|
Given that Pathfinder 14 still hasn't shipped, I decided to run some numbers, to see how long it has been between issues. A quick chart:
- Issue Days
- 1 --
- 2 36
- 3 33
- 4 36
- 5 35
- 6 14
- 7 42
- 8 35
- 9 21
- 10 42
- 11 23
- 12 25
- 13 0
- 14 59+
These numbers come from my subscriptions, and I know they ship over a period of a few days, so numbers may vary slightly from person to person. Still, with these numbers, we're looking at an average of 30.8 days between issues - which is good. But, there's a standard deviation of 14.6 days, nearly half of the average. That is, to be blunt, terrible.
It makes it very hard to plan payments, let alone game schedules. I'm sure this is high on Paizo's priorities - at least, I hope so. Just wanted to actually run and post the numbers to make it clear. Please try to improve schedule regularity.
Drew Garrett
Director of Brand Strategy
|
It makes it very hard to plan payments, let alone game schedules. I'm sure this is high on Paizo's priorities - at least, I hope so. Just wanted to actually run and post the numbers to make it clear. Please try to improve schedule regularity.
I'm gonna pop in here and try to ninja this away from the Paizo staff, who are busy getting back on schedule >.>
Many members of the editorial and management staffs at Paizo have said for a long time that getting on a more stable schedule is one of their top goals. A lot of it started when they were developing the first few modules and the beginning of RotR while Dungeon and Dragon were still going on. This spread them pretty thin, and the increase in demand for the games and resulting increase in products, has continued it. I know that they've been working very hard to be on a schedule, and from what I understand, they are now all caught up, at least from a development end. Once the last stragglers come out off-schedule, they should be good to go from now on.
Beyond the annoyance to customers of not knowing when things are coming out, Lisa and James have both said (and I'm sure others have too) that Paizo as a business can't really run optimally without having the books come out regularly either, as their expenses are pretty regular, being utilities, printing costs, development, and Wes's paycheck (though I hear they're getting rid of that soon).
Also, keep in mind that the current scheduling gap is due to something completely out of their control, and that we would have received PF14 in mid-September had customs not intervened.
Give it til the end of the year, when the modules will also be caught back up, and then see how those averages start to change more toward the ideal.
|
Chief Technical Officer
|
I'll also add that by the time #14 is through customs, it can probably just be shipped at the same time as #15. For some this might be too much at once for their monthly expenses, but for me, I'd gladly save a few bucks in shipping for the combined shipment.
14 is due... well, we're still not sure, but hopefully any day now. 15 isn't due to arrive for three more weeks.
|
|
Acually, because this is a Poisson Distribution and not a Normal or Uniform Distrution, the Standarad Deviation is closer to 5.3. You should also throw out the PF-14 data point (it's invalid due to custom's intervention).
-Your Friendly Neighborhood Statistics Nit-Picker
I have to disagree with your nitpick. I am not counting shipments, I am measuring days. For measurements, a normal distribution is appropriate. I could be more sympathetic to an argument of throwing out the PF14 data point as a statistical outlier (hmmm, no, I think perhaps systematic error might be a better reason ;-) but only if there is a change in process so it doesn't happen again. If they're following a process in which this remains a possibility, then whether or not it's in their control, it remains a valid data point.
Hey, hey, don't look at me that way... Some of us like discussing statistics.
Drew Garrett
Many members of the editorial and management staffs at Paizo have said for a long time that getting on a more stable schedule is one of their top goals. A lot of it started when they were developing the first few modules and the beginning of RotR while Dungeon and Dragon were still going on. This spread them pretty thin, and the increase in demand for the games and resulting increase in products, has continued it. I know that they've been working very hard to be on a schedule, and from what I understand, they are now all caught up, at least from a development end. Once the last stragglers come out off-schedule, they should be good to go from now on.
Actually, it seems that they continue to have issues that compete for their time, (Gen Con, the Beta of PFRPG, hardcover campaign setting, ...), not just the initial start-up craziness. I don't see much of these sorts of things going away (conventions are important and annual and the PFRPG will probably just increase in its complexity and time-consumption), which is troubling. I have said a couple times that I feel the Paizo staff, while talented and well-intentioned, have bitten off more than they can chew and the quality of their product, including shipment schedules, suffers, much to the annoyance of some of their customers. I am actually planning to cancel all my subscriptions here over the next couple months, for a variety of reasons, but among them are the issues I have with product quality and shipping that seem almost entirely the result of Paizo trying to do too much with too few people.
O
|
... I feel the Paizo staff, while talented and well-intentioned, have bitten off more than they can chew and the quality of their product, including shipment schedules, suffers...
I don't doubt that Paizo's staff do have a great deal on their collective plates. Most of it, I believe, is a direct result of forum outcry for additional materials. The forum community wants a deities book. The forum community wants player friendly books (Pathfinder Companion). Not to mention numerous other products from the Chronicles line. It turns out the user community here is pretty demanding.
I guess I don't see the quality issues you you see, but that's a discussion for another thread.
-Skeld
If I elect to have all my subscriptions shipped once a month (which is what I do elect), does that mean I'll probably receive PF#14 and PF#15 together if they both come out in October?
I realize there's a "probably" and an "if" in that question so I understand the answer won't be definitive.
I don't doubt that Paizo's staff do have a great deal on their collective plates. Most of it, I believe, is a direct result of forum outcry for additional materials. The forum community wants a deities book. The forum community wants player friendly books (Pathfinder Companion). Not to mention numerous other products from the Chronicles line. It turns out the user community here is pretty demanding.
I guess I don't see the quality issues you you see, but that's a discussion for another thread.
-Skeld
I agree that the product quality issues are a topic for another thread, but the issues with shipping and timeliness of those products are topics for this thread. The current delay in shipping is certainly not the first time Paizo has had shipping "irregularities" (the month in which everything arrived is one example) and this is the sort of thing I worry about continuing.
I agree that the forum community wants lots of cool shiny things: books, adventures, etc., but I see those more as requests than demands. It is up to the company to understand what is a reasonable amount of material to try to produce at a time and in what amount of time. The staff is certainly not controlled by the customers, without any judgment on its own part. I, personally, would rather have fewer, quality products, on time, than a glut of error-filled material on a schedule that can't be predicted.
O
You should also throw out the PF-14 data point (it's invalid due to custom's intervention).
So you should throw out extremal data points that conflict with your model? You have a great future working on Wall Street, I predict.
;-)
I, personally, would rather have fewer, quality products, on time, than a glut of error-filled material on a schedule that can't be predicted.
To follow up on Skeld's remark...
Why don't you start a new thread about Quality Control issues?
I propose this to you without sarcasm. Merely that we should differentiate the shipping schedule from quality control.
Also, such a thread would actually be helpful feedback and might lead to improvement (especially if it was drama free). I'm also curious what you see as a serious issue.
|
You should also throw out the PF-14 data point (it's invalid due to custom's intervention).So you should throw out extremal data points that conflict with your model? You have a great future working on Wall Street, I predict.
;-)
Actually, if an external event interferes with one point of your experiment, that data point should be discarded, i.e. if you were measuring the effect of hunters on wild deer and one year a disease decimated the population, that data point should possibly be excluded or it would bias the result. As Paizo do not control US customs and that is the cause for the spread with regards to Pathfinder 14, it should be excluded as it is biasing the experiment: Are Paizo shipping things on time?
Additionally Paizo have introduced new systems as they know that this has been a problem for them. It seems unfair to not give those systems a couple months before taking data points.
To follow up on Skeld's remark...Why don't you start a new thread about Quality Control issues?
I propose this to you without sarcasm. Merely that we should differentiate the shipping schedule from quality control.
Also, such a thread would actually be helpful feedback and might lead to improvement (especially if it was drama free). I'm also curious what you see as a serious issue.
I don't mean to sound like a big unhappy-pants. I have done my share of praising Paizo's product, and I truly do love their APs (since I don't have time to write my own stuff, and I much prefer a campaign path than individual adventures stapled together). However, ...
I don't think we need a whole new quality control thread. Just look at the errata/question/clarification threads that pop up regarding every product that Paizo puts out. Now, before people leap up and state, outraged, that "Errata just happen! They can't be prevented!" I should state that I understand that occasionally typographical errors happen. But the map errors, missing monsters, stat block errors, missing abilities, etc. that appear in every Paizo product show me that the editors and writers are working too quickly to meet deadlines that are unrealistic. Paizo needs to either hire more people to doublecheck for quality or push those deadlines back and publish less material. I realize that both these options cost Paizo money, and therefore probably lead to more expensive material (although it's hard to imagine the already overpriced Campaign Guide being any more expensive), but I believe that quality is worth price.
This is just me feeling, and I am reacting in the manner appropriate to me: 1) Voicing my thoughts in order to help a company whose products I appreciate and enjoy buying, 2) Speaking with my $ and canceling my subscription. My hope is that these two actions will have an effect, and I can start buying again at some point.
O
How, then, do you propose Paizo become immune to random Customs inspections?[ The current delay in shipping is certainly not the first time Paizo has had shipping "irregularities" (the month in which everything arrived is one example) and this is the sort of thing I worry about continuing.
My point is that this is NOT the first time that Paizo has had shipping issues. I don't claim that this current situation is their fault, but happens to be another circumstance in which I, as a subscriber, do not have the product that I'm supposed to have in the condition and at the time I'm supposed to have it.
O
I don't mean to sound like a big unhappy-pants.
Its cool. I'm mellow today. :) But what I mean is, I wasn't setting you up to get stomped on by me.
Now I can't say I'm a personal friend of James Jacobs, but I've met the guy and we've talked in the Paizo chatrooms and on the boards.. And he's always wanted to know quality control issues. The caveat to that is that he can't do anything about something in print, except to assist on the boards in coming up with errata (to which he is generous with his time).
But feedback on quality is useful as a preventative measure.
Anyway, I won't insist to you make a thread you don't feel needs to be made.. I just wanted you to know that I wasn't trying to make you feel like you had to defend yourself.
I love Paizo products too, but sometimes there are some glitches.
Actually, if an external event interferes with one point of your experiment, that data point should be discarded, i.e. if you were measuring the effect of hunters on wild deer and one year a disease decimated the population, that data point should possibly be excluded or it would bias the result. As Paizo do not control US customs and that is the cause for the spread with regards to Pathfinder 14, it should be excluded as it is biasing the experiment: Are Paizo shipping things on time?
From my point of view, I work on financial models for a bank and we're not allowed to ignore data points caused by external events when calculating return on investments or historical volatility (say). Otherwise there would be a lot of ignored data points recently. :-)
If Paizo's customs problem is a 1-in-a-thousand issue, then it ultimately wouldn't count for much of the variance.
|
|
Actually, if an external event interferes with one point of your experiment, that data point should be discarded, i.e. if you were measuring the effect of hunters on wild deer and one year a disease decimated the population, that data point should possibly be excluded or it would bias the result. As Paizo do not control US customs and that is the cause for the spread with regards to Pathfinder 14, it should be excluded as it is biasing the experiment: Are Paizo shipping things on time?From my point of view, I work on financial models for a bank and we're not allowed to ignore data points caused by external events when calculating return on investments or historical volatility (say). Otherwise there would be a lot of ignored data points recently. :-)
If Paizo's customs problem is a 1-in-a-thousand issue, then it ultimately wouldn't count for much of the variance.
True. But at the moment it's a 1 in 14 data point. That will skew the analysis something rotten. When operating with small numbers of data points, you need to keep the data as clean as possible.
|
RPG Superstar 2008 Top 32, 2011 Top 16
|
Actually, if an external event interferes with one point of your experiment, that data point should be discarded, i.e. if you were measuring the effect of hunters on wild deer and one year a disease decimated the population, that data point should possibly be excluded or it would bias the result. As Paizo do not control US customs and that is the cause for the spread with regards to Pathfinder 14, it should be excluded as it is biasing the experiment: Are Paizo shipping things on time?From my point of view, I work on financial models for a bank and we're not allowed to ignore data points caused by external events when calculating return on investments or historical volatility (say). Otherwise there would be a lot of ignored data points recently. :-)
If Paizo's customs problem is a 1-in-a-thousand issue, then it ultimately wouldn't count for much of the variance.
I think the difference is that for a financial model to predict ROI, everything is a valid data point, since ROI is a measure of profitability, which doesn't care about causes, it just cares about how much money you'll make (to the extent that the models care in the first place :) For an analysis that is looking at a process, then external causes of outliers should be excluded, since they are not part of the process (assuming that we're talking about the shipping SCHEDULE.) If we're talking about the shipping decisions such as to print overseas, then a customs inspection delay data point should be included. But if you want to analyse that, then you have to factor in a cost factor and decide if eliminating a 1 in a 1000 event is worth the increased printing costs of printing domestically.
As a price conscious customer, I'd say for me, I'll take my chances with the customs delays for a lower price.
How, then, do you propose Paizo become immune to random Customs inspections?
You get ahead of schedule and plan for delays. Plan to have the product delivered 2 weeks before it's due to be shipped. Then you can give customers the option to receive the product as soon as possible or no sooner than the scheduled delivery date.
The problem with this is that it means you potentially carry inventory at your expense for 2 weeks if there are no delivery delays. Most of Paizo's customers would prefer product delivered ASAP but some folks like the OP would probably want them on a more rigid schedule.
Personally... doesn't make much difference to me, this is the sort of reading I do when I get time. I like getting stuff as quickly as Paizo cranks it out. Delays don't bother me. It's not like I can run the stuff as fast as they release it.
|
Chief Technical Officer
|
If I elect to have all my subscriptions shipped once a month (which is what I do elect), does that mean I'll probably receive PF#14 and PF#15 together if they both come out in October?I'm wondering about this, too. I've heard Lisa, Cosmo, Vic, etc. say that they'll ship them separately, but without specifying whether that is for all subscribers or just "ship with each AP as fast as possible" subscribers.
If you have multiple subscriptions, and have elected to "Hold for monthly shipments," you will get them both together at the end of the month.
If you have "Hold for Planet Stories," they'll wait for Hounds of Skaith, which is due in after both Pathfinders.
If you have "Hold for Pathfinder," or "Don't hold anything," or you have only the Adventure Path subscription, we'll ship them to you separately, as they arrive.
|
|
I am just curious what dates are being used for counting as late.
When I look at the store, it says "Preorder Expected [Month]"
It doesn't say 1st day of the month, and it doesn't say shipping day X. It says it is expected during this month sometime.
So if a book says "Preorder Expected October", then I expect to see it sometime during October or even a week or so into November. I would only consider the book Late if it arrives more than two weeks after the expected month, and no one notified me there was a delay.
So, since we are getting into statistics and nit picking, just what exactly are you basing your standard of late on? and is it actually based upon information published by Paizo?
I don't think he's counting anything as late, mostly just noting an inconsistency in the amount of time between AP shipments.
It would be interesting to see how different it would be if he'd selected the shipping option to hold his volumes until the end of the month.
|
|
- Issue Days
- 1 --
- 2 36
- 3 33
- 4 36
- 5 35
- 6 14
- 7 42
- 8 35
- 9 21
- 10 42
- 11 23
- 12 25
- 13 0
- 14 59+
And based upon these numbers which do not actually include dates, when allowing for the 30+ day variance built into the expectations I find exactly twice they have been late, which admittedly is 14% of the time, and perhaps that is too high. But I can live with it.
So, honestly, the data is just too incomplete to draw any conclussions from it.
|
|
My point is quite simply, they are well within their advertised time periods. If they were outside those periods then I could see a valid complaint.
It seems to me the complaint is that product does not ship on exactly the 1st and 15th of each month or something along those lines.
My point is that Paizo has never claimed its shipping would be that precise. I am sure they would like it to be tighter. But they have never claimed it would be.
My mom used to tell me, "Wish in one hand, spit in the other. See which one gets full first."
Wish all we want, but they have fullfilled their stated delivery times very well.
|
Chief Technical Officer
|
On the original topic, we're well aware of the desire for more regularity. In fact, I'd go so far as to say that regularity is even more important to us than it is to you, and keeping things on a tight schedule is one of James Jacobs' highest priorities.
That said, there's an awful lot that's not in his control, like the customs inspection for Pathfinder 14, and printing times and shipping times in general. For example, our schedules assume that, once we turn the files over to the printer, it would take a certain number of weeks for them to print it, and then a few weeks in transit. Well, as our print runs have been increasing, the time it takes to print has also been increasing. (That's one reason that products have been late in the past.) So we're now building that extra time into the schedules, but schedules for products that are already in process were designed with old assumptions about printing times, so even products that were turned over on schedule are going to arrive later than we'd planned.
One thing we're doing to help this is to rework our schedules so that products arrive at the beginning of each month instead of the end of each month; now if things slide a week or two, they'll still at least come out in the same month. (Unfortunately, because we work so far in advance, that change won't have a visible effect until next year.)
|
Chief Technical Officer
|
Acually, because this is a Poisson Distribution and not a Normal or Uniform Distrution, the Standarad Deviation is closer to 5.3. You should also throw out the PF-14 data point (it's invalid due to custom's intervention).
-Your Friendly Neighborhood Statistics Nit-Picker
I have to disagree with your nitpick. I am not counting shipments, I am measuring days. For measurements, a normal distribution is appropriate. I could be more sympathetic to an argument of throwing out the PF14 data point as a statistical outlier (hmmm, no, I think perhaps systematic error might be a better reason ;-) but only if there is a change in process so it doesn't happen again. If they're following a process in which this remains a possibility, then whether or not it's in their control, it remains a valid data point.
Hey, hey, don't look at me that way... Some of us like discussing statistics.
Drew Garrett
I think what he's pointing out is that your method isn't ideal for the task because of data like this:
6 14
7 42
If we knew we were going to be measured exclusively by your method, we could have delayed shipment of number 6 by two weeks so that we could make the data look like this:
6 28
7 28
That would bring the deviation down significantly, but it wouldn't really serve most customers better.
If you have multiple subscriptions, and have elected to "Hold for monthly shipments," you will get them both together at the end of the month.
If you have "Hold for Planet Stories," they'll wait for Hounds of Skaith, which is due in after both Pathfinders.
If you have "Hold for Pathfinder," or "Don't hold anything," or you have only the Adventure Path subscription, we'll ship them to you separately, as they arrive.
Thanks a lot for the information.
My concern wasn't for the sake of idle curiosity, or because I'm chomping at the bit (though I am excited), but because I want to make sure I have money in the bank to pay for the stuff when it does ship since stuff is charged when it ships.