Science

The first few stages of the Tech tree, showing the unlock system

Science, sometimes called science points, is needed to unlock new parts in the tech tree. It is obtained by performing different scientific activities at different locations and then either returning to Kerbin and recovering the craft or sending your scientific analysis home through an antenna.

Science activities

This is an overview of all science activities which can be performed. Most activities are restricted to certain heights. Currently^[outdated] the planet Kerbin and its moon Mun are divided into different biomes. On some heights, some experiments return different results per biome.

Possible activities in different locations

Location	Surface Samples	EVA report	Crew Report	Experiments
				Mystery Goo Containment Unit	SC-9001 Science Jr.	2HOT Thermometer	PresMat Barometer	GRAVMAX Negative Gravioli Detector	Double-C Seismic Accelerometer	Sensor Array Computing Nose Cone
On the ground	Biome	Biome	Biome	Biome	Biome	Biome	Biome (if atmosphere exists)	Biome	Biome (except water biome)	Biome (if atmosphere exists)
Lower atmosphere	—	Biome	Biome	Once	Once	Biome	Once	—	—	Biome
Upper atmosphere	—	Once	Once	Once	Once	Once	Once	—	—	Biome
Near space	—	Biome	Once	Once	Once	Once	—	Biome	—	—
Outer space	—	Once	Once	Once	Once	—	—	Biome	—	—

Surface Samples

Requires	any command pod, one Kerbal
Base value	30
Maximum value	40
Data scale	1
Transmission efficiency	50%

Surface samples can be taken by Kerbonauts when performing an EVA on the ground (or in the oceans) of a celestial body. The results vary by biome.

EVA Reports

Requires	any command pod, one Kerbal
Base value	8
Maximum value	10
Data scale	1
Transmission efficiency	50%

EVA reports are done by Kerbonauts on an EVA. Like crew reports they give different results depending on altitude and biome if performed low enough.

In contrast to crew reports and experiments, a Kerbonaut can memorize a different report for each condition he performed an EVA in during the mission.

Crew Reports

Requires	any command pod
Base value	5
Maximum value	8
Data scale	1
Transmission efficiency	100%

Crew reports can be created in every manned command pod. Each command pod can only store one report at a time (Kerbonauts seem to be unable to memorize more than one at a time), but they can be sent home with an antenna without penalty allowing the crew to create a new one.

Crew reports differ between current celestial body, altitude, and at low altitudes also between different biomes.

Experiments

Experiments require parts like the Mystery Goo Containment Unit and behave similar to crew reports. They can be performed in different flight phases and on different celestial bodies returning different results. The biome, however, does not affect them unless the vessel has landed. Just like crew reports, each experiment part can only store one result at a time, but a result can be sent home with an antenna to allow the experiment to be performed again under the same or under new conditions. There is a penalty for sending it home instead of recovering the performed experiment, but when the craft has a steady supply of electricity to supply its antenna this doesn't matter much because an experiment can be repeated indefinitely resulting in the same science output as would be possible through recovery.

Experiment	Mystery Goo Observation	Materials Study	Temperature Scan	Atmospheric Pressure Scan	Seismic Scan	Gravity Scan	Atmosphere Analysis
Required part (tech tree lvl)	Mystery Goo Containment Unit (2)	SC-9001 Science Jr. (4)	2HOT Thermometer (5)	PresMat Barometer (6)	Double-C Seismic Accelerometer (7)	GRAVMAX Negative Gravioli Detector (8)	Sensor Array Computing Nose Cone (8)
Base value	10	25	8	12	20	20	20
Maximum value	18	35	10	14	23	22	24
Data scale	1	1	1	1	2.5	3	10
Transmission efficiency	40%	20%	60%	60%	60%	60%	90%

Note that the game only considers a ship to be in the atmospheric region if it is on a sub-orbital trajectory. Trying to take upper atmosphere readings while aerobreaking, for example, will return data for "near orbit" instead, even if the ship is within the atmosphere.

Craft recovery

Recovering a craft after it landed on Kerbin also yields science points by itself. Its value depends on how far the craft flew. Recovering a craft after a suborbital flight returns less valuable (but different) results than one which orbited the Mun, for example. The returned value appears to be 5 * (surfaces visited + bodies orbited).

Celestial body multipliers

Celestial body	On the surface	Atmosphere	Altitude	Space	Altitude
Sun	N/A	N/A		11	1 Gm
Moho	9	N/A		8	80 km
Eve	12	7	~20 km	7	400 km
Gilly	9	N/A		8	6 km
Kerbin	0.4	0.7	18 km	1	250 km
Mun	4	N/A		3	60 km
Minmus	5	N/A		4	30 km
Duna	8	7	N/A	7	140 km
Ike	9	N/A		8	50 km
Dres	8	N/A		7	25 km
Jool	N/A	7	~118 km	7	4000 km
Laythe	10	9	~10 km	9	200 km
Vall	10	N/A		9	90 km
Tylo	11	N/A		10	250 km
Bop	9	N/A		8	25 km
Pol	9	N/A		8	~21 km
Eeloo	9	N/A		8	60 km

Each celestial body has different multipliers applied to the collected scientific data, based upon where it is taken in relation to that body. Biomes do not affect these multipliers. The only exception is Kerbin on the surface for the KSC, runway, and launch pad where the multiplier is 0.3 instead of 0.4.

The altitudes given in this table determine the altitude above mean sea level of the boundary between lower and upper atmosphere and between near and outer space. The upper atmosphere extends to the atmospheric height and the outer space extends up to the end of the sphere of influence.

Repeating experiments

Performing the same activity multiple times in the same environment yields additional science points, but with diminishing returns each time the activity is repeated.

Each activity will provide a certain amount of science points, and the subsequent amount will be decreased whenever some science points are already obtained (either through recovery or transmission). This decrease depends on the amount of received points, which means it makes no difference whether a player recovers or transmits data: the maximum amount of science which can be obtained is always a constant.

Science calculations

There are several forms of equations relating the yield to the number of times the experiment has performed depending of the desired result.

Step form

This form calculates the yield of each run (i.e the increase of science). Its advantages are

• being the most straightforward, because the equation that was coded into the save-file is in step form
• easier to calculate the yield of each run than the total form (see below)
• still works when both transmission and recovery are performed

The equation is as follows:

${\begin{aligned}y&=S\cdot M\cdot T\left(1-{\frac {p}{C\cdot M}}\right)=S\cdot T\left(M-{\frac {p}{C}}\right)\\&=S\cdot M\cdot T\left(1-{\frac {S\cdot T}{C}}\right)^{n}\end{aligned}}$

Where:

• $y$ is the yield obtained at that run
• $S$ is the science. Listed as the base value in tables above, it is a constant based on the type of experiment.
• $C$ is the science cap. Listed as the maximum value in tables above, it is a constant based on the type of experiment.
• $T$ is the transmission efficiency, which is 1 for recovering the vehicle on Kerbin.
• $M$ is the science modifier based on the situation (see the celestial body multiplier table).
• $p$ is the sum of previous science gained from this experiment in this situation, and
• $n$ is the number of times the experiment has been performed in the past ($n=0$ for the first run of the experiment).

Total form

This form calculate the total yield for a number of runs. It is best suited for

• calculating the total yield for doing a lot of runs, or
• the amount of runs needed to achieve certain amount of science

, because the calculation is always done only once, in contrary to the step form.

Reduced total form

For experiments that have not been performed before, the total yield equation is as follows:

$p(t)=C\cdot M\cdot \left(1-\left(1-{\frac {S\cdot T}{C}}\right)^{t}\right)$

Where:

• $p(t)$ is the total science obtained from the start (which is identical to the definition above),
• $S$, $C$, $T$ and $M$ is same as above, and
• $t$ is the amount of times the experiment is going to be performed (compare $n$, the amount of times the experiment has been performed).

As can be seen in the total form, there is a asymptotic limit on the science on can get - it approaches $C\cdot M$ as one repeats the experiment, which is the maximum amount of science one can possibly get in a game.

Total form

If the experiment had been performed for some runs before, then the equation is as follows:

${\begin{aligned}P(t)&=p(n+t)-p(n)\\&=C\cdot M\cdot \left(1-{\frac {S\cdot T}{C}}\right)^{n}\cdot \left(1-\left(1-{\frac {S\cdot T}{C}}\right)^{t}\right)\\&=\left(1-{\frac {S\cdot T}{C}}\right)^{n}\cdot p(t)\end{aligned}}$

Where:

• $P(t)$ is the total yield, and every symbol is the same as above.

Note that this equation reduces into the step form when $t=1$, or the reduced total form when $n=0$, since $P(0)=0$.

Ratio form

If one only cares about the percentage $R\%$ of the remaining science that will be extracted, a very simple form can be used, as follows:

$R(t)=\left(1-\left(1-{\frac {S\cdot T}{C}}\right)^{t}\right)\cdot 100\%$

Where:

• $R(t)$ is the percentage of the science one will extract, and
• every symbol is the same as above.

Inverse form

The inverse functions, calculating $t$ for a specific $p(t)$, $P(t)$ or $R(t)$ also exist, by taking log on both sides. Note that step form does not have any inverse, as it should be. The inverse forms are useful for determining the amount of times $t$ a experiment need to be performed to get a target science. For all functions every symbol is the same as in the original function. The logarithmic function $log$ can be performed at any base, provided that both bases are equal (and is a valid base).

Inverse reduced total form	Inverse total form	Inverse ration form
$t={\frac {log\left(1-{\frac {p(t)}{C\cdot M}}\right)}{log\left(1-{\frac {S\cdot T}{C}}\right)}}$	$t={\frac {log\left(\left(1-{\frac {S\cdot T}{C}}\right)^{n}-{\frac {P(t)}{C\cdot M}}\right)}{log\left(1-{\frac {S\cdot T}{C}}\right)}}-n$	$t={\frac {log\left(1-R(t)\right)}{log\left(1-{\frac {S\cdot T}{C}}\right)}}$

Known Bugs

• The Seismic Scan, Gravity Scan and Atmosphere Analysis experiments show more science value in the report window than is retrieved upon recovery or transmission (issue 1578).