Jump to content
punkUser

I tested my dice...

Recommended Posts

35 minutes ago, Darth Meanie said:

Please tell me you just earned a PhD in randomness for this. 🤣

 I have seen shoddier scientific studies that were supposed to be advancing medicine.

I feel like unironically just us all reading this and trying to pick it apart counts as a more vigorous peer review than some journal articles are subjected to

Share this post


Link to post
Share on other sites
31 minutes ago, Darth Meanie said:

Please tell me you just earned a PhD in randomness for this. 🤣

This!

But seriously - excellent work you magnificent nerd! :lol:

If I may ask, regarding identifying the symbol on the dice: I notice in the image captures of the dice that sometimes multiple facings can be seen in a single image. How did your image analysis algorithm determine which ones were not the real results?

Share this post


Link to post
Share on other sites
6 minutes ago, Kieransi said:

I feel like unironically just us all reading this and trying to pick it apart counts as a more vigorous peer review than some journal articles are subjected to

This feels especially true when the author comes back with well thought and researched defenses. Definitely much more academic and professional than some thesis/article defenses I have seen. 

Share this post


Link to post
Share on other sites
3 minutes ago, Arboghasthero said:

This feels especially true when the author comes back with well thought and researched defenses. Definitely much more academic and professional than some thesis/article defenses I have seen. 

That it confirms that the inconsistency with dice that many suspected exists is there, although not to the extent or leaning some claim, which helps to make it palpable. The dice are generally not accurate to what their faces show from what I can tell from their study, but the inaccuracy is random in how it presents itself from die to die. :)

Share this post


Link to post
Share on other sites

I am just curious now if this can be used to see if there is any validity to the salt water tests. It would be preferable that it would disprove that method just to shut people up and so they can't try to quickly and easily get a fix on their dices' biases.

Also, while physical defects are probably fairly easy to identify, I just can't seem myself buying a caliper and a magnifying glass, but I know so many people that would... 

Share this post


Link to post
Share on other sites

I've been sending this to Academic PHD friends and they are just "If this wasn't about X-wing Dice this could be published in a scientific journal."

 

My only gripe. I need to know your 3D print Dice Schematics! Can it be uploaded somewhere? I dig how you did it and the results speak for themselves!

Edited by SharpEdgeSoda

Share this post


Link to post
Share on other sites
1 hour ago, Scrivner said:

If I may ask, regarding identifying the symbol on the dice: I notice in the image captures of the dice that sometimes multiple facings can be seen in a single image. How did your image analysis algorithm determine which ones were not the real results?

The magic of machine learning :) I expected this to be more of an issue but the ML network had no trouble sorting it out. Other than a few outlier results that didn't land flat (and thus even I couldn't tell which was the more "bottom" face), it seemed to have no trouble sorting it out. Whether or not it does it the same way that we do (via lighting/shading and face perspective) is anyone's guess, but it works flawlessly.

1 hour ago, Arboghasthero said:

This feels especially true when the author comes back with well thought and researched defenses. Definitely much more academic and professional than some thesis/article defenses I have seen. 

Thanks :) I've working in academia and industry research labs for most of my career, so some of that shows through. Still, I wanted to make sure to address the anticipated concerns/criticisms up front as they are generally the same ones I would have with many people making claims about dice.

50 minutes ago, Arboghasthero said:

I am just curious now if this can be used to see if there is any validity to the salt water tests. It would be preferable that it would disprove that method just to shut people up and so they can't try to quickly and easily get a fix on their dices' biases.

I think it's highly unlikely that those tests are accurate. Indeed my skepticism about them is largely what motivated me doing this testing in the first place. As mentioned in the Krayt thread, most of the community methods aren't even robust and repeatable enough to test through... the salt water "test" in particular boils down (pun intended) to feelings about your dice in general, and doesn't seem to offer anything more valid than just rolling your dice and acting on your feelings about how they rolled either.

Quote

Also, while physical defects are probably fairly easy to identify, I just can't seem myself buying a caliper and a magnifying glass, but I know so many people that would... 

Agreed, but anecdotally I know a fair number of people who have spent some time and effort trying to pick "good dice" who certainly would have no issues measuring die faces, so I think it's a legitimate concern. Now IMO any sort of attempt to cheat with dice is at the very least "unsportsmanlike conduct" and thus against the tournament regulations, but I doubt that will stop people, particularly as the incentives for winning events increase.

Edited by punkUser

Share this post


Link to post
Share on other sites
21 hours ago, Crimsonwarlock said:

Better question, how would one even begin to enforce against someone measuring the probability of their dice and picking the best ones.

You really can’t, unless they’re doing it at the FLGS during setup.  If someone wanted to do this, they would do it at home and only bring the “chosen ones”.

The best practice, as a community, is to make sharing dice the norm.

Edited by Phelan Boots

Share this post


Link to post
Share on other sites
7 minutes ago, Crimsonwarlock said:

Better question, how would one even begin to enforce against someone measuring the probability of their dice and picking the best ones.

Honestly I don't think it is feasible to do so. Probably why allot of the casters and bloggers that post here are suggesting sharing a dice pool, which works well enough for me.

Share this post


Link to post
Share on other sites

Right, there's no way to know if someone has brought intentionally biased (in one direction) dice. Sharing is the only real practical thing we can do to mitigate the risks I think.

And to be clear I don't want to imply people need to panic or anything here, just that going forward at higher levels of competition making sharing the norm might be something to consider in light of the data.

Share this post


Link to post
Share on other sites
5 hours ago, prauxim said:

giphy.gif

MinwdqZ.jpg

Yeah, I tested a bunch of die and almost all of them were really biased toward blanks (as shown in gif)

Wonder how well salt water result correlate to actual roll performance.

You use a glass of water, then salt the water until the dice are buoyant, then roll one die into the water a bunch of times to record the results. You'll actually see the die spin, slow down, then flip to a side or even a corner. That's how I chose my dice. Obviously they'll roll differently on a table, but why choose dice that are weighted to favor blanks? Like I said earlier, one die NEVER showed ANY paint, another showed a focus ONCE, the rest blanks after 100 rolls.

Share this post


Link to post
Share on other sites

The great part of a shared dice pool is that over the course of a tournament, people will start mixing up their dice. Even if players bring stacked dice pools those pools will be unstacked after a few matches.

I mean. I don't know how we expect to get back the very same dice we started with. Anything that would make a particular die instantly recognizable should be enough to get it banned. 

Share this post


Link to post
Share on other sites
1 hour ago, Punning Pundit said:

I mean. I don't know how we expect to get back the very same dice we started with. Anything that would make a particular die instantly recognizable should be enough to get it banned. 

From the Xwing tournament regulations:

"Players may mark dice with a permanent or indelible marker to indicate ownership in an unobtrusive manner but cannot otherwise alter them in any way."

Share this post


Link to post
Share on other sites
1 hour ago, Punning Pundit said:

The great part of a shared dice pool is that over the course of a tournament, people will start mixing up their dice. Even if players bring stacked dice pools those pools will be unstacked after a few matches.

I mean. I don't know how we expect to get back the very same dice we started with. Anything that would make a particular die instantly recognizable should be enough to get it banned. 

You're assuming players mix their dice together; you use one set of dice.

Share this post


Link to post
Share on other sites

In particular, I was highly skeptical that the “floating dice” method had any merit but I was unable to find anyone who had properly tested it against a known control.

 

The obvious next step, if you still have your dice sorted, is to try the float test on the biased dice to see if it indicates the bias.

Share this post


Link to post
Share on other sites

...I'm very grateful for the detailed breakdown here of the faults of commercial dice, and genuinely wish this knowledge was more widespread.

I would never voluntarily participate in a shared die pool, because I bias my die pool and fully expect my opponents to do the same. I think you aren't playing competitively if you don't do this, and putting yourself at a significant disadvantage. Until FFG mandates this practice, no, sorry - I don't see a good reason to use just whatever dice are lying around instead of ones that give you a 5~% edge while knowing that many other players will be doing the same thing.

I also completely disagree with the framing of this matter as cheating. These are official products, I haven't tampered with them in any way, I'm always up front about the variance I have tested for in my dice and searching through dice lots to find ones with positive variance is not much different (at least in mind) from looking through the different available ships / upgrades to see which represent 'wrong' point values in a direction that makes them more powerful than they ought to be (and let's be real, a 5~% dice edge in a vacuum is nowhere near the edge you will get by playing a top tier list vs an average list).

Edited by President Jyrgunkarrd

Share this post


Link to post
Share on other sites
5 hours ago, Hiemfire said:

Honestly I don't think it is feasible to do so. Probably why allot of the casters and bloggers that post here are suggesting sharing a dice pool, which works well enough for me.

Oh, the tournament organiser would have to provide one set of statistically fair dice for every table, baked into your entry fee. Done.

However unless FFG introduces a quality worth its name or allows non FFG casino dice quality dice, using the app is the only way.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...