iPhone 4 may be dropping as many as 100% more calls than iPhone 3GS

[scaredpoet]

To be honest, whether you notice the increase or not is besides the point.

No, it's exactly the point. End users aren't putting their iPhone 4s on pedestals and pontificating with others about statistics and percentage increases. People are actually using their phones to make actual calls. And whether or not a significant call drop increased is noticed is the whole point.

If your analysis fails to acknowledge this, then it makes even less sense.

And this is a fundamental problem not just with your findings, but with a lot of the videos and other evidence posted out there. We can view videos of death grips galore, and see bars drop on iPhones and non-iPhones. But how many of those videos actually show a call in progress, by an actual person using it in a normal situation to converse with another human being to test whether the call actually drops?

So far I've seen 0 such videos. And 0 is a pretty significant number, too!

 

[Terdinus Asus]

What I'm more interested in knowing is whether the comparison is drawn between data of the 3GS and iPhone 4 over the same time period or if it's the life-long average of 3GS vs. iPhone 4.

If it is life-long vs. 22 days of iPhone 4, the difference could easily be due to how the iPhone 4 has been distributed throughout the country. Due to limited supply Apple caters to big metropolitan centers first, so places with poor reception like NYC and San Francisco tend to be where most iPhones pile around in the first few months of release. This may well skew drop data in a phone's early life.

It may be interesting to see how this data changes as the supply shortage comes to an end and we see a more balanced picture of distribution of the iPhone.

 

[cube]

Given this conversation, yes it is. Steve said "less than one more" and the point of this thread is to say that it could possibly be 100% more. So starting with zero doesn't work.

The title of the thread is wrong. 100% is not the limit. The limit is infinite % more.

1% more dropped calls can mean that a phone is infinite % worse than another one.

Of course, we know that some calls are actually dropped with any phone, so the difference would not really be infinite, but it could actually be a HUGE number.

 

[AZREOSpecialist]

And they try to make it sound less awful then it is. If you drop 1 call per day on a 3gs in your area then with the i4 you should expect to see up to 2. Doesn't sound bad at all. But what if person X doesn't live in such a good service area? If this person use to see 5 dropped calls a day he should expect to see up to 10. Person y lives in a even crappier service area. He use to see 10 drop calls a day. He can now expect to see up to 20 with the i4.

I can keep going but I'm sure you can see how much larger the dispersion gets. People who suffer from dropped calls will notice the difference.

Again, more assumptions. Who said the 3GS drops 1 call per day? I haven't seen any statistics released with regards to how many calls per 100 are dropped by the 3GS. Have you?

 

[TonyC28]

The point was to show that there is a statistically significant difference the rate of dropped calls on the iPhone 4G versus other phones. By showing that there is indeed a statistically significant difference, and a fortiori by showing that this difference is very large, I hoped to debunk Steve Jobs' claim that the iPhone 4's reception problem is one that plagues all phones, the iPhone 4 only slightly more, if at all, than others. I wished to show instead that there is such a difference between the iPhone 4 and other phones that the problem is to a large extent unique to the iPhone 4.

But what you said was that the "iPhone 4 may be dropping as many as 100% more calls than the 3GS." And to say that there is a "statistically significant difference" is kind of misleading unless you acknowledge that while you are technically right in saying that the increase could possibly be 100%, according to Apple then that would mean an increase from 1 out of 100 to 2 out of 100.....which to me, isn't very significant at all.

 

[TonyC28]

The title of the thread is wrong. 100% is not the limit. The limit is infinite % more.

1% more dropped calls can mean that a phone is infinite % worse than another one.

Of course, we know that some calls are actually dropped with any phone, so the difference would not really be infinite, but it could actually be a HUGE number.

It's still to early on Saturday to get into all this math stuff, but wouldn't an infinite increase mean that every call is dropped?

 

[solarein]

What I'm more interested in knowing is whether the comparison is drawn between data of the 3GS and iPhone 4 over the same time period or if it's the life-long average of 3GS vs. iPhone 4.

If it is life-long vs. 22 days of iPhone 4, the difference could easily be due to how the iPhone 4 has been distributed throughout the country. Due to limited supply Apple caters to big metropolitan centers first, so places with poor reception like NYC and San Francisco tend to be where most iPhones pile around in the first few months of release. This may well skew drop data in a phone's early life.

It may be interesting to see how this data changes as the supply shortage comes to an end and we see a more balanced picture of distribution of the iPhone.

This is a great point, there could be other confounding factors that explains the discrepancy rather than actual bad performance of the iPhone 4.

 

[solarein]

No, it's exactly the point. End users aren't putting their iPhone 4s on pedestals and pontificating with others about statistics and percentage increases. People are actually using their phones to make actual calls. And whether or not a significant call drop increased is noticed is the whole point.

If your analysis fails to acknowledge this, then it makes even less sense.

And this is a fundamental problem not just with your findings, but with a lot of the videos and other evidence posted out there. We can view videos of death grips galore, and see bars drop on iPhones and non-iPhones. But how many of those videos actually show a call in progress, by an actual person using it in a normal situation to converse with another human being to test whether the call actually drops?

So far I've seen 0 such videos. And 0 is a pretty significant number, too!

We are talking about actual calls dropped here, my friend.

Think about it this way. If you made a phone and it drops twice as many calls as the previous version (even though the actual number of dropped calls is small), and you tell your boss this, is he going to say, oh that's not a problem at all, or is he going to say that you have a problem on your hands?

 

[solarein]

The title of the thread is wrong. 100% is not the limit. The limit is infinite % more.

1% more dropped calls can mean that a phone is infinite % worse than another one.

Of course, we know that some calls are actually dropped with any phone, so the difference would not really be infinite, but it could actually be a HUGE number.

Yeah, I know. 100% is really my best estimate. I added in "as many as" because it could be fewer if one used different assumptions. I wanted to put my best estimate in the subject rather than some ridiculously large number.

 

[solarein]

But what you said was that the "iPhone 4 may be dropping as many as 100% more calls than the 3GS." And to say that there is a "statistically significant difference" is kind of misleading unless you acknowledge that while you are technically right in saying that the increase could possibly be 100%, according to Apple then that would mean an increase from 1 out of 100 to 2 out of 100.....which to me, isn't very significant at all.

That's not how statistically significant works, however. There's a technical definition for that term which I won't go into here. You can look it up in Wikipedia if you'd like.

 

[TonyC28]

That's not how statistically significant works, however. There's a technical definition for that term which I won't go into here. You can look it up in Wikipedia if you'd like.

Oh, well if it's on Wikipedia then it must be right

So given the fact that for it to be an increase of less than 1 (but we'll say 1 for the sake of making it easier) and to be a 100% increase that would mean going from 1 to 2 out of 100, you think it's fair to call that "statistically significant"?

 

[cube]

It's still to early on Saturday to get into all this math stuff, but wouldn't an infinite increase mean that every call is dropped?

No, the rate of increase in drops would be infinite, not the increase in drops. We know that the increase in drops is 0%<x<1%

If x is the new drop rate and y is the old drop rate, the increase in drop rate is x-y and the new phone is x/y times worse than the previous one.

 

[solarein]

Oh, well if it's on Wikipedia then it must be right

So given the fact that for it to be an increase of less than 1 (but we'll say 1 for the sake of making it easier) and to be a 100% increase that would mean going from 1 to 2 out of 100, you think it's fair to call that "statistically significant"?

Yeah I think it's safe to say that an 100% increase is statistically significant.

 

[eawmp1]

I was going to write a serious response to this, but then I felt rather silly for wanting to do so to someone who's obviously either a troll or too mentally incompetent to comprehend a rebuttal anyway.

Huh? You are delusional. And you need to reevaluate how you use "troll"

 

[thetexan]

I'm on Verizon, what's a "dropped call"? Never heard of it.

I kid, I kid.

Fact of the matter is there are three different types of lies. They are lies, damn lies, and statistics. These numbers can be bended and molded into anyway the presenter would like them to be. Until we see raw data, which we never will, we have no idea of what is really going on. It's very possible <1 per 100 increase is a large increase over the 3GS, and it's possible that the increase is barley there. Bottom line is that statistic is meaningless because we have no base to compare it to. Also, we were promised better reception as it was a cornerstone of the keynote about the iPhone 4, and we've had an increase in dropped calls.

 

[solarein]

Huh? You are delusional. And you need to reevaluate how you use "troll"

Umm...you do realize that you pretty much just admitted that you are mentally incompetent?

To quote my favorite cartoon show.

Amy: How do I look, Professor?
Professor: Like a cheap French harlot.
Amy (dismayed): French?

 

[TonyC28]

No, the rate of increase in drops would be infinite, not the increase in drops. We know that the increase in drops is 0%<x<1%

If x is the new drop rate and y is the old drop rate, the increase in drop rate is x-y and the new phone is x/y times worse than the previous one.

My head is officially spinning. I think we might not be on the same page here but we are probably in the same chapter. I'm disappointed that the 4 has even a .000000000001% increase in dropped calls because they said it was gonna be so much better. I'm operating under the assumption that AT&T's numbers for dropped calls on specific phones (i.e the 3GS) is proprietary like Steve said and that is why he couldn't give the exact numbers to compare to. I'll admit that a 100% increase is possible but if it was going from 1 to 2 that is very misleading and would scare people if they didn't know that first number. I think the way Steve explained it was the best way he could have because using a percentage would have seemed a lot worse than it probably really is. Again, I'm assuming that he really couldn't give the drop call numbers for the 3GS so for him this was the best way to explain it.

 

[atomicbatteries]

Ok so your saying Steve Jobs is lying and he made. Up that whole presentation as some sort of cover up and that antennagate is alive and working clandestinely right this very minute while I write this post?

Do you even have an iphone4?

 

[gloss]

1 in 100 = 1/100 = .01 = 1%

What part of that would you like me to help you with?

Methinks you're confused.

1 in 100 calls means one percent of all calls made on the phone.

Saying that the iPhone 4 drops less than one percent of ALL calls more than the iPhone 3GS is totally different than saying that the iPhone 4 drops less than one percent MORE calls than the iPhone 3GS. Steve was saying the former, not the latter. People who believe the latter need lessons in semantics.

 

[TonyC28]

Yeah I think it's safe to say that an 100% increase is statistically significant.

I'm really trying to see this through your eyes but if you think going from 1 to 2 out of 100 is statistically significant to users I think we're gonna have to agree to disagree.

 

[mdatwood]

"Lies, damned lies, and statistics"

Give me all of the raw numbers and can I just about make them saying anything you would like. It's not very hard.

In order to fully understand a given topic you need to see the statistics and raw number so that the statistics are given context. The problem is that people tend to pick and choose whatever side they want to show in order to prove their point. So few people are really about presenting the facts nowadays.

 

[ashVID]

You can spin stats any way you want. Bottom line, Apple admitted the newer phone with "better" technology dropped more calls. Mine, with a bumper, drops more calls than my 3Gs. Probably 1 out of 50 or so.

I can assure you the less than 1 in a 100 is the best way for them to present it. If it had been .1, they would have said that, my guess is that it is between .9 and .99.

Ultimately, you don't need stats to arrive at the conclusion that Apple has as much to do with dropped calls and network issues as At&t. Most of us have known this all along.



ash =o)

 

[gloss]

I'm really trying to see this through your eyes but if you think going from 1 to 2 out of 100 is statistically significant to users I think we're gonna have to agree to disagree.

It is for what it represents. If 2% of all calls are being dropped it could easily mean that many people experience service that's perfectly fine while a small contingent are dropping half or more of their calls. That, from a consumer standpoint, is a much bigger issue than if everyone were dropping only 2 out of 100.

Statistics are not communist.

 

[dogie678]

OMG it drops 1 more call per 100 it's the end of the fuccing world!!

LOL

 

[cube]

My head is officially spinning. I think we might not be on the same page here but we are probably in the same chapter. I'm disappointed that the 4 has even a .000000000001% increase in dropped calls because they said it was gonna be so much better. I'm operating under the assumption that AT&T's numbers for dropped calls on specific phones (i.e the 3GS) is proprietary like Steve said and that is why he couldn't give the exact numbers to compare to. I'll admit that a 100% increase is possible but if it was going from 1 to 2 that is very misleading and would scare people if they didn't know that first number. I think the way Steve explained it was the best way he could have because using a percentage would have seemed a lot worse than it probably really is. Again, I'm assuming that he really couldn't give the drop call numbers for the 3GS so for him this was the best way to explain it.

He could have kept the secret by saying, "the 4 is 30%, 100%, 1000% worse than the GS", but then the people would not know that that translated in less than 1% more dropped calls. "1000% worse" would lead people to think that it totally sucks (and an engineer would think this is the correct interpretation, in any case).