# “Schätzen”: How to differentiate between a guess and an estimation?

To translate the phrase `"this value is only an estimation"` the most obvious choice would be `"dieser Wert ist nur eine Schätzung"`.

However, I feel there might be a problem with it. In my context it's about physical measurements taken by a device (voltage, current, etc.), and I have to express, that the visualized values are estimated, not 100% accurate. The point is that the estimation is still quite reliable and accurate. Writing `"dieser Wert ist nur eine Schätzung"` in the user manual feels to me too negative, and closer to a "guess" rather than an "estimation". I would like to express that it's only an estimated value, but not to lead people into thinking that it's just a wild guess.

This question discusses the difference between "schätzen" and "abschätzen", and as a gut feeling I think "Abschätzung" is closer to "estimation" on the guess-estimation spectrum than only "Schätzung".

Another possibility would be `"dieser Wert ist nur eine Estimierung"`, but I found very few uses of "Estimierung", mostly in scientific papers written by non-native speakers, and no mention of it in any dictionary I searched.

Edit, based on comments:

I wanted to avoid a lengthy scientific explanation of the phenomena involved, because the English word "estimate" is perfectly suitable. I just feared that "Schätzung" might lie closer to "guess" than to "estimate".

But if you are really interested, think of something like you can change a setting to some position which is used as an input to a current regulation, but the effective current is not itself directly measured (there are differences between average and effective current, RMS, etc.). So the user sets a setting and we can write in the manual that if it's turned to the maximum, it corresponds to x Ampere. It's not that the value is necessarily inexact, the point is that it is only an estimation, because it depends on how the device is used (AC or DC, filtered or only rectified, etc.). Otherwise "ungefähr x Ampere" or "ca. x Ampere" would be OK, but that's not the most important point. Effective AC current values are not necessarily the same as the current values measured by an AC ammeter, this is why, as the device does not measure true RMS directly, the value is an estimate. However, I didn't want to make the question too localized. I thought the question `where does "Schätzung" lie on the "guess" to "estimate" spectrum, and which German word is closest to the English word "estimate"` would in itself be useful to solve.

• It would be good to know what kind of estimation it is. Maybe some other words are suitable (Kostenvoranschlag, Vorausberechnung, kalkulierter Wert, ...). Maybe an additional hint like "Schätzung mit hoher Genauigkeit" or "Schätzung mit der Möglichkeit geringer Abweichungen" will fit? – hellcode Oct 1 '14 at 7:30
• @hellcode : I updated the question. It's not a prediction about the future. It's about the measurement of very real, physically quantifiable and present values. Some can be measured directly, others can be estimated based on other measurements. It's important to know that it's only an estimation, not because it's calculated, but because we cannot make any official guarantee about its accuracy, as it can be affected by many unknown factors. Still, it's a useful-to-know value, which, under normal conditions, is indeed very accurate. – vsz Oct 1 '14 at 7:33
• I wouldn't go with "Abschätzung". Would "Annäherung" work for you? Perhaps even "Approximation"? – Em1 Oct 1 '14 at 7:38
• Maybe "Messabweichungen sind möglich" or "Messungenauigkeiten sind möglich" will fit? This is not a kind of "Schätzung" but normal measurement inaccuracy. – hellcode Oct 1 '14 at 7:45
• There are several good replies on this post now, does one fit your question? – jawo Oct 2 '14 at 10:02

To me, „Schätzung” as well as “estimate” convey an image of human intervention. Where a pure (uneducated) guess would be „raten“ instead of „schätzen“, the word „Schätzung“ implies some amount of knowledge, so it is an estimation or at least an educated guess. If it is a human who is doing an estimation, then „Schätzung” would be suitable.

If, on the other hand, you have some device which all by itself yields some inexact value, then I'd not call this a „Schätzung“, and I (as a native German) would not trust myself to use “estimate” either. Instead, I'd use terms like these:

• Das ist (nur) ein Näherungswert (this is (only) an approximate value)
• Das ist (nur) eine Näherung (this is (only) an approximation)
• Der Wert beträgt näherungsweise … (the value is approximately …)

As you can see, in these cases I'd use a different word in English as well. I'd not concentrate on the one who does the estimating, but instead on the measured value itself, which is characterized by being close (in der Nähe / in proximity) to the actual value.

ich denke "Messgenauigkeit von (z.B.) 98%" und/oder "Messabweichungen von (z.B.) 2%" beschreiben was du sagen willst.

Beispiel:

Dieses Messgerät besitzt eine Messgenauigkeit von 2/100. Alle Messwerte sind daher nur Annäherungen.

Der Voltmeter misst Spannungen bis 450V mit einer Messgenauigkeit 0,02V, was einer durchschnittlichen Messabweichung von unter 1% entspricht

Edit: Meine Antwort bezog sich alleine auf Abweichungen die durch die Bauart des Messgerätes entstehen. Messfehler die durch falsche Anwendung, Einflussnahme von Störquellen oder falsches Ablesen entstehen sind ganz einfach "Messfehler".

I would guess (hehe) that this can't be translated directly. From your description it is also a bit unclear if the measurements or their depiction are inaccurate. Actually, "accuracy" could be the better term for it because they are actually measured and not estimated, right?

Die gezeigten Werte sind nicht exakt / geglättet / mit Messfehlern behaftet / ...

These could be appropriate descriptions depending on the context.

Please correct me, but I have the feeling that the original sentence should be "an estimate" instead of "an estimation" (talking about the result and not the process).

• While I guess your answer is technically correct, the problem is that it doesn't give back the "feeling" of the English text. We don't want to accentuate that it's not exact. We just want to play safe and say that it's an estimate. I don't want to accentuate the negative, this is the primary reason I didn't like "Schätzung". And you are right, "estimate" would be better than "estimation", I'm not a natvie English speaker either. – vsz Oct 1 '14 at 8:23
• @vsz What's your native language? Maybe it would help if you could say how you would phrase it in that language. – Matthias Oct 1 '14 at 13:20
• Based on this answer I'd go with `dieser Wert kann mit Messfehlern behaftet sein`, I feel by using `kann` you alleviate most of the negative taste. – stryba Oct 1 '14 at 13:33
• I'd like to point out that geglättet implies that the accurate values are actually known somewhere in the underlying system and intentionally made less accurate (e.g. by truncation). I don't think that is what is described by the OP. – O. R. Mapper Oct 2 '14 at 7:36

Updating after thinking about the direct relation to measurement:

When talking about estimating the value that is displayed by an analogue instrument, it is about approximation: The instrument is build to approximate the magnitude of some physical value, and the user approximates the instrument value when "reading" it.

In German, there is the word "Näherungswert" for a quantity that approximates another quantity.
I would translate it with "approximation value", (or maybe "approximated value" or "approximating value"?).

It's directly related to the context of measurement, not to general approximation or estimation.

On the scale from guess to estimate, approximate is on the opposite side of estimate than guess, with a similar distance:

A guess is based on no current information about the measured quantity, a estimate based on some partial information, and an approximation on all available information.

How much variation is expressed by "Näherungswert"? I think it expresses merely that there is variation, and the error value can be specified independently. It could be a direct numerical value, but something like "small", "smaller than ...", "irrelevant compared with..." would work too.

### Notes

• Related to "Näherungswert", "Näherung" could express approximation in this context, but that has only a weak relation to measurements. ("Annäherung" is more referring to spacial distance, "Näherung" can be used for both.)

• Comparing "Schätzung" and "estimate", I thinkg there is no difference in terms of being more similar to "guess". At least in a technical context, the words would mean just the same.

• Instead of "ungefähr x Ampere" or "ca. x Ampere", "etwa x Ampere" may be good.

• Regarding the "Estimierung": it's not used in Germany, but possibly in use in Swizerland or Austria.

Inital answer - independent of the relation to measurement:

I agree that "Dieser Wert ist nur eine Schätzung." feels somewhat negative.
But the negative aspect does not seem to be related to "Schätzung", but to "nur". Without that, it becomes "Dieser Wert ist eine Schätzung." which sounds very neutral and clear to me.

The interesting point is whether "this value is only an estimation" has the same negative aspect as "Dieser Wert ist nur eine Schätzung.". If not, it's a correct translation without "nur".

If the English has the same negative aspect, then simply the German variant with "nur" is the correct translation.

In German I'd go with a compound noun for "this value is only an estimation":

Dies ist nur ein Schätzwert

or, if it was even less accurate (rough estimate)

Dies ist nur ein grober Schätzwert

Note that both, the German Schätzung/Schätzwert, and the English estimation/estimate are used for values that could not be measured but were derived by (mostly statistical) approximation from uncertain data, as it apparently also was the case for the device described in the addendum to the question. By this it is better than a guess (vermuteter Wert) but it still is less valid than an inaccuracy of a measurement.

• That does sound quite in accurate to me and I honestly don't think it fits the semantics described by OP. – Emanuel Oct 2 '14 at 9:57
• @Emanuel: see my comment above "Wouldn't, in English, the term for an inaccurate physical measurement be approximation (Näherung, Annäherung) rather than estimation (Schätzung) too?" - Volker's and MvG's answers may be the best for the given context. – Takkat Oct 2 '14 at 10:22

Einschätzung might come closer to what you wish to convey. Its meaning is closer to valuation or judgement. These denote that thought has gone into a decision about the value. Nevertheless, they leave room for imprecision.

I'd say something like

Dies ist nur ein ungefährer Wert

although "ungefähr" is more literally translated into "approximate".

If there is any technique or process followed to derive at a logical reason which would lead to your answer (probability), it should be considered as Estimate.

If there was no steps/ technique followed but if you come up with your own intuition or based on likeliness then it would be considered as a Guess...

Estimation (Schätzung) usually involves a human component, a device is always performing a measurement (Messung), the measurement can of course have an error, sometimes even larger than a human would estimate.

Dieses Gerät misst den FLTR-Faktor in Ihrem System und stellt ihn auf einer ERLG-Skala an. Der angezeigte Wert kann mit Messfehlern behaftet sein.

Last sentence is stolen from stryba. :)

Of course, if you have a trained neural network expert system, it can make an estimation too.

I didn't fully understand the description of the reason why it is just an estimate. However, the "Messfehler" that many answers suggest is something inherent to ANY measurement pretty much so it's not what we need here.

Here are two phrasings that are used if the given measurement is the result of integrating or averaging over some time:

Der Messwert ist gemittelt.

Der Effektivwert (used for oscillations)

If we're using math to calculate a measurement based on some other measurements, then it would be

der errechnete Messwert

• Ist ein `errechneter Messwert` nicht ein Widerspruch in sich? – Harald Oct 10 '14 at 13:34
• @Harald.. rein sematisch wohl schon. Das gilt auch für "gemittelter Messwert". Wenn das Rechnen von einem Computerchip erledigt wird, dann wirds interessant, denn da wird eigentlich nur Strom hin- und hergeschoben und das ist nichts anderes als in den meisten Messgeräten. – Emanuel Oct 10 '14 at 20:12