I made a video about mistakes when timing the roulette ball. More than 80% of predictions were within + or – 2 pockets.
How would you describe that error in time with ms? when clocking ball speed!
Cheers
[quote=“lucky_strike, post:2, topic:1225”]How would you describe that error in time with ms? when clocking ball speed!
Cheers[/quote] like didn’t watch video till end, it was only 2 spins repeated back & forward…
Let’s assume he predicts around 600 ms rotation. I didn’t understood that clicking in different diamonds & stuff… it just spoil the test.
I see on graph 1/4 of wheel full of points with clear pick at the sentre. For first sight, most of predictions are OK, some are revolution of ball back and forward… so around 150 ms Posible time window with most errors inside 20-25 ms to each side …
I , while watching video , with vb3 was right every time even without timer…well my error was expected ( all predictions inside 3 poket).
However it rises other qwestion, how it would perform (rc) without all these fancy tricks of clocking on different diamonds ?
I think Forester have to repeat this demo on one spin only and clocking on one diamond only. And then , make a demo on different spins one after other. So we can compare results of both tests to see clearer picture.
[quote=“serge, post:3, topic:1225”]However it rises other qwestion, how it would perform (rc) without all these fancy tricks of clocking on different diamonds ?
I think Forester have to repeat this demo on one spin only and clocking on one diamond only. And then , make a demo on different spins one after other. So we can compare results of both tests to see clearer picture.[/quote]
Not really because you can look data from first diamond which I clocked 5 times. The other diamonds are just a bonus to show some power. Differed spins are real predictions and it may take long time to see any benefits.
I , while watching video , with vb3 was right every time even without timer....well my error was expected ( all predictions inside 3 poket).
Prediction & Clocking Errors
That’s wrong comparison, equal to it would be if I set to one dd and select tilt play. Then no matter which reasonable error I make predictions would be same. 3 pockets what you see as deviation at 9th rotation it would be equal to ~18 pockets mistake on outcome if you apply same levelled wheel principles. 3 pockets means 3x50ms per pocket =150ms which may be half of accumulated difference in time of 3 ball rotations. So even if you apply 2s time to observe ~3 ball rotations, 3 pockets difference would mean as clocked 2000ms or 2150, which is half rotation mistake similar as I’ve got with deliberately created a mistake.
“How would you describe that error in time with ms? when clocking ball speed!
Cheers”
That is a good question. Put it this way, if I made a 40ms mistake as + or – 20 ms on 4 rotations clocked it would show as a mistake of ~1/10 of rotation or (+ - ) 2 pockets, and it looks I was mostly there. You can see some #16 predications, it was mostly earlier, because it is first prediction where only 2 ball rotations were clocked.
I also know that if play video on a large screen I get better results same applies on a live wheel. The point here is just to show that with reasonable accuracy of clocking if can clock 3-4 ball rotations results are very stable ad clocking mistakes have no significant effect.
However, there is a little help here because of this wheel. On this wheel the difference in between ball rotations is mostly ~100,100,100ms……making it easier for earlier predictions but harder in later rotations, if compared to Huxley roulette wheels. Huxley may have differences as 30,40,60,80,120,200,200… Where earlier predictions because of small differences in between rotations would be far more sensitive to clocking mistakes but later ones would produce far better results.
I were speaking of 3 pokets sector, where my predictions was inside. Like 0 would be predicted as 26 or 32. It makes my prediction inside 64 ms (32 ms to each side posible deviation), assuming 800 ms ball revolution.
That is same what I was explaining. You only don’t understand that 64ms is not 3 pockets error what you see but ~22. Put it this way, if you apply time of 1s at ball rotation where the ball moves one rotation per second, if you see 1 pocket difference it is 22 pockets mistake. If you apply time of 2 sec the one pocket mistake would mean 11 pockets error.
Why?
Because the ball travels 57ms per pocket. It means you see the ball as 1057 ms and not as 1000.
1057ms slower ball on this wheel travels about 18 pocket less then 1000ms/r ball, in that time and rotor moves ~4 pockets less, totalling ~22 pockets mistake. To be able to have 3 pockets error you would need to spot 0.2 pockets difference including start and the end of time frame. So it would be 0.1 pocket accuracy at start of the timer ad 0.1 at the end of time.
Don’t confuse something as VB2 with this test.
If I apply 2s time from DD12 as VB2 I may read same number, perhaps sometimes because of wrong start or observation it may deviate 3 pockets.
But the apply it at DD3. You will not get same number predicted as I did with FF, but you will see still same amount of pockets perhaps 2 pockets less. Which means 2 pockets difference is 9 pockets mistake, but that is with 2s applied time. Therefore mistakes you claim that they are 3 pockets you need to multiply by ~5 and that would be your real error.
If you come across E2 you would see that I used 4.5s for levelled wheel so instead 1 : 4.5 as in this case, I have had 1 : 3.
Vb3 is tricky. It has " autocorrection " built in. It show exactly in wich “revolution” ball is. Or reading that should be there if revolution don’t exist. It should be accounted from same diamond, no feature to click where you want. It use tilt algorithm…
Vb3 error is visual observation error, not the timing one. Timing errors start far from “good rotor speeds”, but can be easy predicted before and auto adjusted during play even without " predicting " them.
I need no reason to multiply anything , difference in rotor speed is adjusted similar to vb2, with move of starting point.
Of course you don’t need to multiply because it has no accuracy to predict levelled wheel for which you would need to be only ~5 times more accurate. But it is about correct, 3 pockets you see as a deviation is about 180ms mistake, ad with RC clocking it was about 40ms. If RC was switched to same tilt principle each time it would predict same number because 3 pockets deviation I have had would reduce to 3/5=0.6 pockets.
difference in rotor speed is adjusted similar to vb2, with move of starting point.Ok , but are also diference in faling place...you looks that nothing do with that... only that backspin and forward spin, what is many times less in acuracy than simply clocking ball.....that means that possible to apply method only on heavy tilted wheels, which really not exsist nowdays.
If to talk about acuracy not foregeth that here is simply physic - something like free fall process… when ball becomes in some speciall speed it start to fall down. And that speed is very stable, maybe on light balls are influence of preasure or gravity , diferent space object,but not looking to that in short period it is really stable.
So if to look to all that like to free fall process, that process must be about the same, but that isnt, because that is not totally fall proces but rolling and ball can roll diferent, depending how dealer throw it.
Sometimes that have big effect sometimes not. But if take that part where - not, here all is very simple - longer clocking have less mistakes and more stable and true result.
And finally we are against such situattion - we clocked wheel , clocked ball and here we made some mistakes , plus are possible diferend distances of moving ball, - in totall we have some possible disperssion.
That disperssion have form Gauss curve . That curve have its paramethers -wide, high, symetry…for more acurate play we must know them - say our play must be diferent when widness is big and when is smaller.
And at all we must know how that curve of disperrsion must look teoretically, because we have small data in most cases and we can have shape very diferent, but that not means that it will stay the same in future.
Main principe is that shape in long term try to stand as theoretical shape and if we have after 20 spins it other - that means that it in future will change.
[quote=“forester, post:8, topic:1225”]Of course you don’t need to multiply because it has no accuracy to predict levelled wheel for which you would need to be only ~5 times more accurate. But it is about correct, 3 pockets you see as a deviation is about 180ms mistake, ad with RC clocking it was about 40ms. If RC was switched to same tilt principle each time it would predict same number because 3 pockets deviation I have had would reduce to 3/5=0.6 pockets.[/quote] Wheel on wich you did demo was titled. Why treat it as level? I have never seen real " level" wheel. Ball overshooting one diamond or under shooting one in 2 pin game don’t count. I have seen true 3 pin with very slight tilt. But never absolutely level. Tilt will quantize timings of rotations,producing similar distances till end. Why complicate?
@ Bebedictus… here is tolk about predicting "“same spin” over and over.
Different spins may, or may not produce different results, wich as you pointed, will be governed by dispersion lows.
BTW, on today training l sou ball overshooting one revolution constantly… it’s Teflon big ball. Pressure changed. Ones it was storm… very windy… ball was making aditional 2 rotations!!! And it’s accounting from 1 sec speed.
If you come across E2 you would see that I used 4.5s for levelled wheel so instead 1 : 4.5 as in this case, I have had 1 : 3.By the way this E2 way is super good , only it needs small changes and multiplicator simply depends on place where we end - so distance till end and from rotor speed... But all that can every corect himself.
E2 is good
For a single spin tested on all diamonds is irrelevant is it tilted or not. The point is that with method you applied you would see 1000 and 1025ms/r ball speeds ~same, when the difference in between them is 9 pockets. For that reason, if from diamond 12 you moved to diamond 3 you will not detect that the ball now will spin ¼ of rotation less.
@ Forester. I find this particular capability of ff fascinating, l just do not see how l would use it, unless for adjustment. In case of adjustment, l would see 9 poket difference.
E2 system is good 
The problem was that for a levelled wheel you need to multiply differences. As Sergiy just pointed out with observation you are always 2-3 pockets wrong multiply it by 3 and it is 6-9. Add some more for not perfect balancing and other factors. I would still swap RC and todays wheels with E2 + wheels on which I was playing E2. Old Huxley wheels with pockets as 20 mm deep :D.
When you are 5 times more accurate in defining roulette ball speed you can use it for anything you want.
It doesn’t mean that you’ll be 5 times more accurate in predictions but only more accurate.
Eve on a tilted wheel if looking overlaps for example ball is predicted 5 rotations but it makes 5.75. With VB you may have close result because of overlap, but better would be to have it predicted as 5.75 and have exact result.
Forester , as i remember one of first your RC worked on principles of E2 method. Can you abit say about theese old times and about E2 computer ?
When you are 5 times more accurate in defining roulette ball speed you can use it for anything you want.
It doesn’t mean that you’ll be 5 times more accurate in predictions but only more accurate.
Eve on a tilted wheel if looking overlaps for example ball is predicted 5 rotations but it makes 5.75. With VB you may have close result because of overlap, but better would be to have it predicted as 5.75 and have exact result.[/quote] Tottaly agree. It would be better to predict exact. I do not know wich diamond gonna hit. I do not know for sure how ball gonna jump. But one thing l may say for sure, l do not rely on 0.75 rotation overlap and hit to adjacent diamond is more likely to produce result lm looking for.Not always, lm not a ninja after all.
Vb3 has its own downsides, like any other system. Needs adjusting for example… needs some samples to be uppled correct way… need to know tilt… require tracking…ets. But on same spin it not miss much. ;D
It has some positive side as well… but that is only for these who help …for now at least.
If at that time I was aware of tilt on roulette wheels, I would be better off. However, don’t know why and how but I did find out that on some wheels if I don’t multiply with 3, I get good results. On some wheels I was actually playing playing badly tuned VB2.
E2 wasn’t so accurate and my roulette understanding wast as it is now, but I bought a house 300m from casino so I could play almost every day. Even with small edge it adds up since it was many spins played.