[gress]

You seem to be ignoring the meaning of average. It's a statistic that you can create to characterize a numeric property of any well defined category. If you agree that there are laptops, and that there are servers as categories, then we can take a sample of each category, and calculate an average HDD size for each, regardless of the standard deviation.

It is true by definition that there are average HDD sizes for servers and laptops. If you disagree with their estimates, do you have reasoning to show that they are wrong?

[joshstrange]

I don't discount that there EXISTS an average but I don't think it would line up at all with their numbers. I fully understand the meaning of "average" and I also understand that there are a LARGE number of times that the average is completely useless due to the input data. In most of those cases it's better to use the MEDIAN not the MEAN (average). For example if you are looking at the following data set:

  $20,000
  $30,000
  $25,000
  $35,000
  $100,000 

The mean is $42,000 but the median is $30,000 which is a better representation of the data IMHO. Same with servers, you have servers that are massive for storing data and small <10GB App servers that sit behind a LB. What I'm trying to say is the numbers don't make sense and it's a BAD comparison. It would be a little better if they gave numbers for either laptops or servers but they way they have it phrased they just as well could have said "They seized 35 apples that is equivalent to 100 oranges". Lastly while there EXISTS averages no doubt I challenge you to FIND these averages, a quick couple google searches produces no such knowledge. Not to mention that while an average exists I believe it to be incalculable unless you want to use sampling due to the fact that the VAST majority of companies do not share that type of information.

[gress]

Do you think the median HDD for servers is larger than the median HDD size for laptops?

[joshstrange]

I'm actually not really sure. I'd probably say that for app servers the HD size for laptops is greater than that for servers but that's my core problem with the comparison. Both servers and laptops vary wildly in HD sizes. Backblaze has hundreds of TB's in one box whereas my digital ocean VM is 20GB. < 100GB HD for laptops would be ludicrous but for servers you only want to allocate what you know you will use which for everything I've ever worked with can be done in under 20GB easy if not under 10GB.

[gress]

If you aren't sure then you are saying that you don't understand the situation well enough to have a view.

You can reasonably ask that they state their assumptions or clarify their categories, but beyond that it seems to me that you just don't know enough to critique their illustration.

[joshstrange]

Um, I know what I'm saying. I'm saying that a Laptop HD to Server HD comparison is stupid. It is and I feel that I have more than enough knowledge to make that statement and stand but it. I'm unsure as to what I've said that makes you think that I don't know what I'm talking about. Their illustration is bad and I called that out. I stand by that and if people want to call it pedantic then that's their right.

What I DO know is that it is 100% IMPOSSIBLE for ANYONE to calculate the median HD size for laptops vs servers. There simply DOES NOT exist the data needed to do the calculations. Period. Not sure how not knowing an impossible to know stat somehow makes me unqualified to discuss this topic. Using your logic I'm expecting a response with the median sizes as you must possess this knowledge or else you wouldn't have commented given your statement.

[gress]

Since nobody has claimed that these numbers are precise, it's obvious that estimation is involved.

If you are claiming that estimation is 100% impossible, then you are indeed unqualified to discuss this topic.