Should you buy metal dice? (Paladin Dice Review)

[@paladinroleplaying2126]

Thanks for the review, and for taking so long making all those rolls - we appreciate it!

[@firestar1056]

Imagine stepping on the d4

[@TheWhoamaters]

@@firestar1056 calm down satan

[@dwainespradling9233]

Ironically a d4 does 1d4 damage.

[@JJR93]

@@dwainespradling9233 IRL caltrops FTW

[@matteussilvestre8583]

Imagine stepping on a metal d4

[@ArcaneForge]

You'd definitely take more than D4 piercing damage!

[@virus6765]

Lego users know nothing compared to d4 users.

[@samconsumespie1012]

So metal dice work effectively as caltrops

[@samconsumespie1012]

My metal dice hurt a little when I hold them

[@sharksuperiority9736]

M Silvestre ohhhhhhh, I don’t want to

[@orHekt]

To be honest 100 rolls is a fairly small sample set especially for a d20. Not really judging it is a super boring thing to do. I am just saying the variance you would see in a set off 100 rolls is not really enough to provide stastical evidence of how the dice are weighted.

[@benjaminbenrachmiel8952]

That's not entirely correct. We can make a decent statistical analysis, as he provides his full list of data and we are comparing to a known, albeit imaginary, uniformly distributed d20. While the paladin die has a higher average, running a t-test, we can see that with 95% confidence, which is a typical benchmark for amateur investigation, the paladin die doesn't vary statistically significantly from an imaginary uniform die. The p-value I came up with was 0.46. I can link you to the data and analysis if you'd like.

[@edsiefker1301]

@@benjaminbenrachmiel8952 p=0.46?

[@benjaminbenrachmiel8952]

@@edsiefker1301 Yep, far too high to indicate statistically significant difference.

[@adameves5970]

@@benjaminbenrachmiel8952 What he's saying is that the more possible results a die can give, the more rolls would be needed. A 20 sided die only gives each face an equal chance to get 5 results in 100 rolls, whereas a D6 would have a chance to get 16+ results. Basically, if you did 60 rolls with a D6, you would have to do 200 rolls with a D20 to get a fair amount of data from each. It's based on possible outcomes, not total rolls. By your logic you could do 100 rolls with a D100 and get an accurate read, which is untrue.

[@jbaidley]

@@benjaminbenrachmiel8952 A t-test only tests your mean, it doesn't test the distribution. Also, the t-test assumes normality and this is a uniform distribution.

[@rodneykelly8768]

The die you sawed in half had a "VOID" in the middle, not an air bubble. When hot plastic in a mold, especially a mold with large cavities, cools, it contracts. Because the walls of a cavity cool first, they form a solid structure. The plastic in the center takes longer to cool and will contract toward the walls. This results in an area at the center of the cavity that is as close to a perfect vacuum as you can get on Earth.

[@holloszaboakos]

It is a very interesting task to eliminate these kind of vacuums during production. Actually you have to keep the top warm so the vacuum sucks more material in..

[@gummostump4217]

Cool, that means my big d20 has enough vacuum in it to clean my house.

[@fenwah1]

Slighty pedantic note: While I don't doubt it is at a pretty low pressure, the void in a die is nowhere near the best vacuum on earth... Hot plastic outgasses like crazy, as evidenced by its strong odour, so there will be tons of volatiles in there floating around. Even ignoring the effects of heat during production, the vapour pressure of plastics at room temperature is at best on the order of around 10^-8 Torr. Ultra high vacuum systems such as you might find at a particle accelerator can reach as low as 10^-12 Torr. There is a reason that vacuum chambers are made of specially prepared stainless steel and not plastic.

[@SeedlingNL]

You get a better vacuum sticking your head out of the ISS, and it's not even outside the atmosphere :P (It's just really, really thin at that altitude.)

[@rickharper4533]

SeedlingNL yes, thats everywhere, there is no such thing as a perfect vacuum.

[@hangarrat]

Our DM has battered his wooden table with using metal dice, after a few choice words from his wife he now uses a gaming mat! We both use Paladin Dice, they’re awesome.

[@loka7783]

I have metal dice (I forget from where) but I find it's easy to avoid damaging the table we game on by simply putting the duo-tang that hold all of my characters and other papers where I am going to roll. Note the particular duo-tang I am using has a solid frame so it doesn't give way some others might. Just my two cents for possibly saving some money rather than buying a gaming mat. :)

[@mynamejeff4883]

That is exactly my worry. I am afraid to damage my table so I have yo invest in a mat or some sort of rolling surface before I buy the metal sets my sons want for their game.

[@etymike]

Don't second-guess yourself, Josh! "one of my dice" is correct: it describes a single die ("one") from ("of") the plural group ("my dice")

[@Thalanox]

"My favourite dice (singular)" is still wrong, though.

[@phalspar]

@@Thalanox exactly it’s **favorite**

[@ericdpeerik3928]

Never say die 🤔

[@Ellanion]

A lot of people are yelling in comments about sample size in an unconstructive manner, and since I too reacted, I thought I'd share some numbers. On 100 rolls a D6 should roll each number about 17 (16,66...) times with a margin of error of about 5. That means if it rolls a one 22 times and a six 11 times that's still perfectly reasonable. With a D20 the acceptable range becomes huge with just 100 rolls. Without boring you with extensive math, I'd conclude that to get a reasonable amateur level confidence that your dice weren't heavily biased (all dice are biased, the goal is just to have it be a low enough level that it doesn't matter) you'd need to roll about 100 times per possible result on the die, and then look at the percentages of each result and decide if they were within an acceptable parameter for you (since, again, all dice are biased, and the odds of actually rolling exactly equal amounts of each face are astronomically low). Also, dice as singular is correct in American English. Die as singular is the British version. One's not superior to the other, no need to feel bad.

[@j2dragon109]

A commenter below ( Samuel Forster ) says the opposite about the whole die/dice debacle.

[@MK-13337]

I would roll them 1000x the number of faces. So 20 000 for the d20. Should be a decent-ish sample for a mathematician like me

[@lukebrookshire1567]

Ellanion That over simplifies it. (J2Dragon) In absolute/literal cases, when referring to an individual die the word will always be die; in looser/nonliteral cases dice (can) refer()s to all of the dice in question, be it zero, one, multiple, or multiples, until a distinction must be or has already been made.

[@MK-13337]

@@lukebrookshire1567 mice - mouse Lice - louse Dice - douse Makes sense

[@konayasai]

dice (n.) plural of die (n.), early 14c., des, dys, plural of dy, altered 14c. to dyse, dyce, and 15c. to dice. "As in pence, the plural s retains its original breath sound, probably because these words were not felt as ordinary plurals, but as collective words" [OED]. Sometimes used as singular 1400-1700. Dice-box "box from which dice are thrown in gaming" is from 1550s.

[@DrinkGameRepeat]

Isn't it salt and water? Sugar won't give you the most accurate results.

[@Alex-yl8dz]

DrinkGameRepeat the idea is that you increase the density of the liquid to be higher than that of the di. Salt is usually used but water can absorb more sugar, so overall the density is increased enough to allow the dice to float and thus to see a bias in the dice.

[@DrinkGameRepeat]

@@Alex-yl8dz From personal experience, and according to friends and other D&D blogs, salt has still worked better. Maybe because it gives it more buoyancy? I don't know the specifics of why, other than changing the density, only seen the results of both.

[@Alex-yl8dz]

DrinkGameRepeat salt may work better, but sugar still works. Salt may make it more buoyant, sugar makes it buoyant all the same

[@Xalor69]

Any soluble compound will work. Sugar is more soluble than salt, so it actually is better in that respect. But since either will allow you to create a solution with density > plastic, then they're actually equivalent from that point of view. Personally, I would use epsom salt, since it's slightly easier to clean up, and isn't sticky.

[@beanscoped6437]

3:17 You put the d8 and d12 in the wrong holes. YOU MONSTER

[@dolphinboi-playmonsterranc9668]

This crime must not go unpunished

[@KitFox1011]

Roll for initiative!

[@smithy1578]

We must revolt I’ll be the cleric of the healing domain

[@arthilas_]

@@KitFox1011 natural 20! Together with my +5 dex, that makes 25! xD

[@DavidMBebber]

You can still do the floating test if you can get enough mercury to float the dice in.

[@davedujour1]

The bar graph is a terrible way to represent this data. A simple line chart, probably with 3 lines for the 3 different type of die, would show the "evenness" of the rolls and compare to each other type of dice. The bars make it difficult to visually compare one value to another. Also, the die number "10/20/30/etc." is a d10, not a d100. It doesn't have 100 sides, it's just another d10 to make a percentile roll easier with a "normal" numbering of d10. A d100 looks like a golf ball.

[@Quandry1]

The correct term for that kind of d10 is a percentile die. It just happens to easily double for a regular d10. The two by technicality are different.

[@adameves5970]

@@Quandry1They both have 10 faces. Means they are the same.

[@MonkeyJedi99]

In out group we call it a "tens die" or "tens-ten"

[@lordwindgrace]

I would have also liked to see at least some statistics... i.e. averages, standard deviation, etc.

[@hopelesspink7177]

The whole video I couldn't stop thinking about stepping on the D4

[@levi1929]

Hi! Ok, I haven’t seen the video yet but the answer to the question “should I buy metal dice?” is YES! Metal dice feel so much better, and the styles are cooler than plastic dice (even stylish chessex dice). I’ve gone metal and all other dice feel like flimsy toys to me now.

[@halelloyd]

You should do the float test in liquid mercury, if the dice are steel it should float (maybe most metals too? not sure). That would honestly be a great video!

[@TumluGaming]

Cody's Lab Cooperation coming up? :D

[@halelloyd]

Then Cody should cast a solid gold d20! lol @@TumluGaming

[@whoaitstiger]

That would be awesome! Aluminum, zinc, iron, copper, silver and lead will all float in mercury because they have a lower density. Gold has a higher density and will sink, but I don't think I'm going to be buying solid gold dice anytime soon. :D

[@IrrevMike]

Personally, I'd be worried about inhaling mercury fumes. That stuff's toxic.

[@whoaitstiger]

​@@IrrevMike My dad used to handle it in school as a kid in the 50s, no one knew any better. It's disturbing to think of now!

[@guamae]

Standard Deviations and Average Values might be more .... enlightening when comparing lots of rolls. In any event, nice to see some good-ol' TEDIOUS Maths done to dice!

[@VulpeculaJoy]

Aggree

[@Xalor69]

If you calculate the deviation from the expected average number of rolls (7 would be +2, 3 would be -2, on 100 rolls of a d20, with average expected number of rolls being 5), then graph using a line graph, the results are easier to interpret. Also, I'd suggest you need 400+ rolls so you can see irregularities beyond standard deviation.

[@apocalypticmoth6040]

Thank you for stating your ethics. I have no doubt in my mind that you are unbiased. Great video as always.

[@bentosan]

He received the item without paying, that makes him at least slightly biased, stating the fact he got them for free is required by law. At best you could call the review honest but definitely not unbiased.

[@ArcaneForge]

I don't legally have to say how I came by them @@bentosan, but if it helps, I plan to give them away to viewers soon.

[@bentosan]

Arcane Forge I used to do blogging (for a US based company) and it was a legal requirement we state when we received items when doing articles that included them. I just assumed you were US based which may be a mistake on my part. Wired has an article on it if you google “FTC TELLS AMATEUR BLOGGERS TO DISCLOSE FREEBIES OR BE FINED”. Stating your association is definitely the morally correct thing to do so I commend you on that.

[@bentosan]

As far as giving the item away that’s up to you to decide. I don’t think it completely removes the bias but it certainly doesn’t hurt and I’m sure your audience would appreciate it. Please excuse my matter of fact approach to the topic, I certainly don’t think any less of you for accepting any items for review and wouldn’t think less of you for keeping it to yourself either.

[@arwo1143]

Never EVER slam you hand on a metal d4 I did it It hurts It hurts ALOT I rammed my copper d4 in the skin between my left hands middle and ring finger Much worse than having a cutter blade in the thumb (did that too, also not fun)

[@cl0udbear]

A plastic one will do that, too. They just tend to have had their edges rounded off in a rock tumbler.

[@denzelpanther240]

dont hurt the alot its a cute creature

[@smithy1578]

Some mistakes where made

[@zetsumeinaito]

3 words: Dice Balancing Caliper The trick to getting cheap ones, is just get a $2 clamp and mirror polish the flats that hold the Di. Spin the dice on corners, and you'll find out quick if there's a bias. Is it work? yes. Still saves you a solid 50 bucks.

[@VincentOak]

I tested a d20 at work when I was bored as there was nothing to do. It strongly evened out at about 1000 rolls

[@TableTopNerdCast]

As a Christmas present for all my players at the Well That Didn't Work Crew Member channel and podcast team, I purchased each a set of metal dice. We found that the dice sound picks up better at the table when we use them, where as before with the standard plastic style dice, we would get questions about our actual game play sessions if we were fudging rolls since people could not hear them!

[@joshuasteel3543]

Spectacular video! We sre always eager to see the next vid.

[@nathanwhitfield7834]

An interesting result from my own testing, my hit dice very often tend to roll a 1

[@voltsoftruthBSbuster]

I just purchased a set of titanium dice. The stupid thing is a set is considered one die, so to get 2 you have to purchase 2 sets, which to me would seem that you should get 4 dice. But anyway just wanted to let you know they are probably the best dice I've used, very durable, solid construction but still not very heavy. They each weight 11g. If you ever get a chance give titanium a try. I also have brass dice, and granite dice. The granite are very nice but each time I roll them and bang them into each other I fear ill break a piece. Personally I've never broke one, but have seen people chip the edges, which is a shame cause they loose their stunning looks immediately once they have chips. Thats another benefit of titanium I don't have to worry about chipping, cracking, breaking, and you can heat them up to 2500F and roll them while they are glowing bright red, your game set with all the pieces will ignite and your table might catch on fire but I'm sure it will be stunning to see them burn through everything they land on, assuming you can figure out a method to roll them at that temp.

[@ashleygillespie-horne6429]

This is pretty cool. I saw Paladin Roleplaying on Amazon by accident, and been perving at their metal sets for a while now. I got my first metal set for Christmas, and after enjoying the heft of them I'll never roll plastic again. My current set is 24k gold plated which is lovely, but the numbers are *also* gold and therefore not very clear. These Paladin ones in the video seem to have very clear numbering and there's a pristine beauty to them, so I might get a set...

[@edsiefker1301]

Metal dice are sharp and heavy. They will leave marks on your table. You will need a rolling tray. Ah, uh, dice rolling tray. ;D You could make this long, arduous, boring project into a long, interesting, and fun one if you spent your time automating the project. A cheap camera and OpenCV would have been a lot more fun to play with.

[@CMDRFandragon]

I just rolled the d6 from this first set of metal dice I got. They seemed to roll fairly even with a slight tendencey to roll 6. In 200 rolls, I got 32 each of 1's, 2's, 4's and 5's. I got 28 4's and 44 6's

[@HermenegildGaido]

If you want to know if a value appears more often than it should, use the chi-square test (there are free programs that can help you to calculate it). Looking at chart by own eye may be misleading.

[@thepenguinhouse]

Your commitment to data collection is one to be proud of

[@occultnightingale1106]

I've been exclusively using metal dice (I got a set for Christmas two years ago) and I've been rather satisfied with the results, in that I have not been noticing a favoritism towards any certain results. The only issue is that they kind of beat up whatever I roll them on, and I needed to invest in a less easily-damaged surface.

[@mortendreier4446]

I roll mine on my sheet. Under the sheet I have the folder I have the sheet in. As of yet I have not caused any scratches or damage to the table we sit at.

[@anonpers0n]

How about float test in gallium or Mercury?

[@thespiritdragon]

I put the D20s through a Chi-Squared formula (helps seeing if your data is statistically abnormal). The paladin D20 rated at 3.667, Red is 7.333, and the Peacock is 9.4. Higher than CV of 30.14 (p=0.05, DF=19) indicates a statistical oddity such as potentially off-balance dice. As such, the Paladin dice are easily the fairest, however, none should be considered off balance by any stretch according to this method. Bonus: When you group them into 4 groupings of 1-5, 6-10, 11-15, and 16-20 (p=0.05, df=3, CV = 7.815) the Paladin = 0.44, Red = 6.88, Peacock = 2.84. While none break the CV, the Red dice have the highest value here (while being middle before) while the Paladin continues to be the lowest by far indicating (to me at least) they are VERY well made for balance and fair-rolling. For those wondering, I don't own Paladin dice or have any affiliation. I'm just a nerd who has recently been doing tests like this myself and got curious what the math said vs pure observation. Been thinking of getting a metal set, definitely going to look into Paladin dice after this. Also, I fully realize I may have miscounted on a couple of numbers. It shouldn't throw the values off by very much though considering none of them were particularly close to breaking the CV anyway.

[@GeekzAnonymous]

What's throwing you off is that it's "dice" no matter the number across the pond. Even game systems like Games-Workshop for 40k and Warhammer use the term "dice" for a singular die.

[@HLR4th]

Nice video. I purchased two sets of Paladin dice as Christmas presents. Both recipients were extremely happy and impressed. I gave a dice tray along with one of them, worried about dents in wooden tables. For the satisfying weight and great sound the dice make, dents, or a tray are well worth it! The Paladin Dice were available on Amazon in the US (and on PRIME). My ONLY enhancement request for Paladin would be a set with two D20's, so we can roll advantage in style!

[@thatoneguy2274]

theoretically, you could do the water/sugar (or water/salt) trick with metal dice, using some other liquid. as long as the die floats in the liquid it'll work. so all you have to do is find a liquid that has a higher density. one that comes to mind is mercury, but if you do that, be very careful and clean thoroughly, and check wether the type of metal the dice are made out of doesnt react with mercury beforehand.

[@RiverSiege]

I use gravity dice for my games, unlike many other dice they aren't mould poured (the same process that creates uneven plastic ones) but instead CNC cut. The ends are chamfered so they aren't sharp and they're made of lightweight aluminium so they won't wreck your table.

[@chrisgriffith1573]

For metal dice, I covered my "die rolling space" with two materials: a smooth 1/8" thick foam which I found from popular kitchen draw covering supplies, and then the well know felt color of my choice.

[@HayNay55]

Woah what’s that D20 glass looking container at 2:40 ?? Where can I get that??!? Looks cool :)

[@matthewenokson4251]

I'm to lazy to do the work but if you wanted to you could follow up with this: A die should add up to a standard number on opposing sides (ex 21 on a D20 or 7 on a D6) and there are fairly standard templates for dice. Arranging data as 1-20 might not show "weightedness" well. If you arranged according to the template it may show more of a trend to the weighted side if there is one. I realized it on the D6 chart when the 1 had higher numbers but the 6 showed standard. If it was weighted to the 1 the 6 should have comparably low numbers but it doesn't. Oh and I found one of my d20s was weighted to the 3 when I did the sugar water test so thanks for showing that!

[@fenwah1]

Advice for doing those dice tests: set up your camera and throw the whole set in at once. Do this a bunch of times, then skip through the footage later and record all the numbers. This is way faster than rolling every dice individually and constantly swapping between rolling and recording data.

[@MonkeyJedi99]

The game systems my group has used for 30-ish years uses percentile rolls. I have been using the same two d10 for around 20 or so years.

[@Sw-nn6le]

I have a set of metal dice and I really like it. I love the heaft to them and they land with a foreboding THUD. I've noticed that they do not roll much, if at all, once they hit the table. Other than that, this may be my long term set.

[@MiniHicks]

yeah, ive noticed with metal dice you need to roll them a lot more in the hand before dropping them than with a plastic dice. although ive found that a metal dice with roll very nicely on neoprene (like a mouse mat) even with a very gentle roll

[@TheLastSoundNL]

I would also take note of the opposite result of a roll, with the exception of a d4 as it has none. Like rolling a lot of 1 on a d6 would result in a small number of 6 if it was a bad die. That didn't happen, it was instead the 4 that was low.

[@flumphflumph6021]

Yo Arcane Forge . In the Uk it is just Dice, like Sheep. So we kind of think people in the USA are kind of weird for saying Die. So float your boat how ever you want.

[@j2dragon109]

Wait, what. A commentator ( Ellanion ) above says the opposite.

[@konayasai]

dice (n.) plural of die (n.), early 14c., des, dys, plural of dy, altered 14c. to dyse, dyce, and 15c. to dice. "As in pence, the plural s retains its original breath sound, probably because these words were not felt as ordinary plurals, but as collective words" [OED]. Sometimes used as singular 1400-1700. Dice-box "box from which dice are thrown in gaming" is from 1550s. In other words, Samuel is full of it.

[@jadedatheist1828]

I had some dice I'd purchased way back in the day. One of the d20's rolls about 60%+ 20's or 1's, and 40% of any other 18 sides.

[@ricktedder4749]

try covering your dice tower with a rubber or high friction material it prevents the dice from sliding and causes them to tumble more but with light plastic dice they stop in the middle

[@grystle3091]

A sample size of 100 rolls for a 20 sided die is not very much. Would have loved to see a larger sample to really confirm if there is a bias or not.

[@jade_capricorn]

You're not supposed to store metal dice in a high-friction environment such as a dice bag or a dice vault. Ideally metal dice should be put in snug cases that separate each die such as the case it comes in. If you don't care about scratches or dulling the dice then feel free to ignore this then I suppose.

[@GameFreak7744]

Hm, running the results of the d4 (as it has the best data) through the chi-squared test it has an χ² of 6.48 (assuming I didn't make any mistakes), which is actually edging into cause-for-concern territory as it would be expected to be

[@flynnmoroney1900]

You should check if the "Class sets" from Random Encounters Dice have problems with balance, I have a set of the Warlock and want to know if the little book imbed is making a difference.

[@DnDnBeer]

I bought the silver set two months ago. Very nice. I would have liked a set of 11 rather than 7, and at present they're sold out of their d6 set and do not sell single d20s. Hopefully Paladin will soon begin selling single dice so I can collect a matching set of 11. Thanks for the review!

[@joshuadrark3207]

I always try to use a set of transparent or semi-transparent dice. To see if the dice are weighted or not.

[@CarlosRodriguez-dd4sb]

You can use a crayon to fill in the numbers on the peacock dice so they are more visible. If you want test how random the rolls are look up comparison of proportions. You could also compare the means.

[@DoubtlessCar0]

This may be a bit too much effort but I would've found out the metal composition of the dice tested their volume then multiplied the average weight of the metal compound with the volume to find what the weight should be then weigh it to find out if it matches.

[@earthobserver1007]

Rolls are quicker with metal dice, they don't roll as much, loud as hell though.

[@HilltreesHarmony]

My metal d20 weighs 26 grams! Is that too much??? I use a dice tray so it doesn't damage my table but hot damn it seems heavy

[@HellGod67]

just depends on the damn metal contents ofcourse

[@Runescythe9852]

I use metal diehard dice from their gemstone sets and they are very pretty to look at.

[@malachigruenhagen2453]

It would be nice to see some statistical analysis of the data. It has been a while since I took statistics, but I have to wonder if the sample size was large enough. (My interest in dice is not so much games but random number generation).

[@phantomaquarist3417]

The get the peacock color on the metal from reheating it and in doing so it causes the metal to become denser

[@ATMOSK1234]

You could probably still do the fluid test with a ferrofluid solution if the dice are made of aluminum or some light weight alloy.

[@Angus-McFife-2nd]

I have those dice. Mine also have "Poleplay" on the tin, not "roleplay". I rolled the d20 1000 times to see how fair it was. Almost perfectly even. Only numbers mine slightly favoured was the 19, 17 and 1. But that d6 really favours the 5 and 6. Love rolling stats with it in D&D

[@rubenbanuelos2699]

If you have several sets of the same plastic die, you can weight them all together to get a better reading.

[@mauriciolacruz]

I'm a newbie in this world and have noticed the following about the 7 dice set I received: 1. The opposite sides on the D6, D12 and D20 dice sum the same, as it should: D6: (1/6) (2/5) (3/4) D12: (1/12) (2/11) (3/10) (4/9) (5/8) (6/7) D20: (1/20) (2/19) (3/18) (4/17) (5/16) (6/15) (7/14) (8/13) (9/12) (10/11) 2. The opposite sides on the D8 and two D10s, however, do not sum the same: D8: (1/2) (3/4) (5/6) (7/8) D10: (0/1) (2/3) (4/5) (6/7) (8/9) D10: (00/90) (10/80) (20/50) (30/40) (60/70) And the sequence on these dice neither follow any logic nor match one another, as the order looks scrambled for instance on both D10s. Here below the numbering on these two dice from left to right by hemispheres: D10: (0-8-2-6-4) and (3-9-1-5-7) D10: (00-40-60-20-80) and (70-30-90-10-50) When, to match one another, it should be: D10: (0-8-2-6-4) and (3-9-1-5-7) D10: (00-80-20-60-40) and (30-90-10-50-70) Or: D10: (0-4-6-2-8) and (7-3-9-1-5) D10: (00-40-60-20-80) and (70-30-90-10-50) Or perhaps better (unscrambled): D10: (0-2-4-6-8) and (1-3-5-7-9) D10: (00-20-40-60-80) and (10-30-50-70-90) Maybe other makers do it correctly, but these (Chinese 😖) are the ones I've got. 3. Both D10s are not numbered starting by "1" (1-10) and (10-100) like all other dice do, but (0-9) and (00-90). Yes, I know these dice are mainly used for percentages (%), but still is somewhat weird starting by "0" and "00" instead of by "1" and "10", despite you can assume when you need the "0" is "10" and the "00" is "100".

[@kharnthebetrayer4867]

Really nice experiment, but i would have compared these three dice also to a number generator. This way we could better compare if the unusual peaks are by chance or by failure in production.

[@ArcaneForge]

Firstly; Blood for the Blood god! Secondly; That would have been an amazing idea! I wish I'd thought of that at the time. I'll make sure to do that if anyone else sends me dice in the future! Cheers for the suggestion!

[@felixbillington6151]

Maybe use a chi squared test

[@emc246]

Can these dice be used by other classes?

[@thomasway0320]

Quick question, is it possible to hand carve a wooden dice set? I know it’ll probably be easier to make a fair dice set with tools, but I want to get into wood carving and also DnD, so I want to try and approach these two things at the same time.

[@draege626]

Thoughtd about bisecting a singular metallic icosahedral gaming result indicator?

[@calebshaffer2283]

I have a gold colored D20 that is about the size of a golf ball. I don't roll it on a wooden table, but I threaten my players with it regularly lol. I have a total of 3 sets of metal dice. I love them for all the same reasons you do.

[@ericpeirce5598]

I used to do this kind of roll test with my dice. His results for the d20 give it a 10.84 average per roll. A true average would be 10.5, so it is not too far off the average for only 100 rolls.

[@EllPro]

I got all three sets of MinMax metal dice from Amazon. The sets are only $10 each, they roll perfectly, they feel nice and solid and perform better than a $50 set ever will.

[@timeshark8727]

i hate to tell you, but you would need closer to 1000 rolls to get good statistics. especially for thing like d20's

[@franciscoalejandro7567]

100 rolls isn'tn that big of a sample, so I think it is ok for the dice to not give perfectly even results, that'd be very unlikely to happen, so as long as there isn't a very noticeable trend towards one particular number (like, rolling ones half the time), it seems like a fairly random result, if you were to repeat it a few more times, that would give you more accurate data, but rolling them 100 times was already boring enough, so I don't expect anyone to do it 1000 times.

[@adameves5970]

Depends on the amount of faces the die has. 100 rolls is pretty decent for a D4, but a D20 should have 500 rolls to equal the data of that D4. It's an uneven sample of data due to the amount of faces. Even though each die was rolled 100 times, the D4 acquired the most data.

[@franciscoalejandro7567]

@@adameves5970 That's very true. I usually do 100 times for my D6 dice, for the D20 I tried testing the other day, I did 300, it was tedious but seemed like a good number at the moment (I didn't wanna keep going)

[@skelitonking117]

I mean as long as it’s from a solid mold and the alloy isn’t shit or impure they shouldn’t be biased or weighted unless purposefully made so; it’s not as if they’re at risk of air pockets like resin mold dice are.

[@steveharrison76]

My wife has some of their dice sets and she loves them! Personally, I can definitely recommend them as she was chuffed when I got them for her. To be honest I’m probably going to pick some up myself. (I also have that exact same kitchen scale...)

[@adamconnor1898]

There is this thing called "statistics" that could be used to have a much sounder analysis of whether the variation is consistent with random chance.

[@noctairedorvoxin0]

Love these videos. Keep up the good work. One critique. Data organization. When comparing the looks and mass of each dice, it's more efficient to group them together. Show all the d20s together. Then all the d12s. Helps people compare differences. Even more important would be how you presented the data. In this case I would group all of the numbers together for each die. What you are looking for is an even distribution for all numbers for each die, to determine how fair it is. Just some ideas.

[@rhiannonmacnair6856]

He rolled a lot of dice 2,400 to be exact

[@KingAm0]

So...uh, that's not how probability works. For this, you would use the chi square test.

[@milhousevanhoutan9235]

Or at the very least an average to see of it regresses to the mean. In theory a fair d6 should give a number close to the mean of 3.5 over 100 rolls a d20 10.5 etc

[@jajwarehouse1]

Could you take the average of the 100 rolls for each die? Which of the averages aligns closest to the center point? Does one set of dice consistently average closest to center across the various die types?

[@thebeardedmike]

Hey brother, great video! Where is that dice tower from? Looks like just what I’m looking for.

[@MisterMagikarp]

Would the test be inconclusive since you did not roll each dice with the same number facing up/etc. when you dropped them into the dice tower each time?

[@ITpanda]

Roll metal dice in Mercury to determine if it is weighted to favor any side

[@Parsalin]

take the rolls, and give the number rolled a color from green to red. then put the colors on the die. if one side is rolled more often (the effect expected from weighted dies.) itll show visually. i like anything about dies though +1 for you sir.

[@evamonet3015]

I just wanna know where to find that glass box he's housing all his dice in🤔🤔

[@ArcaneForge]

Haha! I'm glad you like it! I got it years ago from a shop here in the UK called "paperchase" which makes stationary and things like that. They may still make it but they tend to change their stock seasonally. It might be worth looking up terrariums? Maybe a garden centre might have something in the same neighbourhood?

[@troyty007]

Awesome video! I especially like the one about the sugar water tests. This should help me weed out the bunch of dice i got for bargain hahaha. Hm...I use this for rocks but how about anchoring or gluing a string at one point (with a removable glue of course) and hanging the dice in mid air, then taking a picture and tracing a perfectly straight line from the string to the ground.. Repeat this several times at different points to get the center of mass of the die. The trick is making sure they are on the same "plane" on the d20 for example. The could be done by making sure the same number is facing the camera at all times and rotating about that axis. For rocks we do this by dabbing a small bit of white out. No need for dice though as we have the number to guide us. Then repeat this at a number perpendicular to the the first one to make sure its even on all axises. This should help for metal dice since they dont float o/

[@MrLifeitself]

@arcane Forge, Where did you get your Icosahedron container ?

[@SBVCP]

Do you tought (sorry for bad english) about trying to test them in another denser fluid? i dont know much about physics but perhaps theres one fluid and dense enough to let you

[@stuartbailey5985]

your sample size is about 1/10th the size it really should be... one could argue it's 1/50th of what it should be. Any check of probability of a die (or dice) should include 1000-5000 test rolls to get a good feel for their balance and rolling performance. Good start though

[@supersneakyshrub3975]

I appreciate metal dice because they've got more heft for obvious reasons. Also theres nothing more satisfying for a dwarf barb than chucking heavy objects across your table each turn.

[@wayner7263]

I love my metal dice. I would have gotten a set of gold and silver dice but the sets were so expensive and just settled for the gold dice set

[@danielpierce922]

Thanks for this! Still can’t believe you rolled each 100 times. That’s a lot of nope! They will scratch up a table, but you can always use a dice roller. On a solid wooden table that’s used for games (like DnD, Pathfinder, Starfinder, etc.), the dents of metal dice gives it a nice character. Of course, I’m talking over a long time. Also, HUGE fan of the multicolored dice set.

[@zososldier]

Missed 1 point. Are they billet or cast. If it's billet then the internal structure should be very uniform=fair. If it's cast, it can have air bubbles like plastic=biased.

[@dasoler8385]

your unbiased opinion got me sub for its rarity

[@fakjbf3129]

Oh those graphs are hard to look at. I think a better way to display it would have been a line graph plotting the frequency of each result and then superimposing the three sets of lines over each other. It would drastically cut down on the horizontal length and allow you to just show the entire graph at one time instead of having to scroll to the side.

[@marcw3233]

If I can make and certify casino grade dice machined to .0005 tolerance would there be a market?

[@tristanwilliams7208]

I have those exact same dice and the case too

[@AllYourBaseRBelong2Us]

have you tried floating them in mercury?

[@howardcameron2551]

I wish you would have done a chi square analysis test, but overall very well done and thank you for your role data. 😃

[@wisperingiron3646]

Pretty nice video. Maybe you could do a Chi-squared test to help you interpret the results of these kinds of statistical tests in the future. I'm a bit of math nerd so I think that would be cool. It was great that you also took the look of the dies and potential damage to table into consideration. Personally I love the look and feel of metal dies, but don't use them without a dice tray, as my linoleum table easily gets marks from the points where the dies hit the table when they first land.

[@backonlazer791]

Before watching the video (and never handling metal dice) I assume that the dice with less sides might roll poorly due to the extra mass (like d4, d6 and maybe d8) but others are probably fine. They might also be kinda loud. Edit: Of course, using a dice tower negates this, which is why I wish you had done some testing by throwing them by hand since most people don't have dice towers =/ Edit 2: I guess I expected more talk about the feel of the dice rather than the numbers they roll (at the end of the day no 2 dice sets are the same even if they are from the same manufacturer). Maybe some testing on what kind of marks they leave on the floors if you drop them or what happens if you step on them (of course I'm not demanding such sacrifice from you, haha). But yea, to me this is kinda just saying "just buy them if you like the way they look" and nothing else =/

[@SeanBlader]

Metal dice because of the weight tend to bounce less, whether this makes a difference in hand rolling results, it is a little less satisfying than a light weight plastic die.