Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
iPhone 4 may be dropping as many as 100% more calls than iPhone 3GS
- Thread starter solarein
- Start date
- Sort by reaction score
Well it does. The rub is that the quantification that apple presented was biased to present the phone ambiguously and in the best possible light to downplay the problem. The OP has quite correctly highlighted how the figures could also be validly represented in a far less appealing way. The figures, either percentages or absolute numbers, are absolutely what people take into account when buy a new phone. And a new model performing more poorly on an important criteria like dropping calls (which afterall is the prime reason some people still use their phone) is a balck mark against the ip4.
You can bet your bottom dollar that if the percentages fell apple's way they would be touting it as 50% less dropped calls than the 3GS or other phones on the market.
This appears to be your problem. Although it's been pointed out numerous times now what you consider insignificant isn't what other's think insignificant. Dropped calls is a subjective issue. For people who make a lot of important calls it most certainly is an issue.
1/0 is undefined, but not
lim 1/x
x->+0
which is +infinity
Both phones are similarly worse than the 0 phone, but the 0.000....1 is much better (finite) than the 1 phone.
If you "substract" a finite number from infinity, it stays the same infinity.
You're immediately discounting the increase to being something major, i.e. design flaw. I think Jobs' case theory is very realistic. Not only that, but the concrete increase of less than one dropped call as being an engineering problem isn't a consumer problem per se. As I said before, this is about presenting information the consumer, not internally to engineers. And a consumer should see the information in it's most fairly representable fashion.
Telling a consumer that the iP4 drops up to 100% more calls than the 3GS means that the consumer will assume/believe either a.) the iP4 drops almost all calls (they won't take relativity into account and just think that almost 100% of calls made on an iPhone 4 are dropped [yes, some consumers may be that dumb]) or that b.) the iPhone 4 drops nearly twice as many calls as the 3GS. Point a is for those who don't think correctly. Point b doesn't matter because we don't know the sure statistic of dropped calls per 100 for the 3GS. So either way, the consumer can't make a sound decision against/for the iPhone 4 based on the information presented (unless that potentially 1 extra dropped call is REALLY important to them [compared to the 3GS]). Which is exactly what Jobs was trying to do.
Comparing the iP4 to other smartphones is a fallacy in logic, considering you're using dropped call data from ALL phones on AT&T's network ("smart" and otherwise) and applying it to the additional dropped calls the iP4 has compared to ONLY the 3GS, which is ONE other smartphone. This whole thread is really rather pointless.
I don't believe that Apple is presenting their data with a bias by using a <1 figure. I think that stating the figure as a percentage is much more misleading.
The OP pointed out that the iPhone could be dropping as many as 100% more calls. So in the WORST possible scenario, you are dropping one more call for every 100 that you make. If your calls are that important, then perhaps you should only use landlines.
Who said anything about misleading? They are presenting it in the most appealing light. Which is exactly what you'd expect from a damage control PR event.
As I said if the iP4 was dropping 50% less calls you can bet apple would be shouting the percentage from the rafters. They wouldn't be claiming "less that 0.5 calls per 100 dropped compared to the 3GS". Apple are playing exactly the same game you are trying to discount the OP for.
Which is what the OP pointed out. And went to a lot of effort to lay out his assumptions and make reasonable inferences about how large the range might be.
This is pretty much on par with the quality of the jobs explanations for the deficits in his products.
Trying to turn this into a percentage as opposed to how many calls per 100 is a waste of time. If the 3GS dropped 1% of calls and the 4 2% of calls (this is all made up since we don't know the numbers) then we could say OMG the 4 drops 100% more calls, thats twice as many! The reality is: it's 2 calls out of 100, a complete non issue. It's like a stock that went from .01 to .02 a share. Wow, a 100% increase! Well, whoopdedoo, it's still a worthless stock.
Imagine that? Steve tried to put things in a way that would put his product in the best possible light. Shocking!
I think we are confusing each other. I completely agree with you about dropped calls affecting people's decisions and also agree that OP has spun it in a negative light while Apple did the opposite.
What I meant by "topic having nothing to do w/ people's needs" is that the topic I was discussing was simply about statistics manipulation and nothing more and that I felt that on an absolute scale the rate of dropped calls is less alarming than what the OP is suggesting. At the same time it might be worse than what Apple is suggesting as well.
I never assumed anything about people's considerations when buying phones as a result of the statistical manipulation. I know it's going to affect what people want, and I figured it was effectively a tautology.
And I fully expected Apple to present those numbers in the most positive light they can
- why wouldn't they?
^^^Agreed. As I said to M-5 to bemoan the OP whilst defending the apple is entirely inconsistent. Both are valid ways to present the data. One the most cheery, the other the least attractive. Either might be unpalatable depending on your point of view.
Which it might be for you. But not for everybody. For others it might be enough to make them decide on another phone.
What is x? Isn't it the number of the calls that a phone drop?
If it is zero, it's zero. If it is 0.000203451, it's 0.000203451. How can x approach some number?
Yeah, I know they are worse, but I want to know how much they are worse. Can the 0.000...1 phone be 0.000...1 worse than the 0 phone and the 1 phone be 1 worse than the 0 phone? Or must they be "infinitely worse"? If so, how come the "infinitely worse" phone is still better than some phone?
In which case they would be idiots, because that 2 out of 100 is only comparable to the 3GS. If someone was trying to make a decision between all smartphones offered on AT&T's network, they would be using the comparison data between the iP4 and the 3GS to decide on the iP4 and everything else. The 3GS could drop an average of .5 calls per 100, the iP4 1.5 calls, and the next smartphone drops 2.5 calls on average. The iP4 is better than everything BUT the 3GS. But of course, we don't know exact numbers, so again, this whole thread is pointless.
Well it's a fairly insignificant increase. To pass up what I and many others consider to be the best smartphone out there because of a stat that in the real world doesn't mean much, would be a shame IMO. But each to his own.
It doesn't matter. You can't say that because your call-drop rate increased from 1 to 2 per 100 (100%) that someone else with 30 dropped calls per 100 will see an increase of 60 dropped calls (100%). The large percentages only work for small figures, so it's not accurate to state that dropped calls will increase 100%.
I just need to be convinced that dropping an extra call for every 100 that I make is a big deal.
That's how you compute in calculus things that cannot be computed using arithmetic.
For your second point:
lim x + 1 = +∞
x->+∞
That's how one of those phones can be finitely better than the other, but both infinitely worse than the 0 phone.
Cube's discussion about limits is about calculus.
The definition of a derivative is
f'(x) = lim h->0 (f(x+h) - f(x)) / h
So you learn about limits as the first topic (at least that's the first thing I learned when I took the course)
Great discussion technique
Way to respond to one part of my entire post. Talk about misleading.. Please respond to everything else I stated if you want to contribute.
EDIT: To add, if you're rolling eyes at me saying that someone would be an idiot to consider another smartphone (which encompasses all smartphones that can be used) based on data comparing the iP4 to a 3GS that's not an innate fact such as "the iPhone 4 runs on iOS 4", or "the iPhone 4 has a 5 MP camera", but a relative fact of dropped call rates between the two, then does that mean you think it's intelligent to do otherwise?
Uh, but 1/0 is still undefined "in calculus." There is a limit of 1/x when x->0+ but that doesn't make 1/0 suddenly have some value. It's still undefined.
And by the way, you didn't answer what is x and how can x approach something. How does limit have anything to do here?
Just watched the presentation, I didn't know those stats came from the AT&T users, I thought it was from controlled testing in Apple's fancy facility (bad assumption I guess).
I think you missed the point of my question though. Here is what I was implying...imagine 3 different sitautions: Person A with iPhone 4 holding it without interfering with the antenna, Person B with iPhone 4 holding it with antenna interference (i.e. deathgrip), and Person C holding their 3GS normally. If Person A only drops one more call per 100 than Person C, I don't think it's a big deal (this is what I thought was reported from controlled testing, not the field). If person B (using the deathgrip) drops 2 more calls than person C then I also don't think it's a big deal either. However, if the person B drops 50 more calls then I think it's a problem.
My point was, if deathgriping the iPhone doesn't lead to that many more dropped calls (despite lowering the signal) then it isn't a big problem. However, if they start dropping a ton more calls, then I think it starts to be an issue (hope that makes more sense).