REDi is an acronym for Read, Evaluate, Decide, and Implement. REDi is a system I created a few years ago to logically and consistently analyze poker hands and arrive at the optimal actions that maximize expected value. In this post, I describe the REDi system and from where it derives. I also work through a relatively simple set-mining example to illustrate how REDi works.
All’s Fair: Poker Is Warfare
Poker is played directly against other people. You’re trying to take what they have, and they’re trying to take what you have. In a sense, poker is warfare waged with cards instead of bullets. There are high-level “Geneva Convention” rules in place (e.g., no cheating or collusion allowed), but otherwise pretty much any tactic is allowed and legal. The goal is to make fewer mistakes than your enemy, and capitalize on those that they make. Understanding the battlefield situation, evaluating enemy weaknesses, deciding on tactics, and attacking in a manner that confuses, disorients, and, ultimately, destroys the enemy is what
war poker is all about.
Uh, this all sounds a little extreme, right? Wrong. Poker is warfare, waged on the battlefields of virtual and real green felt tables, on kitchen tables, and in the back rooms of bars and nightclubs. Poker is indeed you pitted against other players, and the goal is to win and reap the spoils— so, you better treat each and every encoumter as a battle. Trust me; the other guy is thinking of you as the enemy, as someone he wants to crush and exploit, so you had better get your head around the idea that you’re in a war—or you will lose.
Which begs an obvious question: If poker is indeed war, then how do you wage effective warfare at the tables? How do the pro’s consistently win? Is there a systematic method in which you can decide on the correct and most effective actions that maximize your profits and minimize your losses?
Answer: You betcha!
Taking A Lesson From the Army: OODA
The US military is the largest and most capable war machine on the planet. Of the many strategic advantages and tactical battlefield superiorities they posses, perhaps the most formidable is their command and control structure— the brains, if you will, of the overall global organization. Said another way, the US knows how to wage war.
A key reason for this advantage is the systematic methods that US battlefield commanders are taught to help them plan and execute attacks and defenses in the most efficient, deceptive, and deadly manner possible. At the heart of this logic is something called “OODA,” which is an acronym for Observe, Orient, Decide, and Act:
- Observe. First, before you can make any informed and logical decisions going into battle you have to gather intelligence. This is the collection of the raw data and facts you need to know about the upcoming battle, your adversary, his abilities, intentions, and capabilities, his location, and so on.
- Orient. Second, you have to take all that raw information collected during the Observe step, and then break it down, analyze it, and evaluate and directly compare your capabilities to that of your enemy. The army calls this step “orienting” the situation. Who has what advantage compared to the other? What is their relative strength in comparison to yours? What are their weaknesses? The bad guys might have the higher ground, for example, but your air force is stronger which negates that advantage. And so on.
- Decide. Third, given your observations and orientations, you have to decide on an appropriate course of action. Do you attack? Retreat? Feint? Set a trap? Or something else? You proactively decide on a course of action, rather than react to the enemy’s intentions.
- Act. Finally, you need to carry out, or execute your decision in the most effective manner. This generally means applying some type of deceptive maneuvers that cause the enemy to react in a way that results in the most harm to them.
The principles of OODA are fairly easy to grasp, and can be applied to essentially any strategic or tactical decisions we face in life, even if it’s not on a literal battlefield. It’s easy to imagine this type of Observe, Orient, Decide, and Act system applied to everything from chess games to business decisions, and, as we’ll see next, it can be tweaked slightly (and renamed) to be entirely applicable to poker—with powerful results.
OODA Becomes REDI
Years ago, when I was struggling to find a way to consistently win at poker, I stumbled across the concept of OODA while on vacation reading Nathanial Fick’s excellent book One Bullet Away: The Making of a Marine Officer. Around that same time, I also read Miller et al’s Professional No Limit Hold’em Vol 1, which introduced a similar, but much more rudimentary system called REM, which stood for Range, Equity, Maximize.
I was immediately captivated by this idea of being able to systematically work through a poker hand and—presumably—arrive at the most profitable action to take by using raw data and the application of logic. This fascination developed into a full-blown obsession as I took an entire year and worked through all 52 hands of the classic poker “Donkey Test” set of questions that players have been using for years to evaluate their general knowledge of the game and understanding of how to play specific (and difficult) poker situations.
I won’t bore you with all the details (or the numerous missteps and blind alleys I took during that developmental year) for morphing OODA and REM and a half-dozen other decision tree ideas I unearthed into my own system for evaluating hands. Instead, I’ll jump right to the final system itself, which I call “REDi” (pronounced “red-eye”). REDi is an acronym for Read, Evaluate, Decide, and Implement. Here’s how it works:
- Read. Like the Observe stage in OODA, this is the initial preflop step (and then on every subsequent betting street) to gather (read) the so-called “game state” and collect as much information you can about your opponents and the detailed status of the hand in general. The information we gather includes such things as players’ positions, their stack sizes, the types and tendencies of the villains, physical and bet sizing tells our opponents exhibit, our own position and cards, and so on. The ultimate goal of this step is to put your opponent on a range of possible cards he or she likely holds. Note that we don’t try to put an opponent on two specific cards, as this is essentially impossible. Instead, we apply our observational skills to eliminate various hands from the villain’s range, and therefore narrow down his or her likely holdings. We do this preflop and then on each subsequent street of action, continually refining and narrowing the assumed range as more information becomes available to us. The other primary goals of this Read step are to a) determine what our opponent thinks about the strength of their own hand and what they want to achieve with it (e.g., bluff, build value, get to showdown cheaply, draw to a winning hand, etc.); and b) determine what our opponent thinks our hand range and intentions are. These latter things are definitely more advanced, but as we progress in learning REDi, this information will naturally fall out of our analyses.
- Evaluate. Like the Orient phase in OODA, this is the step where we apply analysis to the data and information we gathered in the Read step. We calculate things like pot and implied odds from bet, pot, and stack sizes. We also determine our relative hand strength against the range of cards we put our opponent on; this is known as estimating our “pot equity” or “hand equity.” We also try to determine how attached our opponent is to their hand, and whether they could be made to fold or not if we bet; this is known as determining how much “fold equity” we have. In this step, we also determine things like our own pot commitment, the equities that our opponent is putting us on, and so on.
- Decide. Once we’ve gathered the game state information and evaluated/analyzed it, the next step is to determine what, exactly, we want to achieve with our hand. Remember that there are only two ways to win a hand of poker: 1) get all the other players to fold their cards before we get to a showdown; or 2) showdown the best hand on the river. Therefore, if we want to win a hand, we have to actively choose one of these two paths; this is known as selecting a “line” to take. Besides these two primary lines (bluffing and value betting), there are other possible secondary lines we can take, such as drawing to a showdown hand, isolating a bad player on an early street (so that we can attack him without other players involved on a later street), semi-bluffing, and even just folding. It sounds strange, but many beginning and intermediate players struggle with winning at poker primarily because they don’t take the time to actively decide what they’re trying to achieve with their hand. That’s why this step is so critical to success at poker; you have to know what you’re trying to accomplish before you figure out how to achieve that goal.
- Implement. The final step of REDi is the Implementation stage. This is where we take our line decision made in the previous step, and figure out how best to achieve it. This is often done by the sizing of our bets, or sometimes acting in a specific manner that leads our opponents to think we’re trying to achieve something different than we actually are.
A “Simple” REDi Example
To illustrate how REDi works in practice, here’s a simple example: You’re in a $5/$10 NL live cash game with $1000 stacks. You have a tight image. A new-to-the-table unknown player under the gun (UTG) open-raises to $40. A skilled, loose-aggressive player in middle position (MP) calls. You’re on the button with 6h-6d. What do you do?
This is a relatively easy problem, and you probably already know intuitively what the correct answer is, but let’s see what REDi has to say. We’ll also examine what happens post-flop, which might not be quite as intuitive (Note that we’ll ignore the effect of rake to keep the numbers easier to track):
- UTG Villain. We said we don’t know anything about this player, so our default is to put them on the same conservative range of cards that we ourselves would open in that same position. This means 88+, ATs+, KJs+, and AQo+. Yes, they could be opening wider than this, but when we assume a conservative range we tend to keep ourselves out of trouble.
- MP Villain. We said this player is skilled, loose, and aggressive. Instead of raising, he just cold-calls with a number of players yet to act after him. He’s skilled, and has probably recognized the strength of the UTG player’s open raise, so any re-raising the MP Villain might do in this seat would probably be limited to very strong hands. Instead, he just cold-calls, so we should put him on “drawing” hands like 22-88 and suited connectors.
- Hero. We have a tight image. We will also have position if we stick around. Our hand is 6-6. There are two players left to act after us (SB and BB).
- Other. The stacks are moderately deep at 100xbb. The pot is currently $95 ($40+$40+$10+$5).
- Against the UTG villain’s range, our 6-6 hand has only about 35% equity. He’s representing a strong range, so our fold equity is relatively weak. He probably is only folding the weakest part of his range to a re-raise from us. He’s deep stacked, so is not yet pot committed.
- Against the assumed MP villain’s range, our hand is a 60:40 favorite. He’s moderately weak, and our fold equity is probably decent against him. He’s also deep stacked, so his pot commitment is also relatively low.
- Against both opponents combined, our hand equity is ~26%.
- Direct pot odds of us calling are $95:$40, or 2.38:1, which is 29.6%.
- Implied odds (if all the money ends up going in) against either player our implied odds are ~$1000:40 = 25:1.
- Stack-to-Pot (SPR) ratio of calling is: 960/(40+40+40+10+5) = 7.1
- With SPR of 7.1, we won’t be pot committed post flop.
- We’re not getting enough direct pot odds to call and expect to have the best hand enough of the time to be immediately profitable.
- We are however getting 25:1 implied odds to hit our set. Our rule of thumb for set mining is to have at least 15:1 implied odds.
- Because the UTG player has the most assumed strength, he would be the most likely to pay us off if we hit a hit a six on the flop.
- Our line to take is therefore a drawing line to hit a set of sixes against the UTG player.
- We decide to implement the drawing line by calling the $40.
- If either blind re-raises, we will need to stop and reevaluate, noting that we might actually have to fold preflop.
We call and both blinds fold. The flop comes out Jc-2h-6s, giving us middle set. The UTG villain bets $100, and the MP player calls. What should we do?
- Pot size is now $135.
- The UTG and MP stack size is now $860; our stack size is $960.
- The board texture is uncoordinated and relatively dry.
- UTG villain is still showing strength on this dry board. Given his preflop range, he could have hit a set of Jacks on this flop, but most of his self-perceived value is still with the big pairs like QQ-AA, and maybe TT and 99. He might also just be continuation betting with his missed big over-cards, like AK. In other words, he is probably c-betting his entire preflop range here.
- MP villain just cold-called the UTG bet. He’s skilled and probably recognizes that the board is dry, so he could be continuing with any of his middle and small pairs here, thinking the UTG Villain is continuation betting his entire range and his own pair might still be good. We can rule out almost all of the suited connectors from his continuation range, except for QJ and JT, as there are no real draws on the board that align with the other connectors.
- Hero. We hit our set and have position. Life is good!
- We have huge equity against the two villain’s modified ranges— something like 87% or so against both opponents combined.
- Neither opponent is pot committed at this point, with only about 14% of their stacks in the middle.
- We have a monster value hand.
- In general, our goal with a value hand should be to get more money into the middle, and, in fact, with this much value we should try to play for stacks against at least one of these players. We do not want to slow-play and allow either player to get a look at a free turn card.
- We want to bet enough to build value, and give the wrong odds for them to draw to a better hand than ours (e.g., a higher set), but not so much as to scare one or both players off the pot.
- A re-raise between 1.5-2.5 the UTG bet size will look like we’re trying to thin the field, but not appear to pot commit ourselves. It might cause the UTG to come over the top on us, which of course would be fine.
We raise to $250 on the flop. The UTG player calls, and the MP villain folds. The turn card is the 3d. The UTG villain checks to us. What should we do?
- Pot is now $635.
- Remaining effective stacks are $710. The turn card has not changed much if anything; i.e., it does not improve the villain’s hand given his range that we have put him on.
- Villain’s check to us is in response to our re-raise on the flop. He’s probably wary and wants to see what we do. He has to be wondering why we raised the flop, and he could be putting us on a set (we’re unlike to have two pair on this board). He could also be putting us on a TPGK-type hand (Top Pair, Good Kicker) and think we’re trying to push him off the hand with our flop raise; I.e., if he’s halfway good, he realizes that many (too many) players slow play flopped sets, so it might be hard for him to convince himself that we actually have a hand that beats his.
- Our pot equity has just jumped to 95% or so.
- Our fold equity against the villain has also gone up somewhat.
- With an SPR of 710/635 = 1.1 and this much equity, we’re pot committed, even if something like an Ace comes out on the board next.
- We have monster value. Life is
- We have monster value. Life is
- Our default implementation with any kind of big value hand should almost always be to continue building on that value by betting. Only if we’re very afraid of the villain folding should we consider checking or otherwise slowing down. The villain has come along thus far and is essentially as pot committed as we are.
- This leads us to betting as the preferred option. So do we bet big or small?
- Betting large builds the most value now, but also increases fold equity a lot and could maybe cause the villain to get off his hand. Yes, he’s mathematically pot committed, but we can’t assume he’s fully psychologically committed to that idea; good players often have the ability to get away from big one-pair hands when playing for their stacks.
- Betting small gives the villain a good price to come along. It could also look weak and induce a check-shove. We don’t want to bet too small, however, as we still want to build a pot.
- Therefore, we will bet $250, which is ~40% of the pot size. This amount gives the villain the wrong odds to hit a river over-set if he has an over-pair, but it still looks a little weak and gives the appearance of giving him a good price to see a showdown on the river.
The river card is the Js, pairing the board. UTG villain checks again to us. What do we do?
- Pot size is $1135.
- Remaining stacks are $460.
- Nothing materially has changed. If we were beaten before the river, we’re still beaten. If we were ahead (which is very likely) we’re still ahead. If the villain was putting us on TPGK, he might see this J as us improving above his over-pair. In other words, our fold equity just went up, which is not something we really want, but also nothing we can do anything about.
- Given the villains range and the board, our pot equity is sky high at 98%.
- Fold equity is relatively high.
- We’re pot committed. So is the villain.
- We still have massive value, so that is still our line.
- With only $460 behind, shoving is really the only reasonable option. Betting smaller might seem like an option to help ensure the villain doesn’t fold, but the truth is that with this stack-to-pot ratio and the action thus far, the villain is either going to call or not, regardless of our bet size, so it might as well be big.
- We decide the best implementation is to shove all-in.
Okay, you might be thinking, “Wow, this seems like analysis overkill. It’s a lot of work to get to what should be an easy decision on each street.” Yes, but this is in fact what the pro’s do on each hand they’re involved in; they read the situation and put their opponents on hand ranges; they evaluate their pot equity, fold equity, and commitment; they proactively decide what line they want to take; and then they implement that line in a way that best achieves the desired outcome.
Sometimes REDi analyses lead to “obvious” results (like this one), but sometimes we get to non-intuitive lines and/or implementation methods. You won’t actually know until you run through the process. If you don’t analyze a hand in a systematic manner like REDi you’re really just guessing in poker. And guessing is bad– really, really bad.
The Exceptional Bottom Line
There are many different ways to evaluate a poker hand and arrive at the correct decision and action to take (remember, it’s the decisions that matter in poker, not the actual results). I originally developed REDi because I’m an engineer at heart, and really just wanted to see if it was possible to create a simple and systematic means of reducing the glut of information in a poker hand to a simple, defensible decision. I truly wanted something that was easy, straightforward, and logical. No more guessing at the correction decision, no more “psychic soul” reads or other mumbo-jumbo that some experts purport to use. Instead, I wanted linear, math-based and easy. REDi is what I ended up with.
Is REDi perfect? Hell, no; of course not. But REDi is pretty darn good. It is in fact how I play poker, how I teach it, and how I analyze poker hands away from the table. As a result, I’m a consistent winner at medium stakes and below, both online and live. You can be too if you implement something similar. Logic wins in long run; guessing loses.
In upcoming posts, I’ll break down each of the REDi steps in detail and examine them more closely. I also intend to begin (re)working my way through the entire Donkey Test set of questions, applying this new-and-improved REDi system directly wherever possible to show how this works in practice, and posting the analyses here on this blog for you to follow along with me.
And I really do hope you follow along. As you transition from what-are-my-cards level-1 poker to what-cards-do-my-opponents-have level-2, you have to master the art of hand reading and equity analysis. You also need to actively choose lines and select the highest EV manner in which to implement those lines. REDi is a logical framework to do just that.
Heck, if this kind of system is good enough for the US military, it’s more than good enough for you and me sitting at small-stakes poker tables! Stay tuned…
Learn. Master. Crush.
I have a favor to ask of you. If you liked this post, I implore you to please share it on Facebook or Twitter. There are buttons right below this sentence to make it one-click easy. Cheers!