Blackjack Series: Final Analysis - Caleb Jones

In February this year, I made the first post in this blackjack series, stating that I make regular net income from playing blackjack. Since then I have posted five articles here regarding exactly how I play, and exactly how much money I have won and lost since February.

For your convenience, all five articles are linked here:

Part 1: The Winning Move

Part 2: Basic Strategy

Part 3: Table Selection

Part 4: Betting Strategies

Part 5: Counting Cards

You may recall that when I first posted about this in February, several guys commented that I was full of shit, and that winning money in blackjack was mathematically impossible. Since then, as you can see in the above posts, I have won a net profit of $5,307 playing blackjack. This was between February 1st and September 15th, representing an average monthly income of about $707.

Yes, you can win money playing blackjack, provided: A) you play consistently, over a longer period of time, B) you play correctly 100% of the time, in ways virtually no one else plays, and C) maintain absolute emotional control while you play. Easier said than done, but still very doable.

Today I’m going to summarize my results from the last few months playing blackjack in more easy-to-read charts and graphs, to show you with real numbers that making money like this is indeed possible when playing blackjack.

Since I track all this stuff on spreadsheets, here’s is summary of my wins and losses since February. The session numbers (first column) are color-coded. Black=$25 base bet, red=$50 base bet, and blue=$100 base bet.

blackjack

As you can see, when I lose, I lose big, but I don’t lose very often. The vast majority of my blackjack sessions result in small wins. These small wins add up over time to exceed the infrequent, large losses, resulting in net profit for me.

Playing 42 times, I had 7 losses, but 35 wins. This means I win 83.3% of the time I play. If you won 83% of the time you played blackjack, don’t you think it would be pretty easy to make money doing it? (Provided you kept your losses as small as possible of course.)

This is the same exact pattern I have experienced for years and years of playing blackjack. I usually win, but rarely win very much. I sometimes lose, and when I do the loss is much bigger than the average win. (Events like that $1900 win during the final session are rare, and notice even then it was immediately proceeded by a massive loss.)

Yet the wins eventually out-profit the losses. I’ve never had a year of playing blackjack, playing it as I’ve described in this series, where this wasn’t the case. In other words, this year was not an unusual year for me in terms of blackjack winnings. It was typical.

The only difference this year was that I jacked my bets up to $100, something I’ve never done before. But in terms of win/loss percentages and win/loss amount ratios, this year was typical.

Here’s a graph showing the same wins and losses (click to zoom):

blackjack

On February 1st, I started playing with $225. By September 15th, I had turned that $225 into $5,532. That’s a profit of $5,307, or an income of $707 per month.

Because I raised my betting amounts about halfway into the process (at session number 24), this skewed the numerical balance upwards, but the overall pattern and trend is what’s important.

Yes, you can make money playing blackjack like this, provided you do all of the following, 100% percent of the time (as summarized from my previous articles on this topic):

1. You GET UP, STOP PLAYING, and LEAVE THE TABLE  as soon as you start winning a game.

2. You play basic strategy and never deviate even if your emotions are screaming at you to do so.

3. You only play at the tables and casinos that offer the absolute best rules which benefit the player’s odds of winning, and never play anywhere else.

4. You only bet a maximum of one-twentieth (or 5%) per hand of your starting pot, and never deviate from this unless you’re doubling/splitting per basic strategy or if you’re counting cards and you’re 100% sure the card count is insanely good (which is rare). You also constantly pay attention to your running pot balance as you play, and follow very strict stop-loss and stop-win rules based on this balance.

5. You pay attention, at least somewhat, to the composition of the cards in the deck. You don’t have to literally “count cards” via the standard Hi-Lo system (I don’t), but you should be at least paying attention and keeping track to some degree, taking advantage of a good deck composition when it occurs (and it usually won’t).

This is the final installment in this blackjack series. I hope it’s been helpful. One of my goals for 2016 will be to practice real (Hi-Lo) card counting to the point where I get as good as a pro (or at least semi-pro). But again, even if I never do this, I’ll still be making money playing blackjack. 🙂

Category: Fun

22 Comments on “Blackjack Series: Final Analysis

  1. The issue with your first post isn’t that wining is impossible, it certainly is possible if you play right and count cards, but that when you leave the table doesn’t matter. If you are able to have a positive expected value when playing blackjack, leaving the table early hurts you, since you will win more money the more hands you play.

  2. The issue with your first post isn’t that wining is impossible

    The arugment from commenters in that first post was that what I was saying wouldn’t work because if you played blackjack on a computer it wouldn’t work blah blah blah.

    when you leave the table doesn’t matter

    Not only does it matter, it’s one of the most important aspects of this.

    If you don’t pay attention to when you leave the table, you will almost always leave when you are losing. Thus you will never win money over the long haul.

    If you are able to have a positive expected value when playing blackjack, leaving the table early hurts you, since you will win more money the more hands you play.

    Go back and re-read that post. If my pot balance is way, way up, and I’m on a huge, consistent winning streak, then I keep playing until I lose 2 or 3 times, then I leave.

    This doesn’t happen very often though.

  3. The problem in your initial post was that you made it sound like all you had to do was leave the table when you were up and you’d come out ahead. That certainly wouldn’t work. Counting cards and playing the correct strategy does work, and will work no matter when you leave the table. The Ace/Five count gives the player the edge over the house, which means the longer the player plays, the more money they’ll win.

    This is very easy to test in simulations, if you remember the bet I offered in your first post. To make it simple I would word the bet as “Using the Ace/Five counting strategy, and the Basic strategy, and a favorable table, the simulation will come out ahead after playing 1,000,000 hands in a row without stopping”. We could also run that simulation thousands of times and either keep track of the net earnings or percentage of time it came out ahead, as long as those first three assumptions are true more hands will beat less hands.

  4. The problem in your initial post was that you made it sound like all you had to do was leave the table when you were up and you’d come out ahead.

    No. I made it very clear, in the article and in the comments, that that post was the first in a series of many necessary techniques, and leaving while at positive winnings was one of those.

    To make it simple I would word the bet as “Using the Ace/Five counting strategy, and the Basic strategy, and a favorable table, the simulation will come out ahead after playing 1,000,000 hands in a row without stopping”. We could also run that simulation thousands of times and either keep track of the net earnings or percentage of time it came out ahead, as long as those first three assumptions are true more hands will beat less hands.

    I’m not going to bet based on a computer model, but I’d be happy to make some kind of bet based on real-life play.

    I agree that an expert card counter playing basic strategy 100% of the time will come out ahead in the long run. That’s what I’ve been saying and that’s exactly what these guys do.

  5. The problem with basing it off of real life play is that the sample size is too low. I like to make bets I know I’ll win 🙂

    The main disagreement we have is that leaving the table at a certain time makes a difference, why would this not be true with a computer shuffling the deck versus a human?

  6. The main disagreement we have is that leaving the table at a certain time makes a difference, why would this not be true with a computer shuffling the deck versus a human?

    I don’t know for 100% sure, but based on my research, it’s clear that not all computer models 100% reflect real-life play, and the fact you’re dying to bet me via a computer model but you refuse to bet me regarding real-life play clearly shows you agree with that assessment.

  7. Thanks for this series Caleb! I’ve consistently walked away with small profits (<$100) since reading it and studying up on basic strategy. This is under conditions that are not perfect and also odds that are unfavourable – sometimes as low as 46-47%. I agree with your assessment that mathematics do NOT model real life situations accurately.

  8. What research did you do and which computer models did you find don’t ‘100% reflect real-life play’? I’ve no problem coding a simulator such that it does if necessary. I’ll also happily take Tony’s bet against you based on real-life play provided a statistically significant sample size is used.

  9. What research did you do and which computer models did you find don’t ‘100% reflect real-life play’?

    I’ve read about 20 books on blackjack written by professional, full-time blackjack players, as well as a lot of articles. A lot of them, not all of them, but many of them, have said that computers are great for determining odds on specific plays and rules, but not massive amounts of simulated play.

    The great web site WizardofOdds.com is good example of this; lots of computer data there and it’s 100% accurate. But that data is regarding specific odds of specific, isolated moves accomplishing certain things. However the books I’ve read (not all of them, but many of them) also say when you make a computer run thorough thousands or millions of simulated games and compare that to keeping track of real-life play, the results are so different that (they say) you can’t rely on these kinds of computer models. They say the behavior of real life cards shuffled by real life dealers or shuffling machines behave differently than numbers generated randomly by a computer CPU.

    Is this right? I have no idea. There’s no way to know this for sure, and that’s true of both you and me.

    The reason I tend to lean in that direction is because of what I have seen for years with my own eyes. For years, guys have been telling me it’s impossible or near-impossible to consistently make money playing blackjack the way I play because if you ran a computer model blah blah blah…yet for some bizarre reason I have never had an year of playing blackjack where I didn’t make money. I’m always proven right and these guys are always proven wrong, for almost a decade now.

    Moreover, I’m also not the only guy who has reported this. I’ve met plenty of other guys who play a lot, don’t 100% card count, and seemingly violate these worshiped computer models. (See Lachlan’s comment above.) Not to mention all the pro blackjack players who also report results that the computer models don’t.

    Therefore, as I said in the other thread on this, there are only 3 possibilities:

    1. Magic. I’m one of the most magically lucky men on planet earth.

    2. I’m lying about all of this.

    3. Real-life play with real-life cards does not equal random simulated play on a computer CPU.

    I know number 2 isn’t correct, and I really doubt number 1 is correct (though I suppose it’s possible). That only leaves number 3.

    Could I be wrong? Sure. But so for (almost a decade) it doesn’t look like it.

    There could be a 4th option too:

    4. They way I play would report a net gain if you plugged it all into a computer, and it played exactly like me for a long time.

    That’s possible too. But again I have no idea.

    I’ll also happily take Tony’s bet against you based on real-life play provided a statistically significant sample size is used.

    Give me the details of what you have in mind and I’ll tell you if I’m down for it. Just remember I don’t have kind of schedule where I can play blackjack all day. You might need another guinea pig beside me who has more time to play and plays exactly the way I instruct him without ever deviating.

  10. “1. You GET UP, STOP PLAYING, and LEAVE THE TABLE as soon as you start winning a game.”

    What do you mean exactly? If you sit down, place one bet, and you win …. you leave after only one hand? And that counts as a session?

    What would you say the average number of hands is that you play in a session? Considering that a typical dealer with two or three people at the table will get in maybe 100 hands per hour? I’m sure it varies a lot, but what would you say your overall average hands/session are?

  11. There is a little bit of a difference between how a human shuffles and how a computer shuffles, but unless your strategy specifically takes advantage of these differences (which it doesn’t) the difference doesn’t matter. It certainly has no impact on leaving the table early.

    The reason I wouldn’t want to bet on a real life test is two-fold:

    1. The sample size will be smaller, and thus the person who is actually correct could lose by chance.

    2. Any real life player will be biased in favor a certain outcome.

    For the first point, if you have a game that pays $1 for every $1 you bet, and you win that game 51% of the time, you will win money playing that game in the long run. However, even if you played the game 300 times there’s still a 34% chance you’d end up losing money. Even playing it 1,000 times means you’d lose 25% of the time. However, if you could play 100,000 times the chances of you losing money is 1 in 8 billion. The math is similar to counting cards in blackjack. If we did a real life test there’s a chance I could lose, if we did a simulation there’s basically no chance I’d lose.

    For the second point, the only fair way to do a bet without a simulation would be for you to test both strategies (i.e. compare your old results to one where you don’t leave the table based on your winnings), but you would be subconsciously biased in favor of your former strategy, which is not good for the chances of me winning.

    The only way I can think of is if you believe there’s a strategy that works in real life but would fail in a simulation. Or your core claim, that quitting at a certain time makes a difference, which shouldn’t be effected by the difference in shuffling between real life and the computer.

  12. What do you mean exactly? If you sit down, place one bet, and you win …. you leave after only one hand? And that counts as a session?

    More like two bets, but yes. If you go back through my posts you’ll occasionally see sessions as short as 4 minutes.

    However, it’s very rare that you’ll sit down playing blackjack and immediately start winning. I almost always, as in perhaps 95% of the time, start losing as soon as I play. Then over time (but not too much time) I climb back up over my starting amount and then I bail.

  13. Great series! It’s funny, I’ve played blackjack like this for years (just playing basic strategy, and leaving when I was up a little bit) but I honestly never thought of betting more than $5 and trying to make any real money at it. Thanks to your theories I’m going to try to slowly build up and do just that.

    Two questions for you if you don’t mind. When would you usually hit your stop loss? After going down 50%, 75%, 80% of your buy in? Also, at what amount of money in your bank would you try to increase your bet? I’m thinking of having maybe 2 to 3 times the buy in before increasing my bets.

  14. I’ve played blackjack like this for years (just playing basic strategy, and leaving when I was up a little bit) but I honestly never thought of betting more than $5

    That’s how I got started! $5 a hand! I remember those days…

    When would you usually hit your stop loss?

    When I get down to about 30-40% of buy-in, sometimes more. Many times I’ve been WAY down, but eventually, after a painful game, squeaked back up to 10% over buy-in, then I bail. Win!

    at what amount of money in your bank would you try to increase your bet?

    As you said, only if I’m WAY up, as in 150% – 200% of buy-in at least. I play very conservatively.

  15. Interesting! Do you know any winning strategies in other casino games? The only casino in my city only offers tables with 4 decks, so it is not an option.

  16. No. I don’t play anything else in the casino other than blackjack, since blackjack is the only common casino game in which you can (somewhat) overcome the house edge.

  17. I see. Do you advocate for online casinos? A lot of people avoid them because the underlying software may be messed up.

  18. Do you advocate for online casinos?

    No. If I can’t see the real cards, I’m not going to trust it as a source of income.

    But playing around just for fun? Sure, I guess.

Leave a Reply

To leave a comment, enter your comment below. PLEASE make sure to read the commenting rules before commenting, since failure to follow these rules means your comment may be deleted. Also please do not use the username “Anonymous” or “Anon” or any variation thereof (makes things too confusing).

Off-topic comments are allowed, but Caleb will ignore those.

Caleb responds to comments in person, but he only does so on the two most current blog articles.

Related Posts

Begin typing your search term above and press enter to search.