GTO and Exploitative Play

GTO and Exploitative play

Today I’m going to expand on a dichotomy that has been largely subtextual but would be more useful made explicit.

GTO
In poker there is a concept known as GTO, or Game Theory Optimal. In a GTO model every decision is optimal because it cannot be exploited by an opposing player even if he has full knowledge of your gameplan because the risk-reward is entirely and mathematically accounted for.

The best illustration of GTO is the Prisoner’s Dilemma. The Prisoner’s Dilemma is easily solvable. A payoff matrix reveals that without knowledge of the opponent’s decision you should always betray them. This is the best decision precisely because it has the highest reward attached to an outcome that cannot be made worse (punished) by the opponent. Interestingly, humans have a cognitive bias toward cooperative behavior even though cooperation is in this case a mathematically losing strategy. This demonstrates the importance of actually making the matrix to determine the GTO in even this simple scenario.

Now, what if we turn to Rock Paper Scissors?
GTO for RPS is to throw rock 1/3 times, paper 1/3 times, and scissors 1/3 times in a random order. This has the highest reward attached to an outcome that cannot be punished. But if your opponent deviates from GTO then this is a little bit problematic. Let’s say that you face an opponent that abandons randomization and always throws rock. GTO demands that you ignore him and continue to randomize your throws. In the case that there is some knowledge of the opponent, GTO is in practice suboptimal depending on how you define optimal. This is of course what makes the a theoretical GTO so interesting in poker, a game in which results are measured in profit over time. Game Theory Optimal carries the highest profit with the least amount or risk but this does not not necessarily equal Most Profitable— in fact GTO only breaks even. Thus, if maximum profit is the goal then GTO is suboptimal in any case in which the opponent is not also playing GTO! By refusing to open yourself up to exploitation, you cannot exploit an opponent.

Exploitative Play
Exploitative play is, in a nutshell, recognizing risk in an opponent’s gameplan and compensating for it. Let’s say I recognize that my opponent throws rock every hand. Even though rock-only is exploitable, GTO cannot exploit it. In order to exploit rock-only I have to abandon GTO and adopt a more paper-heavy strategy. Once I do, provided that my opponent does not deviate from rock-only, Paper-heavy has an increased profit that is exactly as profitable as it proportionally favors paper. HOWEVER, in abandoning GTO to exploit my opponent’s strategy, I have adhered to a new strategy that is equally exploitable. It is entirely possible for my opponent to counter-adjust and switch to Scissors-only. That is the risk attached to abandoning GTO in search of profit. Your opponent may punish you at least as severely as you sought to punish them.

In summary:
GTO is maximizing profit by eliminating risk.
Exploitative play is further maximizing profit while inviting risk.


So what does this mean for Melee?

Potentially a lot. As I’ve repeatedly discussed, mixups are closely related to RPS. There is an inherent GTO. Adhering to or abandoning GTO for a more exploitative strategy is a judgement call that we always make deliberately, intuitively, or out of ignorance. It might be appropriate, it might not be. It’s a matter for individual assessment.

Here is what we should remember:

* GTO goes even unless you gain an unfair advantage, at which point GTO will always win over time precisely because it eliminates risk. It's specifically designed not to lose.

* Similarly, an optimized GTO model is more profitable than an underdeveloped GTO model.
If you are playing with rock (1pt), paper (1pt), and scissors (1pt) but your opponent is playing with rock (1pt), paper (1pt), and nail-clippers(.25 pts) then you win over the long term even without any exploitative play because you're using better options.

* Exploitative play requires that you understand your opponent’s strategy. You might consider it Attacking your Opponent’s Understanding. Maybe your opponent’s brain honestly believes that rock-only is optimal. Or maybe he’s just leading with rock to try and bait a paper switch. In a fighting game in which prepared reactions can trump a mixup scenario altogether there's a huge difference. In order to be successful, exploitative play requires 1) information and 2) acumen, otherwise it is not strategy, it’s just blind hope and high-risk variance.



Further reading:
https://arxiv.org/pdf/1404.5199v1.pdf and
http://poker.cs.ualberta.ca/publications/IJCAI03.pdf

Faster Improvement

Faster Improvement

Before you start, you have to accept a few base assumptions.
* In a game, “skill” is just your ability to execute winning tactics/strategy.
* In this context, “improvement” is synonymous with “learning skills.”
* Skill acquisition is a function of the accumulation of focus-intensive work, not directly of time.

Remember the Four Stages of Competence? It works nicely with these.
It, integrated as a cyclical model found in Improved Drastic Improvement, is the best methodology known to me. At its heart, this model simply asks that you:

* Identify a problem.
* Identify the solution.
* Practice the solution until it’s in your unconscious.
* Repeat.

Over time these correct solutions accumulate to form your unconscious gameplan. Each skill as learned individually measurably contributes to your results, ideally building on one another to create a juggernaut. With enough of the curated skills worked to the unconscious level, winning is inevitable.

But what does this process actually look like?

As of now, I think it’s best manifested in the following manner.

1) Record a (netplay/tournament/seriouslies) set.
2) Immediately break and identify the ONE most important lapse in your execution/strategy.
3) Identify the best tournament-viable solution to that lapse.
4) Practice executing this solution until it’s locked into your unconscious.
5) Repeat.

I’m sure that sounds a bit repetitive at this point, but in the past year or two the increase in popularity and infrastructure within the community has created a thriving netplay scene. Because you can record the equivalent of a tournament set vs a worthy human opponent at will it is now PRACTICAL to use the above model as your primary way to play, not just to color or elaborate on infinite friendlies. A like-minded friend that is able and willing to go through this process with you is obviously still a godlike asset, but it is a boon that this very rare kind of individual is no longer a requirement to go through it.

Before closing, I would like to make a few points.
* Working on exactly one issue at a time allows you to focus harder on it, increasing the efficacy of your practice.
* This issue could be technical/strategical/mental/health/attentional/etc. Anything that is an issue is an issue.
* Some issues might be completely solved in a matter of minutes, hours, or weeks. It might take some studying with debug mode, google, another smasher, or even a book to find the best possible solution for you. Who knows?
* The most effective practice with the smallest time commitment is two or three half-hour sessions spread throughout the day. Remember, intensity of focus is more important than time spent.
* What you choose to prioritize and what you honestly believe is the best solution will culminate into your personal style. It’s silly to worry about that because it will happen naturally.

Godpuff Reaction Techchase

I've frequently wondered if Puff has a valid reaction techchase similar to sheik/falcon's.
Shortly after researching spikestun rest I posited one that I've since ruled flawed and deleted but now I'm revisiting the idea.

There are a few recurring situations (most frequently upsmash at ~20-30%, pound, and some AC bairs on spacies) where puff can position herself at a tech position as it occurs. This opens up the possibility to reaction techchase. However due to puff's poor speed it is not obvious how she can cover all four tech options. The following sequence can cover all four within a practiced reaction time.

1) dash toward the tech location
2) if MISSED TECH, pivot rest
3) SH
4) if TECH IN PLACE, rest
5) if TECH ROLL OUT/IN, immediately drift toward then use pound's boost to catch it.

This techchase is difficult and unintuitive but feasible. Because it is so infrequent it violates the 80/20 rule and I do not advocate practicing it. But it is pretty funny/cool.

Coaching Abstract

I've decided to start experimenting with structured coaching.

The following is a proposal outlining the overarching design. I am open and expecting to tailor it to meet individual needs. If interested please contact me at acnesler@gmail.com on on fb.


Aim

My goal in coaching is to give you the tools that you need to improve as you want to. I want to be a resource that enormously benefits you by way of greatly simplifying/streamlining and the process of improvement. I cannot do the work for you but I can make sure that you are doing the most relevant work and doing it well, eliminating waste and confusion. It should have dramatic impact.

This all assumes that coaching is distinct from an analyst. Here I would define each as: an analyst dissects your gameplay and offers corrections. I already offer match analysis for $10 a game to fill that role. A coach analyzes your methodology (in and out of game) and offers corrections.

Qualification
alexspuffstuff.blogspot.com lol
To be transparent, my direct experience regarding coaching is limited to working with Soft to construct our methodology. That being said, having tested individual ideas on Soft and myself and having thoroughly researched their legitimacy I am extremely confident in our ideation.

I know first-hand the frustration of a bad teacher or false information. For that reason I make a concentrated effort to purge anecdotal strats and make sure that anything I say can be verified in game or is supported by (usually empirical) research. I’ve spent a huge amount of time in debug mode and have had to read a lot of scientific journals/etc over the past year.

Pricing
I would ask for $30 initially to prepare personalized notes/schema
then $20 per prepared session +$5 for every hour.
I.e. 0-59 min=$20, 60 min=$25, 120 min=$30
This is much less than an LSAT tutor and I think I have higher efficacy.

A session may be done over skype or the equivalent. That time will be used that time to a) identify a systemic problem that needs solved and b) discuss the best way to solve it. Afterward you just have to execute whatever plan we make. When it comes time that you feel that you need more elaboration or come across a different and meaningful problem, we can set another session. This way you are in full control of your pace, we both make the most of the time spent, and you can stop at any time you want.

This is just my first impulse and it reflects me taking your coaching very seriously. Like, preparing you to be a top regional player or more levels of seriousness. If this is too ambitious or if you think there’s a better way to structure for you than sessions then I am naturally open to that conversation.

Working together to find the best possible personal solution is, I think, the core value of successful pedagogy.

Method
We will work together to identify a very clearly defined
* appropriate goal
* subgoal(s)
* optimal methodology per subgoal
* appropriate timeline for subgoals
While these will certainly be flexible and may shift/evolve, they should always remain defined.

Your Responsibilities

* Log your effort. We need to keep track of how long you spend working on anything that we talk about and how effective that effort is, otherwise we can’t truly track your progress.
* Open communication and feedback. If something feels like it isn’t working or is confusing, we need to identify the problem as soon as possible or it can’t get solved.
* Try to stick to whatever schedule we come up with. If you can’t/don’t that’s ok but again needs reflected in your log.
* Anything more that we decide together.

My Responsibilities
* Make quick and accurate assessments.
* Work with you to find the best solutions to specific problems.

* Keep a group of google docs
 - 1 to archive our conversation etc
 - 1 to organize notes tailored to your individual situation.
 - 1 to be your smash bible
* Anything more that we decide together.

Once more, please direct any inquiries to me at acnesler@gmail.com on via fb message.

Mixups

Mixups

Mixups are a unique mechanic that emerges from the inherent design of fighting games as character vs character, button vs button and a 2D surface. Developers have intentionally evolved and expanded the genre by using the concept of a mixup as a— perhaps the—core mechanic. I would personally venture to say that mixups are what make fighting games an enduringly compelling genre. However, the term is frequently used inaccurately or too ambiguously to describe what is actually happening in-game. In the following article we will examine the broader design of a mixup to distinguish mixups and pseudo-random move selection as distinct strategies.

What exactly is a mixup?

A mixup is a game within the game, a key moment most often seen in neutral that takes looks something like a cross between rock paper scissors and a game of chicken. Characters all have different options. Most options will beat some, but lose to others. In a well-designed fighting game, these options create a cyclical yomi system that resembles rock paper scissors. This is simple enough to recognize. But unlike RPS, fighting games are played in real time. They have a nuanced timing aspect based on frame advantage and human reaction speed. For this reason, scissors will only beat paper if the throws are simultaneous. If thrown too late, you lose. If thrown too early, the opponent will react to scissors with rock. In this way, a mixup incorporates aspects of both strategy (weighing the risk, reward, and likelihood of rock, paper, or scissors) as well as execution (perceiving and playing with the correct timing relative to opportunity and reaction times). It’s a singular, beautiful fusion or mental and physical gaming.

What isn’t a mixup?

In a fighting game, it is crucially important to recognize the difference between mixups and randomness. In order to truly work as a mixup, an option has to be systematic. To continue the analogy, let’s say that you have options rock, paper and scissors, as does your opponent. If you choose to instead throw pencil, an unorthodox option that beats paper but loses to both scissors and rock, this should not be considered a mixup. Because pencil does not beat anything that isn’t already accounted for and actually has a greater weakness, it is not truly a mixup. It is an objectively bad play. Although this seems brutally obvious in an RPS format, 90%+ of smashers routinely throw pencil in neutral. Some simply haven’t done their homework in identifying their character’s versions of rock, paper and scissors. This is understandable because it’s a large assignment that demands learned understanding. However, others vehemently defend pencil as a mixup, claiming that unorthodox move selection is a valid strategy. This is partially true, but it is not a mixup.

If a mixup is systematic, then what is outside of the system? In high-level play, unorthodox options are either a) niche or b) suboptimal. Maybe a pencil. Maybe a scissors that only nets you half a point. Maybe a rock that gets you two points but has to be thrown a second early. Assuming the meta is evolved enough, unorthodox moves are always unorthodox for a reason. These moves don’t work well enough in the mixup system to see frequent use. That being said, using seemingly random moves is a real strategy. It takes far less research/practice. It tests your opponent’s execution and/or knowledge of the proper punish. It creates weird, unfamiliar situations that you might be better at dealing with than your opponent. In a phrase, it invites variance. This is a real strategy that works specifically because mixups exist, but it is not in itself a mixup. Remember, by definition, mixups are systematic*. Inviting variance will get you individual wins, but because it is mathematically, objectively a worse strategy for winning than a solid mixup system it is on its own an objectively, mathematically worse strategy for winning the 12 consecutive sets or what have you to win a tournament. To win a tournament you can’t just win 50% of the 50/50s. You have to win more than anyone in the room. You have to eliminate variance. Good design on top of a good punish game can do that.


*at this point it could be argued that any given option can have niche uses as a valid mixup or is at least a bad mixup but still a mixup, This is true, but outside of the spirit of this argument. It doesn’t take many weaknesses for a set of options to start having exploitable holes or insufficient rewards for the risk. This is exactly how people lose games and as such a bad mixup is not deserving of the title.

Soft vs Atma Analysis

soft(puff) vs atma(sheik/fox) at a CA biweekly

http://flowfeedback.com/feedback/qvLWQBk8K6nh5vuHn

will do your sets for $10 a game

breathing exercises

for reference
mindgames blog


Exercise: diaphragm only breathe
Place one hand on your stomach and one on your chest. If your entire breathe is from your diaphragm, then only the hand on your stomach will move, while the hand on the chest remains mostly still.




Exercise: sighing exhalation
Take several deep breaths pausing slightly at full intake. Do a controlled exhale along with a deep sigh. Allow the air to have a slight friction in your throat as it goes out, and make it as deep in your throat as you are able. Pay attention to the moment between exhalation and inhalation as a point of maximum relaxation.


Exercise: rhythmic breathing
Inhale on a count of four, hold for a count of four, exhale for a count of four, and pause for a count of four.


Exercise: rhythmic breathing 1:2
Take a full breath and release it. Repeat each of the following sets four times, then move to the next. Inhale on a count of four, exhale on eight. Inhale on five, exhale on ten. Inhale on six, exhale on twelve. Then go back down the series. Inhale on five, exhale on ten. Inhale on four, exhale on eight. Inhale on two exhale on four. If you run out of breath step back a count and try to breath more deeply and exhale more slowly. Try to control the exhale with your lower back as well and add a deep sigh on the exhale if the friction helps you meet the count.


Exercise: rhythmic breathing countdown
Take a full breath and release it. Visualize the number five. Choose relaxing imagery. Use as many senses as possible. As you get more skilled at imagery or this activity you will more easily incorporate more senses. What does it look like? feel like? What does the environment smell like? What sounds are occurring? What is the taste? When you are ready, on the next inhalation, mentally count down to the number four. When you exhale say internally, “I am more relaxed than I was at 5.” If you prefer, breath several times on number 4, or move directly on to number 3 on the next inhalation. Then on the exhalation say internally, “I am more relaxed than I was at 4.” Let yourself feel the relaxation spread from your chest to your limbs, and deeper into the body. Proceed the same way to number 1. It may take anywhere from thirty seconds to over two minutes to so the complete exercise. The important thing is the effect. If you feel totally calm and relaxed at number 1 then it is going well.


Exercise: rhythmic breathing counting
Breath on a similar pattern to simple rhythmic breathing. For example, four counts in, hold four counts, four counts out, pause four counts. Choose a count that works for you and then do it a few times till it becomes natural. Then start counting your breaths. Focus your mind on the rhythm of the breathing, and the relaxation that accompanies each exhalation. Allow the breath to wash away any intruding thoughts. See how high you can count, and then once you lose count start over. Try to allow your mind to be completely subsumed by your breathing rhythm and the count, keeping away any distracting thoughts or sounds. You can try this exercise in any position. However it may help to start by lying on your back on a flat surface. Place your feet a little wider than your hips, let your feet fall to the side. Hands should be laying alongside you with your palms up, as close to or far from your body as is comfortable. You can also place one hand on your upper stomach to enhance your connection to the rhythm.