The chart is much the same as in the previous post, but the methodology is a bit different. The results are sorted into four different buckets. If they had zero discards when calling riichi, it went into the Double Riichi bucket. If they had less than six, it went into the First Row bucket. If they had more than six and less than twelve, it went into the Second Row bucket, and if they had twelve or more, it went into the Third Row bucket.
Additionally, only first riichis were counted, not chasing riichis.
For an example, let's look at how the win rate of ryanmen shapes changes as we progress through the stages of the game. As a reminder, the columns label how many of the wait are visible, either in discards, the hand, or dora indicators.
Double Riichi |
First Row |
Second Row |
Third Row |
In the Double Riichi section, we can know things more solidly because of how little options there are. A ryanmen shape with 4 out is probably a shanpon, and with 5 out, a shanpon with the dora indicator or an early discard involved. The 4 and 5 outs for the ryanmen shapes only have 71 instances, out of 8966, so these are minor.
Using this data, we can also roughly estimate what the wait type is at each stage in the game.
The "Other" section is probably mostly various shanpon shapes, with some complex shapes in there. The frequency of a single wait (either tanki, penchan, or kanchan) isn't that much higher when you compare the first row with the second row. I was curious how much of those single waits were honor tanki, so you can see that at the bottom. It's drastically higher for double riichi.
Check out all of the data in this spreadsheet. Until next time!
No comments:
Post a Comment