A Paulson color puzzle to pass the pandemic time (1 Viewer)

[One Eyed Dollar]

Maybe this has been answered before, but I didn't find it in a quick search.

Counting base colors and edge spots, what is the fewest number of sample chips you would need to buy to get 1 of each Paulson color?

Which chips would you need?

Just wondering if any of the experts here have assembled a list like that before, out of curiosity.

 

[markleteenie]

Maybe this has been answered before, but I didn't find it in a quick search.

Counting base colors and edge spots, what is the fewest number of sample chips you would need to buy to get 1 of each Paulson color?

Which chips would you need?

Just wondering if any of the experts here have assembled a list like that before, out of curiosity.

What's a sample chip?
Real chips, or made up?

If you could make this chip in different configurations, that's 9 colors each....so 10 chips.

Unless you can individually configure each of the bear claw spots, then that's 13, so 7.

 

[allforcharity]

I'm pretty sure he means real chips with edge spots, in enough variety to represent all (84? 85?) colours available on the Paulson palette.

To be honest, I have no idea. You might be able to pare it down to about 35 chips or so, if you're lucky?

 

[One Eyed Dollar]

Right, like how many individual chips would it take? I guess I mean singles, not samples. Using actual chips.

 

[BGinGA]

What's a sample chip?
Real chips, or made up?

If you could make this chip in different configurations, that's 9 colors each....so 10 chips.
View attachment 437279

Unless you can individually configure each of the bear claw spots, then that's 13, so 7.

^^ Correct answer. Although given the thickness of 1/4" spots vs 1/8", I'd say that a set of ten 8a14-spotted chips would be vastly better to represent all 89 colors. Or eleven chips, if you want the base color to be identical on all of them (probably best for actual use).

 

[allforcharity]

I highly doubt that there are 11 different existing chips that have all 89 colours represented. I was assuming you were trying to get actual casino chips that could represent all colours through their base colours and edgespots.

 

[One Eyed Dollar]

I highly doubt that there are 11 different existing chips that have all 89 colours represented. I was assuming you were trying to get actual casino chips that could represent all colours through their base colours and edgespots.

Right, using existing Paulson chips: casino, fantasy, NAGB, etc.

 

[BGinGA]

actual casino chips that could represent all colours through their base colours and edgespots

That's probably closer to 25 total chips. Figure a combination of three-color and four-color chips, covering 89 unique colors.

 

[FDLmold]

This is a computation-heavy mathematics problem. Needs lots of inputs too, and a computer.

 

[12thMan]

I wonder what the least amount you could do to represent all colors with base colors and spots on chips that are existing right now, not making new ones. Like the big top 5s would knockoff these colors, and maybe a boat chip would knock off some others, and so on until you have all Paulson colors represented over a whole bunch of different casino/fantasy chips.

Would be like a chipper scavenger hunt to have the most unique chips to represent all the different colors.

 

[12thMan]

Right, using existing Paulson chips: casino, fantasy, NAGB, etc.

That's probably closer to 25 total chips. Figure a combination of three-color and four-color chips, covering 89 unique colors.

Lol never mind, apparently I posted without reading the rest of the thread.

 

[justsomedude]

That's probably closer to 25 total chips. Figure a combination of three-color and four-color chips, covering 89 unique colors.

AS is a good example.

I'm not certain, but I do not think there were any repeating colors in that set. If true, you'd have 26 unique colors spread across 8 chips. Extrapolating that to 89, you should be able to get every Paulson color in 27 chips or so. Which is right in @BGinGA's ballpark.

There are probably more efficient ways of doing it, I was just using AS as an example...

 

[allforcharity]

AS is a good example.

I'm not certain, but I do not think there were any repeating colors in that set. If true, you'd have 26 unique colors spread across 8 chips. Extrapolating that to 89, you should be able to get every Paulson color in 27 chips or so. Which is right in @BGinGA's ballpark.

There are probably more efficient ways of doing it, I was just using AS as an example...

View attachment 437431

But can you guarantee that there exists another 19 chips that have exactly the missing 63 colours you need spread among them, without duplicates at all? I highly doubt it. Just looking at the sheer number of blue shades alone, there's no way you could avoid duplications of other colours if you could collect every shade in a real chip. That's why I allowed for some slippage when I guessed in mid-30s, and something tells me you may need a few more than that, because some rare colours might be only in a 2- or 3- combo chip that combines with more common colours.

 

[justsomedude]

But can you guarantee that there exists another 19 chips that have exactly the missing 63 colours you need spread among them, without duplicates at all?

Really not trying to guarantee anything. I was merely spitballing with the first set that came to mind.

Just looking at the sheer number of blue shades alone, there's no way you could avoid duplications of other colours if you could collect every shade in a real chip.

I dunno; it could be possible. Again, look at AS... there are 6 different shades of blue in that set. Pretty impressive given there are only 8 chips. There are also 3 yellows, 5 pinks, and 4 grays/beiges.

 

[redwine]

I spent WAY too much time on this, but I was able to use a greedy approach to get to my current best answer of 39 chips.

I was able to achieve 73 unique colors with just 24 chips, but the next 15 colors each took 1 incremental chip, as I did not have any chip in my database that had 2 or more of those remaining 14 colors.

My database is laughably incomplete derived mostly from this link and various posts on this forum. Obviously, if I had more color information this number will go down. I think the theoretical lower bound would be 22 chips, as a paulson chip can have at most 4 colors(?) and 88/4 = 22.

So the answer is somewhere between 22 and 39.

the 15 colors to be represented by an incremental chip: DayBlue, HawaiiFlower, LimeGreen, DolphinBlue, IceBlue, DarkRed, OceanBlue, PastelGreen, ShowboatGrey, DesertFlower, Carrot, Bronze, DarkBrown, LemonGreen, Sunrise

 

[FDLmold]

Impressive. Improving your result would require taking those 15 "single" colors and getting an exhaustive list of casino chips they are on, then calculating if you can eliminate any of the first 24 chips as a result.

 

[JustDave]

AS is a good example.

I'm not certain, but I do not think there were any repeating colors in that set. If true, you'd have 26 unique colors spread across 8 chips. Extrapolating that to 89, you should be able to get every Paulson color in 27 chips or so. Which is right in @BGinGA's ballpark.

There are probably more efficient ways of doing it, I was just using AS as an example...

View attachment 437431

This is a big part of why i bought a sample.

 

[redwine]

Impressive. Improving your result would require taking those 15 "single" colors and getting an exhaustive list of casino chips they are on, then calculating if you can eliminate any of the first 24 chips as a result.

Yes, that would definitely be one way to improve the result.

The overall problem is basically a set-cover optimization problem which is computationally hard. As I gather more color data over time, I'll be monitoring whether I can get the number lower... If I had to guess, the true minimum is probably in the upper-20s.

 

[CrazyEddie]

Now cross-index against a database of singles market prices and then re-optimize for price rather than number of chips.

I wonder whether the answer turns out to be more expensive than just buying a color sample set off of eBay (or PCF!).

 

[allforcharity]

Now cross-index against a database of singles market prices and then re-optimize for price rather than number of chips.

I wonder whether the answer turns out to be more expensive than just buying a color sample set off of eBay (or PCF!).

Well, the underside of the going rate for a complete Paulson sample set is $400, so if you needed, say, 40 chips to get all the colours, and you can get them for less than $10 each on average, then you could be good to build your own.

Me, I'd rather have the Paulson set.

 

[JustDave]

Well, the underside of the going rate for a complete Paulson sample set is $400, so if you needed, say, 40 chips to get all the colours, and you can get them for less than $10 each on average, then you could be good to build your own.

Me, I'd rather have the Paulson set.

I wanted one until I saw the price.

Made me sad.

 

[redwine]

Progress. Found several 38 chip solutions (previous: 39) by adding PNY chips and searching more combinations. One such:

the 14 colors to be represented by an incremental chip: DolphinBlue, NavyBlue, BahamaBlue, DarkRed, OceanBlue, PastelGreen, LemonGreen, DesertFlower, Carrot, Bronze, DarkBrown, Maroon, Sunrise, Beige

 

[redwine]

It's been a while, but with the information in this really helpful post the solution is trivially reduced to 36 chips!!! Huzzah!

• We can add in "Jack Cincinnati Primary 500" to eat up NavyBlue and DolphinBlue.
• We can add in "Jack Cincinnati Primary 25000" to eat up BahamaBlue and DarkRed. Alternatively, we can add in "Jack Cincinnati Secondary 1000" to eat up DarkBrown and DarkRed.

But wait, there's more!

With some brute-forcing, I was able to find a 35 chip solution. It seems that we are still at the point where even a little bit of extra data helps to reduce the solution quite quickly.

with LimeGreen, PastelGreen, LemonGreen, MossGreen, OceanBlue, DarkBlue, Sunrise being represented by single-color solids.

If anyone knows of any paulson chips that are known to have 2 (or more) of these colors, we can reduce this even further without re-brute-forcing.

"LimeGreen, PastelGreen, LemonGreen, MossGreen, OceanBlue, DarkBlue, Sunrise"

 

[redwine]

To be honest, I have no idea. You might be able to pare it down to about 35 chips or so, if you're lucky?

Nailed it. (for now)

 

[Nanook]

A more useful solution would be one that was the cheapest way to obtain all the colors in the Paulson color palate.

 

[allforcharity]

And "market value" of a complete Paulson colour set has increased to at least $700, with the over going up up up as we have recently seen. So if you can get those 35 chips for less than $20 average each, then you're ahead of the current market curve.

 

[redwine]

A more useful solution would be one that was the cheapest way to obtain all the colors in the Paulson color palate.

For sure.

This is just mental masturbation at its finest.

 

[Marius L]

Not sure what color info you have in addition to the sets in the link you posted earlier, but i believe the colors of several recent chiproom sales are available in his threads. (Not sure if all of them, but if you have some time to search and look through it you might find more, based on what you need) here's two from a few forum searches (although you might have them already)

HSI colors
Lady luck


There also some colorinfo on random chips in this thread

And as linked to in the above thread, there were several instances of pictures like this in this thread as well, if I remember correctly. If you have some time to browse through that long thread it could be a lead at least

 

[redwine]

Not sure what color info you have in addition to the sets in the link you posted earlier, but i believe the colors of several recent chiproom sales are available in his threads. (Not sure if all of them, but if you have some time to search and look through it you might find more, based on what you need) here's two from a few forum searches (although you might have them already)

HSI colors
Lady luck


There also some colorinfo on random chips in this thread

And as linked to in the above thread, there were several instances of pictures like this in this thread as well, if I remember correctly. If you have some time to browse through that long thread it could be a lead at least

I did not have the starlites. I had all the chiproom ones and the ones on those random threads. Thanks for the heads up!

 

[redwine]

Found this mother lode of chip color information at the gaming commission website.

The upper bound of the solution is now at 31 chips. Lower bound is 23. Getting closer.

It is at the point where no color needs to be represented by its own chip, so any further improvements are likely to be difficult to find.