Book review: "Play Razz Poker To Win"

I mentioned Mitchell Cogert's book back in March when the subject of razz first caught my attention. Mitchell has a poker blog here (mostly about razz, but with some stuff on hold'em) and a web site about his book--which appears to be self-published--here. When the book arrived in the mail, I spent an hour or so reading the first couple of chapters, then put it aside and never finished it. Based on that initial reading, I recommended it to a couple of people.

Well, tonight I finally got around to really tackling it head-on. I have read it carefully, except for the final section, which is basically a description of every hand that the author played in a $5/$10 razz session last December. I'm interested in going through those hands, but it won't affect my analysis of the book, so I'm writing up my impressions now.

I should mention at the outset that Mitchell and I have exchanged several emails, and I've found him to be friendly, smart, analytical, and open to suggestions.

First, the minor stuff. This book really needed an editor. It's far from the worst I've seen in terms of typographical and grammatical errors, but enough of them slipped through that it leaves readers--at least a reader as picky as I am--with a sense of unprofessionalism. For example, "odds" as a singular noun on pp. 12 and 44, the strange use of the word "flop" on p. 22, a missing possessive apostrophe on p. 29, confusion between "affect" and "effect" on p. 29, "maybe" versus "may be" on p. 35, "goods odds" on p. 44, "whose" instead of "who's" on p. 47, "weary" instead of "wary" on p. 54, "you was called" on p. 64 and again on p. 65, "there was two or more raises" on p. 115. Stuff like that drives me crazy.

Now for the substantive matters. Here I have to introduce a caveat so large that you may want to disregard the entire rest of my review: I'm not an expert at razz by any means, and it may be laughable for me to be reviewing a book on the subject as if I know what I'm talking about. Furthermore, my understanding of the game is still in rapid evolution, the steep part of the learning curve, and it may be that where I disagree with Mitchell, he's absolutely right, I'm absolutely wrong, and six months or a year from now I'll reread these words and cringe at how foolish I was.

However, as I have occasionally reported here, I have been more successful at the game than I thought I would be, at least for the past month or so. I'm in a much better position to do a book review now than I was when it came in the mail in, I think, early April.

There's another caveat: In many, perhaps most, cases, my disagreements on strategy choice are based on my experiences playing PokerStars razz at $0.50/$1, $1/$2, and $2/4, plus the razz component of my almost-daily $10 HORSE single-table tournaments on the same site. What works at these lowest stakes may not be good advice upon moving up to $5/$10 and above. Similarly, I found Sklansky's book on razz not only quite poorly written, but almost completely useless as an introductory text, because he focuses very specifically on live $15/$30 and $30/$60 games. That's a whole different razz world than the one I inhabit.

I also want to mention that I'm diligent about checking the hand history (either the text-only form or PS's snazzy new animated version) immediately after virtually every hand I'm involved with, so I really do know what people have been playing against me. Unlike on Full Tilt (where I believe Mitchell puts in most of his time), Stars lets you see the cards in the order they were dealt, so you don't have to guess what starting hands players took to battle. In my view, that's such a big advantage that it makes playing razz on FTP a vastly inferior choice to Stars, if you care about understanding your opponents' play.

So with those preliminaries out of the way, I'll plunge ahead.

Because I'm going to list my specific disagreements, it's worth pointing out explicitly that I agreed with maybe 90-95% of the advice and analysis and statements about what he would do in specific cases. It's easy to lose sight of that fact when looking at what will appear to be a litany of differences of opinion.

P. 19: Mitchell advises not defending the bring-in if one's up card is two ranks or more above the up card of the raiser. Frankly, I think this is pretty silly. If I'm defending my bring-in, I don't really care if I'm showing a K, a Q, a J, or a 10. Nor do I care whether my opponent is showing an A or a 7 or a 9. Let me explain why.

There are really only two possibilities. First, I think my opponent in late position is simply trying to steal with a bad hand. In that case, I assume that, regardless of what he's showing, he has at least one stinker card, a pair or a face card. When that's so, we're basically on even footing, even if my bad card is a K and his is a J, because neither of us is going to be pressing on later streets if we're including that bad card in our best five-card hand. In other words, if my bad card on 3rd is still part of my best hand on 7th, I'm not going to be putting any more money into the pot anyway. I'm either going to make a hand in which the bad card doesn't play, or I'm going to give up--so I don't care if it's a K or a 10.

The other possibility is that I actually believe my opponent is starting off strong, with three unpaired small cards, and I'm just hoping that I get lucky and he gets unlucky. Now, this may be utter donkey play, but I think it makes sense, with specific constraints. I have no mathematical justification for this, but my practice has been to defend against a single opponent for a single raise if my hole cards are both 5 or lower. If I catch bad on 4th, I'll shut down and fold to a bet even if my opponent caught bad, too. I'm not going to try to catch three good cards in a row to beat him. But if on 4th I catch good and he catches bad, then we're on roughly equal ground, and I'm often a favorite, since I've probably been pickier about selecting my two down cards to go to war with than my opponent has been. If on 4th I have A-3-K-6, and he has X-X-6-J, I'm very happy with the situation. Yeah, I'm officially behind as the hands are, but most likely neither my K nor his J will be a factor at the end, and my 6-3-A is quite likely to be ahead of his best three cards.

That's why I think Mitchell's "two level" rule doesn't make sense. It keeps coming up in other contexts through the book, and I disagree with it in those spots, too, for similar reasons.

Again, this may be specific to the bad players in the low levels at which I play and not a valid observation at higher stakes. I don't know. But I see a lot of players raising quite indiscriminately, with a hidden bad card and without good position.

P. 24: Mitchell introduces his point system for evaluating razz starting hands based on the cards, position, what other cards are showing, and the action before one makes a decision.

I'm highly suspicious of all attempts in poker to reduce decisions to formulas, and I have to include this one in my doubts. I would feel better about it if Mitchell could tell us that he had run thousands of simulated hands using a variety of differently tweaked point systems, and this is the one that yielded the best results. But as it is, it looks homemade and like guesswork to me.

A point system is undoubtedly better than if somebody does nothing but look at his own three cards, because it does incorporate all of the relevant factors (at least all those that are quantifiable--it doesn't and can't include factors such as whether the raiser is a known maniac). But I'm not at all sure that it's better than a non-quantitative, gestalt evaluation of the situation. I think it would work just as well to give general advice, such as "Be less inclined to call or raise when more of the cards you need to catch are showing," as to try to make a mechanical point system. But if it helps new players force themselves to notice and account for the relevant facts, perhaps it's more useful to them than I'm giving it credit for.

P. 30: "If a player raises with a 4 showing and a player calls with a 6 showing, the raiser may be on a steal but the caller most likely has three cards to an 8 or better."

Well, this is certainly how it should be, but, trust me, it ain't necessarily so at low stakes. People call raises with K-A-3 or 2-2-4 all the time.

P. 37: The example given is me with a 6-7-4-K against an opponent's X-X-7-7. Mitchell writes, "Don't bet here.... Wait till 5th street to decide the relative strength of your hand. Otherwise, if he hits good on 5th street and you hit bad, you have wasted a bet."

I couldn't disagree more strongly here, though yet again this may be because of the bad average quality of opposition that I'm used to. When a substantial fraction of opponents here actually have a hidden bad card (high one or a pair), both of us hitting bad on 4th becomes a great time to discover that fact and take down the pot with a bet. If I get called, then I can be reasonably confident that he is not in that category. In other words, this is a juicy time to set up a screening test by betting. It usually chases away the hands that now have two bad cards. So you either win the pot or you more clearly define your opponent's hand.

Moreover, because I'm so dang tight with my 3rd-street starting requirements, by the same logic I explained above (i.e., that if both my opponent and I have one bad card each now, neither of us will be using that bad card by the end of the hand), I should on average be ahead here, because both of us catching bad on 4th street leaves the status basically as it was on 3rd street. Perhaps if my opponent reraised me on 3rd and I just called that would imply a different situation. But if my opponent just called me on 3rd, he probably doesn't have one of the ultra-premium starting hands.

Even if we're very close at this point, I don't mind getting an extra bet from each of us into the pot, because I believe that I make, on average, better decisions on the later streets, so more money in the pot works to my advantage.

Finally, betting instead of checking projects more strength, and may help make a bluff more likely to succeed on later streets. Checking seems to be saying, "I not only didn't like that card, but I'm no longer as thrilled about my first three." Betting, conversely, seems to announce, "Yeah, 4th street didn't help me, but I started out this race ahead of you, I'm still ahead, and you, my sad little friend, are running out of streets on which to catch up."

P. 43: Mitchell writes, "As a result, you have an outcome you wanted to avoid: putting in three bets on 4th street and still having two opponents."

I don't think it's correct that this is necessarily a situation one wants to avoid. If I'm in a close race with a good opponent, but there's a bad one with an awful hand desperately trying to catch up, I'm delighted to have him stick around and continue contributing money to the pot. Hell, it's best for me if he raises when he's way behind!

It's kind of like in high/low games: if you and another player are going to be taking the high and low, but there's a third player who is drawing nearly dead for both parts of it, by all means let him put as many chips in as he is willing to!

In razz, of course, the pot won't be chopped like that, but the idea is the same: welcome into the hand the player who has the worst chance of winning it, and cap the betting if you can. If one opponent and I are roughly equally likely to take it down at the end, it is in both of our best interests to suck as many chips as possible out of the guy who is trailing. Repeat that scenario a thousand times, and if I win 40% of the pots, my good opponent 40%, and the lagging-behind guy only 20%, it's hugely +EV for both me and the other good opponent to pump the pot as far as the bad opponent will tolerate.

A couple of years ago, Mike Caro opened my eyes about the flawed concept of "thinning the field" in poker. In brief, overly aggressive raising tends to weed out the wrong players. You may keep the strongest players with the best hands, and drive away the weakest players with the worst hands, which is not what you want to accomplish. There is also, for each poker hand, a mathematically optimal number of opponents to maximize long-term profit, and that number is often something different from one. See his eye-opening article from Bluff magazine here. He was writing about hold'em specifically, but surely the same concepts are valid in razz. Yet Mitchell seems to sort of blithely assume that one always wants to contest a pot heads-up. It's just not true.

OK, that's enough for now. Much more to say, but it's after 2 a.m. and I have to take my car in for a couple of new tires in the morning. I'll try to finish up this review tomorrow.

Note: Part 2 of the review is now posted here.