Week 9 - Mentality

 Jared Tendler's book is very good. I think basically everyone should read it.

This Week

I had a streak of bad variance in Expert 1, so it took a bit longer than expected, but I reached Expert 2. Longer than expected? It's only been a week... Well, it felt longer. Here were my stats at the moment I reached it:

Rather than results, let's look at Maka grades. Maka only keeps the grades for a week, and I didn't think to do this until a few days in, so I don't have all of them for last week.

I've really cleaned up the lower end. I'll reiterate that I didn't consciously change anything about my Mahjong. I just focused on avoiding tilt and improving my mentality.

Ultimately I decided to follow my heart and do the Tenhoui-style yakuhai pons. My callrate is pretty high now. I'll have to pay attention to my calling and make sure I'm not being silly and breaking my rules. Here's a couple of unusual calls. Click the images to make them bigger, this is as big as I can make them in Blogger without the overlapping the sidebar.

I thought this call was a little insane, but I suppose if Maka has it at 43, it's a must call for me. The rules also say I can call bad shapes whenever, but triple kanchan sanshoku is definitely outside of the norm.

I was in furiten, and knew if I called kan, I would get out of furiten, on top of showing a kabe to everyone. I could switch my tanki to almost any other tile I draw, but the 1p wait looks strong. Also maybe I could check off that sankantsu achievement. I think this is a good call.

I also had an interesting moment where Maka disagreed with aiming for first.

Only 25 skip?

This ron is 4500, which is 200 points short of getting first. A tsumo or ron from shimocha on 9s gets first. Of course, I'd accept the 6s tsumo with moderate complaints. Mortal also has this as a 100% ron.

26 skip now? Why is it higher with one more visible?

It's still 2 skip?

With the kan, we have enough fu to make this a 5800 ron, which takes first even if we hadn't hit the dora.

I suppose considering the fact that Mahjong Soul's ranking points are based on your points in game, taking a win that doesn't raise your placement has more merit. Winning the 4500 ron is better than tsumo on 6s, even though first is worth a lot more. But, I'm definitely the type to aim for first when there's basically no chance of fourth.

In line with the mental training, let's do some math to figure out how much variance can actually affect us, in hopes of tempering the impulsive feelings with logic. To be kind, let's assume we have a 30% first place rate.

To estimate how long we might go without a first, we can use ln(games) / ln(1/winrate). With that, over 100 games a streak of 13 games without a first is completely normal, and over 1000 games a streak of 20 games is completely normal. This is just variance, unrelated to our skill.

If it was a perfect 25% winrate, it'd be 16 games over 100, and 24 over 1000.

Of course, you can have a bad streak that isn't just losses. Using binomial calculation, over 100 games, someone with a 25% winrate has a one in ten chance of only getting a 20% or worse winrate in those 100. Of ten people with a 30% winrate, one in ten would see only a 24% winrate. Results like these can make you doubt your ability or simply curse the Mahjong gods.

And this goes both ways. You could have a 30% winrate and get a 36% winrate over 100 games, or a 25% winrate and get 30% over 100 games. Results like this can inflate your ego, make you feel more confident and better than you are, and lessen your motivation to improve. In Mahjong terms, 100 is a rather small sample. All my stats might be +/-5% from reality.

In Uotani Yuumi's book, she suggests starting by playing 1000 games to see what your real ability is. That seems to be wise.

For an example of this, look at the Tenhoui （≧▽≦）'s stats in Gold versus Jade.

Gold room games.

Jade room games.

The Gold room results are worse. But, obviously, Gold players are not better than Jade players. That's just how the variance shook out in those 106 games. Though, Gold is probably higher variance than Jade, due to the less frequent folding.

For the opposite side of the variance coin, we can do the same with the triple Tenhoui ないおトン.

Gold room games.

Jade room games.

Their Gold results are completely different, but they normalize in Jade where the sample size is over a thousand games. ないおトン only spent 23 games in Gold.

In terms of my mentality issues, the one I was really thinking over this week was, "Why do I get upset by seeing players make bad plays, even when it doesn't affect me?" Logically, my opponents being weak should help me.

I think this specific case is something the book being poker-focused is difficult for. A major difference between poker and online Mahjong is that poker is played for money. A weak player means more profit for you, which I think would be more motivating. In Mahjong, a weak player is not profit.

And, I'm fairly confident the underlying reason is not entitlement tilt. I never think things like, "Why am I losing to these players?" or such. At the same time, it's not injustice. What is it?

I rubber ducked it with ChatGPT a bit and it made a good observation:

I think that seems quite likely. An annoyance with rank lag, or the length of the grind. I suppose the answer to that is, as always, focusing on improving my play instead of the results. I can train most of my skills (reading, bluffs, assists and such aside) regardless of the quality of my opponents.

Next Week

So, I've been shifting my focus away from ranking, and towards my own play. So far, I've been doing that through the Maka grades. Getting an S- or better is a success for me, and an A+ or worse is a failure.

However, there's another core mental issue that comes along with this shift:

Perfectionism.

If you're focused entirely on results, you might develop some ladder anxiety, fearing a fourth. Switching my focus to Maka's grades, I instead developed performance anxiety, worrying about receiving an A+ or worse, a feeling that grew worse the better my results were. "My grades are so good for this week, I don't want to ruin them..."

I don't do well when under expectations, including my own. That's probably one of the main things that went wrong last time.

So, next week, let's ruin the grades. Let's try to play the worst Mahjong of my life. I'll play when I'm tired, I'll play while the parrots are screaming, I'll play while on tilt. Any mistakes I make during this time will be ones that are buried in my subconscious, revealing what I need to work on to improve the lower end of my game. If I'd truly mastered them, I would do them automatically.

My theoretical understanding of Mahjong is quite high, but I can't apply all of it in every game. It's a perfect example of the wide bell curve that Jared Tendler talks about in the book, where my C game and A game are extremely different. Pushing the C game forward is just as valuable as pushing the A game forward.

The mental book also suggests having one day off a week where you don't do anything poker related and do other things instead, to prevent burnout. Let's make that be Thursday for me. It's a day I go into the office, and having a gap might let me have a fresh perspective when writing the blog post. I tried minimizing my Mahjong exposure yesterday and it was surprisingly difficult, especially with 6 new books being delivered.

New Year's Resolution Progress: 29/1000 (2 ahead of pace)