Jump to content

Rules for Unusual Price Suggestions


polar

Recommended Posts

This thread serves as a list of up to date rules on suggesting prices for unusual items. When a rule is changed, the main post (this one) will be updated, and a changelog post will be added below. If you have questions, please join our discord to ask them.

 

Backpack.tf prides itself in transparency, and the mods all realize that a lot of the rules presented in unusual suggestions are not explicitly laid out in a single, standardized guide. I have prepared this guide to review common rules. The complete list here is laid out not only to help you make excellent unusual suggestions, but also to help your peers in their suggestions through comments and references to this guide. We have tried our very best to capture the essence of unusual trading.

 

Important things to note before getting started:

 

(1) All unusual ranges are estimates. Please don't take our values as 100% accurate representations of the unusual market, especially if the last suggestion is more than 3 months old. Also, keep in mind, all our values are what we think a hat sells for in PURE. 

(2) There are inevitably exceptions to almost every rule listed. If such an exception is warranted, a mod will comment with a rationale. This is meant to be a general guideline with the basic rules. 

(3) Relevant links:

http://forums.backpack.tf/index.php?/topic/2775-psa-unusual-price-update-request-thread/ - for help with an unusual suggestion either post here or PM a mod on the forums.

http://forums.backpack.tf/index.php?/topic/8542-report-a-single-sale-the-forum-version/ - if you want to report your sales without going through the trouble of making a suggestion, please post here. 

http://forums.backpack.tf/index.php?/topic/6817-bpbs-mini-suggestion-thread-for-gifted-duped-hats/ - over 100 mini-suggestions for gifted, duped hats to use as proof should you need them

http://forums.backpack.tf/index.php?/topic/60275-suggesters-etiquette/ - additional things to look out for as suggester if you care about your own and the sites reputation

 

Unusual Suggestion Cheat Sheet:

 

Part 1. What is valid proof?

  • Sales
    • Sales older than 2 months old are not valid proof; this does not mean that they are not useful. Feel free to include them to provide supplemental proof. Just don't rely on them to set your range.
      • Exceptions: For very rare hats, if there have only been 1-2 sales in the last 3 months, a trade up to 3 months old is okay to use as proof to establish more data points. Keep in mind, new sales should take precedence over older ones. 
    • Sales for hats with outdated prices (>4 months old) are not valid proof unless you provide evidence of their current value in the suggestion (commonly known as a "mini-suggestion"). If you can't find proof for them include the links in your suggestion as supplemental proof. Just don't rely on them to set your range.
    • Trades with scammers are not valid proof
    • Duped hats - do not assume prices or take a fixed subtraction based off the clean versions of hats. For taunts duped generally doesn't matter.
      • If the hat you are pricing is clean, but it sold for or was offered a duped hat, you must make a mini-suggestion within your suggestion to justify its value (i.e. find sales on the duped hat to show its value). Several mini-suggestions have been provided in the duped hats thread linked above. 
      • If the hat you are pricing is duped, you cannot lower the value of its clean version based on its sale. Sales of duped hats can be used as supporting evidence or as proof to raise the value on an unusual or to set a price on an unpriced unusual if there are no clean sales.
        • EXCEPTIONS - A hat may be lowered in priced using duped sales if
          • The last price suggestion that induced a raise is based of a duped sale + there is no evidence that a clean one sells for more
          • Both these criteria are met:

            • There are NO sales of the clean version in the last year
            • The owner of the clean versions can not be active. For clarity, this means:
              • No premium on either trading site and no major trades on either trading site for the past 6 months and that they are Inactive/Offline on steam and have very low hours in tf2 (recently)
              • That they are Trade banned/unable to trade.
  • Offers
    • Offers alone are not enough to make a suggestion. This does not mean that offers are not helpful. They are very useful for defining a range and identifying outliers (see section below).
    • Fake, retracted, "collecting" offers are not valid proof. Offers are only valid if there is documented proof of the offer - either in comments or via screenshots of chat logs. Declined trade offers are better than verbal offers in chat for proof.
  • Buyouts (B/Os)
    • B/Os alone can only be used to drop the value on a hat (you cannot raise based on a b/o.) if all three of these criteria are met:
      • There are no usable sales.
      • You must drop the mean value of the hat by at least 15%. (i.e. You shouldn't be dropping a hat from 31 to 30 keys based on a b/o. You SHOULD be dropping the value on a hat from 100 to 50 if 50 can't sell in a month). 
      • The trade must have a reasonable time to mature (and must be reasonably bumped). What is reasonable varies from hat to hat. A good rule of thumb is to go by roughly 25 keys per week. So a trade for a hat <25 keys in value must be up for at least a week. A trade for a hat <50 keys must be up for at least 2 weeks. A trade for a hat <75 keys must be up for at least 3 weeks. For all other hats, the trade must be up for at least a month in order to price them based off the b/o. 
    • How you should use B/Os: All of this does not mean that B/Os are not useful. Use B/Os to give a general sense of your range. If a hat remains unsold for a month with a B/O of 20 keys, a sale at 30 is probably an outlier. Use them to check whether your suggested range makes sense. Just don't use them as the only proof to set both your range.

Part 2. The essence of overpay

  • Overpay in its purest form
    • No sale should be valued higher than a hat's buyout in pure if that buyout was up for a reasonable amount of time. If a person in the comments of a trade states a willingness to sell for "x" pure, the trade should not be valued above "x." 
    • Overpay is supposed to represent the pure value of any item. If an item sells easily (under a week) for a certain amount in pure and regularly gets offered a certain amount in pure, it should not be valued less than that in the suggestion.

Part 3. Determining an optimal range / identifying outliers

  • Unpriced hats - hats that have never had a proper suggestion or have no price
    • Two sales are required to establish a range on unpriced hats. If the two sales are discrepant, a third sale is required to determine the outlier.
      • Exception 1: High Tier Hats
        • Sales for high tier hats (>200 keys) can be based on a single sale if there are no other sales in the last 3 months. Estimations can be made for unpriced/outdated parts of that trade as long as reasonable. 
      • Exception 2: Rare Hats
        • Any hat can be priced based on a single sale if ALL these criteria are met:
          • There are no other recent sales in the last 3 months
          • If the owner has not resold the hat 1 month after the sale occurred OR the hat went to a collector or person who intends to keep the hat. A collector can be identified by looking at the person's trade history, item history of similar unusuals in the backpack.
          • The hat is not a new hat / effect (i.e. new Halloween effects during the Halloween event). You must wait until after the event is over to price these based on a single sale.
  • Price refreshers
    • Price refreshers are suggestions changing the value of a hat by <5%; changes in values of unusuals by <5% require at least 2 months between suggestions (sometimes they are fine if they fix a mistake from a previous suggestion).
  • Setting a range
    • Low End: Set the low end no lower than quickbuyer prices of a hat (more info and exceptions under "What is a quicksell?" below).
    • High End: Set the high end no higher than common b/o's and bounded by the high end sales.
    • Formatting: organize your proof in an easy to read manner including all your links for proof - screenshots, links to trades - then provide a short conclusion justifying your range.
    • Giving the suggested range the appropriate level of precision: all unusual suggestions are approximations. As such, the suggested range should be rounded off appropriately - http://en.wikipedia.org/wiki/Significant_figures. Please use common sense to guide you. This also counts for cash/scm sales since the key to cash ratio is also an approximation in the end.
      • Rule - for all unusuals below 5 keys, rounding to the nearest 0.05 keys is required. No other decimals are allowed.
      • Rule - for all unusuals 5-10 keys, please round to the nearest 0.5 keys. No other decimals are allowed.
      • Rule - for all unusuals above 10 keys, please round to the nearest whole key
      • Guideline - for hats 100-200 keys, please round to the nearest 5 keys
      • Guideline - for hats 200-1000 keys, please round to the nearest 10 keys
      • Guideline - for hats 1000+ keys, please round to the nearest 50 keys
      • When rounding lower tier hats, to determine which group your hat belongs to is based on the average suggested value, e.g. if your range is 8.7-10.3 (average 9.5), treat it as an unusual between 5-10 and round both ends to the nearest 0.5 (so 8.5-10.5); if your range is 9.7-10.9 (average 10.3), round both ends to the nearest key (so 10-11).
  • What is a quicksell?
    • Someone advertising their sale as a quicksell doesn't mean that it is one, sometimes they try to "quicksell" their hats for months without success.
    • Time is only one factor used to determine a quicksell. If a hat sells within 1-2 days for its b/o in pure, one should be suspicious. However, it could just be a lucky sale as well.
    • More important is how a sale relates to other sales. If a hat sells for less or around quickbuyer prices in pure or unusuals or for less than common trade points, that is likely a quicksell.
    • Exceptions when buy orders don't necessarily counter lower sales
      • Numbers Matter - if there is only 1 lonely buy order and 3 sales lower than that it probably doesn't counter. Also generic buy orders (any effect) are usually buying more than 1 copy of a hat.
      • Common Trading Point - if there is a huge number of sales at or around buyer prices then it might be included in the price range. Sellers being close to buy orders is also an indication for that.
      • Market Shifts - timing of buy orders and sales matter: If a sale took place before a buy order was created the seller obviously couldn't just "sell it to the buy order instead".
      • Counter Buy Orders - Buy orders created after a suggestion was made with the sole purpose of countering the suggestion will be structurally ignored.

Part 4. Other important considerations

  • Cash Trades / Market Transactions
    • It's ok to use cash trades as proof, proof of the PayPal payment would be nice (personal information can be redacted of course). Bp.tf price for keys are acceptable now (like all prices it might be outdated tho so it's still recommended to be doublechecked before blindly using it).
    • Use the value of keys on the site at which the item sold to determine the key conversion at the time of the trade. (i.e. use the SCM value of keys on SCM sales, use the marketplace.tf value of keys on marketplace.tf sales, use the opskins value of keys on opskins sales). For most transactions, this means you can use the key value at the exact date of the sale. When key prices are not stable, or when key values are unusually high or low at a certain day, a key average over a longer amount of time can be taken instead; a more detailed explanation can be found here.
    • You still have check for item histories and buyouts on those sales and provide that info in your suggestion.
  • Steam Community Market Sales
    • For each SCM sale, please provide proof (screenshot) that the hat of that particular effect sold for the said price.
    • To determine a trade value, you should take in consideration ONLY what the buyer paid. Take the sale value divided by the value of keys AT THE TIME. 
  • Bot.tf stats page screensots of sales
    • You can also use screenshots of bot.tf stats pages as proof as long as there are 3 sales at a certain value. You can access the page here if you donated at least 1 key to bot.tf.
  • A note about complex sales
    • A hat should not be priced using a sale in which it represents less than 50% of the trade value. (i.e. a 20 key hat and an 80 key hat sold for a 100 key hat, don't use that sale to price the 20 key one)
  • The values of items in a trade should be the "value" of the item at the time of the trade.
    • This can be very tricky as it can be difficult to extrapolate the value of an item on March 1 when the two most recent suggestions/sales were on February 1 and April 1. In most scenarios, the value should based on the accepted suggestion closest in time to the date of the trade. Sometimes you have to use single sales in a mini suggestion to determine the value at the time of the trade.
      • Examples: 
        • If the sale happened on April 1 and there are two suggestions at March 1 and April 2, the April 2 suggestion value should be used.
        • If the sale happened on April 1 and there are two suggestions at March 30 and April 30, the March 30 suggestion value should be used.
  • Modifications
    • Modifications like Paint, Spells, Name/Desc Tags and Killstreaks don't add value for suggestions generally, for some unusual weapons pro ks can matter tho.
  • Pricing of Strange Unusuals
    • Strange unusual items should be considered completely independent of non-strange unusual items. Do not use non-strange sales to set high ends on strange items even if they sold for more than the strange one. Do no use strange sales to set the low end on non-strange sales even if they sold for less than the strange one. Treat them as different items for the sake of the suggestion.
      • Having said that, non-strange and strange sales should give you an idea about setting appropriate ranges and establishing outliers. If a non-strange item sells for 100 keys but strange is sitting on the market with a b/o of 50 keys, it's a pretty good clue that 100 keys is an outlier. If a strange item sells for 50 keys and the non-strange sales are 80-100 keys, it's a pretty good clue that 50 keys is an outlier. Please use common sense.
    • This generally does NOT apply to items with strangifiers (veils, ground controls, laws, etc.). Exceptions to this can be made in very specific cases where the strangifier has a significant value compared to the hat it is applied to. These are treated case-by-case depending on the rarity of the strange versions/strangifiers, available data, and discrepancy between strange and non-strange sales. Examples of exceptions are linked below:
      https://backpack.tf/suggestion/61366de1e00f5021e57e4023
      https://backpack.tf/suggestion/65e265a0b1570d9d620cc3e6 
      A list of hats for which strangifiers exist can be found here for reference: https://wiki.teamfortress.com/wiki/Strangifier#Possible_variants

 

Link to comment
Share on other sites

  • 3 months later...

The moderating team has made a few significant changes to the rules. Outlined below are the changes. Please review and feel free to ask questions.

 

Major Changes: 

 

(1) We no longer allow suggestions based only on b/os. There must be at least 1 sale in every suggestion. Just like we don't like to use offers as the only proof to raise the value on an item, we don't like using b/os alone to drop the value as this has frequently caused bp.tf to lead the market with consecutive suggestions every 1-2 weeks dropping the hat (see Miami Wraith). We also have run into the problem of using these prices off b/os to price other hats. 

 

- Exceptions to this rule:

B/os can be used as the only proof in a suggestion ONLY if there are literally no usable sales in the last 2 months. In these situations, there is frequently a large buyer-seller gap and dropping the b/o might actually result in a few sales that could then be used to refine the value. These will be evaluated on a case by case basis. You must show previous sales and indicate why they are not usable (quicksells, sales for unpriced hats, etc.) 

 

(2) A formalization of the exceptions to the 2 sale rule

  • High Tier Hats

Sales for high tier hats (>20 buds) can be based on a single sale if there are no other sales in the last 3 months. Estimations can be made for unpriced / outdated parts of that trade as long as reasonable estimations are provided.

 

Example of reasonable estimations:

http://backpack.tf/v...cd7b8a53d8b4568

 

Example of unreasonable estimations:

http://backpack.tf/v...cd7b808318b49e3

 

Courtesy of Vincent, "In my personal opinion, I think what matters more here is what it ultimately sold for rather than what it might have sold for or what the seller might have been willing to take in pure a day/a week/a month prior to the sale. This isn't an unusual that regularly trades hands, this isn't an unusual that's going to see more on the market, this isn't an unusual where- let's face it- what we put down here is going to have any influence whatsoever on how anyone trades these things. These high tier unusuals are and always have been their own little playground and what we put down here for them have always served little more than an interesting little record of how they've most recently changed hands. So long as the suggested price accurately matches the final transaction, for an item like this, I don't see any point in fussing over prior pure B/Os or thinking about what he might have alternatively accepted in pure.

  • Rare Hats

Any hat can be priced based on a single sale if ALL these criteria are met:

- There are no other recent sales in the last 3 months

- The effect is not a summer 2013 effect. (robo effects can be used as they are often rarely traded and rarely unboxed)

- The hat went to a collector or person who intends to keep the hat. A collector can be identified by looking at the person's trade history, item history of similar unusuals in the backpack OR if the owner has not  resold the hat 1 month after the sale occurred.

 

Other Addenda

 

(1) Trades with scammers are not valid proof

 

(2) The values of all items in the trade should be the "value" of the items at the time of the sale. This is tricky since the "value" on an item is not always the same as the most recent suggestion. Generally, this means that it should be based on an accepted suggestion closest in time to that of the trade. For unusuals, this is almost always the most recent suggestions since they are not frequently updated. But this is not always the case as many new unusuals have suggestions every 2-4 weeks.

 

If the sale happened on April 1 and there are two suggestions at March 1 and April 2, the April 2 suggestion value should be used.

 

If the sale happened on April 1 and there are two suggestions at March 30 and April 30, the March 30 suggestion value should be used.

 

To further complicate matters, some items (like buds or the bmoc) drop precipitously. So using this method may not be reliable. To the best of your ability, try to use the "value" of the item at the time of the sale. If that is too hard to predict, that proof should be avoided in making the suggestion.

 

(3) Giving the suggested range the appropriate level of precision

 

All unusual suggestions are approximations. As such, the suggested range should be rounded off appropriately.

 

  • For all unusuals, please round to the nearest integer
  • For hats 100-200 keys, please round to the nearest 5 keys
  • For hats 200-1000 keys, please round to the nearest 10 keys.
  • For hats 1000+ keys, please round to the nearest 50 keys.

 

Examples:

http://backpack.tf/v...cd7b82b7c8b472e - value here is really 100 +/- 10. Giving it any more precision, even 97-105, is honestly more precise than we can reasonably tell.

http://backpack.tf/v...cd7b8ae2d8b4661 - instead of 4.1-4.9, a range of 4-5 - it keeps the average constant while giving a decent approximation.

Link to comment
Share on other sites

  • 2 months later...

MAY 5, 2015 EDIT: A good rule of thumb is to convert hats that used to be priced in buds to their current value in keys by multiplying by 10. So when I list below a value of 2.5-3 buds, the rough equivalent of that today would be 25-30 keys.

 

Over the last 6 months, I have on-and-off been gathering data points here - http://forums.backpack.tf/index.php?/topic/6817-bpbs-mini-suggestion-thread-for-gifted-duped-hats - with the intent to formally re-evaluate some of our unusual suggestion rules regarding using sales of duped hats as proof. And then I realized that this is a frequent question in the trading community. Does duping matter on a hats value? How much do you "take off" in value for a gifted hat? These sorts of questions garner lots of empiric data but nothing solid. So with this in mind, I present to you my analysis of a fairly reasonable data set - roughly 200 unique sales for about 120 different hats. The raw data is local for now, but I'll publish it a bit later.

 

DUPED HATS:

 

Overall Value Reduction: 9.8%

25-75 Confidence Intervals: ~5 to 15% (most sales happen in this range)

 

Sorting By Value

2.5-3 buds:  2% <---- Hats under 3 buds in value should no longer need minis if they are duped

3-5 buds:     8%

5-8 buds:     9%

8-10 buds:   12%

10-15 buds: 12%

15-20 buds: 10%

>20 buds:    12%

 

Sorted by Generation

1st:      10%

2nd:     n/a

'Ween: 10%

 

Sorted by Class:

Single:     10%

All-Class: 10%

 

Sorted by Type:

Misc: 9%

Hat:   10%

 

Sorted by ratio of duped:cleans of a particular hat:

TBD

 

GIFTED HATS:

 

Overall Value Reduction: 19.8%

25-75 Confidence Intervals: ~10 to 30% (most sales happen in this range)

 

Sorted by Value

0-2 buds:   20%

2-5 buds:   21%

5-10 buds: 18%

>10 buds:   23%

 

Conclusions

 

Duped hats: Given the data it seems like a hat being duped starts to matter at 3 buds in value and that, for the most part, a similar reduction in value proceeds from hats ~3 buds in value to hats 20 buds in value. Of note, I have not included huge outliers such as the scorching KE. It seems as if the generation of the effect and class of the hat matters little as the overall reduction in value is about 10%.

 

Gifted Hats: There's a high amount of variability in the reduction of value of gifted hats with most trades seeing a reduction anywhere from 10-30%. Value of the hat seems to matter little as ~20% seems to be the reduction at hats ~1 buds in value and at hats ~10+ buds in value. Of note, there are far fewer gifted sales to report, so further analysis will have to wait for more data points.

 

A few more important points:

 

(1) It's difficult to get time-matched prices for hats. If a duped sale happened on April 1 and a clean suggestion passed on May 1, I'm still using those as close enough aged matched data points to compare clean vs duped. I tried my best to capture what the clean value of a hat was at the time of the duped sale, but that's a challenge. As a result, the analysis here is quite crude. Also, some hats like the hearts mancer have sales prior to the misc update. Some have sales prior to a "hype update." I am still including these data points if I can. I'm just using the clean values of these hats prior to those updates.

 

(2) This is a work in-progress. If you have duped / gifted sales, feel free to keep posting here - http://forums.backpack.tf/index.php?/topic/6817-bpbs-mini-suggestion-thread-for-gifted-duped-hats.

A lot of those minis are now outdated, so posting your sales will both help suggesters as well as provide more data points in this analysis.

 

(3) I will be adding more breakdowns over time, including analysis by class, effect, misc vs hat, etc. I have these all in my data set, but just haven't had time to look at that yet. There may not even be enough data just yet to really look at these factors.

Link to comment
Share on other sites

  • 2 months later...

Minor Rule Change

 

In anticipation of Halloween 2014, making a minor rule change. Summer effects have been around for more than a year. It's about time they be allowed to be priced off one sale as long as they meet the other criteria

 

Rule prior to the change:

Any hat can be priced based on a single sale if ALL these criteria are met:

- There are no other recent sales in the last 3 months

- The effect is not a summer 2013 effect. (robo, 2013 halloween effects can be used as they are often rarely traded and rarely unboxed)

- The hat went to a collector or person who intends to keep the hat. A collector can be identified by looking at the person's trade history, item history of similar unusuals in the backpack OR if the owner has not resold the hat 1 month after the sale occurred.

 

Rule after the change:

Any hat can be priced based on a single sale if ALL these criteria are met:

- There are no other recent sales in the last 3 months

- The effect is not a Halloween 2014 effect.

- The hat went to a collector or person who intends to keep the hat. A collector can be identified by looking at the person's trade history, item history of similar unusuals in the backpack OR if the owner has not resold the hat 1 month after the sale occurred.

 

How Halloween 2014 effects will be handled (at least for now):

 

Regardless of whether a hat sold to a collector, you must have at least TWO sales for all halloween effect hats without exception. We'll probably change this rule in December. So if you do see a 1 of 1 hat go to a collector, bookmark the trade and hold onto it until then and you'll be able to price it off one sale before the sale becomes too old to use. 

  • Like 5
Link to comment
Share on other sites

  • 3 months later...

 

Part 2. The essence of overpay

  • Overpay in its purest form
    • No sale should be valued higher than a hat’s buyout in pure. If a person in the comments of a trade states a willingness to sell for "x" pure, the trade should not be valued above "x."
    • Overpay is supposed to represent the pure value of any item. If an item sells easily (under a week) for a certain amount in pure and regularly gets offered a certain amount in pure, it should not be valued less than that in the suggestion.
  • Standardized overpay
    • All OFFERED hats and items should be taken at 90% of their original average price. For a guide on how to determine the original average price, please look here - http://forums.backpack.tf/index.php?/topic/4350-new-rule-use-original-suggested-values-of-unusuals/
    • If unclear which hat was offered, DO calculate overpay.
    • Clean HOUWARs, Max heads (non-level 1), buds, keys, refined, and cash trades are exempt from overpay calculations.
      • 10% overpay should be added to duped HOUWARs and level 1 max heads.

 

The issue of overpay has been brought up many times recently, and I felt it problematic enough to address it again here. Before going into a discussion, I want to emphasize two things:

 

(1) There is no way to make a defined set of rules to apply to every situation. These are GUIDELINES. I give YOU as the experienced trader / suggester the leeway to make suggestions as you see fit. I only ask that if you stray from these guidelines, you provide solid justification for why you are doing so. And if that is reasonable, it will be accepted.

(2) The whole point of overpay is to determine the pure value of an item. Please don't forget this. EVERY item has a number - the value for which it can sell in pure within 1-2 days. For some items like the duped sparkle luger or clean beams modest, that number is actually less than 10% overpay. But for some hats it is a LOT more. 

 

With these two things in mind, I made this edit:

  • Overpay is supposed to represent the pure value of any item. If an item sells easily (under a week) for a certain amount in pure and regularly gets offered a certain amount in pure, it should not be valued less than that in the suggestion.

I will emphasize the relevance of this rule by examples of when 10% overpay is too much:

 

Australiums

An aussie rocket launcher is listed at 67-70. If someone lists one at 66 keys, it sells incredibly fast - http://www.tf2outpost.com/trade/24981348. 10% overpay on 67-70 is 61.6 keys. Valuing an aussie rocket launcher at 61.5 keys pure would be an unfair evaluation of its pure value. Does this mean that no overpay should be applied? No, because aussie rocket launchers are taking a week to sell at 68 keys pure, so it cannot be considered a direct equivalent of pure. Valuing it at ~66-67 (low end of the current range is probably the closest approximation of its pure value)

 

The issue of double overpay

If you make suggestions long enough, you come across the situation where you have to make a mini for an outdated hat. Do you apply overpay in the mini and then again on the main suggestion? Doing so stacks overpay twice, so if all hats were traded 1:1, your final evaluation comes out to 81% of the original hat's value. What if you have to do a mini for the mini - now you get overpay stacked 3 times. And if all hats were traded 1:1, your final evaluation comes to 72% of the original value. This isn't right, and it doesn't capture the meaning of overpay. GENERALLY (though not always), it is best to apply overpay ONCE. When you apply it once, you are calculating the pure value of an item. Applying overpay again would be the equivalent of applying overpay on pure. This is not a black-and-white rule. I encourage you to use YOUR judgment. If after applying overpay on ANY item, you find that the value doesn't seem right, please use common sense and experience to guide your decision making about what is a more reasonable evaluation of the pure value of the items involved (i.e. if after double overpay, a hat is being evaluated at 13 pure when it gets offered 14 pure and sells in a day at 14 pure, you shouldn't take it less than 14)

 

 

Please always remember what the point of overpay is:

We don't make rules for the sake of making rules. We are trying our best to capture what goes on in the minds of traders when completing an unusual trade. Unfortunately, every thing that goes on in the mind of a trader cannot be easily captured into a defined set of rules.

 

Most (not all) traders will take less in pure than they will in items and unusuals. When you trade for items and unusuals, you do a mental calculation in your head for what that item can sell for easily in pure. The rule of 10% overpay is an archaic one that came from a rough figure of overpay BEFORE third gens even came out. Realistically, overpay is on average more than 10%, and it is also NOT 10% for all items. It's impossible to make a more reasonable standardized overpay. So the 10% rule will stay. However, if you can show for a given item that it should be valued differently based on pure sales and offers, that should always take priority.

  • Like 4
Link to comment
Share on other sites

  • 1 month later...

ARCHIVED POST. IGNORE

 

I will merge this into the unusual pricing guide a bit later but feel it's important enough to have its own thread for now. For the record, I don't think this is the end of buds as currency. Due to a limit on the number of keys that can be traded on high tier hats, there will ALWAYS be a need for a higher denomination of currency. The problem is that buds are just so volatile that it's not really fair to have a firm rule to price all hats in buds. Why is that? It opens the door to a huge amount of price manipulation, intentionally or unintentionally, for some hats to be priced when buds are at their highest value compared to when buds are at their lowest value. In the past it didn't matter since buds remained relatively constant in value even over the course of 3 months, the average time it takes to update a hat. But now buds might rise and fall 2-3 times within those 3 months.

 

Things that won't change:

 

- We will continue to price hats in the currency they MOST commonly sell. This means that if a hat sells in buds more often than it sells for in keys, it needs to be priced in buds. If it sells in keys, there is no issue with pricing it in keys, no matter the tier of the hat. This is NOT a change from our prior guidelines.

 

Things that will change:

 

- What WILL change is our traditional bias towards getting things priced in buds. In the past, when there were 50-50 sales in keys or buds, people usually went with buds. In situations where it is 50-50, I encourage you to go off the more RECENT of the trades but will ultimately leave it to the discretion of the suggester. On hats >2 buds in value, suggesters normally converted the key sale value to buds. It is not necessary to do that. Rather, I encourage you to leave the sales and suggestion in the currency that is actually being used whether it is buds or keys.

 

- Paypal / SCM trading - if a hat sells for paypal or through the market, you can consider that a sale in "keys." What is the rationale for this? I, and most paypal traders I know, tend to derive bud values from the key value nowadays given the volatility of buds. Keys have the most stable paypal and SCM value, so it is most reliable to take a sale on paypal or through market as a sale in "keys." Trying to figure out the value of buds at the time of the sale in dollars has been nearly impossible for both moderators and suggesters. Key prices have been almost completely stable.

 

What about NON-PURE trades?

 

- If a hat sells for items priced in keys, the sale can be valued in keys after the appropriate amount of overpay. This goes for unusual : unusual trades too. If it sells for another unusual priced in keys, it can be valued in keys. If it sells for an unusual priced in buds, it should be valued in buds.

- What about mixed trades? Use common sense. If an unusual sells for a hat priced at 5 buds and items worth ~20 keys, it should be priced in buds as the 5 buds makes up the majority of the trade. If it sells for a hat priced at 50 keys and an item worth 1 buds, it should be priced in keys.

Link to comment
Share on other sites

  • 2 months later...

UPDATING THE B/O RULE (has already been incorporated in the op)

 

When we switched currencies, I made a post saying we would allow pricing off b/os for a short time to fix inflated hats. We haven't fixed all inflated hats, but it has been a while, so I decided to make an addendum to what is and isn't allowed. As always, these rules are not 100% firm. Use common sense to guide you. This is meant to be a guide to help you make good suggestions. 

  • B/Os alone can only be used to drop the value on a hat if all three of these criteria are met
    • ​​You are dropping the value on the hat (you cannot raise the value on a hat based on a b/o
    • You must drop the mean value of the hat by at least 15%. (i.e. You shouldn't be dropping a hat from 31 to 30 keys based on a b/o. You SHOULD be dropping the value on a hat from 100 to 50 if 50 can't sell in a month). 
    • The trade must have a reasonable time to mature (and must be reasonably bumped). What is reasonable varies from hat to hat. A good rule of thumb is to go by roughly 25 keys per week. So a trade for a hat <25 keys in value must be up for at least a week. A trade for a hat <50 keys must be up for at least 2 weeks. A trade for a hat <75 keys must be up for at least 3 weeks. For all other hats, the trade must be up for at least a month in order to price them based off the b/o. 

You may be wondering where 15% is coming from. That has always been the degree of uncertainty I have had in my head regarding the accuracy of our prices. In the past, when hats got a price, the site would automatically add a 15% range around the mean. Remarkably, I've still stuck to that belief even if we are now way past that pricing system.

  • Like 5
Link to comment
Share on other sites

  • 6 months later...

Happy New Years! With the new year, we usher in some much needed changes to the unusual suggestion rules. The idea here is to make things simpler, not more complicated than they already are.

 

Rules on Overpay

 

We are getting rid of the 10% overpay rule entirely. Over the last three years there have been many arguments both for and against having this rule in place. I won't go into a full discussion here, but you are welcome to ask about our rationale in the comments below. Our major reasons for the 10% overpay rule are no longer valid, so it's time to remove it entirely.

 

This does not mean we are getting rid of overpay entirely. We are still trying to capture the pure values of items. This means that a hats value cannot exceed what the seller was willing to sell it for in pure (pure buyout caps are still in place). Paying more than the pure buyout is overpay by definition and is an overpay for convenience.

 

One last thing. This means that pure sales will be even more heavily weighted compared to unusual sales. Whether the pure sale is low or high, you need a really strong basis for not including it in your range (i.e. if a pure sale is at 100 and two unusuals sales are at 150, don't immediately dismiss the pure sale as the outlier.)

 

Outliers and Quicksells

 

This is not a rule change but a clarification because there is a lot of misinformation going on. Outliers are defined by deviation from the mean NOT the time it takes to sell. This does not mean that time to sell is not useful. If a hat sells in an hour for the first offer it gets, do you really think the seller got full value? Probably not. If you look for more sales, including sales by sellers who were patient and waited for a month do you think you would find higher value sales? Probably. But sometimes traders just get lucky. In order to identify if a hat is an outlier, you must compare it to other sales and offers.

 

A few other common misconceptions:

 

- A seller had no buyout, so it can't be a quicksell

- A seller sold for unusuals, so it can't be a quicksell

- A seller had no trade, so it can't be a quicksell

 

Let's think about this logically for a second. There are factors intrinsic and extrinsic to a trade that determine how much a hat can sell for. Extrinsic factors include, among others, how many of that particular hat are on the market. Keep this in mind because this can fluctuate from month to month. A sale one month ago when a hat was 1 of 1 on the market might have gone in one day for 50 keys, but a sale now when the hat is 1 of 5 on the market might have taken 2 weeks to sell for 40 keys. Just because a hat sold in 1 day for 50 keys does not mean that any sale less than that is an outlier.

 

Intrinsic factors include how well a hat is advertised. A trade that gets bumped 48 times a day is obviously going to get more views than a trade that gets bumped once in 3 days. A hat that sells in 5 days without getting bumped may be as much of an outlier as a hat that gets bumped 5 times in one day and sells in 3 hours.

 

Further, trades without any buyouts are generally turn offs for a lot of traders who are unsure about what the seller expects, especially when multiple hats are listed. So not having a buyout makes it more likely instead of less likely that the sale may be an outlier compared to other sales.

 

Whether a hat sells for unusuals or pure has no real factor in whether the sale is an outlier / quicksell. It was up for a fixed amount of time and got a certain number of views. Whether it sold for unusuals or pure doesn't factor in at all into whether it is an outlier. What matters is how it relates to other sales.

 

Finally, the one that makes the least amount of sense to me - sales off market. If a hat was never advertised, it has the least likely chance of getting full value. If a hat sold in one day on the market, at least it has one day of views. Off market sales don't even have that one day of advertising. These sales are MOST LIKELY to be outliers. These are frequently made by banned users, and buyers are even more likely to stay away from them regardless of their ban reason.

 

Summary

 

In short, we have just made things simpler by getting rid of our contrived correction for overpay. No more arguments about overpay stacking, reverse overpay, who offered the hats, etc.

 

Please use common sense. We have our rules in place for a reason - we feel they best represent the values of items as they are in the eyes of traders.

 

What this means for ongoing suggestions ----

 

Please don't go back to make corrections on suggestions that have already been accepted. These changes apply only to NEW suggestions. If you have a suggestion that is up currently, you are welcome to resuggest. If we get to an old suggestion, based on the old rules, we will continue to accept them to clear out the backlog.

Link to comment
Share on other sites

  • 1 year later...

Minor Rule Change

 

Steam Community Market sales no longer get a range. You just take what the buyer paid (which also shows in the graph) and divide it through the key price during that day.

 

Rules prior to the change: 

Steam Community Market Sales

  • For each SCM sale, please provide proof (screenshot) that the hat of that particular effect sold for the said price
  • To determine a trade value, you should take in consideration BOTH what a buyer paid and what the seller sold (i.e. you will get a range from a single sale. low end = what the seller received; high end = what the buyer paid)
    • Buyer value: Take the sale value divided by the value of keys AT THE TIME. 
    • Seller value: First factor in steam tax. Divide the sale value by 1.15. Then divide the result by the value of keys at the time of the sale as appropriate. 
  • Treat SCM sales as you would other sales - compare the sale value to other sales and b/os to determine if it is an outlier)

 

Rules after the change: 

Steam Community Market Sales

  • For each SCM sale, please provide proof (screenshot) that the hat of that particular effect sold for the said price
  • To determine a trade value, you should take in consideration ONLY what the buyer paid. Take the sale value divided by the value of keys AT THE TIME. 
  • Treat SCM sales as you would other sales - compare the sale value to other sales and b/os to determine if it is an outlier)

 

Reasoning for the change: 

The old argument was that sellers would only get 85% of what the buyer paid cause of the market fees. But if you think about it, the buyer had to sell items on the SCM to get the required funds in the first place, and he would have to pay 15% fees everytime he sells something, so if you factor that in then both fees even themselves out. Our old rule regarding scm sales were kinda unfair towards the buyer so I felt like this change was necessary. For marketplace.tf sales we dont consider the 10% commision that sellers have to pay either, SCM is similiar to that as well, you pay the fee to have a secure place to sell your items without having to fear a chargeback or requiring trust. Also I think our new ruling is more intuitive to new suggesters, so they dont have to do even more calculations and get a range with one sale. Now it's basically always: "Can't have a range with one sale."

Link to comment
Share on other sites

  • 1 month later...

Once a year for the last three years we have made significant changes to our unusual pricing guidelines. Included below are two important changes that will apply to suggestions moving forward.
 
Pricing of Dual Quality Items
 
First important note: This does NOT apply to items with strangifiers (veils, ground controls, laws, etc.). We are aware that prices have been wiped on these hats and are working on it. Do not make suggestions for these items. This will be fixed and they will be updated to the value of the non-strange version of the item.
 
Now that that is out of the way:

  • Strange unusual items should be considered completely independent of non-strange unusual items. Do not use non-strange sales to set high ends on strange items even if they sold for more than the strange one. Do no use strange sales to set the low end on non-strange sales even if they sold for less than the strange one. Treat them as different items for the sake of the suggestion.
    • Having said that, non-strange and strange sales should give you an idea about setting appropriate ranges and establishing outliers. If a non-strange item sells for 100 keys but strange is sitting on the market with a b/o of 50 keys, it's a pretty good clue that 100 keys is an outlier. If a strange item sells for 50 keys and the non-strange sales are 80-100 keys, it's a pretty good clue that 50 keys is an outlier. Please use common sense.
  • For a short, one-month "grace" period, we are allowing suggestions up to 6 months old on dual quality items.
    • We understand that some of these items are very rare and may have a single sale over 3 months old. And we understand that people have not given people adequate opportunity to price these. Given this, until MAY 1, we will allow sales up to 6 months old on rare dual quality items with only a single sale that happened any time in the last 6 months. Normal suggestion rules will apply after May 1.
    • In situations where there have been sales in the last 3 months (even unusable sales), sales older than that cannot be used.
    • I know this is frustrating for people with 1 of 1 items with sales more than 6 months old. We are working on a way to add pricing for unpriced items based on very old sales. Please wait on suggestions like these until we have implemented this. Look out for a post in this thread or PM me in the forums for more information.

Rule Changes for All Suggestions
 
Hats with prices up to 4 months old are not considered outdated for the purpose of minis within suggestions. This was previously 3 months.

  • Sales older than 2 months (3 months on rare hats) are still too old to make new suggestions. This only applies to hats used as proof in suggestions (popularly known as "mini-suggestions" or "minis").
  • Please use common sense. This does not mean that a 1 month old price cannot be outdated. Just look at any unusual taunt. Any taunt price over a month old (sometimes even 2 weeks old) is usually outdated and should not be used as proof in a suggestion. However, if b/os on the market and your trading experience suggest that the 4 month old price should be roughly accurate even now, then it is okay to use that as proof in a suggestion to price another hat.
  • Long term, the goal here is to push this out to 6 months. But we'll try a trial period with 4 months before making further changes.

Look out for another update for updating unpriced hats using old sales in the near future.

Link to comment
Share on other sites

  • 2 months later...

We have updated a small part of the rules a couple weeks ago, and after some deeper discussion and adjustments to it today I feel that it's time to give it an extra post to inform you guys about it.

 

New update:

 

EXCEPTIONS - A hat may be priced using duped sales if  

  1. The last price suggestion that induced a raise is based of a duped sale + there is no evidence that a clean one sells for more
     
  2. Both these criteria are met:
  • There are NO sales of the clean version in the last year
  • The owner of the clean versions can not be active. For clarity, this means:

                   a.) No premium on either trading site and no major trades on either trading site for the past 6 months and that they are Inactive/Offline on steam and have very low hours in tf2 (recently)

                   b.) That they are Trade banned/unable to trade.

 

Reasoning/Deeper explanation:

If the last suggestion raising the price was based off a duped sale it's safe to assume that duped and clean share the same values (unless proven otherwise of course). Note that I'm talking about the last suggestion raising a price here, not the last accepted suggestion. Also pay attention if the item was duped already at the time of the suggestion or not.

That means as long as at least 1 owner of any clean version is active you can't drop prices based on dupes (unless the point 1. is in effect). It would simply be unfair towards them.

No premium: If they have premium on a trading site, they are clearly active, even if they are not actively trading themselves, they could use premium on bp.tf to search for new hats etc.

Inactive/Offline on Steam and low hours recently: If the steam account looks abandoned (last online 50 days ago for example), and TF2 has not been played recently then its safe to assume that they are inactive. 

Tradebanned/unable to trade: If a clean version is in a permanently trade banned account then it won't be accessable to the public. Not all tradebans show public on the profile. Community bans inherit trade bans so they also count (not all community bans are permanent tho). 15 day being unable to trade because of mobile authenticator shenanigans DOES NOT count.

Now I understand that some clean owners wouldn't care about their hat dropping, but for consistency sake we should not allow exceptions from the exception here.

 

Raising a hat with duped sales is still fine btw, just don't cherry pick duped sales tho, make sure you catch all sales just like you would with clean ones to determine outliers etc.

 

 

 

Other:

I've also heard the rumour going around that we would up our treshhold for "duped doesn't matter" to 50 keys from 30 keys. Yes we have discussed that but decided not to push for it. Instead we want to emphasize that you should use common sense. For some hats dupes are equal, for others not (especially when it comes to pure sales), no matter what their prices is.

Link to comment
Share on other sites

It is me again, so with our next pricing event coming up shortly, I have reviewed this guide completely and made some smaller changes.

  • Updated formatting
    • Fixing the broken symbols as you see in the quote below
    • Made the Guide consistent with its use of bullet points.
  • I shortened some points a bit, and added some small things or examples to others.

  • I also replaced all mentions of "common offers" to "quickbuyers prices" cause that's usually how we determine whats too low on many hats, especially since public offers are a rarity these days.

  • I removed all mentions of buds from the guide as they are no longer relevant for unusual trading.

Other changes:

Quote

 

  • ​Raising using quickbuyers
    • ​Buyer trades alone can be used to raise the value of a hat only if these two criteria are met:
      • ​You are raising the value by at least 15%
      • The trade must be up for at least 1 week

 

I removed that part from the guide because: It's been never/rarely used anyways and the room for manipulation is just too big, just wait for a sale in those cases and suggest then.

 

 

On 1/9/2014 at 8:01 AM, polar said:

Modifications

  • Modifications like Paint, Spells, Name/Desc Tags and Killstreaks don't add value for suggestions generally, for some unusual weapons pro ks can matter tho.

I added this part and the rules (this was mostly commen sense anyways already but I wanted it to be written down).

 

 

On 1/9/2014 at 8:01 AM, polar said:

Pricing of Strange Unusuals

  • This does NOT apply to items with strangifiers (veils, ground controls, laws, etc.). 
  • Strange unusual items should be considered completely independent of non-strange unusual items. Do not use non-strange sales to set high ends on strange items even if they sold for more than the strange one. Do no use strange sales to set the low end on non-strange sales even if they sold for less than the strange one. Treat them as different items for the sake of the suggestion.
    • Having said that, non-strange and strange sales should give you an idea about setting appropriate ranges and establishing outliers. If a non-strange item sells for 100 keys but strange is sitting on the market with a b/o of 50 keys, it's a pretty good clue that 100 keys is an outlier. If a strange item sells for 50 keys and the non-strange sales are 80-100 keys, it's a pretty good clue that 50 keys is an outlier. Please use common sense.

 

I also added this part about pricing strange unusuals (that were in polars post a few above) to the guide itself.

 

 

On 1/9/2014 at 8:01 AM, polar said:
  • For all unusuals, please round to the nearest integer (this one is a RULE, all following suggestions are just a guideline).
  • For hats 100-200 keys, please round to the nearest 5 keys
  • For hats 200-1000 keys, please round to the nearest 10 keys
  • For hats 1000+ keys, please round to the nearest 50 keys

I've updated the wording on that to make it clear that rounding to the nearest integer is a rule (that means no prices like 8.5-9 keys or something like that, in that case either go 9 flat or 8-9 (usually preferable), while the rest is just a guideline. To note tho, if you chose to not round on a suggestion where rounding is recommended and you miss something in your sale then your suggestion has a high chance of being closed because of it, instead of being accepted as close enough.

 

 

 

PS: If you have any questions or suggestions for this guide or rules in general feel free to PM me on the forums.

 

  • Like 5
Link to comment
Share on other sites

 

Marketplace.tf/OPskins (Cash Sites) Unsolds and Sales:

Be careful with using them. Reasons: Item might be duped (double check), sale might be a quicksale and you dont know how long a listing has been up at a certain price.

If a cash site listing goes against the general market then you should rather focus on the market.

Reasons why Cash Sites are different from normal market:

-Accessability: Not everyone is familiar with them or trusts them, there are less users there. Also you should be 18+ years old to use paypal or credit cards. In theory you could sell keys on those sites to get the required credit, but thats a bit convoluted and makes you pay 10% fee.

 

Edit: About using Marketplace Buyouts:

Buyouts listed in keys (that includes the dual listing in the classifieds) will always take priority over cash buyouts. Key Buyouts usually should be up for at least a week, just a rule of thumb tho.

If there is no key buyout available and there is proof that someone was selling a hat on marketplace for a month at a certain price then I don't mind capping at that. Everything that takes longer than 1 Month is hardly labled as a quicksale. I wanna stress tho that there needs to be a screenshot of the cash buyout and it needs to be clear that its 1 month at that price.

 

How to look up marketplace.tf sales: 

 

 


1dWgs4W.png

Either here if it has sellers or you click on create new suggestion and click on the link here:

y3528ye.png

Then you see a graph with sales, please pay attention to the year in the date:

6DPThT1.png
 

 

To look up the bp.tf history of a marketplace item you simply have to click the button:

PL3jUNh.png

 

 

How to look up a sale/sellers/item histories on opskins:

 

 


To have access to the search feature you have to register/login to the site, alternatively you can get lucky and find it via google.

Lets look up the item history of the burning unusual mask of the shaman for example: https://opskins.com/?loc=shop_view_item&item=103283821

You need to click here:

Jo86ccx.png

Then you get linked to the bots steam inventory: https://steamcommunity.com/profiles/76561198254499310/inventory/#440_2_5674711858

You can look that bot up on bp.tf by copy its steam ID and paste it in the search bar on bp.tf :

Hlt6fYf.png

https://backpack.tf/profiles/76561198254499310

In the  bots inventory on backpack you have to search for the mask to finally get to the masks history: https://backpack.tf/item/2485727592

Also keep in mind you only get to see the last known owner to bp.tf history, not necessary the real seller.

(The mask has since then been taken down from opskins, but it works to same for every item.)

 

If you wanna see past sales you have to look at the graph:

zjwmSlj.png

Sadly the graph is for all effects so you need additional info that it was a certain effect or you can't really use it.

Link to comment
Share on other sites

Minor Rule Change:

 

On 1/9/2014 at 8:01 AM, polar said:
  • Cash Trades / Market Transactions
    • Do not use bp.tf USD values when converting cash trades to keys. Use commonly traded values for keys on the market. If you need help, please ask in the forums.
    • Use the value of keys on the site at which the item sold to determine the key conversion at the time of the trade. (i.e. use the SCM value of keys on SCM sales, use the marketplace.tf value of keys on marketplace.tf sales, use the opskins value of keys on opskins sales).

This can now be changed thanks to: 

 

The changed rules look like:

On 1/9/2014 at 8:01 AM, polar said:
  • Cash Trades / Market Transactions
    • It's ok to use cash trades as proof, proof of the paypal payment would be nice (personal information can be redacted of course). Bp.tf price for keys are acceptable now.
    • Use the value of keys on the site at which the item sold to determine the key conversion at the time of the trade. (i.e. use the SCM value of keys on SCM sales, use the marketplace.tf value of keys on marketplace.tf sales, use the opskins value of keys on opskins sales).

 

 

 

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

  • Teeny Tiny Cat locked this topic

Some clarifications and additional infos:

 

Quote
  • What is a quicksell?
    • Someone advertising their sale as a quicksell doesn't mean that it is one, sometimes they try to "quicksell" their hats for months without success.
    • Time is only one factor used to determine a quicksell. If a hat sells within 1-2 days for its b/o in pure, one should be suspicious. However, it could just be a lucky sale as well.
    • More important is how a sale relates to other sales. If a hat sells for less or around quickbuyer prices in pure or unusuals or for less than common trade points, that is likely a quicksell.
    • Exceptions when buy orders don't necessarily counter lower sales
      • Numbers Matter - if there is only 1 lonely buy order and 3 sales lower than that it probably doesn't counter. Also generic buy orders (any effect) are usually buying more than 1 copy of a hat.
      • Common Trading Point - if there is a huge number of sales at or around buyer prices then it might be included in the price range. Sellers being close to buy orders is also an indication for that.
      • Market Shifts - timing of buy orders and sales matter: If a sale took place before a buy order was created the seller obviously couldn't just "sell it to the buy order instead".
      • Counter Buy Orders - Buy orders created after a suggestion was made with the sole purpose of countering the suggestion will be structually ignored.

^This has been added to the guide, a more detailed explanation of those can be found here: https://backpack.tf/suggestion/5b15768e44325a5c6d075852#comment-5b435572cf6c7558a4778c3c

 

Quote
  • Cash Trades / Market Transactions
    • It's ok to use cash trades as proof, proof of the PayPal payment would be nice (personal information can be redacted of course). Bp.tf price for keys are acceptable now (like all prices it might be outdated tho so it's still recommended to be doublechecked before blindly using it).

^Currently the $ price of keys on backpack.tf is quite outdated and another ref-suggestion is needed to update it. 

 

Quote

^Many people asked if an item has a strangifier or not, so here is a good way to look this up yourselves! Searching for a strangifier on the SCM also works.

  • Like 6
Link to comment
Share on other sites

  • 11 months later...

Halloween 2018 weapons are finally fixed and priceable, to make up for the delay we will grant another 1 month of using older sales.

 

For a short, one-month "grace" period, we are allowing suggestions up to 6 months old on halloween 2018 decorated weapons.

  • We understand that some of these items are very rare and may have a single sale over 3 months old. And we understand that people have not given adequate opportunity to price these. Given this, until August 1, we will allow sales up to 6 months old on rare halloween 2018 decorated weapons with only a single sale that happened any time in the last 6 months. Normal suggestion rules will apply after August 1.
  • In situations where there have been sales in the last 3 months (even unusable sales), sales older than that cannot be used.
Link to comment
Share on other sites

With the upcoming pricing event, it's time to update some minor topics in the guide.

 

1. Change to rules on rounding: We no longer require rounding for unusuals below 5 keys in value. Unusuals at this low value are new and didn't exist when we last updated the rounding rules, to go with the time this change was necessary.

 

  • Rule - for all unusuals below 5 keys, rounding to the nearest 0.05 keys is required. No other decimals are allowed.
  • Rule - for all unusuals 5-10 keys, please round to the nearest 0.5 keys. No other decimals are allowed.
  • Rule - for all unusuals above 10 keys, please round to the nearest whole key
  • Guideline - for hats 100-200 keys, please round to the nearest 5 keys
  • Guideline - for hats 200-1000 keys, please round to the nearest 10 keys
  • Guideline - for hats 1000+ keys, please round to the nearest 50 keys

 

Clarification what do when the low end is below 5 keys for example and the high end above 5 keys: What counts here is the average unusual value. If it's above 5 keys, round to the nearest 0.5 keys, if it's below 5 keys, round to the nearest 0.05 keys on both ends. Same for the 10 key treshhold.

 

 

2. Using screenshots of bot.tf sales as proof: We now allow screenshots of bot.tf stats pages as proof if there are 3 sales at that value. We didn't allow them in the past because an item could have also been sold for items and not the pure buyout. With that being said, a majority of bot.tf sales still happen at their pure buyout so if there are 3 sales for a certain value, chances are incredibly high that at least 1 of them was for the pure buyout, or overpay in items that would be capped at the buyout. You can access that page here if you donated at least 1 key to bot.tf. 

 

 

3. Checking of cash sales is mandatory: This is now a rule. You have to check for item histories and buyouts on cash sales (this also includes marketplace sales as well as all other sites like that). If you can't find them state that in your suggestion. You can use bot.tfs listing snapshots for buyout finding or google cache for example.

  • Like 2
Link to comment
Share on other sites

  • 11 months later...
  • Administrators

We have received a lot of negative feedback on resuggestions only containing a link to their original suggestion (e.g. https://backpack.tf/suggestion/5f14e3938fc18043ea20d588 ). While it is clear to us as moderators where the value is coming from, it may not be as clear to an inexperienced observer. In addition, it requires unnecessary extra effort for anyone that wants to see where the price of a hat is coming from, especially when the original suggestion had a lot of comments/discussion on it.

We have generally granted leniency when this only happens occasionally, but I do not want this to become the de facto standard for resuggesting. That is why we now require the gathered proof to be present on the resuggestion. This means you will have to copy the evidence from your original suggestion - along with any potential new data provided - to your resuggestion, rather than just copying the link to the original suggestion. When you do this, don't forget to manage the spacing in your suggestion.

Also, be aware that any BBcode from your original suggestion will be lost. While it is not mandatory to fix this, it will help you retain your original layout.

Link to comment
Share on other sites

  • Administrators

We have recently seen a few cases where a cash trading site (e.g. marketplace, mannco etc.) has had a large fluctuation in their key values, resulting in unwanted inflation/deflation on hat that sell on days with particular low or high key values respectively.
An example is this:
https://backpack.tf/suggestion/5f063882ef79b65e3b2eb685
Keys were $1.76 on that day according to the graph. This graph however presents the median of sales that day, and is probably low because more keys were instantly dumped on that particular day than the ones that sold for their regular ~$1.9 (which was common at the time).
To compensate for such sudden fluctuations, we will now occasionally ask for an average key value rather than the value on the exact day itself. 

Because it is very difficult to make this a set-in-stone rule with a hard percentage cutoff, we are going to assess this on a case-by-case basis. In many situations this does not change anything: under normal circumstances, you can still continue to use the key value at the day of the sale.

However, if you notice anomalies such as the one in the suggestion linked and see it will cause a large difference in value, a good rule of thumb would be to use the average key value from that week. For the purpose of this guideline, the week is defined by the period stretching from 3 days before to 3 days after the sale (thus spanning a period of 7 days in total). Exceptional situations may call for other time frames, but this should give you an indication of how to go about it.

For the example suggestion, this means the following:
https://gyazo.com/747a1bbbb6da4df9c006c856dab616fd
These are the 7 values (seen in the gif) listed in chronological order
1.99
1.99
1.94
1.76
1.92
1.98
1.98
The average of these numbers is ~1.94, and this will be the key value used in the suggestion.

If you had not spotted the anomaly, we may request you to resuggest with a different key price based on this if it fits the situation, to keep the pricing as accurate as possible.

Link to comment
Share on other sites

  • 1 year later...
  • Administrators

In conjunction with a requirement change in non-unusuals, I also want to make use of the opportunity to use the change to enhance clarity for unusual suggestions. While not as commonly bungled like non-unusuals nowadays, we still get single-link suggestions with no explanations whatsoever. For these too, we will require you to list what the item sold for in writing on the suggestion. Even if your evidence screenshot contains all the information necessary, we still want you to write what the item sold for, so that even those that do not wish to click all sorts of links can still see where the value comes from.

https://backpack.tf/suggestion/614ce7657d70922fa55eb989 
A suggestion like this should, to this end, also include "sold for 65 keys" in text on it. The screenshot and item history should still be present in the case the compare links themselves are inconclusive. Theoretically, you may still give the compare links (given that they contain the item history), but since linking the item history is easier, we also allow that.

- The screenshot functions as evidence

- The item history is there so that we can check it for potential other exchanges

- The text indicating what the item sold for is there so that everyone can instantly see from looking at the suggestion itself where the range is coming from.
 

Link to comment
Share on other sites

  • 1 year later...
  • Administrators

Due to some recent controversy about sales in crypto and their accompanying mini, the pricing staff had discussed the use of crypto for suggestions, and we've concluded that it is best to NOT use sales consisting of only (or primarily) crypto transactions for price suggestion purposes.

The reason for this is mostly related to the uncertainty in crypto value. All crypto currencies have wildly fluctuating values making it very difficult to label them with one particular value. Even over the course of a few days (or even a few hours in some very specific cases), the changes can be so significant that their values will not by default fall within our uncertainty threshold. The less common the crypto in question is, the larger these fluctuations, but we've found that even fpr the most traded ones, fluctuations are too significant to rely on them.

Closely related to this, the low frequency with which crypto is used is an important factor in this decision. It leads to a lot of effort having to be put into something that is so rarely used. More importantly, due to the low traffic in crypto, it is very difficult - if not straight up impossible - to properly mini crypto for suggestions. For cash trades, this is already a hot button issue, but for cash, we have enough data to work with, and this data is significantly more stable, meaning a broader spectrum of data can be used in terms of time. So not only is there less data available within the same time span, but due to their erratic nature, the data quantity is further limited by the shorter time the data has before it "expires".

Another factor that weighed in our discussion was the fact that crypto is - in nature - anonymous in use, which makes it very difficult to track, document and prove the validity of your trades. While this is an issue that other such trades also suffer from, its worse for crypto. The issues alluded to above are more pressing, but this is something that was also taken into consideration.

Link to comment
Share on other sites

  • 1 month later...
  • Administrators

Our price moderators have no access to Mannco anymore, and along with the recent changes that do not allow the viewing of their sales graphs without logging in, we are no longer able to verify Mannco sales. Unfortunately, that means we can no longer allow trades that occur on Mannco for price suggestions.

  • Like 1
  • Confused 1
Link to comment
Share on other sites

  • 5 months later...
  • Administrators

Due to some confusion caused by unusual weapon and warpaint suggestions and the pages they can be found on, we've decided to require unusual weapons to be suggested on the "Decorated Weapon" Quality page. This has been the case for the majority of the weapons already, but some displayed as "Unusual" as quality, and were therefore suggested on that page; it will now be obligatory to suggest those on the Decorated Weapon stat page as well.

This standardization is meant as a QoL update to allow for easier access to unusual weapon and warpaint prices, as well as avoid confusion to all potential suggestors that wish to price their weapon. The unification does not influence already accepted suggestions, so there is no need for 'transfer' suggestions, but for any future updates it will be a requirement.

  • Like 1
  • Thanks 1
  • Haha 1
Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...