ICMIZER poker calculator celebrates 1 year birthday
I am pretty happy to announce that today ICMIZER celebrates its first birthday. 1 year has passed since beta was over and it had become a commercial program. In this post I will share some ICMIZER history, provide up to date comparison with SNG Wizard, share some future plans and show how the new version will look, so keep reading.
I am a professional programmer. I hold a master's degree in Computer Science and got years of experience in IT. I also happen to be a professional poker player.
I was mostly grinding and making money playing poker sit and goes and only dreaming about creating some big personal software project, but I didn’t have the time or money till the June of 2011. I wasn’t certain about it but I had decided to put my effort into creation of a better poker calculator, which would replace a de facto standard tournament calculator SNG Wizard.
Goal was pretty crazy when I think about it now, but I thought I had a shot at it. I happen to have an important combination of skills required for such task. That is both excellent programming and poker skills. So at the same time I know how to create great software, and what professional SNG players need from ICM poker calculator. I thought this unique position could give me a chance on the market, if things go right. And so far, after a year I can say they went pretty good.
ICMIZER vs Sng Wizard
This is a pretty popular and reoccurring topic. For so many years Sng Wizard had been a standard no brainer choice for anyone who wanted to study ICM in SNG or MTT tournaments. A program hasn’t been updated for years now though, and ICMIZER has been specifically built with Sng Wizard flaws in mind.
It would be impossible to have a shot at the market if program was worse than Sng Wizard, no one would want to hear about it. So ICMIZER stayed free for a long time while I believed it didn’t have enough value compared to Sng Wizard, and it became commercial at the point where I have dealt with all major Sng Wizard flaws.
A year has passed since that time and here I will state some simple facts on how ICMIZER is actually better than Sng Wizard
1) As a calculator it has to be about precise results. It sounds pretty strange to have a calculator that produces wrong results, so people still wonder why Sng Wizard and ICMIZER produce different results for the same tournament spots, with the same opponent ranges.
Sng Wizard uses simplified formulas and produces close to correct results, while ICMIZER uses highly optimized algorithms and most precise formulas.
Sng Wizard does several types of calculation simplifications
Two way spot simplification. When two players go all-in preflop against each other there exist 3 potential outcomes. First player wins, second player wins, or they tie and split the pot. It is pretty simple, but Sng Wizard makes an assumption that first player either wins or loses, ignoring the tie outcome. It is extremely important in any ICM tournament situation. For example, busting on the bubble of SNG tournament is always a bad thing.
Our actual bust probability is always smaller than Wizard assumes, because in case we tie we do not bust, and remain in tournament. This seriously affects each hands equity even in cases when our tie probability against opponent range may be as low as 2%.
Three way spot simplification. This is a kind of spot where someone pushes, other player calls and another player overcalls. So 3 players go all-in preflop. In this scenario there are 13 potential outcomes. There are six different ways players can end up if they do not tie. You can think first > second > third, or third > first > second. There also are 6 different ways all-in can play out if two of the players have a tie outcome, for example first and second tie, and third loses, or third wins, and first and second splits. And the last outcome is when all players tie, for example when they play the board straight.
Here again, each time we do not bust from tournament our equity increases, so we fear this all-in situation less, and get better expected value of shoving. ICMIZER considers all 13 outcomes and Sng Wizard simplifies this, ignores ties again and only assumes there are 4 different potential outcomes here. This adds up to mistake from two way calculation making results even farther from reality.
When we consider our folding equity and what will happen behind us our hand often seriously affects chance that players behind us get dealt better hands, say Ax and shove, call and bust each other. So when we hold 32o and fold we will experience a pleasant view of players busting against each other behind us more often, compared to cases when we fold an Ax hand. This effect is usually called card removal effect and ICMIZER takes it into account when it does the calculation. Opponent call probability is basically rarely equal to his/her vacuum range dealt probability, because it is always affected by hero hand. Card removal effect plays role in two way spots, three way spots and two way spots when hero folds and other players go all-in against each other
So to sum it up, I have paid a lot of time creating a fast and precise algorithm for push and fold equity calculation. When you run calculation in ICMIZER you will get correct result. It will be basically never the same as Sng Wizard, except for trivial heads up spots where none of effects described above affect the results.
2) Another annoying limit of Sng Wizard is limited ability to edit hand range. You are given two sliders which allow some control over the range, but you cannot just enter the range you want. For example you cannot paste range from PokerStove or add or remove a hand from it. You also cannot input a so called polarized range. By polarized range I mean a range that includes some powerful hands and some suited connectors and other worse hands in it.
This could go completely unnoticed by novice players, who often get overwhelmed by complexity of task to analyze a single hand when they see this kind of software for the first time. However professional poker players require a finer precision to edit a range and ICMIZER gives complete control over range editing. I know it myself because I am one of them. You can include or exclude any hole cards combination in any range, be it a push, call or overcall range.
3) A major update in ICMIZER since its release last year has been addition of ICM Nash equilibrium calculator. Making it was a real challenge for me. Finding nash equilibrium requires running hundreds of calculations adjusting player ranges here and there, in order to find the closest to optimal game solution. Naïve approach would require minutes or even hours to complete the task, but in ICMIZER it takes some seconds to find Nash equilibrium ranges.
Some players wonder why this is important, in Sng Wizard you get some ranges assigned to players, and then you can hit calculate button and get some results. Thing is that it is 2013 now, and games have become very hard. Poker is insanely competitive and every year games become more difficult. These days you hear GTO (Game Theory Optimal) acronym more and more often. So when you find Nash equilibrium you basically get an idea on how perfect players would play in this spot. I always say that you should be playing your opponents and assigning them their most likely ranges, when you analyze a certain tournament spot. Thinking that they are some Nash equilibrium bots can be a costly mistake for professional player.
I believe the most popular question beginner poker players ask in terms of tournament spots is: “What is opponents pushing or calling range in this spot?” They do not have experience of hundreds of thousands hands yet, but they need some baseline foundation. Nash equilibrium can provide this. It is always a good start from where you can edit the ranges according to your feel of your opponents. Nash equilibrium calculation in ICMIZER takes into account current tournament payout structure, so for the same stacks you would get different ranges if you use say default nine max payout or six max payout structure.
Sng Wizard doesn’t have any of this. All it has is a button, which opens an external website page and presents a Nash equilibrium solution. You can review it, but as I said you cannot paste ranges in Sng Wizard so if you tried to apply observed equilibrium to current Sng Wizard spot, you would need to spent a few minutes just copying it back.
ICMIZER has it as built in function so it is far more convenient and will save you a lot of time you would waste otherwise. I honestly do not expect this to be available in Sng Wizard any time soon. Also from what I could learn they don’t even promise to add this important feature.
GTO is a very curious concept but it has its limitations. The main limitation is that you aren’t playing against players who follow GTO. So blindly following Nash equilibrium suggestions can be devastating to one's bankroll. Always consider the spot from common sense perspective, and adjust opponent ranges from Nash equilibrium suggestion. Usually they call tighter than Nash suggests, and also push tighter in the most spots. However a maniac or a player who thinks that he/she had enough of you could go wider than Nash equilibrium range for given spot
4) MTT ICM calculator. In Sng Wizard you can only analyze final table tournament spots. If you wish to know what is the correct mathematical play in situation when there are 2 or more tables left in MTT you cannot use ICM calculations, and are forced to use Chip EV model.
Now this had been standard for a long while, because alternatives just didn’t exist. In last August I have added ability to analyze MTT spots with up to 60 players still playing in ICMIZER using ICM calculations.
Since no MTT ever pays full prize to first place finisher Chip EV calculations are always flawed. In reality MTT also has its bubbles, also limits your ability to play versus big stack, because you are afraid to bust and so forth. Chip Ev ignores all those factors, and just gives you raw range versus range equity.
If you run ICM calculations for MTT spots in ICMIZER you will observe that actual range is always much tighter than Chip EV suggests. It is a rather new feature, but a lot of MTT players are picking it up and start to base their play less on feel and more on math.
I believe those 4 are the main differentiating features of ICMIZER. There are more differences, and ICMIZER has some other smaller neat features here and there, but those four are most fundamental and important.
One more thing, those can be viewed as static features. There is also a dynamic difference, ICMIZER is alive and in active development. Sng Wizard didn’t change for years, and only promises to fix the simplest bugs, basically since ICMIZER beta release in September 2011, yet they are still there.
I am pretty excited about ICMIZER and I am working on it. It is pretty awesome right now, but I can guarantee that it will only get better in future, more on that later.
Preview of next big ICMIZER update
1. Initially ICMIZER supported only simplest tournament spots. Open push, calling an open push and resteal spots were supported in first beta version. Some more spots have been added over time, but this new update will take tournament spot support to a completely new level.
Basically ICMIZER will now be able to calculate any preflop spot from a tournament you could imagine. It will include but will not be limited to new spots like squeeze spots, where there is a raise and a caller; cold 4 bet or 5 bet spots, where there is a raise and a reraise before you; situations where players are automatically all-in because their stack became so small they actually blind out in this hand; or attacking single or multiple limpers from any position.
No matter how odd or complex situation is, if hero decision is either push or fold, ICMIZER will be able to analyze it and provide you with a correct +EV pushing range.
2. Another important update will affect ranges editing. To get your results you need to enter ranges for all the players first, and I have made sure it will be easier in new version.
There are the following changes
- If someone has a raise, limp or other kind of non-all-in range, it will be placed next to his/her bet, so you can clearly see that he/she raised with 37% of hands for example
- Push ranges will be wider than call ranges, and call ranges wider than overcall ranges, forming a clear hierarchy and making ranges visual comprehension easier
- Editing overcall ranges will be easier. Before they were always located on special tab and it was not so easy sometimes to relate, which range should go where. Now if you are in open shove spot, where no other player is all-in they are still on the special tab, but you can actually see and edit the push range and the call ranges, which are parents to those overcall ranges. If there is an all-in player situation is seriously simplified visually, tab control is not displayed anymore and all ranges can be seen on single screen.
3. Hero range will now be visible. Right now hero range is treated as something special and unknown by ICMIZER, so you can only observe it on results tab after you perform an ICM calculation. However it is often helpful to have hero range displayed, for example in situations when you calculate Nash equilibrium ranges. Before you had to run Nash equilibrium calculation, and then hit calculate, to see the hero pushing range.
Now you can see hero range alongside with other ranges. In case when hero range is unknown, which is when you didn’t assign Nash ranges or run the calculation yet, hero range will be displayed as question mark.
I hope this update will be available as beta later this week and probably released next week maximum.
Now this should be pretty exciting but much more is coming soon.
The next major feature will be Future Game Simulations based calculations. That will allow to take into account potential future actions after current hand, which can often seriously affect our decisions in certain situations. Say, there are few players left and their stacks are very shallow, and posting a big blind in near hand becomes very expensive. This may force someone to shove wider, than he/she would if he/she didn’t take that future action into account, like ICMIZER does currently.
I am really sorry that this one takes so long, I have been promising it for a while now.
Other features, which I plan for near future are:
- Nash equilibrium ranges calculations for all kinds of supported spots. Right now Nash ranges can be calculated for open shoves, calling open shoves and overcalling spots. I plan to add ability to find Nash ranges for all those spots where we face a raise or other kind of non-all-in bet
- Quiz. I got some curious ideas how to make it fun and efficient. This will help players who want to test their knowledge or to obtain knowledge in push fold spots.
- Nash calculations for MTT spots. Right now it is not possible to use Nash equilibrium calculations in MTT ICM calculator mode, but I think making it possible for say 20 players or less should be doable.
Of-course I also plan some pretty exciting changes, which I wouldn’t announce yet.
I want to thank all players from all over the world who chose ICMIZER, who provided ideas, support, help and other valuable input.
Some players had found the time to write to me their success stories. Their ROI has increased since they have been using ICMIZER and they have earned some good cash and moved up in stakes. That is ultimately the goal of ICMIZER as a program, to help players get better and make more money. It requires a lot of work to run all those calculations, and apply gained knowledge in game, but evidently it pays doing that. I thank you guys for letting me know about your successes and always welcome this kind of feedback by email or on forums.
Q, ICMIZER developer