|
 |
| | | Author: | RuiBraga  | | Posted: | Apr 11, 2022 04:27 | | Subject: | Ridiculous prices \ AVG Price's broken | | Viewed: | 324 times | | Topic: | Price Guide | |
|
| Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
|
| | | | | |
| | | | | Author: | LeeGo73 | | Posted: | Apr 11, 2022 05:43 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 113 times | | Topic: | Price Guide | |
|
| Looks like seller has entered prices in USD instead of RUB
In Price Guide, RuiBraga writes:
| | Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
|
|
| | | | | | | | | |
| | | | | | | Author: | RuiBraga | | Posted: | Apr 11, 2022 06:02 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 152 times | | Topic: | Price Guide | |
|
| Probably.
I've sent him a msg.
Let's wait.
In Price Guide, LeeGo73 writes:
| | Looks like seller has entered prices in USD instead of RUB
In Price Guide, RuiBraga writes:
| | Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
|
|
|
| | | | | | | | | | | | | |
| | | | | | | | | Author: | tons_of_bricks | | Posted: | Apr 11, 2022 12:31 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 97 times | | Topic: | Price Guide | |
|
| In Price Guide, RuiBraga writes:
| | Probably.
I've sent him a msg.
Let's wait.
In Price Guide, LeeGo73 writes:
| | Looks like seller has entered prices in USD instead of RUB
In Price Guide, RuiBraga writes:
| | Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
|
|
Nothing else to do. It's probably an error like LeeGo73 mentioned. But if
not, he's free to price his items whatever he wants, no matter how ridiculous
they may be.
|
|
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | Author: | RuiBraga | | Posted: | Apr 11, 2022 14:19 | | Subject: | (Cancelled) | | Viewed: | 61 times | | Topic: | Price Guide | |
|
| | (Cancelled) |
|
| | | | | |
| | | | | Author: | WildBricks | | Posted: | Apr 11, 2022 12:21 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 109 times | | Topic: | Price Guide | |
|
| If you look at all of their sets the prices are crazy high.
I was the kid who quit ballet because it interfered with watching Scooby Doo,
but not even I would pay $12k U.S. for one of the sets. 
In Price Guide, RuiBraga writes:
| | Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
|
|
| | | | | |
| | | | | Author: | Sadler_Bricks | | Posted: | Apr 12, 2022 17:05 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 102 times | | Topic: | Price Guide | |
|
| Could be a mistake. But besides this i have seen other stores do this. The prices
have been broken for a good while in my opinion.
Sadler_bricks
|
|
| | | | | | | | | |
| | | | | | | Author: | Gaston.La.Brick | | Posted: | Apr 13, 2022 09:35 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 50 times | | Topic: | Price Guide | |
|
| In Price Guide, Sadler_Bricks writes:
| | Could be a mistake. But besides this i have seen other stores do this. The prices
have been broken for a good while in my opinion.
Sadler_bricks
|
Correct.
There have been (and will be) numerous discussions on how "Average Price" is
calculated. Currently, it's the mathematical correct way.
As I mentioned before, the correct way for it's intended purpose would
be to exclude the lowest and highest 10% of the prices. This would filter out
the extremes (both downwards and upwards). A new calculated variable "Avg. Price
Percentile (10th-90th)" could be introduced.
|
|
| | | | | | | | | | | | | |
| | | | | | | | | Author: | yorbrick | | Posted: | Apr 13, 2022 14:39 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 47 times | | Topic: | Price Guide | |
|
| In Price Guide, Gaston.La.Brick writes:
| | In Price Guide, Sadler_Bricks writes:
| | Could be a mistake. But besides this i have seen other stores do this. The prices
have been broken for a good while in my opinion.
Sadler_bricks
|
Correct.
There have been (and will be) numerous discussions on how "Average Price" is
calculated. Currently, it's the mathematical correct way.
As I mentioned before, the correct way for it's intended purpose would
be to exclude the lowest and highest 10% of the prices. This would filter out
the extremes (both downwards and upwards). A new calculated variable "Avg. Price
Percentile (10th-90th)" could be introduced.
|
Why exclude the extreme 10%, what is correct about that? Why not 2%, 5%, 16.67%
or 49%?
The intended purpose is to report the mean, which is what it does.
|
|
|
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | Author: | Gaston.La.Brick | | Posted: | Apr 13, 2022 15:08 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 46 times | | Topic: | Price Guide | |
|
| | | | | As I mentioned before, the correct way for it's intended purpose would
be to exclude the lowest and highest 10% of the prices. This would filter out
the extremes (both downwards and upwards). A new calculated variable "Avg. Price
Percentile (10th-90th)" could be introduced.
|
Why exclude the extreme 10%, what is correct about that? Why not 2%, 5%, 16.67%
or 49%?
The intended purpose is to report the mean, which is what it does.
|
I don't agree. The mathematical value is indeed the mean (I think, English
is not my native language). However, I do think the intended purpose was not
to give the mean, but to give an indication to both buyer and seller what a realistic,
average price is for which an item has been sold and is on sale now. I can't
be sure, I wasn't there when it was discussed. As long as extremes are in
that equation, the meaning behind that value has lost its purpose.
Do note I didn't say to change the value of the "Average" variable. It should
remain the mathematical correct one. But instead introduce a new one ("X Percentile
Average", or what you want to call) that actually has a better meaning to it's
value. When you ask what's correct about that: it is an statistically correct
and commonly used way to filter out extreme values polluting the meaning or trend
value. It's also referred to as "removing outliers".
https://www.statology.org/remove-outliers/
https://statisticsbyjim.com/basics/remove-outliers/
On the exact percentage: could be 1%, 2%, 10%. Or it could be filter out the
bottom 3 and top 3 values. Whatever would be decided (I'm not statisticus),
it would give me a better meaning to some of the values.
|
|
|
| | | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | Author: | yorbrick | | Posted: | Apr 13, 2022 16:28 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 51 times | | Topic: | Price Guide | |
|
| In Price Guide, Gaston.La.Brick writes:
| | | | | | As I mentioned before, the correct way for it's intended purpose would
be to exclude the lowest and highest 10% of the prices. This would filter out
the extremes (both downwards and upwards). A new calculated variable "Avg. Price
Percentile (10th-90th)" could be introduced.
|
Why exclude the extreme 10%, what is correct about that? Why not 2%, 5%, 16.67%
or 49%?
The intended purpose is to report the mean, which is what it does.
|
I don't agree. The mathematical value is indeed the mean (I think, English
is not my native language). However, I do think the intended purpose was not
to give the mean, but to give an indication to both buyer and seller what a realistic,
average price is for which an item has been sold and is on sale now. I can't
be sure, I wasn't there when it was discussed. As long as extremes are in
that equation, the meaning behind that value has lost its purpose.
Do note I didn't say to change the value of the "Average" variable. It should
remain the mathematical correct one. But instead introduce a new one ("X Percentile
Average", or what you want to call) that actually has a better meaning to it's
value. When you ask what's correct about that: it is an statistically correct
and commonly used way to filter out extreme values polluting the meaning or trend
value. It's also referred to as "removing outliers".
https://www.statology.org/remove-outliers/
https://statisticsbyjim.com/basics/remove-outliers/
On the exact percentage: could be 1%, 2%, 10%. Or it could be filter out the
bottom 3 and top 3 values. Whatever would be decided (I'm not statisticus),
it would give me a better meaning to some of the values.
|
None of that is better, just different. There is usually a very important thing
if you remove data from any analysis because you believe it doesn't fit -
you report what you remove and why, and still report the true mean before you
bias it by removing data.
Filtering out things like the bottom and top three results is not a good idea.
For items where many sales are made, it makes little to no difference. Where
only a few sales are made, it can have a huge effect. For example, if 10 sales
are made, you throw away 60% of your data. If just 7, you are taking the median.
For six or fewer sales, you throw away all the data.
|
|
|
| | | | | | | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | Author: | Gaston.La.Brick | | Posted: | Apr 14, 2022 03:36 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 29 times | | Topic: | Price Guide | |
|
| In Price Guide, yorbrick writes:
| | In Price Guide, Gaston.La.Brick writes:
| | | | | | As I mentioned before, the correct way for it's intended purpose would
be to exclude the lowest and highest 10% of the prices. This would filter out
the extremes (both downwards and upwards). A new calculated variable "Avg. Price
Percentile (10th-90th)" could be introduced.
|
Why exclude the extreme 10%, what is correct about that? Why not 2%, 5%, 16.67%
or 49%?
The intended purpose is to report the mean, which is what it does.
|
I don't agree. The mathematical value is indeed the mean (I think, English
is not my native language). However, I do think the intended purpose was not
to give the mean, but to give an indication to both buyer and seller what a realistic,
average price is for which an item has been sold and is on sale now. I can't
be sure, I wasn't there when it was discussed. As long as extremes are in
that equation, the meaning behind that value has lost its purpose.
Do note I didn't say to change the value of the "Average" variable. It should
remain the mathematical correct one. But instead introduce a new one ("X Percentile
Average", or what you want to call) that actually has a better meaning to it's
value. When you ask what's correct about that: it is an statistically correct
and commonly used way to filter out extreme values polluting the meaning or trend
value. It's also referred to as "removing outliers".
https://www.statology.org/remove-outliers/
https://statisticsbyjim.com/basics/remove-outliers/
On the exact percentage: could be 1%, 2%, 10%. Or it could be filter out the
bottom 3 and top 3 values. Whatever would be decided (I'm not statisticus),
it would give me a better meaning to some of the values.
|
None of that is better, just different. There is usually a very important thing
if you remove data from any analysis because you believe it doesn't fit -
you report what you remove and why, and still report the true mean before you
bias it by removing data.
Filtering out things like the bottom and top three results is not a good idea.
For items where many sales are made, it makes little to no difference.
|
100 items are sold for $1.00
1 item is sold for $100.00
Mean or Average = $1.98, which is 200% of the expected and actual meaningful
average.
Let's say I use this for setting my selling price. I will be the most expensive
one on BrickLink (except for the $100.00 one of course), but I will believe I'm
offering at the average price, thinking I offer at a competitive price.
I understand you will never even consider this, but for me the average value
has no purpose this way, as I can't use it for deciding if an offered price
is actually competitive or not. Without of course checking item per item, by
manual effort.
Where
| | only a few sales are made, it can have a huge effect. For example, if 10 sales
are made, you throw away 60% of your data. If just 7, you are taking the median.
For six or fewer sales, you throw away all the data.
|
Why 60% is thrown away? As I said, the exact formula should be well thought.
I could be top/bottom 10%, so for 10 sales only 20% is thrown away. And even
then, it shouldn't always be that case, but only for situation where there
is a very large deviation at the bottom and top prices.
Of course the algorithm needs to be a bit more intelligent/smart, as to decide
when to drop which outliers. Again, I'm not a statisticus, but I'm sure
there are intelligent formulas that cover these problems to deliver actual usable
data.
|
|
|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | Author: | yorbrick | | Posted: | Apr 14, 2022 03:52 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 40 times | | Topic: | Price Guide | |
|
| | |
100 items are sold for $1.00
1 item is sold for $100.00
Mean or Average = $1.98, which is 200% of the expected and actual meaningful
average.
Let's say I use this for setting my selling price. I will be the most expensive
one on BrickLink (except for the $100.00 one of course), but I will believe I'm
offering at the average price, thinking I offer at a competitive price.
|
It is obviously fake data, such wide variations in sold data are highly unlikely.
Especially as these are sold prices, they are meaningful. That someone paid $100
presumably means it was a genuine sale. Of course we don't know what was
correct, was it a seller pricing high and getting a buyer willing to pay 100x
too much (unlikely), or was it a seller error, listing 100 at $1 instead of 1
at $100. The latter would likely be snapped up instantly.
This made up data shows one thing though - know what you sell. If you had a $100
item and someone else messed up and listed it as 100 at $1, would you blindly
follow their error price because that is what the biased average is, or would
you check sales data? Either way the $1 or $2 average should seem way off if
you understood the value of the item.
You won't be the only one selling at that price. Why would you be the only
person on bricklink that uses the average sold price? If you are, they should
remove the ability to populate with price guide data, as it is not being used
by any other seller. More likely, many other people will be listing at the same
average.
|
|
|
| | | | | |
| | | | | Author: | Yo_Yo_Flamingo | | Posted: | Apr 13, 2022 09:46 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 49 times | | Topic: | Price Guide | |
|
| In Price Guide, RuiBraga writes:
| | Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
The price guide only takes into account sold sets, and as nobody in their right
mind would buy these, it doesn't affect the average.
|
|
| | | | | | | | | |
| | | | | | | Author: | tons_of_bricks | | Posted: | Apr 13, 2022 10:15 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 44 times | | Topic: | Price Guide | |
|
| In Price Guide, Yo_Yo_Flamingo writes:
| | In Price Guide, RuiBraga writes:
| | Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
The price guide only takes into account sold sets, and as nobody in their right
mind would buy these, it doesn't affect the average.
|
There's two lists on the price guide, "sold" items and "for sale" items.
So while these listing don't mess up the "sold" average, they greatly change
the "for sale" average (though I don't ever look at that anyways).
|
|
|
| | | | | | | | | | | | | |
| | | | | | | | | Author: | Gaston.La.Brick | | Posted: | Apr 13, 2022 11:15 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 51 times | | Topic: | Price Guide | |
|
| | | There's two lists on the price guide, "sold" items and "for sale" items.
So while these listing don't mess up the "sold" average, they greatly change
the "for sale" average (though I don't ever look at that anyways).
|
For items that aren't sold that much (and have a larger number of items in
the "for sale" as in the "sold" category), you often need to rely on the "for
sale" average price.
And even for the "sold" values, they sometimes include extremely low or high
values. That's of course fine, but it makes the mathematical average useless
when it's being used to base your own price upon.
Let's say an item is sold 10 times for 1.00 dollar but once for 10 dollars.
The current (mathematical correct) average would be 9.90$ (9x 1.00$ + 1x 10.00$
= 99.00$ / 10 items).
Better would be to exclude the extremes, which would result in an average of
1.00$ as the sale for 10.00$ is omitted.
|
|
|
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | Author: | Yo_Yo_Flamingo | | Posted: | Apr 14, 2022 00:28 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 54 times | | Topic: | Price Guide | |
|
| In Price Guide, Gaston.La.Brick writes:
| | | | There's two lists on the price guide, "sold" items and "for sale" items.
So while these listing don't mess up the "sold" average, they greatly change
the "for sale" average (though I don't ever look at that anyways).
|
For items that aren't sold that much (and have a larger number of items in
the "for sale" as in the "sold" category), you often need to rely on the "for
sale" average price.
And even for the "sold" values, they sometimes include extremely low or high
values. That's of course fine, but it makes the mathematical average useless
when it's being used to base your own price upon.
Let's say an item is sold 10 times for 1.00 dollar but once for 10 dollars.
The current (mathematical correct) average would be 9.90$ (9x 1.00$ + 1x 10.00$
= 99.00$ / 10 items).
Better would be to exclude the extremes, which would result in an average of
1.00$ as the sale for 10.00$ is omitted.
|
I agree with your position of excluding outliers from the calculation, but the
engineer in my is going to drive me mad until I correct that math.
in that example, 10$ for one item + $10 for another ten at a buck a piece = $20/11,
for an average of $1.81.
|
|
|
| | | | | | | | | | | | | | | | | |
| | | | | | | | | | | Author: | yorbrick | | Posted: | Apr 14, 2022 03:35 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 30 times | | Topic: | Price Guide | |
|
| | |
Let's say an item is sold 10 times for 1.00 dollar but once for 10 dollars.
The current (mathematical correct) average would be 9.90$ (9x 1.00$ + 1x 10.00$
= 99.00$ / 10 items).
Better would be to exclude the extremes, which would result in an average of
1.00$ as the sale for 10.00$ is omitted.
|
Your calculation is wrong. 9x1 + 10 = 19, not 99. The average would be $1.90.
Yes you could exclude the high value, but why is that better? You have excluded
useful data. Those 9 $1 sales could all be from the same seller that undervalued
the item, or purposely manipulated the price with fake sales.
|
|
| | | | | |
| | | | | Author: | legoman77 | | Posted: | Apr 14, 2022 13:46 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 36 times | | Topic: | Price Guide | |
|
| In Price Guide, RuiBraga writes:
| | Hello guys,
Can someone check this store 613 Bricks (https://www.bricklink.com/store.asp?p=613)
The prices are insanely high and it will break the purpose of the avg prices.
Example:
Queen Anne's Revenge
Lot ID: 288191095 Item No: 4195-1
Price:
US $64,990.00
(~EUR 59,700.8515)
Please investigate.
Thanks.
|
I never used the average price function. I pulled up what was listed and priced
accordingly. Price depended on whether I wanted to be the lowest price and sell
quickly or somewhere in the middle if I wanted to reply on my reputionation and
feedback and no rush to sell.
The price guide suck in my opinion. How do you use it to value a set with defects
or was the AVG price had all defected sets and yours was not? Look at the current
listings.
John P
|
|
|
| | | | | | | | | |
| | | | | | | Author: | 1001bricks | | Posted: | Apr 14, 2022 13:52 | | Subject: | Re: Ridiculous prices \ AVG Price's broken | | Viewed: | 67 times | | Topic: | Price Guide | |
|
| | | The price guide suck in my opinion. How do you use it to value a set with defects
or was the AVG price had all defected sets and yours was not? Look at the current
listings.
|
Price guide is like the Argus for automobile.
It doesn't mean ANYTHING but it's very helpful.
Hoping it helps - even if I doubt it does
|
|
|
|
|
Portugal
Netherlands, Zuid-Holland
USA, Missouri
Belgium
United Kingdom, England
France, Provence-Alpes-Côte d'Azur