RJT said:
I agree with Bojack 100% on this one and to be totally frank about it, i wouldn't play with any player who didn't have the dedication to pass the sort of test Bo is suggesting and the MIT team required. If you can't show a very high level of accuracy on the kitchen table, then you are making far too many mistakes when playing live.
As to level 2 and beyond counting systems i have 2 issues - a) the vast majority of counters are not nearly as accurate as they think they are making in my opinion too many mistakes, multi-level systems just serve to amplify the volume of mistakes and b) the extra advantage you gleen from using a multi-level system is small and easily wiped out by making extra mistakes. What's the point in the extra complexity?
AM, if you're saying that one error per deck is not enough to wipe out your advantage, i think you need to take another look. In a shoe game that could potentially place your running count 5 out each shoe! And that's just running count mistakes. If you even make one RC error and hold that for the rest of the shoe you could potentially be making dozens with all the TC errors you are going to make.
Any player that doesn't look to play as close to perfection as humanly possible is in my experience playing a far weaker game than they think they are.
Another point that intrigues me is that i've seen it being said on here that people are only estimating decks to the nearest full/half deck?!! It might not make a dramatic difference to counting, but i've seen the same posters that have played down quarter deck estimation claim to be winning advanced techniques player. If you can't accurately estimate the discard tray to a 1/4 deck you're not going to win playing any sort of shuffle tracking or ace sequencing game. That simple.
RJT.
All right just to clear up terminology, the standard ace sequencing techniques don't require any deck estimation, just memorization of key cards. Most casino shuffles use half-deck grabs so quarter deck estimation won't do you much good in standard shuffle tracking either. The power of the zone type of shuffle tracking is increased with the resolution of your deck estimation and division, but variance goes up with inaccuracy of dealer grabs when you get too fanatical with resolution. In some 8D shuffles dividing the shoe up into four 2D segments gives you plenty of advantage.
Then there is the subjective, visual type of shuffle tracking that requires merely following a pack of cards through the shuffle. No deck estimation of any kind is required for that, just excellent visual skills. My visual skills and deck estimation aren't that good, so I concentrate on sequencing and binary cutoff tracking. But full deck estimation is more than enough for counting. Every simulation ever done proves this, and also that rounded indices are almost as good as exact.
While it's true that a counting error per deck could add up to 5 by the end of the shoe, the odds are only 1 out of 16 of that happening. The odds of the high cards being on the other side of the cut card, leaving you betting into a non-advantage situation are much higher. Counting errors like that leave you nowhere near negative EV.
Nonetheless, the fact that the variance we are playing against dwarfs the effects of random errors is no reason to not reduce errors. I'd be the last one to tell a player to play carelessly or using a technique he is not ready for. But let's keep it real. It is very important to me to know the magnitude of both the advantages and disadvantages we encounter. I don't buy it when people overstate the advantage we can play with nor when they overstate the risks.
Not to tell anyone else how to play, but this is the Prime Directive of card counting- "Wait until there are a lot of aces and 10's left, then bet a lot." Everything else is just a detail, particularly the exact definition of the word "lot." But if you follow the Prime Directive, you are playing with an advantage. Even players of weak systems like Ace-Five are playing with an advantage.
