Suggestions Today for Items are Often Highly Lazy

[Baloo]

Back when evidence for suggestions first started becoming a requirement for suggestions; users were given a predetermined limit of up to 5 links to provide. However, when user Hugh Jazz began expanding his suggestions by posting additional links in the comments; it led to others following suit and the creation of the first post of every suggestion being the place for proof and explanation.

 

The reason I bring up this slice of backpack.tf history is because now it has been increasingly common for some backpack.tf users to simply screenshot a single listing from the classifieds, and use it as the sole basis for a price raise or drop. I understand if an item does not have many listings, or evidence is sparse, then it makes sense that this would be the only proof, but this often leads to an issue people would commonly call suggestors out for years ago. This was labeled as cherry-picking in which a suggestor may take a quicksell out-of-context and say it represents a larger shift in the market value of an item.

 

Nowadays I see suggestors who have multiple suggestions such as this one for the Strange Dead Head which only holds one screenshot of a close up of one listing from the classifieds. My assumption is if the user would look at a current screenshot of the classifieds, then he would again lower the price based solely on the listing of one individual who has only owned the expensive item for 2 days, and even less time when the original suggestion was made.

 

I personally find this lazy and damaging to the integrity of backpack.tf to allow users to provide minimal evidence for price changes. When I first started suggesting, I would see potential listings on TF2Outpost which would make me believe a suggestion may be possible. However, I was told by fellow suggestors who more seasoned at the time; if you cannot substantiate a suggestion, then its not worth suggesting. This site is meant to be a guide for prices, so reflecting actual changes meant more than one data point in order to draw conclusions.

[Findiculous]

Back when evidence for suggestions first started becoming a requirement for suggestions; users were given a predetermined limit of up to 5 links to provide. However, when user Hugh Jazz began expanding his suggestions by posting additional links in the comments; it led to others following suit and the creation of the first post of every suggestion being the place for proof and explanation.

 

The reason I bring up this slice of backpack.tf history is because now it has been increasingly common for some backpack.tf users to simply screenshot a single listing from the classifieds, and use it as the sole basis for a price raise or drop. I understand if an item does not have many listings, or evidence is sparse, then it makes sense that this would be the only proof, but this often leads to an issue people would commonly call suggestors out for years ago. This was labeled as cherry-picking in which a suggestor may take a quicksell out-of-context and say it represents a larger shift in the market value of an item.

 

Nowadays I see suggestors who have multiple suggestions such as this one for the Strange Dead Head which only holds one screenshot of a close up of one listing from the classifieds. My assumption is if the user would look at a current screenshot of the classifieds, then he would again lower the price based solely on the listing of one individual who has only owned the expensive item for 2 days, and even less time when the original suggestion was made.

 

I personally find this lazy and damaging to the integrity of backpack.tf to allow users to provide minimal evidence for price changes. When I first started suggesting, I would see potential listings on TF2Outpost which would make me believe a suggestion may be possible. However, I was told by fellow suggestors who more seasoned at the time; if you cannot substantiate a suggestion, then its not worth suggesting. This site is meant to be a guide for prices, so reflecting actual changes meant more than one data point in order to draw conclusions.

 

It is rather frustrating to see laziness like that pass with no issues and yet you spend a load of time on your own suggestion to make sure its correct.

My favorites are the Unusual Suggestions where the links are wrong or the prices of individual hats are incorrect but by some fluke, it evens itself out and the hat is "close enough" based off the comments of what more experienced suggestors figured out which is basically giving a pat on the back to the suggestor who didnt put in the research required for an accurate suggestion and saying "do it half arsed and somebody else will finish it for you but you can have the credit"

[Zcrab]

Is it really that bad though? He made a suggestion to drop the price of the item, and as you can see the item is definitely going down. Which means his suggestion was true and it even needs to go down further, so his suggestion actually made backpack.tf a better price guide for this item. It is just one screenshot, that suggestion did not take a long time to check and accept. Just like mods say that no function needs to be added to prevent proofless suggestions; because they only take five seconds to close, then why should a rule be made to avoid these kinds of suggestions when they don't take a long time to accept?

 

I think this has more to do with the fact some users are lazy and you do not like that.

[polar]

Can't compare how suggestions are done now to how they used to be done because trading is very different. Bots dominate a lot of the non-unusual market which means that buyer and seller gaps are often only 1 scrap apart. Bots are fine with the 1 scrap profit. That means that accepting a suggestion off buyers or sellers will usually put the value close to where it should be.

 

Obviously not always the case, but the non-unusual market is very different now. Classifieds are so heavily used for non-unusuals that they usually drive the market value on unique items.

 

Unusual guidelines / suggestions are rigorous as they have always been. Actually, the unusual suggestions now are quite a bit more rigorous with the introduction of compare links. It's actually expected to look through all hat histories. Definitely wasn't before.

[Baloo]

He made a suggestion to drop the price of the item, and as you can see the item is definitely going down.

 

The only evidence the price is really going down is the one person who was selling for 16 keys in the suggestion is now selling for 15 keys after the suggestion passed. Items worth multiple keys do not necessarily sell in as a high a volume as items worth a few refined, and you have to expect items to take longer if they're worth more as there are less people willing to buy them. The entire drop was based around a single person. If you would like to expand this further to another suggestion made by the same suggestor, then look at the Strange Potassium Bonnet. Another suggestion based solely on one seller, but was counter-acted by people pointing out other copies in the classifieds had sold. In both cases its even the same trader selling the respective items below backpack.tf price, and is the only one doing so.

 

Which means his suggestion was true and it even needs to go down further, so his suggestion actually made backpack.tf a better price guide for this item. It is just one screenshot, that suggestion did not take a long time to check and accept.

 

A single person being the sole influence to a price suggestion for an item with multiple copies for sale would often be closed for insufficient evidence in the past. The seller, who was the sole basis for the first suggestion could theoretically just keep listing the item lower while someone trailed behind and kept lowering the price making the item seem undesirable and 'cancerous' for traders to want to buy. It was a common problem with many expensive genuine items a while back because the price would drop every day or two based on sellers trying to undercut the next suggestion, and it just spiraled from there. Its no secret that even though backpack.tf tries to maintain a belief of being solely reflective of the market, and not influencing the market; it certainly plays a role in the mentality of buyers and sellers when it comes to the value and price changes related to items

 

why should a rule be made to avoid these kinds of suggestions when they don't take a long time to accept?

 

I am not at all advocating for a rule to be put in place, but rather I am here to voice my opinion, and explain my reasoning behind my belief. I said nothing about implementing a rule nor forcing suggestors to provide more information.

.

[apollothelion]

Can't compare how suggestions are done now to how they used to be done because trading is very different. Bots dominate a lot of the non-unusual market which means that buyer and seller gaps are often only 1 scrap apart. Bots are fine with the 1 scrap profit. That means that accepting a suggestion off buyers or sellers will usually put the value close to where it should be.

 

Obviously not always the case, but the non-unusual market is very different now. Classifieds are so heavily used for non-unusuals that they usually drive the market value on unique items.

 

Unusual guidelines / suggestions are rigorous as they have always been. Actually, the unusual suggestions now are quite a bit more rigorous with the introduction of compare links. It's actually expected to look through all hat histories. Definitely wasn't before.

I feel like that is untrue in regards to unusuals. I've been seeing a lot of suggestions pass that honestly have miscounts in the compare. As well as just skipping over USEABLE sales just because 'too hard'. 

[Baloo]

Can't compare how suggestions are done now to how they used to be done because trading is very different.

 

I am aware trading has changed over time. I do not believe suggestions should now rely on severely less evidence due to this change. When it comes to prices being driven up or down based solely on one individual listing in the classifieds, then I think there is a flaw in the site, and its ability to accurately record measurable market shifts in value for items. For non-unusual items a single data point does not encompass a larger picture of information.

[polar]

I feel like that is untrue in regards to unusuals. I've been seeing a lot of suggestions pass that honestly have miscounts in the compare. As well as just skipping over USEABLE sales just because 'too hard'. 

 

There will always be mistakes, especially with paypal traders like yourself who move around items in one-sided trades. And we check all sales to see if they are at least in the ballpark of sales that are usable. Sometimes you simply can't use a sale if the items in the sale were last priced a year ago. Makes no sense to use a year old price to update something priced 6 months ago. If a suggestion doesn't pass the eye test, it gets left up for a month with the hope another sale pops up. Sometimes our hands are tied after leaving a suggestion up for a month, and we accept it reporting the sale as is.

 

My point here being that there are mistakes still being made, but I can guarantee there are fewer mistakes than there used to be and a lot more sales being accounted for in each suggestion.

 

I am aware trading has changed over time. I do not believe suggestions should now rely on severely less evidence due to this change. When it comes to prices being driven up or down based solely on one individual listing in the classifieds, then I think there is a flaw in the site, and its ability to accurately record measurable market shifts in value for items. For non-unusual items a single data point does not encompass a larger picture of information.

 

Two other major changes in trading that I haven't already mentioned. For rare, marketable items, SCM sales happen more often than item sales. We definitely look at SCM values when we accept suggestions. Most buyers set their buy marks based on SCM trade values. So SCM was definitely factored into that suggestion and SCM sales are all around 16 keys.

 

Another major change now is that there are so many more items than there used to be that rarer items like the ones you are pointing out frequently don't sell more often than once a week in items. SCM sales are more frequent, but even there you can see only about 10 total sales in the last month. For some of these items, especially elite grade items or halloween items, there are frequently only a few sales over the course of weeks. A single sale / seller might actually be fairly representative of value. And when you see buyers at 14 and sellers going unsold for a long time at 18, 16 is no doubt closer to where that hat should be than 18.

[apollothelion]

There will always be mistakes, especially with paypal traders like yourself who move around items in one-sided trades. And we check all sales to see if they are at least in the ballpark of sales that are usable. Sometimes you simply can't use a sale if the items in the sale were last priced a year ago. Makes no sense to use a year old price to update something priced 6 months ago. If a suggestion doesn't pass the eye test, it gets left up for a month with the hope another sale pops up. Sometimes our hands are tied after leaving a suggestion up for a month, and we accept it reporting the sale as is.

 

My point here being that there are mistakes still being made, but I can guarantee there are fewer mistakes than there used to be and a lot more sales being accounted for in each suggestion.

 

I understand that, but I'm not talking about the ones that can't be used, I'm talking about the useable sales that are deliberately not done because 'too lazy' or 'too hard'. Afaik, it was used prior to go make a mini and actually update each thing and not just deem it 'too many hats' and oh its too much work so lets not do it.

 

If it is truly not useable, I'm all for ignoring it.

 

Furthermore the use of 'out-liar' sales, there needs to be a better guideline for that because they have been going about it rather strange.

Example:

 

Sale 1 : 38 2 months ago

 

Sale 2: 50 1 month ago

 

Sale 3: 45 1 month ago

 

Sale 4: 55 1 week ago

 

55 was considered the out-liar, rather than the old sale at 38. That doesn't make sense to me, wouldnt the oldest sale that was decently lower than the rest be considered the out-liar rather than a new sale that is in the realm of being in the correct suggestion. 38-50 vs 45-55.

[Baloo]

Another major change now is that there are so many more items than there used to be that rarer items like the ones you are pointing out frequently don't sell more often than once a week in items. SCM sales are more frequent, but even there you can see only about 10 total sales in the last month. For some of these items, especially elite grade items or halloween items, there are frequently only a few sales over the course of weeks. A single sale / seller might actually be fairly representative of value. And when you see buyers at 14 and sellers going unsold for a long time at 18, 16 is no doubt closer to where that hat should be than 18.

 

I am well aware of the TF2 economy as a whole, but one thing I must say is this:

 

 

backpack.tf currently values the Steam Market price around 18 keys, not 16 keys. As far as trying to legitimize quicksell buyers prices being relative to actual price, I find that a bit faulty. Quicksell buyers often gravitate around a 20% to 30% discount when it comes to items worth multiple keys. Its nothing new for a quicksell buyer to be multiple keys below backpack.tf price, especially on items which do not move fast. However, that isn't justification to say just because an item sits longer means the market equilibrium must be lower. I believe you should personally be aware of this.

[umer]

I well aware of the TF2 economy as a whole, but one thing I must say is this:

 

 

backpack.tf currently values the Steam Market price around 18 keys, not 16 keys. As far as trying to legitimize quicksell buyers prices being relative to actual price, I find that a bit faulty. Quicksell buyers often gravitate around a 20% to 30% discount when it comes to items worth multiple keys. Its nothing new for a quicksell buyer to be multiple keys below backpack.tf price, especially on items which do not move fast. However, that isn't justification to say just because an item sits longer means the market equilibrium must be lower. I believe you should personally be aware of this.

 

SCM sales have a 15% tax, which means the seller gets only 15.88 keys if that listing gets sold. Which is ~16 keys.

[Foamy the Fearsome]

I understand that, but I'm not talking about the ones that can't be used, I'm talking about the useable sales that are deliberately not done because 'too lazy' or 'too hard'. Afaik, it was used prior to go make a mini and actually update each thing and not just deem it 'too many hats' and oh its too much work so lets not do it.

 

If it is truly not useable, I'm all for ignoring it.

When not all sales are mentioned, we do check the rest of the sales. especially with single-sale suggestions. The only exceptions to that are suggestions that already contain 5 or more (pure) sales at or around a logical value (of course depending on how common unusuals are and how frequent they sell, as well as if there's any counter evidence or indications that even with a multitude of sales, the suggested value will still be off).

 

 

 

I understand that, but I'm not talking about the ones that can't be used, I'm talking about the useable sales that are deliberately not done because 'too lazy' or 'too hard'. Afaik, it was used prior to go make a mini and actually update each thing and not just deem it 'too many hats' and oh its too much work so lets not do it.

 

If it is truly not useable, I'm all for ignoring it.

 

Furthermore the use of 'out-liar' sales, there needs to be a better guideline for that because they have been going about it rather strange.

Example:

 

Sale 1 : 38 2 months ago

 

Sale 2: 50 1 month ago

 

Sale 3: 45 1 month ago

 

Sale 4: 55 1 week ago

 

55 was considered the out-liar, rather than the old sale at 38. That doesn't make sense to me, wouldnt the oldest sale that was decently lower than the rest be considered the out-liar rather than a new sale that is in the realm of being in the correct suggestion. 38-50 vs 45-55.

What case was this?

 

Outliers are determined based on a variety of different indicators. As you mentioned, time of the sale is one of them. With those specific values and no other info Id go for 38-55 or 45-50 (range with all, or a range with 1 high and 1 low outlier). Id see no point as to why 55 would be the only outlier there. Taking the time of the sales into consideration, Id probably have excluded 38.

 

However, there may have been other things in play. Sometimes, sales dont have accurate minis, or sometimes theres multiple sellers below a certain value, making it hard to justify certain sales as part of a range. Outlier sales always require a rationale in order to exclude them.

 

The problem with making a firm ground rule for determining outliers is that it varies depending on the situation. There are so many factors in play there that would obstruct default rules, that nearly every case should be considered as an exception to the 'basic rule' that says you to include every sale in the range unless there are compelling reasons to tighten it or to remove certain sales.

 

The only evidence the price is really going down is the one person who was selling for 16 keys in the suggestion is now selling for 15 keys after the suggestion passed. Items worth multiple keys do not necessarily sell in as a high a volume as items worth a few refined, and you have to expect items to take longer if they're worth more as there are less people willing to buy them. The entire drop was based around a single person. If you would like to expand this further to another suggestion made by the same suggestor, then look at the Strange Potassium Bonnet. Another suggestion based solely on one seller, but was counter-acted by people pointing out other copies in the classifieds had sold. In both cases its even the same trader selling the respective items below backpack.tf price, and is the only one doing so.

 

 

A single person being the sole influence to a price suggestion for an item with multiple copies for sale would often be closed for insufficient evidence in the past. The seller, who was the sole basis for the first suggestion could theoretically just keep listing the item lower while someone trailed behind and kept lowering the price making the item seem undesirable and 'cancerous' for traders to want to buy. It was a common problem with many expensive genuine items a while back because the price would drop every day or two based on sellers trying to undercut the next suggestion, and it just spiraled from there. Its no secret that even though backpack.tf tries to maintain a belief of being solely reflective of the market, and not influencing the market; it certainly plays a role in the mentality of buyers and sellers when it comes to the value and price changes related to items

 

 

I am not at all advocating for a rule to be put in place, but rather I am here to voice my opinion, and explain my reasoning behind my belief. I said nothing about implementing a rule nor forcing suggestors to provide more information.

.

I personally rarely handle non-unusuals (especially not lately) so I do not read or check the majority of them. I sometimes leave comments on the rarer items, saying that they should let the sellers mature depending on the item's value.

 

Some of the points made are valid concerns that I share, especially with regards to high-value non-unusuals being priced off of a single, young, unbumped classified seller. When I see those, I close them, especially if there are no other supportive indicators that the item should drop (or that the inevitable drop should not be as severe as suggested). In some specific cases I check whether or not the seller is an active trader, or whether or not they have trade holds. In more extreme cases, some sellers do not even have a trade link in their sell order, but just the option to add them. Items being sold by those people rarely ever sell, so they should not influence the range.

 

However, in many cases, sales are scarce and the market is small. In those cases there's not much we can do than to go by single sellers, as long as no counter evidence is provided.

[Foamy the Fearsome]

SCM sales have a 15% tax, which means the seller gets only 15.88 keys if that listing gets sold. Which is ~16 keys.

Not only that; the displayed value there is based on sellers, not sales. That often leads to discrepancies:

[Adolf Storms]

Yes. It's quite frustrating to watch as bots kill the price for lower unique and strange items often multiple times in a week. The finger of blame shouldn't be pointed at the people making the 1 item proof suggestions, or the mods for whatever reason that pass them... the blame points straightly at BOTS. Big Old Trolling Shits. B.O.T.S. Who seem to enjoy manipulating pricing up and down with no policing or restraint , the entire time using the listing system that was designed for personal traders to simply advertise. It started with auto-trades. It was molested by spirit (to prove a point? i'm told) and has permutated to over 2000 suggestions a month, over half of them being based on 1 or 2 listings with no proof of unsolds at current range only a listing that's matured more than a day. The thing is when the B.O.T.S. go down and you can't send trades, some other B.O.T.S. just step in and put a listing 1 scrap less. Or one weapon less. Want to make suggestions better? make tf2 better. Ban the bots from classifieds and force them to their own dedicated sites where they have to pay the upkeep and mntnc and scripting fees. But bp.tf wouldn't do this, cuz it's a cash cow for them with countless new premium users as B.O.T.S. gotta have that auto-rapid-bumping to keep up with them other bots.

 

Ahhh.... and I'm spent

 

just to sum up everything i've said here is 1 that's fresh in my mind - http://backpack.tf/stats/Unique/Iron%20Lung/Tradable/Craftable  3 drops in a week, 2 with only 1 item proof. ffs just wow

[polar]

55 was considered the out-liar, rather than the old sale at 38. That doesn't make sense to me, wouldnt the oldest sale that was decently lower than the rest be considered the out-liar rather than a new sale that is in the realm of being in the correct suggestion. 38-50 vs 45-55.

 

Outliers are determined in the context of (1) all sales in the last 3 months, (2) b/os on the market, current and past, and (3) offers, with priority given in that order. Further, we consider pure sales the single best indicator of item value and prioritize those above all else. If pure sales are 38, 45, 50 and b/os are all sitting at 50, then an unusual sale at 55 may very well be the outlier in that situation given it is least likely to reflect the pure value on the item.

 

At the end of the day, the beauty of the site is that all sales are recorded in suggestions for any user to see. And if any user has an issue with a suggestion accepted, they are more than welcome to fix it.

[polar]

backpack.tf currently values the Steam Market price around 18 keys, not 16 keys. As far as trying to legitimize quicksell buyers prices being relative to actual price, I find that a bit faulty. Quicksell buyers often gravitate around a 20% to 30% discount when it comes to items worth multiple keys. Its nothing new for a quicksell buyer to be multiple keys below backpack.tf price, especially on items which do not move fast. However, that isn't justification to say just because an item sits longer means the market equilibrium must be lower. I believe you should personally be aware of this.

 

As indicated above, the listed SCM price is the price of the lowest current seller, not sales which is a serious issue you have mentioned in an earlier thread that we're hoping fiskie will fix soon. Plus the tax thing.

 

The quickbuyers are paying 14 keys for the item you mention. I have never said the item should be priced at 14 keys. What I'm saying is that the value on the item lies somewhere between the unsuccessful sellers at 18 and the quickbuyers at 14. There was a seller at 16 in there. That matches the value of the sales on SCM. 16 to me is more reflective of its value than the previous price of 18. The price is not based on just a single seller. It's based on looking at all the other buyers and sellers on the market and looking at sales on SCM in addition to the sitting seller at 16.

[polar]

Yes. It's quite frustrating to watch as bots kill the price for lower unique and strange items often multiple times in a week. The finger of blame shouldn't be pointed at the people making the 1 item proof suggestions, or the mods for whatever reason that pass them... the blame points straightly at BOTS. Big Old Trolling Shits. B.O.T.S. Who seem to enjoy manipulating pricing up and down with no policing or restraint , the entire time using the listing system that was designed for personal traders to simply advertise. It started with auto-trades. It was molested by spirit (to prove a point? i'm told) and has permutated to over 2000 suggestions a month, over half of them being based on 1 or 2 listings with no proof of unsolds at current range only a listing that's matured more than a day. The thing is when the B.O.T.S. go down and you can't send trades, some other B.O.T.S. just step in and put a listing 1 scrap less. Or one weapon less. Want to make suggestions better? make tf2 better. Ban the bots from classifieds and force them to their own dedicated sites where they have to pay the upkeep and mntnc and scripting fees. But bp.tf wouldn't do this, cuz it's a cash cow for them with countless new premium users as B.O.T.S. gotta have that auto-rapid-bumping to keep up with them other bots.

 

Ahhh.... and I'm spent

 

just to sum up everything i've said here is 1 that's fresh in my mind - http://backpack.tf/stats/Unique/Iron%20Lung/Tradable/Craftable  3 drops in a week, 2 with only 1 item proof. ffs just wow

 

Honestly, I hear this every now and then and I just don't understand it. 

 

Bots offer competitive pricing. I think they are exactly the reason why everyone uses classifieds for buying and selling non-unusuals. If you are an interested buyer of an item, you can instantly find the cheapest seller and get your item instantly with automatic. No more waiting for users to come online. Competition among bots guarantees the lowest prices. If you are a seller, you can instantly find the most competitive buying prices. Competition among bots - bots that are fine with profits of 1 scrap - allows you to sell your item instantly for just 1 scrap less than the cheapest seller and you get to sell your item instantly. 

 

It's great for most buyers and sellers. 

 

So who is negatively affected by the bots? Really it's just profiteers in non-unusuals. But I really don't have sympathy for these users. Show some initiative, learn some coding, take advantage of your knowledge of the market and make your own bots with your own unique niche in the market. There is plenty of room when you consider how many items, especially rare items there are. Consider the huge halloween item market or even the strange part market. For the record, I don't mean you specifically. 

 

And regarding the influence on suggestions. Bots have helped tighten our confidence in our listed values because the buyer-seller gap that was once a gaping chasm has now shrunk to 1 scrap for many items. I've honestly considered at times to have pricing range from lowest seller to highest buyer. It works for most items but not all of them. Our pricing based on sellers alone is actually quite biased towards one side of the trade. Unfortunately, can't be done yet since there aren't yet competitive buyers for all items. 

[Kevin the Chicken God]

-snip-

In my experience bots are fine on smaller, ref items since the buyer-seller gap is so small, but when you get to larger items they are inconsistent proof at best. As an example, I listed a Spec KS Australium of mine for 1 key under every other seller, both Spec KS and normal. There were nearly 5 buyers at my price or above at this time, but since sending offers to bots with multiple keys almost never get accepted I decided to post the sell listing and wait it out. Within 6 hours 3 bots dropped their prices to match mine on the Spec KS page alone, and within a day a suggestion was made dropping this item despite the fact that each listing was young. (even though the suggester called them all 'mature' in his suggestion.) Since only the Spec KS bots dropped their prices to match mine, the normal item's prices (listed by the same bots) remained at 1 key above the Spec KS listings. There were still multiple bots buying above my price, so I saw this and sent them all a trade offer. After a few days all the trade offers I sent were declined or the items became unavailable, and the bots lowered their prices as soon as they saw the listings. The bot's programming is pretty rudimentary and it isn't difficult for one user to drop an item with dozens of classified listings. 

 

The fact that people can have their bots make hundreds of buy orders, then have to manually lower them once they see a trade offer exploiting their old price leads to a lot of invalid buyers, and the quick, reactionary listings and immediate suggestions lead to a domino effect of drops in the market value of items, which could be caused by a single user - just look at australium prices recently. I think mods have done a reasonable job at identifying buyers which no longer buy, but bots have really killed non-unusuals worth 12+ keys with how they immediately must be lowest seller on the market.

 

I just realized this rant isn't really that relevant to the thread, I don't really think it contributes a whole lot but I'm not deleting this text because screw that 

[Adolf Storms]

Honestly, I hear this every now and then and I just don't understand it. 

 

Bots offer competitive pricing. I think they are exactly the reason why everyone uses classifieds for buying and selling non-unusuals. If you are an interested buyer of an item, you can instantly find the cheapest seller and get your item instantly with automatic. No more waiting for users to come online. Competition among bots guarantees the lowest prices. If you are a seller, you can instantly find the most competitive buying prices. Competition among bots - bots that are fine with profits of 1 scrap - allows you to sell your item instantly for just 1 scrap less than the cheapest seller and you get to sell your item instantly. 

 

It's great for most buyers and sellers. 

 

So who is negatively affected by the bots? Really it's just profiteers in non-unusuals. But I really don't have sympathy for these users. Show some initiative, learn some coding, take advantage of your knowledge of the market and make your own bots with your own unique niche in the market. There is plenty of room when you consider how many items, especially rare items there are. Consider the huge halloween item market or even the strange part market. For the record, I don't mean you specifically. 

 

And regarding the influence on suggestions. Bots have helped tighten our confidence in our listed values because the buyer-seller gap that was once a gaping chasm has now shrunk to 1 scrap for many items. I've honestly considered at times to have pricing range from lowest seller to highest buyer. It works for most items but not all of them. Our pricing based on sellers alone is actually quite biased towards one side of the trade. Unfortunately, can't be done yet since there aren't yet competitive buyers for all items. 

 

TF2 just released! A first person shooter with team-based initiatives using 9 unique classes with distinctive class attributes!   Do you even tf2?    Trading is a side-game... I think you need to give your reality a check. I understand your position is solely based on the marketing aspect of the game, but them some pretty oblivious peepers ur battin XD I'm not throwin jabs just at you. I meet people on a daily basis who couldn't top-frag once in a casual map if they tried. Trade has changed a lot as you've aluded to, but prolly for reasons not as obvious as some players lack of tf2 gaming hours. Trading used to be a lot more 'social' of a thing. You'd pull up your groups list or perhaps your favorites list, hop onto a server and shoot the shit while lookin for some items you wanted perhaps in trade for some items you didn't use as much anymore or maybe some metal even if you had some. "Show some initiative... learn some coding" ?? DUDE! my brain shut off when they put a gun in my hand. kill kill kill

 

You speak of 'profiteers'... i assume these are the people who strictly look to profit from the game rather than add to the many gr8 communities or play in the casual or competitive maps? People who don't actually play tf2? ya, some people miss the point of the game. Guess the sad fact of the matter is some people forget the main focus of tf2 is a FPS not wall-street-market-simulator. #officially_off-topic_now

[apollothelion]

When not all sales are mentioned, we do check the rest of the sales. especially with single-sale suggestions. The only exceptions to that are suggestions that already contain 5 or more (pure) sales at or around a logical value (of course depending on how common unusuals are and how frequent they sell, as well as if there's any counter evidence or indications that even with a multitude of sales, the suggested value will still be off).

 

 

 

What case was this?

 

Outliers are determined based on a variety of different indicators. As you mentioned, time of the sale is one of them. With those specific values and no other info Id go for 38-55 or 45-50 (range with all, or a range with 1 high and 1 low outlier). Id see no point as to why 55 would be the only outlier there. Taking the time of the sales into consideration, Id probably have excluded 38.

 

However, there may have been other things in play. Sometimes, sales dont have accurate minis, or sometimes theres multiple sellers below a certain value, making it hard to justify certain sales as part of a range. Outlier sales always require a rationale in order to exclude them.

 

The problem with making a firm ground rule for determining outliers is that it varies depending on the situation. There are so many factors in play there that would obstruct default rules, that nearly every case should be considered as an exception to the 'basic rule' that says you to include every sale in the range unless there are compelling reasons to tighten it or to remove certain sales.

 

I personally rarely handle non-unusuals (especially not lately) so I do not read or check the majority of them. I sometimes leave comments on the rarer items, saying that they should let the sellers mature depending on the item's value.

 

Some of the points made are valid concerns that I share, especially with regards to high-value non-unusuals being priced off of a single, young, unbumped classified seller. When I see those, I close them, especially if there are no other supportive indicators that the item should drop (or that the inevitable drop should not be as severe as suggested). In some specific cases I check whether or not the seller is an active trader, or whether or not they have trade holds. In more extreme cases, some sellers do not even have a trade link in their sell order, but just the option to add them. Items being sold by those people rarely ever sell, so they should not influence the range.

 

However, in many cases, sales are scarce and the market is small. In those cases there's not much we can do than to go by single sellers, as long as no counter evidence is provided.

 

As I said previously, yes some hats are priced accurately and 5 pure sales is fine.  I am not talking about the suggestions that are done by experienced people and take the time to put in the effort, I am talking about the ones done wrong and have been passed in a way that makes me wonder WHY????

 

Some of this was based off a few hats I'd seen suggestions on, they might have been fixed in the new suggestion but some of the ones I was mentioning were the hearts baker, hearts muffs, and gfetti veil. IIRC. 

 

I was just looking at the Gfetti Veil again because someone asked me for a price on one, and I noticed that MY sale for the duped hearts tyrants was priced at ~95 through a bad mini, when in fact if you look it seems a bit later an actual suggestion was done which priced it better around ~80. 

 

 

Outliers are determined in the context of (1) all sales in the last 3 months, (2) b/os on the market, current and past, and (3) offers, with priority given in that order. Further, we consider pure sales the single best indicator of item value and prioritize those above all else. If pure sales are 38, 45, 50 and b/os are all sitting at 50, then an unusual sale at 55 may very well be the outlier in that situation given it is least likely to reflect the pure value on the item.

 

At the end of the day, the beauty of the site is that all sales are recorded in suggestions for any user to see. And if any user has an issue with a suggestion accepted, they are more than welcome to fix it.

Yes, but its not always heard lol. I've seen multiple suggestions that should be waiting a month, for more sales, just get accepted anyways.

 

sales at 38 45 50 55, and b/o has ranged from 50-55. Then should the LATEST sales be considered to reflect the pure value on the item. I'm talking about PURE sale at 55. if 38 was done 2 months ago, that should be the out-liar right? (They are all pure sales)

 

There should be a flat rule for suggestions that have 1 sale on a NOT rare hat, it should wait 1 month for sure. Make it blanket so there isn't any wishy-washyness.

[Spirit]

Yes. It's quite frustrating to watch as bots kill the price for lower unique and strange items often multiple times in a week. The finger of blame shouldn't be pointed at the people making the 1 item proof suggestions, or the mods for whatever reason that pass them... the blame points straightly at BOTS. Big Old Trolling Shits. B.O.T.S. Who seem to enjoy manipulating pricing up and down with no policing or restraint , the entire time using the listing system that was designed for personal traders to simply advertise. It started with auto-trades. It was molested by spirit (to prove a point? i'm told) and has permutated to over 2000 suggestions a month, over half of them being based on 1 or 2 listings with no proof of unsolds at current range only a listing that's matured more than a day. The thing is when the B.O.T.S. go down and you can't send trades, some other B.O.T.S. just step in and put a listing 1 scrap less. Or one weapon less. Want to make suggestions better? make tf2 better. Ban the bots from classifieds and force them to their own dedicated sites where they have to pay the upkeep and mntnc and scripting fees. But bp.tf wouldn't do this, cuz it's a cash cow for them with countless new premium users as B.O.T.S. gotta have that auto-rapid-bumping to keep up with them other bots.

 

Ahhh.... and I'm spent

 

just to sum up everything i've said here is 1 that's fresh in my mind - http://backpack.tf/stats/Unique/Iron%20Lung/Tradable/Craftable  3 drops in a week, 2 with only 1 item proof. ffs just wow

I agree with you Adolf =D. Yup I am in it to prove a point =D. As always i am more than willing to offer my Trade Histories of any bot, for a low price of 1 Ref per page of history, or 1 Ref for a week history of ONE Item) =P

I agree that all bots should be banned from site designed for human use, such as backpack.tf, tf2outpost and bazaar.tf

They may be causing pricing to be accurate, due to some of them...."Spirit-Bots" being capable of displaying the TRUE supply and demand of an item.

This true price, with a single scap difference might be great for price suggestions, but is causing new traders to give up, feel hopeless, and wont get anywhere.

So with new traders out of the picture, there is only going to be less and less traders. And i am really excited for that

 

 

In my experience bots are fine on smaller, ref items since the buyer-seller gap is so small, but when you get to larger items they are inconsistent proof at best. As an example, I listed a Spec KS Australium of mine for 1 key under every other seller, both Spec KS and normal. There were nearly 5 buyers at my price or above at this time, but since sending offers to bots with multiple keys almost never get accepted I decided to post the sell listing and wait it out. Within 6 hours 3 bots dropped their prices to match mine on the Spec KS page alone, and within a day a suggestion was made dropping this item despite the fact that each listing was young. (even though the suggester called them all 'mature' in his suggestion.) Since only the Spec KS bots dropped their prices to match mine, the normal item's prices (listed by the same bots) remained at 1 key above the Spec KS listings. There were still multiple bots buying above my price, so I saw this and sent them all a trade offer. After a few days all the trade offers I sent were declined or the items became unavailable, and the bots lowered their prices as soon as they saw the listings. The bot's programming is pretty rudimentary and it isn't difficult for one user to drop an item with dozens of classified listings. 

 

The fact that people can have their bots make hundreds of buy orders, then have to manually lower them once they see a trade offer exploiting their old price leads to a lot of invalid buyers, and the quick, reactionary listings and immediate suggestions lead to a domino effect of drops in the market value of items, which could be caused by a single user - just look at australium prices recently. I think mods have done a reasonable job at identifying buyers which no longer buy, but bots have really killed non-unusuals worth 12+ keys with how they immediately must be lowest seller on the market.

If a bot lowered their price to match a human, that is a bad bot.

 

You speak of 'profiteers'... i assume these are the people who strictly look to profit from the game rather than add to the many gr8 communities or play in the casual or competitive maps? People who don't actually play tf2? ya, some people miss the point of the game. Guess the sad fact of the matter is some people forget the main focus of tf2 is a FPS not wall-street-market-simulator. #officially_off-topic_now

Yeah some bot owners don't even play tf2 anymore (I.... On the other hand.... am a really good sniper =P and participate in Highlander =P as well as was highly involved with in-game NW American Matchmaking-Community =P)

I TOTALLY still play tf2

 

 

Back when evidence for suggestions first started becoming a requirement for suggestions; users were given a predetermined limit of up to 5 links to provide. However, when user Hugh Jazz began expanding his suggestions by posting additional links in the comments; it led to others following suit and the creation of the first post of every suggestion being the place for proof and explanation.

 

The reason I bring up this slice of backpack.tf history is because now it has been increasingly common for some backpack.tf users to simply screenshot a single listing from the classifieds, and use it as the sole basis for a price raise or drop. I understand if an item does not have many listings, or evidence is sparse, then it makes sense that this would be the only proof, but this often leads to an issue people would commonly call suggestors out for years ago. This was labeled as cherry-picking in which a suggestor may take a quicksell out-of-context and say it represents a larger shift in the market value of an item.

 

Nowadays I see suggestors who have multiple suggestions such as this one for the Strange Dead Head which only holds one screenshot of a close up of one listing from the classifieds. My assumption is if the user would look at a current screenshot of the classifieds, then he would again lower the price based solely on the listing of one individual who has only owned the expensive item for 2 days, and even less time when the original suggestion was made.

 

I personally find this lazy and damaging to the integrity of backpack.tf to allow users to provide minimal evidence for price changes. When I first started suggesting, I would see potential listings on TF2Outpost which would make me believe a suggestion may be possible. However, I was told by fellow suggestors who more seasoned at the time; if you cannot substantiate a suggestion, then its not worth suggesting. This site is meant to be a guide for prices, so reflecting actual changes meant more than one data point in order to draw conclusions.

AND NOW.... for the point of the topic =P

 

A. Don't ever.... ban me again.... i will rip you a new one.

B. That suggestion is absurd... A suggestion that only contains one piece of proof (at a selective price)... when many more are available, should never be accepted as a new price for an item.

If i ever see... that someone suggests a new price for an item, witch wildly different from the current price or the price that i am trading the item at. I always offer to release my trade history of that item for may it be a week or a month.

Unfortunately users don't always care for the price of an item enough to spend 1 ref on my weeks history of an item. Although it may have been traded 10+ times within the 1 week =P

I am glad that you have brought this issue up Ballo, I see your concern and agree with it.... Just don't touch me again.

[Baloo]

-snip-

 

First, thank you for understanding my point of view on this subject. Secondly, umm...I suppose a written warning on your trade would have been better in retrospect. Brokering is a service, so offering to broker would be an infraction of Outpost's rules, but I do understand it is frustrating to receive a ban for it immediately. My apologies. 

[Spirit]

-Snippity snap-

My respect for you has gone up. And i am sorry that my understanding / comprehension of the rules was flawed. I did try my best to not break them =P

If anything else that i do is wrong let me know =D Thank you for making it only 3 days at least =D

[Zeus_Junior]

I read through like 3/4 of all this and I must say that SOME items require more proof than just one seller, for example the strange executioner, strange steel shako and the strange dark falkirk helm.

Those three dropped SO MUCH in a matter of days. But items that are pretty stable in price Should just be able to use one seller IF IT'S A MATURE ONE

[ѕιи]

Hey Baloo, you mentioned suggestions getting lazier and lazier, but now you're doing the same thing. I really don't want to start anything, but can you explain this?

 

Examples:

http://backpack.tf/suggestion/584a0c200e2cad51814244a7

http://backpack.tf/suggestion/5842df5886674714122b4eb2

http://backpack.tf/suggestion/5842d99f83de6b1dff6e200f

http://backpack.tf/suggestion/5842d4e0866747405a15b2d2

http://backpack.tf/suggestion/5842d41183de6b77897eeffb

http://backpack.tf/suggestion/5842d3a7c440451b2e7d33ff

http://backpack.tf/suggestion/5842d33f83de6b6c3d6957ae

http://backpack.tf/suggestion/5842d2968667477b635f9eb0

http://backpack.tf/suggestion/5842d228c4404565f241ac22

http://backpack.tf/suggestion/5842d16bc440452578798e82

http://backpack.tf/suggestion/5842d134c4404566183d5be5

http://backpack.tf/suggestion/5841ebfd83de6b63b5644150

 

Thanks! :3