Zone Frequencies

A good bit of my coronavirus pandemic poker activity has consisted of playing Zone poker on the Ignition site. Ignition poker was spun out from Bovada a few years ago. (Hit me up for a referral code if you want to join and create a new account.)

Zone poker is Ignition’s version of a “fast fold” game. If you don’t like your hole cards, clicking a Fold Now button instantly transfers you to a new table with a new group of villains. You have to treat every hand as if it is the first hand you’ve ever played with this particular group of strangers.

Ignition offers Zone poker for no limit Texas Hold’em, on 6-max tables, at blinds ranging from $0.02/0.25 up to $2.50/5.00.

This game is beatable, but not by developing a profile of each other player at the table. You cannot figure out ‘that player alway does this’ and ‘this player never does that’ as you would attempt to do in a normal game, and then adopt strategies to exploit their weaknesses. Every player is anonymous, and every hand brings in a new, random selection of opponents.

Instead, a game theoretical approach is needed, based on a composite profile of the player pool. Think of each attribute as having a normal distribution across a bell curve. There is a bell curve for looseness/tightness. Passive/aggressive. C-bet frequency. Bluffing frequency. A little over 68% fall within one standard deviation of the mean, and 95% within two standard deviations. A typical bell curve looks like this:

So we want to know what the players in the fat part of the bell curve are doing, and adjust our strategies to combat their tendencies. It’s imperfect, but hey! It’s poker.

Starting the the $0.10/0.25 blind levels, I began assembling a composite profile. Here are the initial results.

First, I wanted to know the VPIP for each position at the table when there was no prior action. This will provide a good sense of opening ranges, and how those ranges changes according to position. I only included hands where the only action prior to each position was folds. As expected, wider ranges are found in later positions.

Position	Limp %	Raise %	Sample Size
UTG	3.7%	15.3%	300
Hijack	4.1%	20.0%	245
Cutoff	5.3%	23.5%	187
Button	3.7%	33.1%	136
Small Blind	6.4%	27.7%	47

These frequencies do not include my own hands. Now let’s convert these figures into ranges.

UTG is raising 15.3% of the time. I’m going to assume that 80% of those raises are made using the strongest possible starting combinations (12.2% of the time) and the rest are a random assortment of outliers from maniacs, donkeys and tilted players. A top 12% range looks like this and we should only call or 3-bet with hands that play well against this range.

 

This isn’t exact. Some players might prefer A8s to QTs, or prefer 55 over JTs, as combos that are desirable for raising under-the-gun in these 6-max games. But it’s useful as a general composite, while reminding ourselves there will be some outliers too. If these are the raising hands, the limping hands must be the next strongest. While my software omits the lowest pocket pairs, I’m inclined to include them as possible limping hands.

Next to act when UTG folds is the Hijack seat. Now the range widens to 20% raises and 4.1% limps. Again using the 80/20 rule, we get the following top 16% range for the Hijack to open raise. This allows us to widen, slightly, our calling and 3-betting ranges.

Next comes the Hijack limping range… not quite good enough to raise, nor bad enough to fold, at least in the minds of the 25NL Zone player pool. Some of these players will raise with these hands; others simply fold. but limping is at least plausible.

Rinse, repeat. Here is a sample open raise range for the Cutoff seat:

Which means the limps from the Cutoff seat probably look close to this:

When everybody folds around to the Button, that seat is given a license to steal, right? The open raise range gets even wider here. Where would you draw the line? But the wider range has its drawbacks, as the Button’s frequent raises invite the blinds to 3-bet and re-steal. More on that soon.

Button limps, whether they are smart plays or not, do happen. Here’s a possible limping range from the Button:

And last but not least, the Small Blind also raises with a wide range when everyone in front has folded, albeit not quite as wide as the button. One reason for this is the risk of having the play the remainder of the hand out-of-position when called. Another reason is the Small Blind can simply fast fold his mediocre hands and move on to anther table without waiting to find out of everyone else is going to fold.

I also wanted to know a few other frequencies. When the Small Blind raises (after everyone else has folded) and the Big Blind calls, how often does the Small Blind make a continuation bet (C-bet) after the flop? When the Button open raises with their wide range, and gets re-raised (3-bet) from one of the blinds, how often does the Button fold? How often does the Cutoff fold to a 3-bet?

	%	Sample Size
Small Blind C-bet	37.5%	8
Cutoff Fold to 3-bet	0.0%	6
Button Fold to 3-bet	50.0%	6

These are very small sample sizes and cannot fully be relied upon. More data is needed for any sort of statistical validity. But until then, we can avoid making light 3-bets against the Cutoff seat.

I want to accumulate these same data at the next blind levels of Zone poker – $0.25/0.50 and $1/2. From playing a lot at all of these levels, it definitely feels like the Button raises, 3-bet frequency from everywhere and overall aggression is higher at $1/2. As the only way to gather the data is by simultaneously playing the games, I wanted to start at the lower level just in case the distraction from multi-tasking led to sub-optimal play.

And I need to fine tune my own ranges for opening raises, calls and 3-bets based on what I now know about the middle of the bell curve.