Retired Sets by Month / Year Data - any help would be appreciated

[scatttcat]

Hi,

I have a quick question. Is there any form of data that has the past 3 years (lets say to 2014) of which sets retired each year? Like sets from 2014, 2015, etc.

I have searched here and I know we have a running list for the 2016 sets, but I was wondering if anyone had an excel sheet they'd possible be willing to share that went back further. I am basically looking for sets that have retired since 2014 and each year they've retired in (down to the month would be even better!). I will start building moving forward, but for the past 3 years I was just hoping something out there existed.

I know I can check brickset for the "last sold date" and get a sets retirement time frame, but I am more looking for somewhere that tracks them by month. Doesn't seem to be a comprehensive website for that unfortunately. Closest thing I found was retiring sets.com but that is really just blog posts, not a database that can be filtered by year.

Hopefully someone has something current on here who has also tracked these. Just a simple spreadsheet would be amazing! Something to get me started!

Thanks! ?

[Pseudoty]

@Grolim did this. I think someone has  updated version in the performance analysis thread perhaps @Val-E

https://community.brickpicker.com/applications/core/interface/file/attachment.php?id=3544

[scatttcat]

@Val-E I saw the other thread you started with 2016, and this has sets to 2014 and earlier... I guess I am just missing 2015

Any help?

[Val-E]

2 hours ago, scatttcat said:

@Val-E I saw the other thread you started with 2016, and this has sets to 2014 and earlier... I guess I am just missing 2015

Any help?

Yeah, I have some stuff. I´ll look for it when I get home from work.

[feed]

 

Something like this?
I did it for a while for SH and SW, but stopped, as i didn't find it really helped all that much.  Whilst it was useful in understanding price bracketing and to some extent marketing strategy and value (GBP/USD/EUR don't align so it's possible to spot under/over regional pricing – and where discounts are likely to occur).  What it didn’t show was set availability as Lego EOL dates are largely meaningless with stores in the UK holding EOL stock at RRP for up to a year after the Lego EOL date.  

Edit to add: I pulled all the data from brickset manually 
 

Edited May 27, 2017 by feed

[fossilrock]

Your best bet is to go on brickset.com and look through the sets.  They have the retirement data from when they dissappeared from lego.com.  It's been very accurate too so they are your best option to collect that.  

I started one of these as well, but gave up.  There's just too much to do it manually.

[scatttcat]

 
Something like this?
I did it for a while for SH and SW, but stopped, as i didn't find it really helped all that much.  Whilst it was useful in understanding price bracketing and to some extent marketing strategy and value (GBP/USD/EUR don't align so it's possible to spot under/over regional pricing – and where discounts are likely to occur).  What it didn’t show was set availability as Lego EOL dates are largely meaningless with stores in the UK holding EOL stock at RRP for up to a year after the Lego EOL date.  
Edit to add: I pulled all the data from brickset manually 
 

Your best bet is to go on brickset.com and look through the sets.  They have the retirement data from when they dissappeared from lego.com.  It's been very accurate too so they are your best option to collect that.  

I started one of these as well, but gave up.  There's just too much to do it manually.

Thanks both of you! My goal is more to just track the sets that retired and track the growth each month, not necessarily how long they have been retired.

Problem is, it is crazy how many websites DONT have this. Like brickset has the last sold date, but that doesn't even have the ability to be filtered. Kind of ridiculous in my opinion, I feel like date retired is a decent stat to keep track of with collectibles.

I have also noticed this will ultimately take a lot of manual entry, I just wish there was an easy way to grab all retired sets for let's say, 2015.

I reached out to brickset, maybe there is some trick I didn't think of.

Anyways, thanks for the help!

[Pseudoty]

Yeah the brickset date is not that helpful IMO unless it is a set like the EEE that only was available from S @ h and the lego store in the USA IIRC. Then you have store exclusives that can sit on the shelf for a year after retirement at half off MSRP and still end up doing well like this 

https://www.amazon.com/Star-Wars-Palpatines-Arrest-Exclusive/dp/B0088Y01S6/ref=sr_1_1?s=toys-and-games&ie=UTF8&qid=1495941525&sr=1-1&keywords=lego+palpatines+arrest

[c_rpg]

If you just want to see how fast a set grows after retirement it's generally adequate to keep track of price data and quantity sold per month. This will give you a pretty good indication of when the set became unavailable. You'll see a spike in sold quantity around the point it becomes unavailable and generally a slight increase in price post retirement. Even just price data might be enough for most sets.

Here are some examples. Sydney Opera House on the left and Wall-E on the right:

       

It's not always very clear, but for 90% of sets it is enough. For example, Scooby's Mansion is a bit less clear because there weren't as much sales and the peak is spread out over several months:

The data is from bricklink.

[scatttcat]

If you just want to see how fast a set grows after retirement it's generally adequate to keep track of price data and quantity sold per month. This will give you a pretty good indication of when the set became unavailable. You'll see a spike in sold quantity around the point it becomes unavailable and generally a slight increase in price post retirement. Even just price data might be enough for most sets.
Here are some examples. Sydney Opera House on the left and Wall-E on the right:
       
It's not always very clear, but for 90% of sets it is enough. For example, Scooby's Mansion is a bit less clear because there weren't as much sales and the peak is spread out over several months:

The data is from bricklink.

That is exactly what I want to do. Just having the initial trouble of finding retired set lists from each year. That's where it's frustrating where there is no simple filter feature to show "sets retired in 2015", etc.

Now those graphs, is that a section of bricklink that plots those? Or are those your graphs that you created after pulling BL data?

[c_rpg]

20 hours ago, scatttcat said:

That is exactly what I want to do. Just having the initial trouble of finding retired set lists from each year. That's where it's frustrating where there is no simple filter feature to show "sets retired in 2015", etc.

Now those graphs, is that a section of bricklink that plots those? Or are those your graphs that you created after pulling BL data?

Unfortunately, they don't have such a feature. You'll have to look at release dates and try to estimate what sets have retired. For example, all SW sets from 2014 have now retired. Brickset also has a handy tag system which you can use to browse through sets by year and by theme. https://brickset.com/sets/theme-star-wars/ownedBy/year-2014/page-1

You can quickly browse through themes and years and see what was released (not retired unfortunately). There's a similar system on bricklink, but I find that brickset is a little more user friendly.

As for the data: to be honest, you probably won't see anything useful on bricklink for older sets as they only show the last 6 months of data. It's something you have to keep track of manually or via some script every once in a while. Though, technically you only need to pull the data once every 6 months so it's not really a lot of work if you only have to update a couple of sets each month.

 

All this information is useful to have and it's good to know a lot about the performance of older sets, but in the end you'll still have to guess what set prices will do post retirement. I have tried all kinds of algorithms to estimate which sets will do well, but the fact of the matter is it usually depends on what Lego decides to do. You can look at the popularity of a set, the prices, etc. but those things are offset by the number of units they produce and what sets are released in the future. If we knew that, it would be much easier to pick sets.

[scatttcat]

9 hours ago, c_rpg said:

Unfortunately, they don't have such a feature. You'll have to look at release dates and try to estimate what sets have retired. For example, all SW sets from 2014 have now retired. Brickset also has a handy tag system which you can use to browse through sets by year and by theme. https://brickset.com/sets/theme-star-wars/ownedBy/year-2014/page-1

You can quickly browse through themes and years and see what was released (not retired unfortunately). There's a similar system on bricklink, but I find that brickset is a little more user friendly.

As for the data: to be honest, you probably won't see anything useful on bricklink for older sets as they only show the last 6 months of data. It's something you have to keep track of manually or via some script every once in a while. Though, technically you only need to pull the data once every 6 months so it's not really a lot of work if you only have to update a couple of sets each month.

 

All this information is useful to have and it's good to know a lot about the performance of older sets, but in the end you'll still have to guess what set prices will do post retirement. I have tried all kinds of algorithms to estimate which sets will do well, but the fact of the matter is it usually depends on what Lego decides to do. You can look at the popularity of a set, the prices, etc. but those things are offset by the number of units they produce and what sets are released in the future. If we knew that, it would be much easier to pick sets.

Also for @Val-E @Pseudoty who helped me with this

 

So I spoke to brickset and found a solution. Now for this solution you'd have to accept that Brickset's "LAST SOLD ON LEGO.COM" = retirement date. If you accept that, then this works to filter sets by year for retirement.

You simply make your own search query, you need to be logged in for this.
https://brickset.com/queries

Create a new query and dictate the And/Or statement as need be, in this case I did "last sold on lego.com on and after 1/1/16 and before 12/31/16"... that grabs all the "retired" sets from 2016. Repeat for each year you want. I also modified a few other specifications like no Lego Dimension Figures. Create CSV and boom you have a spreadsheet.

Here is my 2016 query (I did the same with 2014 & 2015)

https://brickset.com/sets/query-3605

 

Now, to your point about set values, yes I will still have to pull them manually via bricklink. I plan to grab every 6 months for older years and for 2016 I will do every month this year. Tedious? Yes of course. Worth it? I am sure many of you will say no haha

But in the end I love organization and I don't mind spending the time on it. Plus it helps me play catchup on all the sets the past few years. See attached screen for where I am taking this. ?

 

 

 

[Jeff Mack]

On 5/27/2017 at 9:38 PM, scatttcat said:

Problem is, it is crazy how many websites DONT have this.

It's really not that hard to understand why.  LEGO doesn't want to give out exact info.  It is all a guestimate.  Brickset does a pretty good job at it, but it is not always accurate. Sometimes it can jump by a month or two if for some reason it was out of stock, doesn't come back, but then comes back for a few days.  Eventually it irons out and stays retired, but it's not exact.

[c_rpg]

14 minutes ago, scatttcat said:

Also for @Val-E @Pseudoty who helped me with this

 

So I spoke to brickset and found a solution. Now for this solution you'd have to accept that Brickset's "LAST SOLD ON LEGO.COM" = retirement date. If you accept that, then this works to filter sets by year for retirement.

You simply make your own search query, you need to be logged in for this.
https://brickset.com/queries

Create a new query and dictate the And/Or statement as need be, in this case I did "last sold on lego.com on and after 1/1/16 and before 12/31/16"... that grabs all the "retired" sets from 2016. Repeat for each year you want. I also modified a few other specifications like no Lego Dimension Figures. Create CSV and boom you have a spreadsheet.

Here is my 2016 query (I did the same with 2014 & 2015)

https://brickset.com/sets/query-3605

That's pretty awesome you actually got help from them and got something working.

14 minutes ago, scatttcat said:

Now, to your point about set values, yes I will still have to pull them manually via bricklink. I plan to grab every 6 months for older years and for 2016 I will do every month this year. Tedious? Yes of course. Worth it? I am sure many of you will say no haha

But in the end I love organization and I don't mind spending the time on it. Plus it helps me play catchup on all the sets the past few years. See attached screen for where I am taking this. 

Yeah, I also enjoy cataloging my sets and keeping track of all the stats. To me, it's part of the fun of buying/selling. If anything, it helps you remind what you have in stock and when might be a good moment to sell it.

There's all sorts of data you can get from brickpicker, brickset and bricklink. Prices, demand (wanted lists), set/box information, availability, release dates...

[scatttcat]

2 hours ago, c_rpg said:

That's pretty awesome you actually got help from them and got something working.

Yeah, I also enjoy cataloging my sets and keeping track of all the stats. To me, it's part of the fun of buying/selling. If anything, it helps you remind what you have in stock and when might be a good moment to sell it.

There's all sorts of data you can get from brickpicker, brickset and bricklink. Prices, demand (wanted lists), set/box information, availability, release dates...

Speaking of, is there anyway to pull data from sets from bricklink into excel? Like their 6 month averages sales? Or do you literally have to search for each set and view it? I am thinking the latter

[Jeff Mack]

Just be careful with the values you get from Bricklink.  I spent a lot of time trying to work with the data to add into the price guide here (not sure if I will or won't yet) and found it to be pretty unreliable for hundreds of sets.  For example, take a look at the UCS Falcon, right now it is being priced at under $1700 and only $930 for May.  Really???  40 sold in the last 6 months and that is the price.  It is like this for hundreds of sets.

[DadsAFOL]

51 minutes ago, Jeff Mack said:

Just be careful with the values you get from Bricklink.  I spent a lot of time trying to work with the data to add into the price guide here (not sure if I will or won't yet) and found it to be pretty unreliable for hundreds of sets.  For example, take a look at the UCS Falcon, right now it is being priced at under $1700 and only $930 for May.  Really???  40 sold in the last 6 months and that is the price.  It is like this for hundreds of sets.

It's been a while since I've poked around that part of the API, but I believe you can exclude "incomplete" sets, which is what is showing up in the history as those $400 sales.  The API lets you get some interesting granular data at the transaction level, like country of buyer and seller.

[Jeff Mack]

It's been a while since I've poked around that part of the API, but I believe you can exclude "incomplete" sets, which is what is showing up in the history as those $400 sales.  The API lets you get some interesting granular data at the transaction level, like country of buyer and seller.

Sure. I can show you my API calls and it doesn't make a difference.

[scatttcat]

3 hours ago, Jeff Mack said:

Just be careful with the values you get from Bricklink.  I spent a lot of time trying to work with the data to add into the price guide here (not sure if I will or won't yet) and found it to be pretty unreliable for hundreds of sets.  For example, take a look at the UCS Falcon, right now it is being priced at under $1700 and only $930 for May.  Really???  40 sold in the last 6 months and that is the price.  It is like this for hundreds of sets.

Where would you recommend is the most accurate values then? Brickpicker (eBay)? or an average of the two?

[scatttcat]

3 hours ago, Jeff Mack said:

Just be careful with the values you get from Bricklink.  I spent a lot of time trying to work with the data to add into the price guide here (not sure if I will or won't yet) and found it to be pretty unreliable for hundreds of sets.  For example, take a look at the UCS Falcon, right now it is being priced at under $1700 and only $930 for May.  Really???  40 sold in the last 6 months and that is the price.  It is like this for hundreds of sets.

Looks like Brickpickers graph of values the last 12 months is a little more accurate. Do you agree? Like when you hover over each month and get price. Well its a lot more accurate for the Falcon at least haha

[Val-E]

3 hours ago, scatttcat said:

Looks like Brickpickers graph of values the last 12 months is a little more accurate. Do you agree? Like when you hover over each month and get price. Well its a lot more accurate for the Falcon at least haha

Jeff does a lot of work filtering out scam sales, keyword abuse, clones and other stuff that infect Ebay and distort final sales. For US data, BP is probably the best source but you have to be patient to get the updates and some sets have null values.

You can do the same on Ebay for individual sets yourself but you are limited to the last 60 days.

Brickset has the most comprehensive data set on set info like availability dates, RRP prices and can be exported to Excel and customised.

Bricklink is the best place for parts and minifigs and also allows you to value sets on part out value. That is very useful for spotting sets that are greater or worse than the sum of their parts. As mentioned, the number of set sales is very low so price data is not always accurate (some sales are clearly fake or part of a superlot and fake sales that get cancelled are not wiped from the stats). It is still a good indicator of real market prices as the commissions are lower and it´s a lego only platform so buyers tend to be better informed and won´t pay such crazy money for sets as on Amazon, for instance. You also get real time sales and 6 month hustory and there is no currency exchange pollution, like on Ebay  where everything gets put into $.

Brickowl has even less traffic so I wouldn´t bother with that.

Another option is to use 3xcamel to take a look at amazon price evolution and the sales ranking of individual sets. This helps determine the popularity of the set and at what price your rivals have been getting it for.

For my purposes, BL is the least bad option as it is instant, in Euros and all the sets are there. The worst point is you do need to filter out the lowball and higball sales and maybe some months, or indeed semesters, there are 0 sales on some items. Another good point is you get an idea of sales to hoarded ratio and you can easily see which sets are rare.  If a set is selling well on ebay or Amazon and is not highly stocked on Bricklink, chances are you have found a gem.

In summary, there is no one best solution and you probably need to work with 2 or 3 sites to get reliable and useful info.

Edited May 30, 2017 by Val-E

[HandyHand]

11 hours ago, c_rpg said:

That's pretty awesome you actually got help from them and got something working.

Yeah, I also enjoy cataloging my sets and keeping track of all the stats. To me, it's part of the fun of buying/selling. If anything, it helps you remind what you have in stock and when might be a good moment to sell it.

There's all sorts of data you can get from brickpicker, brickset and bricklink. Prices, demand (wanted lists), set/box information, availability, release dates...

Same for me about keeping track of certain stats of the sets I bought, which I use for determining value, (potential) profit, stock, etc.

Coming back to the list of Retirements of 2016, it's interesting to look at the difference between "Own this set" vs "Want it", on Brickset. I know it's very arbitrary and just a check that people voluntarily make on Brickset, but when looking at the entire list of 2016 retirees these numbers can give an idea of the demand for a set. Specifically sets where the Want it number is higher than the Own it number. There are a few limited release City sets that were not widely available where that's the case: 

- 60096 - Deep Sea Operations Base

- 60112 - Fire Engine

- 60071 - Hovercraft Arrest

They might turn out to be future desirable sets. Too bad I didn't get any of them when they were sold at the biggest discounts. 

Edited May 30, 2017 by Haay

[c_rpg]

8 hours ago, Haay said:

Same for me about keeping track of certain stats of the sets I bought, which I use for determining value, (potential) profit, stock, etc.

Coming back to the list of Retirements of 2016, it's interesting to look at the difference between "Own this set" vs "Want it", on Brickset. I know it's very arbitrary and just a check that people voluntarily make on Brickset, but when looking at the entire list of 2016 retirees these numbers can give an idea of the demand for a set. Specifically sets where the Want it number is higher than the Own it number. There are a few limited release City sets that were not widely available where that's the case: 

- 60096 - Deep Sea Operations Base

- 60112 - Fire Engine

- 60071 - Hovercraft Arrest

They might turn out to be future desirable sets. Too bad I didn't get any of them when they were sold at the biggest discounts. 

I got a few Fire Engines when they were discounted to around 25€. It's a pretty nice set actually, though a remake seems highly likely. Seems like they got a LOT of stock of them for some reason over here. I'm talking well over 30 in one store. I'm pretty sure it had limited availibility in the US, which might explain the high wanted/owned ratio.

Wanted vs owned is a useful statistic along with supply vs demand.

18 hours ago, scatttcat said:

Speaking of, is there anyway to pull data from sets from bricklink into excel? Like their 6 month averages sales? Or do you literally have to search for each set and view it? I am thinking the latter

Bricklink does not provide any ready to use feature that I know of. There are ways to pull the data from bricklink, however crawlers etc. are forbidden by their TOS, so basically it's not allowed unless you ask permission. Have a look at robots.txt on wikipedia.

I guess they might allow it if you don't overload their servers with page requests. Some script that pulls the data for a set once every minute might be tolerated.

18 hours ago, Jeff Mack said:

Just be careful with the values you get from Bricklink.  I spent a lot of time trying to work with the data to add into the price guide here (not sure if I will or won't yet) and found it to be pretty unreliable for hundreds of sets.  For example, take a look at the UCS Falcon, right now it is being priced at under $1700 and only $930 for May.  Really???  40 sold in the last 6 months and that is the price.  It is like this for hundreds of sets.

Yeah there are lots of problems with the bricklink price guide. It's driving me insane sometimes. I don't know if they simply don't care or are too stupid to fix the issues, but to me it's pretty obvious what needs to happen:

• Provide additional filters for new and sealed sets in the price guide along with the current "new" and "used" catagories
• Provide an option to exclude super lots from statistics --> this would correct minifigures prices among other items.
• Exclude sold data from banned sellers / scammers. Right now, if some scammer decides to list a UCS falcon for 100$ and someone buys it, it appears in the statistics. Even if the actual sale never goes through, the stat remains. This has a HUGE impact on the perceived value of rare sets. It can drive the price down by hundreds of dollars.
• Warn / ban sellers that do not stick to the item description. A complete set of 16 minifigures should not be listed under a "box of 60", even if it's mentioned in the description.
  • Better yet: provide additional options for selling these kinds of lots. Make an item listing for complete sets of collectible series or provide some type of filtering for buyers.

Those things alone would fix 99% of the problems with the price guide.

End rant.