I'm sure there have been threads on this, and I realize that many people believe dice setting is of little to no help unless mastered with tens if not hundreds of hours of practice and even then is susceptible to not working when used amid the chaos of real-life situations. But that all having been said, I'm sure some of you - like me - still opt to set the dice when you roll craps.

For those that do set, what sets do you use and when/why? I could never get the flying V to work at all - it did give me a decent share of 6s and 8s, but I was 7ing out too much, so I settled upon the crossed sixes and then buying the 4 and 10 and placing the 5 and 9.

Also, when you set and assuming you throw with your right hand, do you prefer stick left (where you throw across your body) or stick right? Or do you like to roll from the end of the table for a perfectly straight path?

Just curious how others approach this.

I prefer a 5's up inverted hillbilly set off the top but often switch up to a 3 left snake grip (low dangle) if the bounce is favourable.

You should read Cranston's paper. You gain 1.2% on standard felt going a 3/5 split on the inverted hillbilly.

Yeah. I think his work on the Duckfield Torsion method is great. Not too sure of the validity of that particular paper though.

Are you guys trolling? The names you are using sound made up.

Not sure why you'd troll though, as I believe dice setting has been empirically shown to be an executable (although extremely difficult) strategy over a significant sample size.

Influencing one die is close to impossible. Influencing 2 dice probably is.

That's a quite interesting statement.
Could you elaborate?

Seems intuitive to me that trying to do one task successfully is easier than doing two simultaneous tasks successfully. It is possible that each of the simultaneous tasks are independently easier than the single task. But without a strong correlation of outcome I don't think that trying to influence 2 dice to a level of statistical significance AND yielding a profitable betting strategy is worth pursuing. This is why anytime I hear of dice influence methods to throw fewer sevens than the expected mean are almost certainly nonsense.

A successful method to beat craps only requires influencing one die, and has nothing to do with lowering the expected frequency of throwing sevens, which IMO is practically impossible. That being said developing and executing such a method would probably require physical skills possessed by less than 5% of adults, tens of thousands of repetitions training, advanced analytical tools such as high speed video cameras, and specific table conditions not always available. (e.g. specific table size, specific open spot to throw from, no chips near the landing area, etc.)

I prefer the triple lindy bellyflop...overhand method obviously preferred but underhand can work if the heat is on.

Thanks for the answer!
I think this too.

I used to do Cross 6's.

In-line 6's on the comeouts...

Now I don't care anymore and just throw it.

With all of the knowledge on influencing dice, these smart guys who have figured this out should go to Silicon Valley where smart guys get so much money so much easier than this.

There's some crap called machine learning that makes dice control look like a freaking ridiculously hard problem.

All you gotta do is train the computers now and not even program them and you get $300 an hour.

You geniuses who figured out dice should just go and work where all the other smart people are with that machine learning and stuff. It probably pays better than 1.2% on a $5 bet.

10's of hours you say...

Should probably teach a course.

I like the V set 2s on top more than 3s hardway set sometimes with 4s on top I shoot right handed and want to be right of stick. that said I think it does little if any good but very impressive when you get that rare long roll.