BP.TF Unusual pricelist - Dies a death unless...

[AwesomeMcCoolName]

When bp.tf has so few prices in it's list it becomes redundant.  Waiting for proof will help revalue items of course but by relying on that process alone is why it can't keep up.  Voting is vital to keep items from being hyped or deflated but waiting for sale proof just aint keeping up.

 

Question. If bp.tf cant keep prices on old or new hats upto date now, how is it going to look after Halloween and Xmas?

 

Answer. More like this I imagine:

http://backpack.tf/u...ner's_Pompadour

http://backpack.tf/u...sgow_Great_Helm

http://backpack.tf/u...30135/Wet_Works

http://backpack.tf/u.../Whirly_Warrior

http://backpack.tf/u...14/Valley_Forge

http://backpack.tf/u.../Plumber's_Pipe

http://backpack.tf/u...Burning_Bandana

http://backpack.tf/u...tual_Viewfinder

http://backpack.tf/u...43/Virus_Doctor

http://backpack.tf/u...nfernal_Impaler

http://backpack.tf/u.../Tartan_Spartan

http://backpack.tf/u...54/Bunsen_Brave

http://backpack.tf/u...yrantium_Helmet

 

....and those are only new-ish hats.. there's hundreds of outdated prices on old hats too.

and most of those have traceable sales. People just don't know what they should suggest on, and because of the current system, people need to focus on updating already priced unusuals rather than pricing an unpriced unusual as the former as easier. 

[F CHARLIE]

There is one valid point here. The volume of effects that will have been released will add 50% to the amount available in one yesr. It has created valuation problems. Many traders will not accept new effects regardless. I have always said that we should be concerned about owners who are considering selling their hat.

[Brad Pitt]

If you wanted to create a formula, you'd need to take into account the effect depending on the classes the hat can be used on (single, multiple) -- engy should have a bonus with nuts and bolts, sniper with flies, etc. So you'd need to maintain a bonus ratio for each effect for each class. Then some hats get bonuses because they are "themed" with specific effects. There are probably other exceptions to take care of... And these bonuses may change at any point depending how the trading community decides which effects they like more as time goes on.

 

It's an interesting idea and something I contemplated but in reality it's pretty hard. That's why the USD storage seemed to make sense, especially with the steam market allowing unusual sales. I don't think the TF2 economy will ever be logical enough to create algorithms to predict accurate prices at any point... 

[Meltron]

If you wanted to create a formula, you'd need to take into account the effect depending on the classes the hat can be used on (single, multiple) -- engy should have a bonus with nuts and bolts, sniper with flies, etc. So you'd need to maintain a bonus ratio for each effect for each class. Then some hats get bonuses because they are "themed" with specific effects. There are probably other exceptions to take care of... And these bonuses may change at any point depending how the trading community decides which effects they like more as time goes on.

 

It's an interesting idea and something I contemplated but in reality it's pretty hard. That's why the USD storage seemed to make sense, especially with the steam market allowing unusual sales. I don't think the TF2 economy will ever be logical enough to create algorithms to predict accurate prices at any point... 

True. Still you apply bonus prices to items which have special parts or paints

[AwesomeMcCoolName]

True. Still you apply bonus prices to items which have special parts or paints

thats different. Unusuals are highly subjective, pricing unusuals at all doesn't work all that well by the very nature of unusuals. 

[F CHARLIE]

If you wanted to create a formula, you'd need to take into account the effect depending on the classes the hat can be used on (single, multiple) -- engy should have a bonus with nuts and bolts, sniper with flies, etc. So you'd need to maintain a bonus ratio for each effect for each class. Then some hats get bonuses because they are "themed" with specific effects. There are probably other exceptions to take care of... And these bonuses may change at any point depending how the trading community decides which effects they like more as time goes on.

 

It's an interesting idea and something I contemplated but in reality it's pretty hard. That's why the USD storage seemed to make sense, especially with the steam market allowing unusual sales. I don't think the TF2 economy will ever be logical enough to create algorithms to predict accurate prices at any point... 

 

 

A grid could be used based on the avg price for that effect on other hats.  It could even give a range based on class.

[+SPONGE+ NUTTER]

A grid could be used based on the avg price for that effect on other hats.  It could even give a range based on class.

 +1

 

It shouldn't be too difficult to put scale together considering all the current known values.  Once you have the formula referencing a table/grid then estimates should be pretty accurate.  If the table/grid contains hats based on a simple points system then multiplying by additional points/value for effect based and then rarity scale should work quite nicely imho.  For any hats that fall outside the norm (themed hats) you simply reference a modified effect scale (copy paste edit).

 

This should only be for new or outdated prices so as long as the value is labeled as 'estimate' then traders will expect some variation. Once a trade has been record of a 'new' hat type it will set a baseline to automatically determine the value of that hat type with all the particle effects.  Let voting and sales proof continue to tweak prices further and you end up with a low maintenance and accurate price list.

 

Figures to work with:

Unusual Hat type points table. Allocate points for each hat type e.g. (TC=80, Saumr-eye=40, Dead Cone=10 etc...).  New hats can be given an arbitrary value in points to set a minimum guide price.  As soon as a sale is recorded it will set an accurate baseline.  The hat point table should be all you need to adjust and votes carried on valuations set this reference,  This way you only need votes on one hat rather than for every effect and the whole system will become fluid, like any other stock market index (which is what it is, only for virtual stock).

x

Calculate Rarity Scale based on current average difference for all unusuals. e.g.1of1=10points 1of2=4points 1of3=2 points etc.  Should be easy to calculate from db query.

x

Calculate a Particle Effect Multiplier from average percentage differences for each different effect. db query.

=

Output Value in keys then convert to buds (conversion exists).

 

 

As there are many hats that have up to date values it will make it very easy to test for accuracy and tweak where needed and adding new effects and hats will be quick.  There's nothing to stop you tweaking the estimated values down to set a minimum price instead of going for all out pinpoint accuracy on new hats. either way there's a real opportunity to set a benchmark 

[F CHARLIE]

+1

 

It shouldn't be too difficult to put scale together considering all the current known values.  Once you have the formula referencing a table/grid then estimates should be pretty accurate.  If the table/grid contains hats based on a simple points system then multiplying by additional points/value for effect based and then rarity scale should work quite nicely imho.  For any hats that fall outside the norm (themed hats) you simply reference a modified effect scale (copy paste edit).

 

This should only be for new or outdated prices so as long as the value is labeled as 'estimate' then traders will expect some variation. Once a trade has been record of a 'new' hat type it will set a baseline to automatically determine the value of that hat type with all the particle effects.  Let voting and sales proof continue to tweak prices further and you end up with a low maintenance and accurate price list.

 

Figures to work with:

Unusual Hat type points table. Allocate points for each hat type e.g. (TC=80, Saumr-eye=40, Dead Cone=10 etc...).  New hats can be given an arbitrary value in points to set a minimum guide price.  As soon as a sale is recorded it will set an accurate baseline.  The hat point table should be all you need to adjust and votes carried on valuations set this reference,  This way you only need votes on one hat rather than for every effect and the whole system will become fluid, like any other stock market index (which is what it is, only for virtual stock).

x

Calculate Rarity Scale based on current average difference for all unusuals. e.g.1of1=10points 1of2=4points 1of3=2 points etc.  Should be easy to calculate from db query.

x

Calculate a Particle Effect Multiplier from average percentage differences for each different effect. db query.

=

Output Value in keys then convert to buds (conversion exists).

 

 

As there are many hats that have up to date values it will make it very easy to test for accuracy and tweak where needed and adding new effects and hats will be quick.  There's nothing to stop you tweaking the estimated values down to set a minimum price instead of going for all out pinpoint accuracy on new hats. either way there's a real opportunity to set a benchmark 

 

Rarity has less effect in the long run. Desire is more relevant. Some hats are common but never on the market. I have a PE Berliners, the only one available I think, and no one will pay what I am asking; 46 keys which is well below list.

 

An example would be any new NnB. The range would be 12 - 40 keys aprox ( I don't know what the highest price NnB is). While it seems like a large range it could be made better by filtering by class. It could be usefull in cutting back on hype price but Hype pricing is an incentive in trading and traders will complain even more if we have unproven prices. The grid would have to be a seperate page for estimates only.

 

I do think it would be helpfull to bring sanity to the U market but in reality I think most people intuitively arrive at sane price offer price anyway. This is why we don't allow hype sale as proof. I used to judge the effect compared to others and use that to calculate quickbuy offers.

[+SPONGE+ NUTTER]

Rarity has less effect in the long run. Desire is more relevant. Some hats are common but never on the market. I have a PE Berliners, the only one available I think, and no one will pay what I am asking; 46 keys which is well below list.

 

An example would be any new NnB. The range would be 12 - 40 keys aprox ( I don't know what the highest price NnB is). While it seems like a large range it could be made better by filtering by class. It could be usefull in cutting back on hype price but Hype pricing is an incentive in trading and traders will complain even more if we have unproven prices. The grid would have to be a seperate page for estimates only.

 

I do think it would be helpfull to bring sanity to the U market but in reality I think most people intuitively arrive at sane price offer price anyway. This is why we don't allow hype sale as proof. I used to judge the effect compared to others and use that to calculate quickbuy offers.

 

Agreed, rarity has less effect in the long run but every aspect that affects value should be included in the calculation.  I think we can all agree that hat type and effect are the most important.  I appreciate that class does affect value but given the hat type itself dictates the class it may be unnecessary to include both.

 

Regarding your PE Bucket Helm.  You can see you're already affected by the current problem.

http://backpack.tf/unusuals/303/Berliner%27s_Bucket_Helm

Out of the 56 Unusual versions listed:

9 of them have old/outdated prices (including your PE).

10 of them have no prices at all!

This is an old hat too and is already close to having values listed on only 50% of them.  As Halloween and Xmas adds even more you can appreciate it's going to get a lot worse and soon.  If prices can be calculated based on the above suggestions it may stop the list failing altogether.

 

 Regarding the bullshit term 'ON THE MARKET' (and this is not aimed at you personally). Many hats sell outside tf2op tf2WH etc., like on trade servers and privately between friends/traders and other Unusual trading servers and forums all over the world. Those hats\trades are also part of 'THE MARKET' The amount of times I see and hear "only one on the market" the reality being only one on tf2op in 99% of cases

 

The unusual market is WORLDWIDE. I for one trade on Korean, Russian, US, UK, French, Australian and Italian servers regular    Google translate and TF2 wiki loaded into Steam browser helps

If traders assume it's only whats sold on 2 or 3 trading sites/forums then they're sorely mistaken.

 

So back to the topic...

 

Are there any mathematicians out there with some db query experience?  I'd be happy to put up a few buds for anyone that can crack this

[Cedarium]

Sorry Cedar, I can only really respond properly to your comments if they relate to either the OP or subsequent posts. Might help if you read and understood the thread before asking irrelevant and unrelated questions.

 

My responses:

 

• Yes,
• This is why the pricelist is failing,
• Obviously not?
• No one is discussing craft hats (read OP),

I read your posts, but very simply algorithms aren't successful, ever. We had the lowcrafts and vintages, also effect tiers don't always effect this. Would an effect tier sugguest that a scorching Tc would be 50ish when burning is 140ish (same for kes), or that a scorching tossle would be above a burning? Effect tiers never work.

[+SPONGE+ NUTTER]

I read your posts, but very simply algorithms aren't successful, ever. We had the lowcrafts and vintages, also effect tiers don't always effect this. Would an effect tier sugguest that a scorching Tc would be 50ish when burning is 140ish (same for kes), or that a scorching tossle would be above a burning? Effect tiers never work.

 

Both the TC and Tossle are not only old hats with new effects but both have been valued based on sales then votes therefore reflect market price.  We're not suggesting that system changes.

 

What we're trying to establish is a mechanism for pricing new hats and effects coming onto the market and valuation on out of date prices.  Once a basline price has been set for new effects and hats, voting continues to allow prices to reflect actual sales. For out of date values, each can be estimated based on their relativity to already valued hats.  As long as these 'estimated' values are flagged as such then traders wont expect pinpoint accuracy until a vote has been passed.  This will provide values for everything that doesn't rather than having a pricelist with no values at all.  Hats that have votes passed and reflect true market value won't be affected by the auto valuation.  The formula is only applied to un-valued or out of date items.

[AwesomeMcCoolName]

As long as these 'estimated' values are flagged as such then traders wont expect pinpoint accuracy until a vote has been passed. 

people don't read. If you want estimations with absolutely nothing of any actual value to support it, go to tf2pricecheck.net or ask for a pc on a server. 

[Cedarium]

Both the TC and Tossle are not only old hats with new effects but both have been valued based on sales then votes therefore reflect market price. We're not suggesting that system changes.

 

What we're trying to establish is a mechanism for pricing new hats and effects coming onto the market and valuation on out of date prices. Once a basline price has been set for new effects and hats, voting continues to allow prices to reflect actual sales. For out of date values, each can be estimated based on their relativity to already valued hats. As long as these 'estimated' values are flagged as such then traders wont expect pinpoint accuracy until a vote has been passed. This will provide values for everything that doesn't rather than having a pricelist with no values at all. Hats that have votes passed and reflect true market value won't be affected by the auto valuation. The formula is only applied to un-valued or out of date items.

These algorithms could skew the market though. Instead of actually getting to see what out hat sells for, maybe the price was much too high so it never sold, or too low so it was snatched up in an instant. Algorithms aren't good.

[+SPONGE+ NUTTER]

These algorithms could skew the market though. Instead of actually getting to see what out hat sells for, maybe the price was much too high so it never sold, or too low so it was snatched up in an instant. Algorithms aren't good.

 

LOL.  Algorithms work for every other global market economy and your opinion is based on something that has not been tried.  What skews the market in reality is when traders have no reference point so make up their own values rather than the opposite. This is what used to happen when trading first began with craft items until spreadsheets started appearing

[AwesomeMcCoolName]

LOL.  Algorithms work for every other global market economy and your opinion is based on something that has not been tried. 

Craft #s, odd lvl'd vintages, paints and strange part bonuses. None of those algorithms are correct, and they've all skewed the market (craft #/lvld vintages especially).

[+SPONGE+ NUTTER]

Either way, if neither a formula nor a voting panel can be established to update old values or set prices on new Unusuals or effects then the Unusual pricelist will cease to function because it can't keep up.  Why not try suggesting something that might 'help'.  So far all we have here are reasons why not to try anything.

[Cedarium]

Either way, if neither a formula nor a voting panel can be established to update old values or set prices on new Unusuals or effects then the Unusual pricelist will cease to function because it can't keep up.  Why not try suggesting something that might 'help'.  So far all we have here are reasons why not to try anything.

Revealing the secret belt

[+SPONGE+ NUTTER]

Revealing the secret belt

 

Not helpful.  If you have nothing useful to add try the flame/troll bucket.

[Cedarium]

Not helpful.  If you have nothing useful to add try the flame/troll bucket.

Do you even know what the secret belt is? It's the prize for 1000 accepted suggestions.

[+SPONGE+ NUTTER]

Do you even know what the secret belt is? It's the prize for 1000 accepted suggestions.

 

So what's that got to do with this thread?  Neither helpful or relevant.

[Professional Map Painter]

Why not try suggesting something that might 'help'.  

http://www.tf2pricecheck.net/index.php

[+SPONGE+ NUTTER]

http://www.tf2pricecheck.net/index.php

 

Referring Unusual traders to TF2pricecheck instead of proposing a solution to help seems a little lame at best.  You're a donator too so seems pretty to be a dumb way of showing your support. encouraging traders to go elsewhere.

[Liddojunior]

Referring Unusual traders to TF2pricecheck instead of proposing a solution to help seems a little lame at best.  You're a donator too so seems pretty to be a dumb way of showing your support. encouraging traders to go elsewhere.

 

Price check from several traders opinions can be helpful for sellers as much as seeing a price defined by known past sale averages is helpful to buyers

[+SPONGE+ NUTTER]

So...

 

8 new Halloween effects

Approx 260 hats that can have these effects.

Estimated number of recorded sales (proof) required to update the Unusual Pricelist for all new effects = 2080

 

One thing for sure, voting on sales evidence is not gonna clear the backlog by Xmas, when there will be even more effects to add

Needs to change before it dies.

[AwesomeMcCoolName]

So...

 

8 new Halloween effects

Approx 260 hats that can have these effects.

Estimated number of recorded sales (proof) required to update the Unusual Pricelist for all new effects = 2080

 

One thing for sure, voting on sales evidence is not gonna clear the backlog by Xmas, when there will be even more effects to add

Needs to change before it dies.

There really isn't an alternative. I don't know a single person who didn't completely shit on tf2pc; even more so than they do on bptf. Bp.tf is at least based on proof so when someone cries about their unusual being priced incorrectly you can easily show them proof as to why it was priced that way. Unfortunately with what your suggesting there wouldn't be any proof and bptf would become the new tf2pc (which is completely dead). 

 

And even if you put a warning saying this is just an estimation it wouldn't matter. People are too stupid to figure that out (e.g. look at all the people who accuse you of scamming if you sell for more/offer less).