Quote:
Originally Posted by Peter_Spaeth
It's interesting that when Snow has a point to make, it's OK to use a very simple correlation -- more eyes equals higher prices. But when someone else has a point, such as that examination of a reasonable sampling of high dollar auctions could be indicative of hanky panky, HIS anaylsis is way too simplistic to prove anything and is meaningless buffoonery absent a full deep data dive. Hmmmm.
Let me guess, Snow will say apples to oranges.
Or maybe he'll say that since he alone understands data, he alone can determine how much is needed for any given question.
|
Yes, I see this as being an apples to oranges comparison. I don't believe that anyone here would honestly argue that having more eyes on an auction would not bring about higher prices. I think anyone arguing to the contrary is simply being disingenuous. I don't believe for one second that you don't accept that premise as true. Surely, the relationship is not linear though, as there would be an element of diminishing returns (i.e., having 40 million eyes on an auction is effectively equivalent to having 5 million eyes on that auction, but having 40,000 is still significantly better than having 5,000).
I shouldn't need to provide data to support that claim. It is obvious. However, I could certainly prove that it's true if I cared to mine the data. Maybe I'll do that for fun in my spare time. It could be an interesting project to understand and measure the impact that it actually has. It would probably only take a few hours to do. The most difficult and time-consuming part would be building the dataset. Doing the math from there is the easy part.
Quote:
Originally Posted by BobC
I know what you are saying and understand exactly what he does. I did point out his own posted figures and asked why his own evidence doesn't support his statement that PWCC has the highest prices because Probstein actually has more followers. And there is no reason to go easy on him as he is obviously knowledgable, just doesn' t really say anything that basically ends up as anything other than he's right and you're wrong. Same thing you're pointing out. And just because he's an engineer he gets no pass from me. I've worked with many engineers and architects in the past, and engineers are typically very knowledgable and pay attention to details and facts. And my daughter is an engineer as well. Feel I'd be letting them all down if I didn't call him out.
|
For the record, I have always claimed that Probstein auction also outsell the competition. I don't know about PWCC vs Probstein though, but from my readings of the various forums and other online communities, most people share this belief that both Probstein and PWCC sell for more than the others. Many of the data tools out there even have options to exclude PWCC and Probstein listings from their comps because of it.
I'm not an engineer. I'm also not a data analyst or a computer programmer or a developer. There is a bit of overlap with my field to each of those skills, but a data scientist, in general, is someone with multiple degrees in either mathematics, statistics, computer science, or physics. Typically, they will have majored in at least one of those 4 disciplines in undergrad and then will have a graduate degree in at least one or more of these fields as well. There are many sub-specialties within the data science space, but my primary area of expertise is in building predictive models. I basically use complex mathematics to teach a computer to learn how to solve specific problems.
When I get the time, I will explain in more detail why this skill set is relevant to the problem of identifying shill bidders, and why I claim it would yield a far superior solution to that problem than hiring a team of data analysts or accountants to do it manually.