Your top tier status folks may expect every upgrade to clear, however most mid and low level elites see it as a perk.
As a delta plat, my upgrade percentage is about 75% now, and I am quite happy.
Worst case, I know that I can select Comfort Plus/economy comfort seats for free, and will have free drinks on the flight along with the free premium movies.
If it is a long flight and I must absolutely get the upgrade, i am ok with paying for it or using a certificate.
If there were no unlimited upgrades.... I would not be flying delta almost exclusively.
As a delta plat, my upgrade percentage is about 75% now, and I am quite happy.
Worst case, I know that I can select Comfort Plus/economy comfort seats for free, and will have free drinks on the flight along with the free premium movies.
If it is a long flight and I must absolutely get the upgrade, i am ok with paying for it or using a certificate.
If there were no unlimited upgrades.... I would not be flying delta almost exclusively.
Quote:
Furthermore, you cannot say a priori how large a sample size is needed to make a probability statistically meaningful. There are many factors involved, and hundreds of cases may well be sufficient. There are many statistical parameters that allow one to determine whether a given probablility is meaningful - what the "margin for error" is, for example. One such parameter is standard deviation.
These are just a couple of examples of why the above statements by the OP therefore reflect a fundamental lack of understanding of the principles of statistics and probability. Or if he/she does understand, then instead he/she is intentionally misusing or distorting his interpretation to arrive at an inaccurate conclusion.
I don't know if "standard deviation" is the same as margin of error. I don't think it is.Originally Posted by JerryFF
The above comments are simply contradictory to basic statistical thinking and analysis. Of course, for every individual event, with two options, it is either A or B. But conventional thinking is based on the concept that over a certain number of trials, A occurs a certain % (=X) and B occurs (100-X) %. This allows people to make decisions in a wide variety of situations. To say it has no meaning because, for an individual event, it must be either A or B simply fails to understand the basic concept of statistics and their applications.Furthermore, you cannot say a priori how large a sample size is needed to make a probability statistically meaningful. There are many factors involved, and hundreds of cases may well be sufficient. There are many statistical parameters that allow one to determine whether a given probablility is meaningful - what the "margin for error" is, for example. One such parameter is standard deviation.
These are just a couple of examples of why the above statements by the OP therefore reflect a fundamental lack of understanding of the principles of statistics and probability. Or if he/she does understand, then instead he/she is intentionally misusing or distorting his interpretation to arrive at an inaccurate conclusion.
Actually, in most fields they do determine how large a sample size is needed to make something statistically significant. If they didn't do that, I don't know that there would be many jobs for statisticians. In many fields the p value is arbitrarily set at 0.05.
I don't think you understand statistics. Ever heard of "n=1"?
Quote:
I have to agree that I have no idea how'd I explain it to someone who doesn't know what probability means.
Statistically significant means that you can reject the null hypothesis to your assumption with a certain confidence level which is itself a probability. The number of samples is nowhere nearly as weighing in this process as is the randomness of the sampling, the strength of the tested hypothesis, and the required confidence interval, and the ability to find the distribution of the stochastic events.
To claim that from a sample of 100 flights, one cannot draw statistically significant conclusions is mind bogglingly far removed from any mathematical reality
.
Actually, it is pretty simple. If a person were not upgraded for any flight out of their 100 flights, would you say they can draw the conclusion that there are no upgrades? How would they speak to someone else who has been upgraded all of their flights?Originally Posted by weero
Haha.I have to agree that I have no idea how'd I explain it to someone who doesn't know what probability means.
Statistically significant means that you can reject the null hypothesis to your assumption with a certain confidence level which is itself a probability. The number of samples is nowhere nearly as weighing in this process as is the randomness of the sampling, the strength of the tested hypothesis, and the required confidence interval, and the ability to find the distribution of the stochastic events.
To claim that from a sample of 100 flights, one cannot draw statistically significant conclusions is mind bogglingly far removed from any mathematical reality
.
I haven't gotten any upgrades for my last 10 AA flights. Probably 50% for my last US Air flights. And 3/5 for my last Delta flights. I can speak to someone and compare stats, but it really doesn't mean anything. I would comment that AA never upgrades me. They would say something like AA often upgrades me.
If I flew 100, you are saying that means my statement would carry so much worth that invalidates his statement? I don't think so.
Quote:
I haven't gotten any upgrades for my last 10 AA flights. Probably 50% for my last US Air flights. And 3/5 for my last Delta flights. I can speak to someone and compare stats, but it really doesn't mean anything. I would comment that AA never upgrades me. They would say something like AA often upgrades me.
If I flew 100, you are saying that means my statement would carry so much worth that invalidates his statement? I don't think so.
I wonder what is the probability that you are writing all this rubbish just to get a rise out of someone, As we all know , satistics can prove whatever you want them to !!Originally Posted by s0ssos
Actually, it is pretty simple. If a person were not upgraded for any flight out of their 100 flights, would you say they can draw the conclusion that there are no upgrades? How would they speak to someone else who has been upgraded all of their flights?I haven't gotten any upgrades for my last 10 AA flights. Probably 50% for my last US Air flights. And 3/5 for my last Delta flights. I can speak to someone and compare stats, but it really doesn't mean anything. I would comment that AA never upgrades me. They would say something like AA often upgrades me.
If I flew 100, you are saying that means my statement would carry so much worth that invalidates his statement? I don't think so.
I'll just go out on a limb and say that upgrades aren't normally-distributed random events. They're driven by algorithm.
So I may be 0 for 100 and you may be 10 for 10, and luck (or distribution of random events) has absolutely nothing to do with it.
In a similar vein, it's comparable to how data about IDB or lost baggage is lumped together to make those events appear incredibly rare and randomly distributed across all events, when in fact they aren't.
So I may be 0 for 100 and you may be 10 for 10, and luck (or distribution of random events) has absolutely nothing to do with it.
In a similar vein, it's comparable to how data about IDB or lost baggage is lumped together to make those events appear incredibly rare and randomly distributed across all events, when in fact they aren't.
FlyerTalk Evangelist
walk up / standby / change is possible
but hotel upgrades (including walk ups) are more common and easier
but hotel upgrades (including walk ups) are more common and easier
Quote:
While generally a very good lecture ^ ... I disagree on this part.Originally Posted by JerryFF
..Furthermore, you cannot say a priori how large a sample size is needed to make a probability statistically meaningful. There are many factors involved, and hundreds of cases may well be sufficient..
If you do not know the distribution of the events, you cannot even conclude a posteriori how large the likelihoods are. And if you know the distribution, you can calculate with decent accuracy how large your sample size must be.
But for most natural cases with Normal Distribution, you are overall spot on.
Quote:
I see. "No upgrades" is a very strong hypothesis - much like for gravity that all heavy bodies attract each other - it would need immense testing to get it meaningful.Originally Posted by s0ssos
Actually, it is pretty simple. If a person were not upgraded for any flight out of their 100 flights, would you say they can draw the conclusion that there are no upgrades? How would they speak to someone else who has been upgraded all of their flights?..
But if you fly on times of the day and on all days of the week, and in many booking classes and you did not get upgraded once in your last 100 flights, to say that upgrades happen in less than 10% of all cases would hit very hight on a Student's t or F test.
You can illustrate it from a trivial comparison: of you assume that the anti-hypothesis is true (>=10% upgrades), then for the worst case, you chances of NOT getting upgraded on a flight are 90% or 0.9. The chance that you do not get upgraded once in 100 flights is hence 0.9_to_the_power_of_100 or 0.0026%.
For the 5% hypothesis this grows to 0.6% and for the 1% hypothesis to 37%.
Evidently to say 0% would be nonsensical. But 1% is - while unscientific - already quite plausible.
I get what OP is saying in that people expect upgrades and don't get them feel worse than if they never even had a chance at one.
I think the biggest issue is time... what I mean is there are people here that started flying pre-consolidation and load factors were in the 50% range on most flights and upgrades were practically automatic, and are now upset when their upgrade percentage is far lower.
Then there are the more recent fliers who may have only been true road warriors the last few years and are more accustom to sitting in Y and think of upgrades more as a treat than an expectation.
Obviously there are people who have been flying for years that have adjusted just fine to the consolidated industry and people who just hit the lowest tier status for the first time ever and can't understand why they aren't being upgraded on their Monday morning flight from LGA to ORD.
Just my thoughts on the matter
I think the biggest issue is time... what I mean is there are people here that started flying pre-consolidation and load factors were in the 50% range on most flights and upgrades were practically automatic, and are now upset when their upgrade percentage is far lower.
Then there are the more recent fliers who may have only been true road warriors the last few years and are more accustom to sitting in Y and think of upgrades more as a treat than an expectation.
Obviously there are people who have been flying for years that have adjusted just fine to the consolidated industry and people who just hit the lowest tier status for the first time ever and can't understand why they aren't being upgraded on their Monday morning flight from LGA to ORD.
Just my thoughts on the matter
Well, to someone who flies often it is obvious that upgrades are dependent on which flight. Or to someone who stays in lots of hotels.
If you want a good shot at an upgrade take some random US Airways flight with AA status. But if you are flying Delta out of Atlanta, just give up before you even start.
The issue with a lot of these companies is that upgrading isn't effortless. Because sometimes you aren't even put on the upgrade list unless you specifically request it (like with the AA/US merger, and code-share flights). With Marriott you have to ask for an upgrade-they don't like to hand them out, apparently.
This makes people unhappy, because they have to conscious remember to ask for an upgrade, and then get rejected.
If you want a good shot at an upgrade take some random US Airways flight with AA status. But if you are flying Delta out of Atlanta, just give up before you even start.
The issue with a lot of these companies is that upgrading isn't effortless. Because sometimes you aren't even put on the upgrade list unless you specifically request it (like with the AA/US merger, and code-share flights). With Marriott you have to ask for an upgrade-they don't like to hand them out, apparently.
This makes people unhappy, because they have to conscious remember to ask for an upgrade, and then get rejected.
Quote:
And hotels are well.
They want to be able to use their leftovers for goodwill/rewards for loyal members.
I feel it is like salaries. There is no such thing as making everybody happy. There is only distributing it in such a way that most people aren't unhappy.
You think that and you feel that? Did you do that thing, you know, where you spend some time researching to see if there's any data to back up the ideas that you're pulling out of your ahhh thin air?Originally Posted by s0ssos
I think the possibility of upgrades makes people madder than happier. But airlines love to sell everybody on the possibility of getting upgrader to first-class (dependent on availability, whether your flight is leaving on-time, where the sky is clear, whether the baggage crew has showed up on time that ).And hotels are well.
They want to be able to use their leftovers for goodwill/rewards for loyal members.
I feel it is like salaries. There is no such thing as making everybody happy. There is only distributing it in such a way that most people aren't unhappy.
Quote:
Probably more than you. Do you have any research? Or is it just your "feelings"?Originally Posted by ou81two
You think that and you feel that? Did you do that thing, you know, where you spend some time researching to see if there's any data to back up the ideas that you're pulling out of your ahhh thin air?