Yomi Goals?

I can affirm that lots of cpu matches teaches hand management and matchups. Playing against the computer helped me learn that stuff faster than against people by sheer virtue of the ludicrous numbers of games I would play against the computer daily. I have about maybe 2-3 thousand games against the computer and less than 200 against people.

4 Likes

I have another diversion. I am currently reading Range by David Epstein. It is about generalized learning vs specialized learning. There is an interesting section on unique human skills versus what machines can do. For instance, in the game of chess, Grandmasters learn the game through the process of chunking. They learn through grinding the game and chunking the game states. Usually, they know their move within seconds.

However, if the rules are slightly changed, they are worse than novices. Machines can replicate this chunking type of learning. Machines cannot adapt to rules that are more dynamic. I am still trying to wrap my head around this, but I would recommend the book for anyone interested in game theory and more broadly in learning in general. I am sure there is some application to Yomi in there.

3 Likes

Frankly Chessā€™ static nature is part of why I donā€™t like it.

Iā€™ve already achieved my yomi goals of being able to do reasonably well in competitive settings playing only characters I enjoy. As much as I would like to aspire to be a ā€œtopā€ player I think the yomi community is way too small in its current state to feel like itā€™s really competitive. Gone are the days of 100+ entrants in IYL.

4 Likes

we can play yomi for $$ or some half yomi half poker whenever you want :wink:

3 Likes

I limit myself to the characters I like as well. However, two of my favorite characters are Troq and Setsuki, so that isnā€™t exactly a big handicap in my case.

2 Likes

A chess grandmaster is a worse fairy chess player than a novice? That canā€™t be right. It makes sense that having a lot of knowledge about a very similar, but distinct game would be misleading and detrimental to your skill but there is still a lot that remains applicable - I canā€™t see a grandmaster losing to someone whose knowledge of the game stops at how the pieces move.

1 Like

I would have to think about that. Unfortunately, I would be breaking rule one. Only take on risk when the odds are in my favor.

@ArthurWynne , I will check for the exact chess quote at a later time.

3 Likes

I believe that (without additional training) an AI trained on chess using something like AlphaGoZero would be worse than a novice when moved to a different game. That same statement seems much less true of a human player.

Personally, my goals are actually about doing what I can to revive the scene so that I keep having players to play with. Currently, thatā€™s taking the form of keeping tournaments running so there are always games to be found, but I have some more ambitious long-term plans.

If anyone has any ideas in that regard, Iā€™d love to hear them.

10 Likes

I do. Iā€™ll send you a PM with my thoughts.

3 Likes

I said I would update this when or if I accomplished my goal. When I was a kid, I loved baseball. I used to set a mitt up against the concrete and throw the ball into it. If it landed in the mitt it was a strike and if it did not, it was a ball. I created the hits with my imagination. I would simulate an entire season. My point is that I loved baseball. I would read the USA Today to keep up on the latest stats, as well.

To be considered a really great team, there is the 100 win season. The MLB season is 162 games long. If your team finished 100-62, then the team would be considered a greatā€¦maybe historically great team.

Early on, I had the sense that the Yomi win probability was similar to baseball (I remember the great Deluks917 agreeing with this in the chat). I knew winning 80 percent was not going to happen. I created the goal of the 100 win Yomi career before/at game 162. Today, I accomplished that feat. I am now 100-59. I did it. I remember being 53-38. I thought it wouldnā€™t happen, but it did. It does feel good to accomplish something I set out to do five years ago.

Thanks @mysticjuicer for being the Yomi stats and data guy. I probably would have forgot about this idea if you didnā€™t input the data.

11 Likes