# BBO Discussion Forums: If zeros are in red, tops should be in green. - BBO Discussion Forums

Page 1 of 1

## If zeros are in red, tops should be in green.

### #1gwnn

• Csaba the Hutt
• Posts: 12,863
• Joined: 2006-June-16
• Gender:Male
• Location:Heidelberg, Germany
• Interests:matching LaTeX delimiters :(

Posted 2017-December-24, 07:10

Why not? It would make self-scoring much easier.
... and I can prove it with my usual, flawless logic.
George Carlin
1

### #2barmar

• Posts: 18,097
• Joined: 2004-August-21
• Gender:Male

Posted 2017-December-25, 02:56

Why would green be easier than black?

It's not just zeros that are red, it's anything below 50%. We don't use special colors for bottoms.

Are you talking about BAM (as in MP challenges), where there are just 3 results, 0, 50, and 100%? It probably would be reasonable to have 3 different colors for this. On the other hand, I find that the scores stand out pretty well by themselves, since they have different numbers of digits.

bridgewebs.com, a site that many bridge clubs use to publish their results, uses a whole spectrum of colors for different score ranges. When I suggested doing the same thing here, I was told not to bother until we get around to completely redesigning myhands.

### #3gwnn

• Csaba the Hutt
• Posts: 12,863
• Joined: 2006-June-16
• Gender:Male
• Location:Heidelberg, Germany
• Interests:matching LaTeX delimiters :(

Posted 2017-December-25, 03:06

barmar, on 2017-December-25, 02:56, said:

Why would green be easier than black?

Because green is easier to distinguish from black than black from black. I thought this wasn't a controversial stance, but OK I guess it is. Maybe we should do a poll? "is green more different from black than black from black?"

Quote

It's not just zeros that are red, it's anything below 50%. We don't use special colors for bottoms.

Are you talking about BAM (as in MP challenges), where there are just 3 results, 0, 50, and 100%? It probably would be reasonable to have 3 different colors for this. On the other hand, I find that the scores stand out pretty well by themselves, since they have different numbers of digits.

Yes, I meant the BAM results (100, 50, or 0). I also kind of expected the 100 to be easy to distinguish from 50 based on the third digit. But in practice, for me, it's quite tough. Of course it's not like I spend 15 minutes to count them, but counting the zeroes is much faster than the tops.

Quote

bridgewebs.com, a site that many bridge clubs use to publish their results, uses a whole spectrum of colors for different score ranges. When I suggested doing the same thing here, I was told not to bother until we get around to completely redesigning myhands.

Yes maybe dividing it by fifths would be a good idea.

If you ever decide to do something like this, you can get some inspiration from check out this website:
http://colorbrewer2....scheme=RdBu&n=5
... and I can prove it with my usual, flawless logic.
George Carlin
0

### #4hrothgar

• Posts: 13,807
• Joined: 2003-February-13
• Gender:Male
• Location:Natick, MA
• Interests:Travel
Cooking
Brewing
Hiking

Posted 2017-December-25, 03:59

Red/Green color blindness is one of the most common forms

Might want to steer away from this particular combination.
Alderaan delenda est
0

### #5gwnn

• Csaba the Hutt
• Posts: 12,863
• Joined: 2006-June-16
• Gender:Male
• Location:Heidelberg, Germany
• Interests:matching LaTeX delimiters :(

Posted 2017-December-25, 04:52

hrothgar, on 2017-December-25, 03:59, said:

Red/Green color blindness is one of the most common forms

Might want to steer away from this particular combination.

Fair enough! RdYlBu could be a good one. http://colorbrewer2....eme=RdYlBu&n=11

Now that I have a data/MATLAB genius on the line, do you have any suggestions for a sequential color scheme that is perceptually uniform but doesn't blend in with a white background? Parula and others are great if I want to plot a map, but not so good when I have sparse data on white background. Or should I manually just set the lightness to a constant value?
... and I can prove it with my usual, flawless logic.
George Carlin
0

### #6hrothgar

• Posts: 13,807
• Joined: 2003-February-13
• Gender:Male
• Location:Natick, MA
• Interests:Travel
Cooking
Brewing
Hiking

Posted 2017-December-25, 05:49

gwnn, on 2017-December-25, 04:52, said:

Fair enough! RdYlBu could be a good one. http://colorbrewer2....eme=RdYlBu&n=11

Now that I have a data/MATLAB genius on the line, do you have any suggestions for a sequential color scheme that is perceptually uniform but doesn't blend in with a white background? Parula and others are great if I want to plot a map, but not so good when I have sparse data on white background. Or should I manually just set the lightness to a constant value?

I left MathWorks 5 years back and even then, data visualization was never my strong point.
(These days, I am Akamai and using a lot more R)

However, here's a couple thoughts

1. Can you shift your background color to black?

2. My go to color scheme is Viridis. I can almost always find something in there that will work.
Alderaan delenda est
0

### #7gwnn

• Csaba the Hutt
• Posts: 12,863
• Joined: 2006-June-16
• Gender:Male
• Location:Heidelberg, Germany
• Interests:matching LaTeX delimiters :(

Posted 2017-December-25, 08:50

Swapping the background color to black is a great idea for talks and in fact I'm seriously considering it for the future. But for papers or reports it's quite iffy (even if I don't care about ink, printers can make a mess of it).

Yes Viridis is quite pleasant.

It's just a bit weird that there's no obvious solution to this problem IMO. I'm happy to tweak existing stuff, but I'm more comfortable using something already developed and accepted by others.
... and I can prove it with my usual, flawless logic.
George Carlin
0

### #8hrothgar

• Posts: 13,807
• Joined: 2003-February-13
• Gender:Male
• Location:Natick, MA
• Interests:Travel
Cooking
Brewing
Hiking

Posted 2017-December-25, 10:24

gwnn, on 2017-December-25, 08:50, said:

Swapping the background color to black is a great idea for talks and in fact I'm seriously considering it for the future. But for papers or reports it's quite iffy (even if I don't care about ink, printers can make a mess of it).

Yes Viridis is quite pleasant.

It's just a bit weird that there's no obvious solution to this problem IMO. I'm happy to tweak existing stuff, but I'm more comfortable using something already developed and accepted by others.

Hope that you can fine something that works and sorry that I couldn't be of more direct help.
Alderaan delenda est
0

### #9gwnn

• Csaba the Hutt
• Posts: 12,863
• Joined: 2006-June-16
• Gender:Male
• Location:Heidelberg, Germany
• Interests:matching LaTeX delimiters :(

Posted 2017-December-25, 11:41

Thanks anyway! And my plots are usually already much prettier than those of others (who don't have any time to think about this stuff), I'm mostly just interested for the heck of it.
... and I can prove it with my usual, flawless logic.
George Carlin
0

### #10smerriman

• Posts: 1,113
• Joined: 2014-March-15
• Gender:Male

Posted 2017-December-25, 13:30

hrothgar, on 2017-December-25, 03:59, said:

Red/Green color blindness is one of the most common forms

Might want to steer away from this particular combination.

I always find it strange how despite this, red and green are universally important opposites, such as on traffic lights..
Feel free to send me a friend challenge of any format as often as you like. I'll always accept :)
0

### #11barmar

• Posts: 18,097
• Joined: 2004-August-21
• Gender:Male

Posted 2017-December-26, 09:49

smerriman, on 2017-December-25, 13:30, said:

I always find it strange how despite this, red and green are universally important opposites, such as on traffic lights..

Traffic lights also have standard non-color clues -- red is on top, green on bottom. And in many other contexts, color is a secondary indicator that simply accompanies numeric data -- it's just there to make patterns easier to see, which is beneficial for the 95% of people who aren't color-blind.