[whojammyflip]

Just a suggestion, rather than having the simple mean as the average price in the review section, what about having something like the median so that outliers are removed. I've just seen a review posted in the M 1.4/50 where its obvious the reviewer has never tried the lens and has spammed the price with a low ball value.

Often, I use the prices on the forum as a guide, but I always end up doing something like taking the prices individually and weighting them according to the number of posts of the member. On the K2.5/135 it gives a price of USD 144 instead of the simple mean of USD 200.

I don't like putting prices in myself, as I tend to low ball on Ebay, so the value on the forum would get skewed.

[Ira]

You basically have to ignore the prices in that section, for the older lenses anyway.

Most of the reviews are old, and the prices have skyrocketed over the last 6 months.

For example, I paid around 35 bucks for my Super-Tak 50 1.4 a year ago, and that was an average price.

Now, you're lucky to get one for under $100.

[interested_observer]

At first blush it seems like a reasonable idea. However, you need to consider all the potential situations.

• A lens with just a couple of prices. Small sample.
• A lens with a reasonable number of samples, but some one who purchased the lens in 1978 and posted that price.
• A price from someone new to the board with just a few postings.
• How about time based weighting, where the last price posted would reflect what the current market is - but what if that again was a 1978 price?

Then there is the case of the static price, but the number of postings are dynamic, so based on some weighting, the "average" price would need to be be recalculated on each viewing.

So, in the end there really is not a best way to post an average price, as they all have their own pit falls.

[GeneV]

I really don't see a way to make the average prices relevant. The market changes all the time. Median prices won't really help when many of the lenses were purchased 20 or 30 years ago. It might help to indicate when the lens was purchased.

[dgaies]

As other have suggested, the average price in the lens review section has very little meaning in terms of correlation to current prices and/or fair market value.. This is in part due to the recent increase in Pentax glass prices across the board, but also due to many other factors (such as small sample size, inclusion of below market value "great deals", combining both used and new prices, etc.) which make it impossible to use the average price for anything meaningful.

[whojammyflip]

OK, what about a Kalman filter estimation of price, just kidding. Just that if an average price is being calculated somewhere in the html jungle, then a median would be just as easy (instead of sum/N its the (N+1)/2 price when the prices are ranked) to calculate, and would filter out moronic prices being occasionally slapped in.

You basically have to ignore the prices in that section, for the older lenses anyway.

Most of the reviews are old, and the prices have skyrocketed over the last 6 months.

For example, I paid around 35 bucks for my Super-Tak 50 1.4 a year ago, and that was an average price.

Now, you're lucky to get one for under $100.

This could be something to do with the Fed, the Weimar republic inflation and the USD devaluation. Lenses could be the ultimate inflation/currency devaluation hedge as many of the buyers are in the far east. I collected a MX off a chap locally recently, and he told me he sells most of his camera kit to buyers in the far east for relatively high prices (but thats because he is selling in Sterling, and Merv the swerve King is trying to emulate Benny B. I digress.)

[dgaies]

. Just that if an average price is being calculated somewhere in the html jungle, then a median would be just as easy (instead of sum/N its the (N+1)/2 price when the prices are ranked) to calculate, and would filter out moronic prices being occasionally slapped in.

Since most lenses have between 5-20 reviews, just scrolling down the list gives people a decent idea of the prices that people paid and from there can "weight" them as they see fit based on the reviewer's description of whether they bought it new, got a great deal, etc.

If people were really interesting in determining current market value for used lenses, I suppose you could add a couple line to the marketplace feedback process that would allow you to enter the final price paid for a lens and link it directly to the lens data in the review section. That way a running list of the prices lenses sold for, and the date they sold, would be more easily accessible. Obviously this would only capture lenses being sold on pentax forums, but you could also make a provision to add price data for outside transactions, even if the buyer didn't want to leave a full review.

[Ira]

This could be something to do with the Fed, the Weimar republic inflation and the USD devaluation. Lenses could be the ultimate inflation/currency devaluation hedge as many of the buyers are in the far east.

Nah:

It's mostly old farts like me who love the quality and manual experience from our youth.

But it's getting to the point with prices for many of the old Taks that they just aren't going to be worth it compared to the new stuff.

Who's gonna pay $400 for an S-Tak 18mm Fisheye, when you can get a DA 10-17 for around $500? Especially when that full frame 18 is in effect a 27 on an APS-C?

[whojammyflip]

If people were really interesting in determining current market value for used lenses, I suppose you could add a couple line to the marketplace feedback process that would allow you to enter the final price paid for a lens and link it directly to the lens data in the review section. That way a running list of the prices lenses sold for, and the date they sold, would be more easily accessible. Obviously this would only capture lenses being sold on pentax forums, but you could also make a provision to add price data for outside transactions, even if the buyer didn't want to leave a full review.

Sure, but my idea was to keep it simple, just by changing the calculation from an arithmetic mean to a median. Nothing needs changing other than that, and it will filter outliers. I can stick a price in of 100k at the moment for a lens and it will lift the arithmetic mean massively, whereas it has no effect on the median.

Interestingly, you can enter the price in orders of magnitude, like 1+e20, which gives average prices like this:
https://www.pentaxforums.com/lensreviews/showproduct.php?product=84

Will remove that shortly, its just there to show its possible for an individual to distort the prices shown in the forums.

EDIT: I've deleted my 1e20 price from the M 1.7/50 review, but for a short while, the lens value represented more than the entire cash in the world (probably)

[Stringmike]

The median would be a lot more accurate indicator of prices paid than an average and is very easy to compute. Accurate historical data is still useful, even when prices are rising.

Mike

[gofour3]

This web site tracks lens selling prices from various mfgs.

jcolwell.ca - home

Phil.

[paperbag846]

I have an idea... price by year posted.

I mean, any reader can do this him/herself, but it gives the buyer a rough estimate of what they should expect to pay for the lens in the current market.

E.g. show:

-Avg. / Median Price
-Avg. / Median Price for Past 12 months (should display number of reviews included).

It would be a simple extra calculation that would sample all of the reviews less than 1 year old, and give buyers a more accurate idea of what a lens is worth these days. If you show both, then readers can see if the value of the lens is decreasing or increasing with respect to the average price paid.

I prefer the median idea, but that is a different debate.

I find Ebay a bit of a circus... and now Craigslist is going that way as well...

[dgaies]

Sure, but my idea was to keep it simple, just by changing the calculation from an arithmetic mean to a median. Nothing needs changing other than that, and it will filter outliers. I can stick a price in of 100k at the moment for a lens and it will lift the arithmetic mean massively, whereas it has no effect on the median.

I understand, and agree, that using a median would create a more stable measurement that isn't (as) affected by high and low outliers.

And in the end, I wasn't actually suggesting that it become more complicated, was more just thinking out loud.

I just know that having bought and sold 100+ lenses in the last 12 months, the average prices in the review database don't reflect the true market price for the majority of lenses.

[openyourap]

I agree with the general sentiment:

• Median - would be the biggest improvement with the least effort.
• Time weighted median - would more acurately reflect the market, but only where there are sufficient reviews
• Marketplace median - would be ace, but hard to impliment, as the marketplace isn't cross referenced to the lens database

[Jewelltrail]

While we are at it, using the median for the score of a lens would be a great idea. It really is silly that one person can score a lens with a 2 or a 3 when mostly everyone else in the scoring is giving 9s--but this happens, often. Also, it seems these very low scores are often submitted by people with very few posts.

Using the median here would be a great idea. The mean is unduly affected by the extremeley low scores in the lens database.