pace - make it possible to transmit 90% of the pool and return ALL remaining 10% in a final return mission.

[boolybooly]

Yes when you factor in the money economy acting as a hurdle in combination with science it is going to be quite interesting, much trickier, (especially if they add range dependant recovery costs and recovered vehicle salvage value etc).

When you are having to scrape together a few bucks to buy one sensor it will play very differently.

Makes me wonder just what methods we can use to earn money. As games go it has the potential to be pure poetry, thanks Harv & Squad, its going to be fun.

[Aragosnat]

One way would be the rocks we collect could be worth money. Even if they end up being worth not science points later on as materials to build with will always be in demand. Expecially with the way some of us build our ships.

[RSwordsman]

Mining various resources for money would be freaking sweet. I can see mineral/Kethane? collection being there in addition to government grants for scientific discovery.

That said I'd be relieved to see that no science potential can be lost forever if you do things the wrong way. Because I'm so far terrible at the game AND playing with Deadly Reentry, so trying to bring back a capsule and science module on a ballistic trajectory home has not been very fruitful.

[MalfunctionM1Ke]

This Thread changes all my Craft designs

Someone should make it sticky