[pran]

barmar, on 2020-March-09, 09:18, said:

Scientists often solve this by writing most numbers using scientific notation. 9.876543210 x 10^4 versus 9.9 x 10^4. The number of significant figures is always decimal places + 1.

Indeed.
But I didn't want to go into too much details and introduce even another notation

[Zelandakh]

pran, on 2020-March-09, 01:29, said:

or 123.45 is a figure which has 5 significant digits and 2 decimal places.

No pran, we are talking here about mathematical terms and the correct term is significant figures, not digits. Have you ever corrected someone to say that Double is a call and not a bid? Same thing.

[Trinidad]

Zelandakh, on 2020-March-10, 04:16, said:

pran, on 2020-March-09, 01:29, said:

or 123.45 is a figure which has 5 significant digits and 2 decimal places.

No pran, we are talking here about mathematical terms and the correct term is significant figures, not digits. Have you ever corrected someone to say that Double is a call and not a bid? Same thing.

Go figure.

Rik

[barmar]

Zelandakh, on 2020-March-10, 04:16, said:

No pran, we are talking here about mathematical terms and the correct term is significant figures, not digits. Have you ever corrected someone to say that Double is a call and not a bid? Same thing.

I never have. There are a small number of Laws where the distinction is important, so it's important when discussing those situations. But in ordinary conversation at the bridge table, I consider them synonymous.

It always grates me when my partner asks "Is it my call?" or "Whose call is it?". Practically everyone else uses "bid" in phrases like this. But of course I won't say anything because he's technically correct.

[pescetom]

barmar, on 2020-March-11, 09:14, said:

It always grates me when my partner asks "Is it my call?" or "Whose call is it?". Practically everyone else uses "bid" in phrases like this.

It would grate me because he should know, irrespective of which term he chooses.
Although I'm grateful (isn't language odd) when he occasionally reminds me that it is (or is not) my turn.

[barmar]

pescetom, on 2020-March-11, 12:24, said:

It would grate me because he should know, irrespective of which term he chooses.

Yes, the fact that he loses track also annoys me, although not as much as when he forgets agreements we've had for nearly 2 decades.

[sanst]

barmar, on 2020-March-12, 08:58, said:

Yes, the fact that he loses track also annoys me, although not as much as when he forgets agreements we've had for nearly 2 decades.

Could be an indication that there’s something badly wrong with him. Dementia is becoming an issue in bridge clubs over here, at least there’re discussions about how to handle it.
Rarely I’ve seen a topic changing as much as this one.

[PeterAlan]

pran, on 2020-March-09, 08:51, said:

Examples:
98765,43210 is a 10-digit number with 5 decimals.
98765,4321 is the same number with only 4 significant decimals (note the omission of the last zero!).
98765 is the same number with 5 significant digits.
and 99000 is again the same number with only 2 significant digits. (note the rounding of the thousands!)

No. Both 98765 and 99000 are different numbers (and differ from each other), being the (different) results of rounding the same number (98765.4321 to use UK / US style) to different amounts of precision.

[pran]

PeterAlan, on 2020-March-13, 07:29, said:

No. Both 98765 and 99000 are different numbers (and differ from each other), being the (different) results of rounding the same number (98765.4321 to use UK / US style) to different amounts of precision.

maybe I should have written "state the same quantity with different precision"?

[barmar]

PeterAlan, on 2020-March-13, 07:29, said:

No. Both 98765 and 99000 are different numbers (and differ from each other), being the (different) results of rounding the same number (98765.4321 to use UK / US style) to different amounts of precision.

"amounts of precision" is essentially the same as "significant digits" -- the latter is a numeric description of the former.

[Trinidad]

pran, on 2020-March-09, 08:51, said:

Indeed.

A significant number to me refers to one (or more) 'significant' number(s) within a set of numbers, not to the internal properties of any particular number.

So if you for instance consider the population (expressed as a number) of the US states you might end up ranking the states according to their significance.
There you have an example of significant numbers.

Examples:
98765,43210 is a 10-digit number with 5 decimals.
98765,4321 is the same number with only 4 significant decimals (note the omission of the last zero!).
98765 is the same number with 5 significant digits.
and 99000 is again the same number with only 2 significant digits. (note the rounding of the thousands!)

And how would you write a number that is between 98999.5 and 99000.5?

99000 has 5 significant figures. Round numbers do occur, from time to time. If you want to write a number with 2 significant figures that lies between 98500 and 99500, you write "99 k" or "99^.10³".

Rik

[pran]

Trinidad, on 2020-March-13, 12:57, said:

And how would you write a number that is between 98999.5 and 99000.5?

99000 has 5 significant figures. Round numbers do occur, from time to time. If you want to write a number with 2 significant figures that lies between 98500 and 99500, you write "99 k" or "99^.10³".

Rik

In a forum where this is relevant I would use the scientific notation 0,99000E+5

(I cannot imagine better than 1% precision being relevant in any bridge relation and then write 0.99E+5)

[Trinidad]

pran, on 2020-March-13, 14:58, said:

I cannot imagine better than 1% precision being relevant in any bridge relation

I will tickle your imagination: Why do you think MP scores for a typical bridge club evening are reported with 4 significant figures?
Your 1% precision will be good for 24 boards at 3 tables (96 top). Bigger events will need better precision.

I can easily imagine that in a big XIMP tournament one pair beats another with 99000 over 98999... (or the equivalent accuracy when the scores are normalized).

Rik