The 20 most recognizable constellations?

[Ags]

I was wondering which constellations are the brightest and most recognizable. Ursa Major, Orion, Cassiopeia, Crux, Scorpius are obvious picks but I made a little algorithm this afternoon to make the list more science-y. I score all constellations by their stars that are mag 2.5 or brighter, weighting the list linearly with the brightest star scoring 5 times more than the faintest (mag 2.5) star. I think the list is about right. Lots of Southern Hemisphere constellations hogging the top of the list 🥲

Rank:  Const. : Score
1    :    Ori    :    2990
2    :    Sco    :    2816
3    :    Cen    :    2474
4    :    UMa    :    2435
5    :    CMa    :    2108
6    :    Car    :    1707
7    :    Vel    :    1612
8    :    Cru    :    1322
9    :    Gem    :    1296
10    :    Leo    :    1249
11    :    Cyg    :    1202
12    :    And    :    1197
13    :    Cas    :    1148
14    :    Peg    :    1123
15    :    Tau    :    879
16    :    Aur    :    873
17    :    Boo    :    842
18    :    Sgr    :    829
19    :    Gru    :    827
20    :    Per    :    824

Edited June 19, 2022 by Ags

[JeremyS]

Interesting. Would never have thought of Grus, Agnes

Can you do the same for us borealites at ca 52degN? 

[Franklin]

What about constellations that actually look at bit like what they're supposed to be? Leo definitely looks like a lion and Crux is obviously a cross but some of them leave a little too much to the imagination. Take Canis Minor for example, it's just two stars? How does a line look like a small dog?

Looking at your list I would put Taurus the bull ahead of Pegasus and Andromeda, I can definitely see the horns of a charging bull when looking at Taurus.

Some of the more modern additions are quite vague really. I think the older, established patterns that the ancients recognised and named are the most prominent.

[Gfamily]

I'd say there are two measures in your query - "brightest" and "most recognisable", and although there is a certain amount of overlap, I think that your list is fair as regards brightness, but not so good as regards recognisable. 

My guide here is the "if you were out under broken clouds, and saw constellation 'X' through a gap, would you know what it was?"

Here, there are some smaller constellations that would be higher up the list, doing better than their brightness would suggest - I'm thinking particularly of Lyr and Del as good examples, and I'd say  that CrB is another. 

As for the constellations that are on your list, I would put Cas and Sgr far higher than 13 and 18 respectively. 

 

Edited June 19, 2022 by Gfamily

[Ags]

@JeremyS here is the top "20" for the northern hemisphere, we only get to 19 constellations with mag. 2.5 stars:

1    :    Ori    :    6210
2    :    UMa    :    5195
3    :    Gem    :    2676
4    :    Leo    :    2629
5    :    Cyg    :    2582
6    :    And    :    2577
7    :    Cas    :    2528
8    :    Peg    :    2503
9    :    Tau    :    1799
10    :    Aur    :    1793
11    :    Boo    :    1762
12    :    Per    :    1744
13    :    UMi    :    1734
14    :    Lyr    :    915
15    :    CMi    :    912
16    :    Aql    :    908
17    :    Ari    :    871
18    :    CrB    :    854
19    :    Dra    :    851

Here is the southern division (also only 19 long):

1    :    Sco    :    6036
2    :    Cen    :    5234
3    :    CMa    :    4408
4    :    Car    :    3547
5    :    Vel    :    3452
6    :    Cru    :    2702
7    :    Sgr    :    1749
8    :    Gru    :    1747
9    :    Oph    :    1698
10    :    Eri    :    911
11    :    Vir    :    904
12    :    PsA    :    902
13    :    TrA    :    878
14    :    Pav    :    876
15    :    Hya    :    873
16    :    Cet    :    869
17    :    Pup    :    849
18    :    Lup    :    845
19    :    Phe    :    839

@Franklin I can't score the aesthetics, but I agree Taurus looks Bull-like.

This algorithm favors larger constellations that simply have more stars that can score. Lyra would score highly in my personal estimation, but it is to small to compete. Same with Canes Venatici, very easy to recognize, but its just two stars.

[Ags]

@Gfamily I agree with you to a point, but for me if a constellation does not have a bright star it is just a blank area of light-polluted sky. Lyra pops out instantly for me, but Delphinus is only recognizable in my binoculars.

[Zermelo]

An interesting exercise.

There are other factors I would include, but are harder to score, such as:

- the degree to which the brighter stars of a constellation are in an area of sky without much "competition", and so are less easily mistaken (I believe that some mount control software takes this into account when offering lists of alignment stars that need to be identified by the observer)

- the background sky brightness: many of us have been to dark sky sites and had trouble identifying otherwise familiar constellations because of the appearance of so many fainter stars. Does a change to the sky brightness affect all constellations equally? If not, then the recognizability order might change accordingly.

[Spile]

But are we talking about constellations or the familiar asterisms that make up the bits that stand out? The saucepan for example I know but the rest of Ursa Major ? I’m not so sure.

[Ags]

Well, every constellation is going to have faint bits that are a challenge to spot (UMa is indeed a good example). With Bortle 9 skies, seeing any of the stars is a win!

When I moved to Amsterdam from semi-rural Wales in 2010 I was shocked by the light pollution, it is simply never truly dark here. But... it's even worse now. I could make out the complete "virus" shape of Bootes in 2011, but now I can pick out just a few of the stars. Even the Big Dipper is a strain to see clearly. So the faint bits are not on my menu

[badhex]

A question: what's the specific use case for which you are designing the list? I would say that the weighting and therefore output you need would depend on that use case. If it's for humans to use, say to find stuff, it might be hard to factor in the human element - in that we recognise some asterisms more readily either because our brains a hard wired to recognise certain patterns more readily than others, or that we've been seeing/recognising certain bits of those asterism for a very long (pointed out to us as children say), or bits of are just quite bright and recognisable. As mentioned above sky brightness may also rearrange the list probably due to main asterism stars disappearing under bright skies, or conversely getting lost in a sea of other bright stars under dark skies. 

So for example, if the use case of the algorithm is to output a list of the most 20 recognisable for a new book project of some sort e.g. helping people find their way around or star hop, Lyra would in my mind be pretty essential because it's very readily recognisable even thought it is tiny and only has one really bright star. Ursa Major is also obviously very easily recognisable just due to the Plough part, even if the rest of the stars aren't so easily known. Often Aquila appears to me as just two stars until I squint a bit. Delphinus doesn't even get a look in! 

You could add maybe two more elements to the algorithm to account for this. One would be a modifier that represents the sky brightness, meaning that lots of medium brightness stars in a given constellation score less well than one or two bright ones (e.g. Delphinus vs Lyra). It sounds like you have something already like this but the threshold is fixed rather than variable. 

The second is much less sciencey and maybe less important - a score that you would have to manually assign to each constellation in your source data, and would represent a "human" factor. This further weights constellations that are incredibly well known, or that contain shapes we as humans recognise. Cygnus as a cross, Cassiopeia as a W (although this would only apply to western alphabet users, I guess), Orion and Hercules as (sort of) humans etc. Obviously this score is very subjective so more difficult to come up with a score for, and may require some sort of survey of humans to gather the data.

If this is just for fun btw, feel free to ignore everything I have just said 😂

Edited June 20, 2022 by badhex

[Ags]

@badhex yes, for a book project. The list is to help overcome my northern hemisphere cultural bias, it's probably not the final word on the "best" constellations 😀

[jjohnson3803]

Given that they're part of the Summer Triangle, I'd definitely include Lyra and Aquila in there.

Edited June 26, 2022 by jjohnson3803

[Voyager 3]

I too agree that recognisability doesn't just depend on the no of bright stars. 

If I show the stars of Orion to a person jogging in the street, he might come close to saying a man holding something in the hand. One really can't think of any other thing. Crux is literally a cross. Now that's what I'd call recongisable. 

But the man who created the constellations Vulpecula and Canis Minor must have had god-like imagination for sure.

Edited June 26, 2022 by Voyager 3

[Ags]

On 26/06/2022 at 16:22, Voyager 3 said:

If I show the stars of Orion to a person jogging in the street

...they would jog a little faster 🤣

[Voyager 3]

6 hours ago, Ags said:

...they would jog a little faster 🤣

Good one 🤣