Lupe Soto Interview
Lupe Soto has advocated for Women in Poker since the late 1990. Starting as a “weekend warrior” mixed games player,
Preflop Strategy: How to Adjust Your Play as MTTs Deepen
During the early stages of a tournament, our decisions are based on a system of risks and rewards that value winning and losing chips completely symmetrically. Just like in a cash game where every chip has direct monetary value, winning 1 chip is ”good” in the same way as losing 1 chip is “bad”. When we study the opening ranges generated by a solver for these early stages of the events, it uses the mathematical model of symmetrical risks & rewards called “Chip EV”, or cEV for short.
However, as the tournament progresses, the field thins down to 50%, then 25% and we are cruising towards the money bubble, the risk/reward system changes. It’s a tournament, afterall – and the value of the chips is no longer measured directly in how many of them we have, but in how well the chip stack enables us to win prize money.
The negative impact of busting – thus losing any chance of making money is “worse” than the benefit of doubling up, which still guarantees nothing! This phenomenon is known as “risk premium”, and it changes rather significantly how the players should think about their decisions – particularly preflop. The mathematical model that takes the chips distributions and payouts into account to compute the optimal preflop ranges in each spot is called “Independent Chip Model”, or ICM for short.
Risk premium-based models tell us we must adjust our play based on how deep in the tournament we are. So, how can we go about studying the relevant theoretical ideas that we can also implement successfully in practice?
The Practical Way to Study Using ICM50 / ICM25 ranges
In real life, our decisions to open, raise, shove, call or fold should be heavily based on the distribution of chips amongst the players, how close we are to the money, and the payout structure. That’s a lot of factors to consider! A precise ICM model for each decision in each spot is totally custom every time. Needless to say, that’s an impractical way to study poker.
While it’s always a good idea to look at some specific tournament spots, run a custom ICM model through a solver, analyze the optimal play for each participant and fill in your own bank of ideas that way, there are other ways to go about it.
In this article we will present a systematic way of studying how the size of the field left affects common preflop decisions. We will be using the three sets of ranges generated using very similar parameters:
- Positions/scenarios: all positions, all reasonable preflop scenarios;
- Stack depth: 20/30/40/50 bb;
- Even stacks only: meaning that all 8 players at the table have the same stack;
- Field left: 100% / 50% / 25%
- ITM: 12% of the field
- Payout structure: typical of large online MTTs
These ranges can be found at www.OctopiPoker.ai, along with the Range Comparison tool we’ll be using today to glean the strategic insights.
By looking side-by-side at how the preflop ranges change as we get deeper, we can develop a list of heuristics that helps to guide our decisions in real life. Let’s delve in.
Early/Mid Positions RFI @20bb, ChipEV vs. ICM25
This is a side-by-side comparison of UTG-8 and High Jack open range in ChipEV and ICM25 (with 25% of field left) at 20bb. The strategic differences are outlined for clarity. What patterns can we observe?
Diagram 1: UTG-8 and HJ RFI @20bb, ChipEV vs ICM25
Presented by: OctopiPoker.ai
Initial observations:
- We open and fold about the same % of the range. However…
- The deeper in MTT we are, the more we value Ax.
- Low pairs drop in value. UTG-8 goes from pure opening 66 to pure opening 88!
- Suited Kx/Qx/J9/T9/98x and lower suited connected holdings that have good playability postflop lose their value fast.
What is the meta-pattern that unifies these observations?
- We are not too interested in playing postflop!
- We want to win the pot right here, right now. This is why we favor Ax above all else – it blocks their continues.
- If we are forced to play postflop, we want 2 broadways that can make big pairs. Thus QJo/KTo opens, J9s/T9s doesn’t, and neither do smaller pairs. The value of non-nutty flushes and straights is largely diminished as we don’t want to play multiple streets to realize our value.
Let’s look at the later position opens at similar depth to confirm our findings.
Blind vs Blind @20bb: chipEV vs. ICM25
This is a side-by-side comparison of Small Blind open range in ChipEV and ICM25 (with 25% of field left) at 20bb,and the Big Blind defense, in ALL-IN and NON-ALL-IN scenarios. The strategic differences, once again, are outlined for clarity.
Diagram 2: SB RFI @ 20bb, and BB response to NON-All IN, chipEV vs. ICM25
Diagram 3: SB RFI @ 20bb, and BB response to ALL-IN, chipEV vs. ICM25
Presented by: OctopiPoker.ai
What do we observe?
- Small Blind changes its strategy dramatically: Folds nearly double. Calls are cut more than in half. All-ins also double!
- In response to NON-All IN, Big Blind folds more, calls less, and raises more
- In response to ALL-IN, Big Blind puts his tournament life at risk and calls 30% less! – even though SB jams TWICE as wide!
The same overall concepts of avoiding post-flop play all together, as well as calling off jams way tighter as we get deeper is on full display.
What about in-position Play vs. Open?
Button vs. UTG-8, CO @30bb: chipEV vs. ICM25
Let’s see how the Button response changes when facing an open from UTG-8 or CO at @30bb in ChipEV vs. ICM25. We have already seen the FRI ranges. We can recall that they will open slightly tighter, and the shape of their range will be geared towards Ax and away from small pairs and middling connectors.
Now let’s focus our attention on the Button response.
Diagram 4: BTN response to UTG-8, CO RFI @30bb, chipEV vs. ICM25
The strategic changes in the response should not surprise us by now. They are very consistent with the previous themes:
- The willingness of the In-Position player to flat falls by over 300%!
- Only best suited broadways are willing to make calls.
- Lower pairs are discarded, along with suited connectors
- Lower suited Ax, such as A3s – A8s are only entering via 3bet (one pip lower facing CO than UTG-8)
- BTN vs CO, we are also no longer willing to jam any of our hands.
You can look at a number of other preflop scenarios, only to find the same themes. If you were to add ICM50 range side-by-side range analysis, you would see that those ranges fall right in the middle between ChipEV and ICM25.
In other words, the strategic adjustments that come from thinking in terms of payouts (ICM) rather than value of chips (chipEV) starts literally at the beginning of the tournament with very light risk premiums, and gradually increases as we get closer to the money. The risk premiums are the highest on the bubble, and then again on the Final Table. Looking at those scenarios would be the subject of another article!
Summary and Takeaways
Just a small sample of the preflop ranges at different stack depth and positions reveal a consistent patterns:
- Open tighter
- Fold more
- Call less
- 3bet more
- Jam less – unless you can put extra pressure on the small number of stacks behind you, like SB vs BB.
- Favor Ax and big cards over small pairs and suited middling cards
I hope this has been a short and sweet introduction to strategic ICM adjustments in late stages of the tournaments. I would like to invite you to explore more ICM50/ICM25 ranges at OctopiPoker.ai.
Are WSOP Circuit Rings Worthless?
Two things can be true at the same time: There can be an excess of poker tournaments that award gold
Remembering Mike Sexton: The Irrepressible Voice Of Poker
In January 2019, myself and my Chip Race co-host Dara O’Kearney interviewed a poker player and broadcaster whose popularity had
