Why Are the Numpad and a Telephone Pad Inverted?

[SamirD]

Tell me what genius thought it would be smart to make these two things similar enough that anyone using both on a regular basis would be entering 1 vs 7 and 9 vs 3 on a daily basis--why?

Seriously, does anyone know the answer to this?

[rowdy]

I wonder this all the time.

I have difficulty dialling a telephone due to this exact issue (and because I've used a computer keyboard numpad many, many more times than a phone keypad).

[SamirD]

I wonder this all the time.

I have difficulty dialling a telephone due to this exact issue (and because I've used a computer keyboard numpad many, many more times than a phone keypad).

Yes!  It's not just me, lol.

Now if we could only find the person responsible...

[fanpeople]

Interesting, never even thought about this but then again I don't do data entry. Just one of those things that happens and gets fixed as is.

http://abcnews.go.com/Technology/story?id=119296&page=1

[Oobly]

I never used the keyboard numpad enough to get used to it, so I'm more used to phone, credit card terminal and ATM numpads, actually. Was very happy to discover the existence of TKLs.

It came from claculator and adding machine layouts being predominantly low numbers on the bottom. So it made sense for data entry to be the same on computers. The telephone layout is more logical, but again we got stuck with the "traditional" layout simply because of familiarity at the time of designing the keyboard layouts. Same reason we all type QWERTY.

[SamirD]

I never used the keyboard numpad enough to get used to it, so I'm more used to phone, credit card terminal and ATM numpads, actually. Was very happy to discover the existence of TKLs.

It came from claculator and adding machine layouts being predominantly low numbers on the bottom. So it made sense for data entry to be the same on computers. The telephone layout is more logical, but again we got stuck with the "traditional" layout simply because of familiarity at the time of designing the keyboard layouts. Same reason we all type QWERTY.

Interesting that the cc terminals and atms also have it laid out like the phone.

Yep, I would be super fast on a calculator and it translated very quickly to the computer.  But since touch-tone phones also existed at the time, I wonder what debates existed if any.

It would be interesting to see a heat map of a numpad laid out both calculator and then telephone style to see what gets hit the most.

[SamirD]

Interesting, never even thought about this but then again I don't do data entry. Just one of those things that happens and gets fixed as is.

http://abcnews.go.com/Technology/story?id=119296&page=1

Thank you!  This settled it for me.  It would be interesting if someone revisits this again.  If in fact, 123 at the top is faster, maybe remapping the keys on the num pad make sense for data entry.  Pretty easy to swap the keycaps too.

[engicoder]

Why is the Ten Key numpad upside down? Same reason most keyboard layouts look like the Model M's Everyone chasing the leader. In 1914, Sustrand introduced an adding machine that became the one everyone copied.

http://retrocalculators.com/sundstrand.htm

[SamirD]

Why is the Ten Key numpad upside down? Same reason most keyboard layouts look like the Model M's Everyone chasing the leader. In 1914, Sustrand introduced an adding machine that became the one everyone copied.

http://retrocalculators.com/sundstrand.htm

Interesting.  I wonder if Bell ever contacted them.  And if so, maybe they didn't want to share the secrets of speed?

[SamirD]

And then of course this beauty gets posted in here today:


Check out that numpad layout!  And is it just me or does that look like phone controls?

[Oobly]

And then of course this beauty gets posted in here today:

Show Image

Check out that numpad layout!  And is it just me or does that look like phone controls?

Hmm.... look at the arrows. I think someone put the caps on in the incorrect positions. Looks like they were designed to be low numbers at the bottom.

[njbair]

Bell Labs doesn't get enough credit for their pioneering work in usability. Those guys did nothing by accident.

I've never thought about this question, but my first thought when I read it here was, "I'm sure the Bell Labs guys did a ton of research."

[SamirD]

And then of course this beauty gets posted in here today:

Show Image

Check out that numpad layout!  And is it just me or does that look like phone controls?

Hmm.... look at the arrows. I think someone put the caps on in the incorrect positions. Looks like they were designed to be low numbers at the bottom.

That's a good observation.  I thought that to be quite odd, but then again, the whole thing looks a bit odd, lol.

[fanpeople]

And then of course this beauty gets posted in here today:

Show Image

Check out that numpad layout!  And is it just me or does that look like phone controls?

Hmm.... look at the arrows. I think someone put the caps on in the incorrect positions. Looks like they were designed to be low numbers at the bottom.

That's a good observation.  I thought that to be quite odd, but then again, the whole thing looks a bit odd, lol.

Wouldn't the 5 remain in the same position though. Even so the arrows would make more sense if they were low at the bottom.

[SamirD]

And then of course this beauty gets posted in here today:

Show Image

Check out that numpad layout!  And is it just me or does that look like phone controls?

Hmm.... look at the arrows. I think someone put the caps on in the incorrect positions. Looks like they were designed to be low numbers at the bottom.

That's a good observation.  I thought that to be quite odd, but then again, the whole thing looks a bit odd, lol.

Wouldn't the 5 remain in the same position though. Even so the arrows would make more sense if they were low at the bottom.

It would and that's why I'm not sure if the arrow theory is correct.

[rowdy]

Numpad on ATM is upside down (compered to computer keyboard) too.

I had to enter my PIN on a computer keypad recently, and my muscle memory is totally upside down for that.  I even have to think what the digits are now.

[SamirD]

I hate thinking about the digits.   This is especially hard when typing in phone numbers I know since my hand remembers the numbers, not me.

[Kola93]

I am wondering if it was perhaps due to a particular patent that prevented others from designing and using another ten key pad in the same layout, so others circumvented the patent by inverting it.

[unoab]

One thing that may influence things is that computer entered numbers have a much higher frequency of 1's, 0's, and the use of decimal points.  So the pad may have been flipped for computer use to minimize finger motion due to occurrence of commonly input numbers.

I remember seeing some statistics on it before showing that lower digits appeared more often in real world data sets, but all I could find with a quick google search was Benford's law about the frequency distribution of leading digits (probability that a numerical input will start with a given digit), which says that in normal use numbers will disproportionately start with smaller digits:
https://en.wikipedia.org/wiki/Benford%27s_law

[rowdy]

One thing that may influence things is that computer entered numbers have a much higher frequency of 1's, 0's, and the use of decimal points.  So the pad may have been flipped for computer use to minimize finger motion due to occurrence of commonly input numbers.

I remember seeing some statistics on it before showing that lower digits appeared more often in real world data sets, but all I could find with a quick google search was Benford's law about the frequency distribution of leading digits (probability that a numerical input will start with a given digit), which says that in normal use numbers will disproportionately start with smaller digits:
https://en.wikipedia.org/wiki/Benford%27s_law

I would suspect that are far more data entry operators using numpads than programmers.  And they would be entering much more than just 0 and 1.

[davkol]

I would suspect that are far more data entry operators using numpads than programmers.  And they would be entering much more than just 0 and 1.

Eh? Benford's law was proposed before there were any digital computers.

[Melvang]

I would suspect that are far more data entry operators using numpads than programmers.  And they would be entering much more than just 0 and 1.

Eh? Benford's law was proposed before there were any digital computers.

How does that change the logic and meaning of the law.  If it addresses real world number frequency, then wouldn't it apply today as much as then.  How many programmers work in binary?

[davkol]

This is confusing.

Although, unoab might have a point, because of variable initialization or incrementation in loops.

[SamirD]

I remember seeing some statistics on it before showing that lower digits appeared more often in real world data sets, but all I could find with a quick google search was Benford's law about the frequency distribution of leading digits (probability that a numerical input will start with a given digit), which says that in normal use numbers will disproportionately start with smaller digits:
https://en.wikipedia.org/wiki/Benford%27s_law

This theory seems to make sense with me.  It would also explain why this worked on calculators.

[unoab]

I would suspect that are far more data entry operators using numpads than programmers.  And they would be entering much more than just 0 and 1.

Oh, I may have wrote that a bit weird, the 0's and 1's part, I didn't really mean binary, I only ever had to write machine code in binary once, and was happy to transition to assembly later in that class, I just meant the normal occurrence of those numbers in the natural use, so lots of 10's 100's and seemingly whole numbers and other ones that you would normally write.

Now, what I would have asked was why on older terminals, like the F122, was the tab key (when available) at the top of the numpad if the more commonly input numbers would be distributed closer to the bottom of the keypad...  That would lead to a lot of top to bottom action filling out data fields if we were assuming they were laid out according to digit frequency and ergonomics.

[SamirD]

I would suspect that are far more data entry operators using numpads than programmers.  And they would be entering much more than just 0 and 1.

Oh, I may have wrote that a bit weird, the 0's and 1's part, I didn't really mean binary, I only ever had to write machine code in binary once, and was happy to transition to assembly later in that class, I just meant the normal occurrence of those numbers in the natural use, so lots of 10's 100's and seemingly whole numbers and other ones that you would normally write.

Now, what I would have asked was why on older terminals, like the F122, was the tab key (when available) at the top of the numpad if the more commonly input numbers would be distributed closer to the bottom of the keypad...  That would lead to a lot of top to bottom action filling out data fields if we were assuming they were laid out according to digit frequency and ergonomics.

Ah, I read it as binary as well.

I think the reason tab may have been there vs anywhere else is because enter was used far more often than tab.  But it is good to know that's on an f122--makes spreadsheets really nice to enter data into.

[Oobly]

Benford's only relates to starting digits, not overall digit frequency, so it's not particularly relevant to numpad design for data entry. For instnce, how many prices nowadays don't have a "9" in them? Ideally you'd have "0",".","1" and "9" in the easiest places to reach, with "5" next easiest.

The research done by Bell labs was the best I've yet seen for this and agree that the low numbers at the top is most usable and practical. That said, however, two handed data entry on the top row has the capability to be faster and more efficient than either numpad orientation

[Dernubenfrieken]

One thing that may influence things is that computer entered numbers have a much higher frequency of 1's, 0's, and the use of decimal points.  So the pad may have been flipped for computer use to minimize finger motion due to occurrence of commonly input numbers.

I remember seeing some statistics on it before showing that lower digits appeared more often in real world data sets, but all I could find with a quick google search was Benford's law about the frequency distribution of leading digits (probability that a numerical input will start with a given digit), which says that in normal use numbers will disproportionately start with smaller digits:
https://en.wikipedia.org/wiki/Benford%27s_law

This is how I always saw it. With the orientation of most numpads relative to the alpha numeric, its more comfortable to have the lower digits, zeros, decimal and enter at the bottom so you can swing your arm out and hit it correctly.

[SamirD]

The research done by Bell labs was the best I've yet seen for this and agree that the low numbers at the top is most usable and practical. That said, however, two handed data entry on the top row has the capability to be faster and more efficient than either numpad orientation

Hmmm...that's an interesting point about the top row of numbers.  If I cover all of them with my fingers, I literally only have to move the fingers to hit 5 and 6.  With each number already under a finger, that's a tremendous savings in time right there.  I wonder if I can train myself to type numbers this way.  If so, then I'll be on the road to a tkl when I never thought I could live without the numpad.

[shibataken]

Isn't it because the calculator has the same layout as a numpad so accountants that would use a numpad much more than the average user could quickly transition from calculator to numpad depending on what they have to do?  So it would be do the calculations on the calculator and then data entry on the numpad.  Would probably be a headache if those two were inverted to be like a phone or ATM.

[SamirD]

Isn't it because the calculator has the same layout as a numpad so accountants that would use a numpad much more than the average user could quickly transition from calculator to numpad depending on what they have to do?  So it would be do the calculations on the calculator and then data entry on the numpad.  Would probably be a headache if those two were inverted to be like a phone or ATM.

I think that's probably the case since computers evolved from calculators, but still when the research showed the telephone pad was a better layout, why wasn't that adopted is a bit of a huh moment.

[Melvang]

Isn't it because the calculator has the same layout as a numpad so accountants that would use a numpad much more than the average user could quickly transition from calculator to numpad depending on what they have to do?  So it would be do the calculations on the calculator and then data entry on the numpad.  Would probably be a headache if those two were inverted to be like a phone or ATM.

I think that's probably the case since computers evolved from calculators, but still when the research showed the telephone pad was a better layout, why wasn't that adopted is a bit of a huh moment.

My guess would be due to adoption principles similar to why we never made the transition to Dvorak.