#89 Foreign Legions

[macrobo]

Hi All

Another fun filled one

Working backwards - great tips from Toby on encirclement - very very informative
Dave and Martin battle the Foreign Legions - I had a different opinion how the Republicans could win but alas Martin did not do this! - even with Dave's unsuper tank tactics
An analysis of Foreign Legions from a 1/35 modelling perspective - I always think of Paul Weir and how he would have added so much more wisdom when discuss military vehicles and tactics - I miss him
A Rv of Double One the tournement just fresh off the press

 

[RandyT0001]

An analysis of Foreign Legions from a 1/35 modelling perspective - I always think of Paul Weir and how he would have added so much more wisdom when discuss military vehicles and tactics - I miss him

As do I.

 

[JoeArthur]

Hey Gents,

Some comments on your game:

• Ian Ainsworth has a great name for when you surround the VC hex with bodies. He calls it "the donut of death".
• Did not see Martin try for vehicular smoke grenades with those tin cans. It might have provided some cover for the infantry?
• Mike Rudd (Toby's long time opponent - now retired from the game) always used to say tanks belong behind the enemy infantry to stop them routing or hopefully eliminate them for failure to rout. Which ties in with Toby's encirclement tip.

Double One was great. Pleasure to play you in Ragnarok Dave. Looking forward to your episode where you examine whether it is possible to manipulate the dice rolls to lower the bell curve whilst still using a dice tower

Thanks to Morris and Gary for a great tourny.

Many thanks for another great episode

 

[bendizoid]

I’ve enjoyed playing this scenario several times. I like the mix of units, especially the Moroccans ie ‘Italian Gurkhas’. Unfortunately the scenario heavily favors the defender, I’d say 85-15. Primarily because of the wall of bodies defense to protect J4 and the other is 6 rampaging mk1 tanks. Consider giving the attacker the balance.

 

[daveramsey]

Looking forward to your episode where you examine whether it is possible to manipulate the dice rolls to lower the bell curve whilst still using a dice tower

It's already been done - and without a dice tower (which by definition adds randomness). Check out this video.

Michaela is a scientist and tests human rolling where you try to roll the same way vs randomised human rolling vs a repeatable roll by a machine, loading a die in the same way. The results are pretty definitive.

 

[RandyT0001]

It's already been done - and without a dice tower (which by definition adds randomness). Check out the video.

Michaela is a scientist and tests human rolling where you try to roll the same way vs randomised human rolling vs a repeatable roll by a machine, loading a die in the same way. The results are pretty definitive.

The video does not address Joe's question. He asked if one can manipulate the outcome of dice rolls using a dice tower. The video tries to answer the question, do you need a dice tower for truly random rolls and the answer is no. It does not try to simulate manipulation of the dice rolls, with and/or without a tower, to alter the outcome as the person manipulating the rolls intends.

 

[daveramsey]

The video does not address Joe's question. He asked if one can manipulate the outcome of dice rolls using a dice tower. The video tries to answer the question, do you need a dice tower for truly random rolls and the answer is no. It does not try to simulate manipulation of the dice rolls, with and/or without a tower, to alter the outcome as the person manipulating the rolls intends.

The question came up because Joe likes to have dice rolled into a tower from a dice cup to avoid stacking the dice in a way that distorts the curve that he thought might be possible without the addition of the cup.

The video shows that rolling dice in an identical way onto a table generates completely random results. My comment was if you add a dice tower to the table that can’t make the roll less random because the tower can be considered a table that adds more randomness to the roll (provable by the fact that if I drop a die onto a table from a small height I will not get random results but I will get less predicable results if I do the same with a tower).

if he still thinks it’s possible to influence the outcome with the use of a tower I will do the experiment but it looks pretty conclusive to me, already.

 

[JoeArthur]

That bloke who was placing his dice in his dice tower with a sheepish look on his face rather than dropping them in has a lot to answer for

That was when I went out and bought two dice cups.

Maybe he was deluding himself..............be interesting to find out.

 

[DVexile]

My comment was if you add a dice tower to the table that can’t make the roll less random because the tower can be considered a table that adds more randomness to the roll (provable by the fact that if I drop a die onto a table from a small height I will not get random results but I will get less predicable results if I do the same with a tower).

That's not necessarily true. The video makes the point that randomness in small height drops is down to the hardness of the table being rolled onto. If the landing surface of the dice tower is softer than a hard table (and in fact many of them are) then indeed a dice tower could make the results less random than just rolling onto a harder table. Whether the baffles inside the tower overcome the reduction of randomness due to the softer landing zone is of course dependent on the tower design.

So certainly we'd hope a well designed tower is plenty random (and I suspect all but the worst designed would be) but that isn't guaranteed. Viewing the tower as another random process inserted into the rolling "algorithm" isn't an accurate model unless the exit height of the tower is at least as high as the non-tower die roll and the die tower landing surface is the same (or at least same hardness and texture) as the non-tower die roll.

Or, you know, we could all realize "bad dice" is usually 99% false perceptions and 1% reality and just get on with enjoying the game!

As usual, thanks for the fun episode!

 

[daveramsey]

Ok! I’ll do the test. Got to get my gadgets out though. Look out for it on the next episode!

 

[DVexile]

Ok! I’ll do the test. Got to get my gadgets out though. Look out for it on the next episode!

Lol! Well certainly don't do it for me though! I'm 99% certain you'll find the dice tower, no matter how you use it, produces random rolls to any reasonable statistically significant level you choose for the test. I was just nitpicking that assuming that adding something to a process makes it "better" isn't always true. In reality I'd be quite surprised if you found the dice tower hurts in any way.

But of course if you think your viewership in general will be entertained by watching you roll endlessly then by all means!

 

[daveramsey]

But of course if you think your viewership in general will be entertained by watching you roll endlessly then by all means!

I mean… how much worse can we get?!

 

[Tuomo]

I want to see a time lapse, with Dave getting progressively more exhausted as the rolls add up. Mopping his sweaty brow, white headband, the works. Until around DR #2657, his metacarpal explodes and he crumples in agony.

 

[JoeArthur]

You got to love this game

Try placing them in the dice tower Dave, do not drop them from a height. Also try manipulating the faces so that they are both the same. For instance both one's face up. See what you get after a good sample size - Indy said 100 rolls?

It is worth noting that the six's are on the opposite side of the one's and that opponent looked unhappy when he rolled a boxcars (but then don't we all).

The problem with this game is that as Martin Vicca says, to win you just have to roll low. Which is why the dice drive us all nuts.

Got to hand it to you - I just went out and bought a set of dice cups like the backgammon players.............

 

[daveramsey]

Phew. I've done it!

So - some assumptions:

1. In order to try to disprove Joe's concern, I built a dice roller. I don't see the difference between placing and dropping dice, and presumably we want to try to get repeatable results. So my contraption takes one die which is placed so that it's the same facing for each drop and is dropped with the exact same speed and position.
2. One die rather than two should show us if there is a bias because we remove an influencer in the other die - if there's no bias after rolling one it's unlikely that the introduction of a second will change things.
3. 100 dice rolls is enough to test with. Scenarios range from about 150-250 dice rolls, so we're on the low side but unless we want a 95% confidence rate we need something like 6000 rolls.
4. In order to concede that there is an issue, I'll present the results to Martin. If he spots the dice-roller's results, we'll consider that a win for the cheaters.
5. I'll test against rolling dice in a tower vs dropping them vs auto-rolling them.

Next episode - ASL's version of the Terminator...

 

[Tuomo]

"I'm sorry, Dave..."

 

[JoeArthur]

Got to admire you Dave.

Bonus points if you write an academic paper on the subject.

But I would place the dice on the ledge rather than drop them.

Good luck!

 

[daveramsey]

But I would place the dice on the ledge rather than drop them.

Hmm - I'll see what I can knock up. This way though is, in theory, repeatable and completely consistent. If you're talking about placing them on a sliding baffle thing, that's almost impossible to do and probably an unrealistic expectation of what someone might try to do.

The key thing here is if you can repeat a throw with very high accuracy can we see a difference in the dice results. And what would you do if someone does this to you? Tune in next week...

 

[JoeArthur]

If you're talking about placing them on a sliding baffle thing, that's almost impossible to do and probably an unrealistic expectation of what someone might try to do.

That Dave is exactly what happened. He placed them very slightly above the baffle.

In fact, if my memory is correct, on certain crucial dice rolls he placed them in the dice tower trying to reach the second baffle.........

 

[MTL]

Phew. I've done it!

So - some assumptions:

1. In order to try to disprove Joe's concern, I built a dice roller. I don't see the difference between placing and dropping dice, and presumably we want to try to get repeatable results. So my contraption takes one die which is placed so that it's the same facing for each drop and is dropped with the exact same speed and position.
2. One die rather than two should show us if there is a bias because we remove an influencer in the other die - if there's no bias after rolling one it's unlikely that the introduction of a second will change things.
3. 100 dice rolls is enough to test with. Scenarios range from about 150-250 dice rolls, so we're on the low side but unless we want a 95% confidence rate we need something like 6000 rolls.
4. In order to concede that there is an issue, I'll present the results to Martin. If he spots the dice-roller's results, we'll consider that a win for the cheaters.
5. I'll test against rolling dice in a tower vs dropping them vs auto-rolling them.

Next episode - ASL's version of the Terminator...

View attachment 26242

Can it fish out and put away my counters before and after play?