Question: How many 1 of 1 hats do you estimate there are?

[Flamadin]

And as a bonus, what do you think the minimum retail sale price would be for one, and what do you think a typical sale price would be for an all-class 1 of 1?

 

[Slam]

It's hard to actually figure that out but I'm bored enough and have time to kill so fuck it I'll use basic statistics and take a shot in the dark. If someone's better at math or there's literally an API or spreadsheet that has this info readily available already then lemme know I wasted 20 minutes of my time lmao.

 

Using https://backpack.tf/unusuals and not including any glitch / spelled / untradable unusual hats as being "unique 1 of 1", I'll take 20 random hats, 10 of each "era", and count the 1-of-1's that exist for each. The first era being pre-case-era (PCE), the last era being case-era (CE). The reason being there's significantly more effects available to case-era hats than non-case-era hats and thus a higher chance for 1-of-1's, so it's good to do an even split of both.

 

Total number of 1-of-1 versions according to bp.tf:

(Note 1: Not including Community Sparkle either, as that for some reason appears on some of these.)

(Note 2: I'm including ones that say "0 exist" here since they're almost certainly 1 of 1's that are just in private inventories or the SCM or something)

 

Killer's Kabuto: 13

Team Captain: 8

Rather Festive Tree: 10

Trippler's Tricorne: 8

Gym Rat: 15

Madame Dixie: 10

Officer's Ushanka: 12

Spook Specs: 8

Brainiac Hairpiece: 12

Brotherhood of Arms: 7

 

Caribou Companion: 19

Misfortune Fedora: 10

Head Prize: 18

Brass Bucket: 18

Citizen Kane: 8

Cranial Carcharodon: 28

Patriot Peak: 2

Big Topper: 22

Brain Warming Wear: 9

Bat Hat: 11

 

Adding all of those numbers together you get 238, divided by 20 you get an average of 11.9 1-of-1's per hat. Might be high or low but whatever close enough for this.

 

The total number of hats that can be unusual according to bp.tf is 514. We get this from https://backpack.tf/unusuals, counting 57 rows times 10 per row then adding the last 7 left over = 577, then CTRL+F "taunt:" and subtract that number for taunts to equal 514 total unusual hats.

 

So from this you simply do 514 times 11.9 to get about 6,117 total 1-of-1 unusual hats that exist in TF2.

 

But I'm gonna say that's high just because there's probably a good chunk of hats that only have a couple of 1-of-1's like the Patriot Peak but not a lot of hats that have a huge amount of 1-of-1's like the Carcharodon, so probably ~5500 total 1-of-1 hats is a safer estimate.

 

I also didn't factor in 2-of-2's where one is locked to a dead/banned account and thus the other one is effectively a 1-of-1, nor did I factor in duped 1-of-1's technically not being 1-of-1's, but both cancel each other out I guess so statistically it wouldn't make a difference, and not everyone cares about either of those anyways.

[🔥Master Throne Crimson🔥]

1, cause its 1 of 1

[SoundDrout]

1 hour ago, Slam said:

It's hard to actually figure that out but I'm bored enough and have time to kill so fuck it I'll use basic statistics and take a shot in the dark. If someone's better at math or there's literally an API or spreadsheet that has this info readily available already then lemme know I wasted 20 minutes of my time lmao.

 

Using https://backpack.tf/unusuals and not including any glitch / spelled / untradable unusual hats as being "unique 1 of 1", I'll take 20 random hats, 10 of each "era", and count the 1-of-1's that exist for each. The first era being pre-case-era (PCE), the last era being case-era (CE). The reason being there's significantly more effects available to case-era hats than non-case-era hats and thus a higher chance for 1-of-1's, so it's good to do an even split of both.

 

Total number of 1-of-1 versions according to bp.tf:

(Note 1: Not including Community Sparkle either, as that for some reason appears on some of these.)

(Note 2: I'm including ones that say "0 exist" here since they're almost certainly 1 of 1's that are just in private inventories or the SCM or something)

 

Killer's Kabuto: 13

Team Captain: 8

Rather Festive Tree: 10

Trippler's Tricorne: 8

Gym Rat: 15

Madame Dixie: 10

Officer's Ushanka: 12

Spook Specs: 8

Brainiac Hairpiece: 12

Brotherhood of Arms: 7

 

Caribou Companion: 19

Misfortune Fedora: 10

Head Prize: 18

Brass Bucket: 18

Citizen Kane: 8

Cranial Carcharodon: 28

Patriot Peak: 2

Big Topper: 22

Brain Warming Wear: 9

Bat Hat: 11

 

Adding all of those numbers together you get 238, divided by 20 you get an average of 11.9 1-of-1's per hat. Might be high or low but whatever close enough for this.

 

The total number of hats that can be unusual according to bp.tf is 514. We get this from https://backpack.tf/unusuals, counting 57 rows times 10 per row then adding the last 7 left over = 577, then CTRL+F "taunt:" and subtract that number for taunts to equal 514 total unusual hats.

 

So from this you simply do 514 times 11.9 to get about 6,117 total 1-of-1 unusual hats that exist in TF2.

 

But I'm gonna say that's high just because there's probably a good chunk of hats that only have a couple of 1-of-1's like the Patriot Peak but not a lot of hats that have a huge amount of 1-of-1's like the Carcharodon, so probably ~5500 total 1-of-1 hats is a safer estimate.

 

I also didn't factor in 2-of-2's where one is locked to a dead/banned account and thus the other one is effectively a 1-of-1, nor did I factor in duped 1-of-1's technically not being 1-of-1's, but both cancel each other out I guess so statistically it wouldn't make a difference, and not everyone cares about either of those anyways.

 

Funnily enough the OP never mentioned it had to be an unusual. You could look at collector's hats and other 1 of 1 rare hats like the blood nosed caribou and count them too (however small the number may be).

[RockSolidWood]

Funny thing is, using the theoretical idea that 5500 hats exist from 8,784,092 

 

5500 is 0.062613187566797% of 8784092.

Meaning

If you own one of these hats, you're supposedly amongst this percentage of players.

Its a worthless and probably wrong statistic, and I'm pretty sure thats quite wrong but even so I think its not far off from being entirely correct

[NotAliasing]

13 hours ago, Slam said:

Caribou Companion: 19

Misfortune Fedora: 10

Head Prize: 18

Brass Bucket: 18

Citizen Kane: 8

Cranial Carcharodon: 28

Patriot Peak: 2

Big Topper: 22

Brain Warming Wear: 9

Bat Hat: 11

 

the head prize would somewhat skew the average, no? the only assassin grade in a list of mercenary grade makes 1:1s wayyy more common on the head prize than any of the others.

[Norah]

Hard to gauge since how old it is, what case the hat is in, among other things all change things. If I had to guess? Probably about 7-10 per cosmetic, but I think the per effect information would be more interesting since that's gonna be much more skewed. Like, Gen 1-3 1:1s are probably only in the double digits.

[Slam]

11 hours ago, NotAliasing said:

the head prize would somewhat skew the average, no? the only assassin grade in a list of mercenary grade makes 1:1s wayyy more common on the head prize than any of the others.

 

I don't think so because with the higher grades, especially elite grades, you have a higher chance of 1-of-1's being common (1st/2nd/3rd gen) effects which balances out them having lower amounts of unique effects from events that weren't related to their case and less overall unusual variants. I.E. an elite grade may only have 20 unusual variants in existence, of which you have 6 1st gen 1-of-1's but only a couple of halloween/christmas/summer effects due to the hat's rarity. Again though my methodology isn't great, but I had time to kill so I said fuck it and didn't take it super seriously lol

 

Really the only one that skews the results is the Patriot Peak with only 2 unique's, but at the same time it helps be kind of the representative of the various hats where there's only a couple 1 of 1's that exist. Either way I did 20 hats, so 1 hat being a bad pick that skews the data would only change my final number by at most 5%.

[LaughingLollipop]

4 hours ago, Slam said:

Really the only one that skews the results is the Patriot Peak with only 2 unique's, but at the same time it helps be kind of the representative of the various hats where there's only a couple 1 of 1's that exist. Either way I did 20 hats, so 1 hat being a bad pick that skews the data would only change my final number by at most 5%.

 

Mathematically speaking, that last sentence is cringe.

 

1 hat can skew the final by way more than 5%.

 

For an extreme example, if my sample was 19 hats with 5 uniques and a 20th hat with 50 uniques, the average is.... 7.25. Without the 'mistake hat' the average is 5. So a 45% skew with 1 bad hat.

 

Numerically 1 bad data point can do damage approaching infinity from an averages perspective (if the bad point is infinitely large or small the skew can be pushed infinitely from a percentage perspective)

[Slam]

19 hours ago, LaughingLollipop said:

 

Mathematically speaking, that last sentence is cringe.

 

1 hat can skew the final by way more than 5%.

 

For an extreme example, if my sample was 19 hats with 5 uniques and a 20th hat with 50 uniques, the average is.... 7.25. Without the 'mistake hat' the average is 5. So a 45% skew with 1 bad hat.

 

Numerically 1 bad data point can do damage approaching infinity from an averages perspective (if the bad point is infinitely large or small the skew can be pushed infinitely from a percentage perspective)

 

yeah but what I meant is in the context of what we're talking about. Not sure why I even said 5%, but there's no such thing as a TF2 unusual hat that can have 50 unique 1-of-1's (well, TECHNICALLY there is, but the chances of that happening are too high for it to ever happen), and the average overall is low enough that having one be a 0 wouldn't skew the data that badly, so having one of the hats I use be an outlier / a "bad choice" wouldn't actually affect the data enough for it to be worth re-doing everything from scratch. I'm shit with percentages and math though