literature on first base vs. third

Related to an earlier discussion thread, there is an
interesting paper discussing the advantage of sitting
at first base in blackjack.

Although most of the math is over my head, it seems to confirm what I think DD' was saying -- that the expected true count after removing any number of cards from the deck is the same as the actual true count before they are removed (see "True Count Theorem").

However, it claims the standard deviation from the expected true count for the third baseman is larger. I had some trouble interpreting the language used in the paper but they seemed to be saying that this standard deviation meant the third baseman would have to play longer than the first baseman to achieve the same financial gain and would experience larger flucuations in his bankroll.

The article is at

http://front.math.ucdavis.edu/math.PR/0006017

Like I said, I don't understand it very well but one of you might.

> it claims the standard deviation from the expected true count for the third baseman is larger.

And that is why third base is better - more gain from playing strategy variations. I wonder if this paper was peer reviewed?

If you are not making playing variations, then there is no difference between first and third base in conventional BJ.

Pretentious math cant can cover up the fact that the assertions are flawed. It has been demonstrated numerous times via simulation that playing variation alone separates first from third.
The argument has been settled, and continues to be settled on an almost bi-monthly basis here.

If there are not enough cards to complete the round without reshuffling.

Its also worth noting that if playing variations are used that will give you slightly different results for betting advantage also.
But the point is that the "standard deviation" effect the paper refers to just does not exist.

FWIW, if adding my voice to the argument matters, everything John, Prof, and T-hop have written above is correct.

Don

I thought that Karel had put these guys in their place and they had retracted this non-sense. This doesn't take any great amount of math education, just some common sense. If 20 cards are going to be dealt on the next round, and the true count is +5 (or whatever), the count will not in any way influence the order of the cards that are going to be dealt. If these 20 cards contain 12 high and 8 low then they have an equal probability of falling to any hand on the table. Same expectation and same variance.

Don S: I guess we should count this new thread towards the over/under from the last thread. What's your estimate now? :)

x

A tip: You guys should archive all your responses. It will save a lot of work when we do the thread again in a couple of months.

And what's with that over/under line? Not even close!

This issue seems to pop up once a year or so. And not surprisingly, it is always answered with the same explanation. Perhaps someone with absolutely nothing to do could assign numbers to all these blackjack not so frequently asked questions (as they do with jokes in prison) and then we could just answer by saying: see #336. (turtle, are you working full time?)

But there is some benefit to them continuing to pop up. This way, as the older BJ experts pass on, they can be assured that a younger "apprentice" is still around to take over steerage duties.

cheers,
adhoc

I certainly understand the theory behind this.
Normally a gambler would question why first base does not have the advantage over third base, but here we have a question of- Does SD vary even though EV stays the same!?...very interesting to me and a rather good question that has compelled me to post.
DD commenting that if the next 20 cards are dealt at + 5 everyone has an equal chance of recieving good cards. I do agree but lets go a step further and presume that 30 cards are to be dealt at a +5. It appears to me that it is possible for fluctations to logically change as more cards are dealt out by the time they arrive at third base.

One thing that is a bit disturbing to me is many patrons assume there is no more research to be made
in regards to all aspects of Blackjack.

Mott

First base bets gets his card. We then have many cards dealt and have a huge fluctuation in the count. Third base gets his card.

You may think that first base bet more accurately but this is incorrect. The true count that 1st base had prior to the deal was only an illusion. Had he known what the several cards to follow his would be he would have bet differently. The true count is based on the number of cards you have seen prior to making a decision. Nothing that happens between your decisions has any relevance. Only when it is time to make a new decision does the new information come into play. The edge, or variance of the edge, is not different for third base unless he gets to make his bet based on the new information.

You can relate this to the sports betting example I give in a post in the original thread below. Bettor#1 thinks he has a certain edge and makes his bet. Now a new list of injuries come out. He was not better off for making his bet before this information was known. His edge was an illusion, or the best estimate he could make with what was known at the time. Same with the bet for first base. Just because those cards that will cause a swing in the count are dealt after his it does not mean that their effect on the pool of cards from which his could have been drawn are any less than they are for the counter at third. You could deal third base his card from the very bottom of the deck and it would make no difference.

You have ten cups. 9 have nothing underneath. 1 cup has a $100 bill under it. 5 players each get a pick. They can pick the same cup if they like. You pick first. Based on what you knew, you have a 10% chance of picking the $100 bill. That last player picks. They now expose you choice, and all other choices between yours and the last guy to pick. The $100 bill is gone before we even see under the last guy's cup. He has a zero % chance of winning now. Does this mean that he made a less accurate choice, that he had less than a 10% chance, that the first picker had more certainty of edge? No, they will both pick correctly 10% of the time based on having the same information at the time they made the choice. The additional information about the losing and winning cups would have made a difference to the last bettor only if he had the information in advance.

Why should it make a difference?

What if they were all face down on the table, sort of like Scrabble tiles. Everybody gets to choose one in order, with first base going first, then shortstop, etc. Is there a reason to believe that 3rd base, who draws last has a lessor chance of getting one the goodies?

If you were thinking of lumping me in that subset which thinks there is no need for further research, please remove my name. I have never subscribed to such a belief. I am simply saying that there is nothing new about the present challenge to the topic of this thread, so why would we expect a different answer? New approaches, ideas, etc., are what lead to new discoveries. Nothing new here that I can see. Same cards sitting there in the shoe waiting to be dealt.

regards,
adhoc

"One thing that is a bit disturbing to me is many patrons assume there is no more research to be
made in regards to all aspects of Blackjack."

One thing that is very disturbing to me is that, even when we tell you, ad nauseam, that the research has already been done and that we already know the answer to this question, many patrons assume that if they don't like the answer and they pester us enough, we'll change our minds!

Don

this article as i scanned through it is that Sonia is in the acknowledgements. i have not been around BJ scene long, but is this the same Sonia that likes to make some pretty crazy posts around BJ21. is this another one of his jokes.
As far as the articles goes, no one seem to address the
specific problems with it for the newcomers (like myself) who have never seen it before. like T-Hopper said, yes the standard deviation from the expected true count (given some reference point) to the actual true count(after n more cards come out) increases with n . the article is correct about this.
OBVIOUSLY, which is why we like deep pen and make better playing decisions at third. but everyone places their bets at the same time with the same knowledge so third base cannot take advantage of this fact when it comes to betting. so purely vodoo to take this then make the statement that third base has longer long run (implying that the standard deviation of his ev larger).
if there is any consequence of this (other than for playing) it must affect all players alike which is where my question comes in.

actually, gotta go, will post question later.