Nerd Alert: MORE about CR 5!

Greetings and salutations from Bubbe, who wants to digress for a moment about CR 5.

That’s my short-hand name for the fifth hand in the Consecutive Runs section of this year’s card. Perhaps you’ve seen it:

                                        FFF 1111 2222 DDD

Bubbe’s just kidding. Of COURSE you’ve seen it. Everyone’s seen it, it comes up all the time. It’s a very straightforward hand: you have two numbers in a row, but you need something else to make a total of 14 tiles. Some Flowers, some of the matching Dragons, and voila. 

The required quantity of each tile varies from year to year. There have been cards like 1995 where the hand involved kongs of the numbers and Dragons, completed with a pair of Flowers:

                                       FF 1111 2222 DDDD

There have been cards like 2013 where the hand included kongs of Flowers and numbers and the Dragons were the required pair:

                                       FFFF 1111 2222 DD

And then, of course, there have been cards like 2016 when there was NO “Consecutive Run with matching Dragons” hand, and everyone moaned and sighed because they missed it.

This year's iteration—pungs and kongs—is the first version that I can recall that doesn’t use any pairs. Open hands that don't require any singles or pairs are MUCH easier because you can use Jokers for any exposure. They can’t go “dead”: for example, even if someone exposed all four of the Dragons that you need, you could still use three Jokers to complete the hand.

As I said in my article about tournament statistics, CR 5 was the winning hand in nine of the 38 non-wall games I witnessed—that’s 24%. It was far and away the most prevalent winner out of dozens of options. In general, the number of each type of winning hand from my small sample was:

                Two pungs, two kongs: 16  (42%)

                Three kongs and a pair: 10   (26%)

                 Four pungs and a pair (all closed hands): 4  (11%)

                 A kong, two pungs, two pairs:  3  (8%)

                 Any kind of quint:  3  (8%)

    Two kongs, three pairs: 1 (2.5%)

     Singles and pairs ONLY:  1 (2.5%)

               

In previous years, I have written about “old reliable” hands. There is one always listed at CR 2, the second hand in the Consecutive Runs section. It involves two pungs and two kongs, in two suits; sometimes it is two pungs in suit A, followed by two kongs in suit B; sometimes it is pung/kong in suit A, then pung/kong in suit B. It is a very simple hand: no pairs, open, easy to call for, never goes dead. Easy-peasy.

I think that this year, because it involves the Flowers and only one suit, CR5 has surpassed CR2 in popularity. 

What do you think? Is anyone else willing to keep a list on their weekly game and let me know which hands won, throughout the day? You can include how many ended as wall games.  You don’t have to track Jokers and Flowers, just write down the appropriate letter for the section and which hand in that group (e.g. Y3 is the third 2019 hand, E5 is the fifth 2468/EVEN hand).

Send me your results (photo or typed out) at bubbefischer@gmail.com and we’ll see what kind of patterns we find.

Thanks for participating—I think I’ll have a free giveaway of both of my novels (Small World and Take Two), chosen randomly among the respondents!

I look forward to hearing from you soon!

Bubbe Fischer

PS Many thanks to Nina H for the above photo. She recently played a game where ALL FOUR PLAYERS were going for CR 5!!!

Bubbe in Boston, Part TWO!

Greetings and salutations from Bubbe, talking more about score stats from my Boston weekend.

It wasn’t a typical tournament for me. Not at all. I usually get about 250-325 points in twelve rounds—falling somewhere in the 3^rd quadrile--because I have no poker face. I make it obvious when someone discards a tile I need, which limits me as a tournament player. In my book, Searching for Bubbe Fischer, I say that on average, over time, among well-matched players, you should expect to win 20% of the games and end another 20% as wall games. So, based on that, what kind of a tournament score would be “average”?

In 48 games, generously rounding the 20%, each player would win 10 times with 8 wall games. Some are 25 points, some are more; we’ll call it 30 points per hand. In addition, the tournament I attended awards 10 points for self-picking and 20 points for a jokerless “non-Singles and Pairs” hand, plus 10-25 point penalties for throwing winning tiles. 

Going by those standards, the “average” player (before penalties and bonuses) should get:

10 wins @ 30 points/hand           300 points

8 wall games @ 10 points/hand   80 points 

Every winning game ends with a self-pick (bonus of 10 points), or calling for a discarded tile (10-25 point penalty to the person who threw it). The winner had a little more than a 25% chance of picking her mahj tile and a little less than 75% chance of calling for it—she only gets 25% of the picks, but her chance increases because she might also pick a tile to exchange for a joker. Let’s say she has a 31% chance of self-picking, on average.

Meanwhile, each of her opponents has slightly less than 25% chance of incurring the “throwing” penalty, which can range from 10-25 points. The majority of penalties are going to be 10 points for throwing to zero or one exposure, but there will be the occasional 20 or even 25 point penalties for throwing to multiple exposures. On average, let’s call it a 11 point penalty, and each player would be expected to throw to 23% of the 30 wins made by other players, incurring 7 penalties (rounding up).

                31% chance of self-picking           31 points for 10 wins

                23% chance of 11 point penalty -77 points for throwing to 7 opponents’ wins

In addition, in my stats the 20 point jokerless bonus was very rare (only 5% of all wins that I saw). Let’s be generous and say 10% of 40 games, or four hands, are jokerless winners, so we can expect an average of one jokerless win per tournament per player, resulting in a 20 point bonus.

I would estimate that the average score for 12 rounds should be something on the order of 

                                300 + 80 + 31 – 77 + 20 =  354, or just under 30 points per round

This estimate was a bit low. Per Sheryl Perry, the scorekeeping maven of DMJ tournaments, the mean (average) score was 380, the median (center ranked) score was 370 and the mode (most common) score was 420! Note: a previous tournament, with almost twice as many competitors, had a mean score of 382.

In my previous article, I vaguely alluded to my performance: I was dealt the average amount of Jokers (2/3 per hand, or 32 overall) and a few extra Flowers (3/4 per hand, or 36 overall). When I got bonanza deals of three Jokers or four Flowers, I ended up winning the hand. I also said that 31%, or a sizeable chunk, of my winning hands were “CR5” (the fifth hand in the Consecutive Runs section):

                                                 FFF 1111 2222 DDD

31% is 4/13; in other words, CR5s were four of the thirteen hands that I won, out of 48. That’s a fairly good performance, a little better than one game per round. I also hung on for 10 wall games. In total, I “didn’t lose” 23/48, almost half the games (48%), significantly higher than the expected 20% win and 20% wall games. I only threw the mahj tile in four games, so suffered 40 points’ worth of penalties. I had 4 self-picked hands (above the “average” 3) and one jokerless “regular” hand.

Most importantly of all, I didn’t just stick to 25 point hands. I took advantage of Jokers to make Quints:

                                                 

                                                Q1: 11 222 3333 44444

                                                Q2: FFFF 11111 NNNNN

and made the best of having no Jokers at all with the sixth Singles and Pairs hand, two suits of consecutive numbers with matching Jokers:

                                                SP6: FF 11 22 DD 11 22 DD

In addition to the Singles and Pairs hand, I made another closed hand, in Like Numbers:

                                                L2: 11 DD 111 DDD 1111

Compared with my "expected average," I won three more games, completed two extra Wall games, played several high-scoring hands, finished one “regular” hand without jokers, and had three fewer penalties for throwing mahj tiles. These were all things of which I was quite proud.

Strategy-wise, tournament play can be different than your regular game. I generally love to play difficult hands with my group (they joke that I’m always going for Quints). I only play once a week, so I challenge myself to try interesting hands. After all, life’s too short to play boring mah jongg.

However, a tournament can be a different story. Let me explain:

I played a steady, regular game during the Friday evening session and landed in a five-way tie for 12^th place (out of 84). In the Saturday morning session, I got the jokers for Quints and made the Singles and Pairs hand, and by noon I had moved all the way up to a tie for 2^nd place!!

It made me think about actually finishing “in the money,” so at that point, halfway through the tournament, I went for simpler hands as the most direct way to score points. After all, you only get points for mahj-ing first: if I wanted to continue to do well in the tournament, I would have to go for the most attainable hand, rather than the most interesting. I still couldn't force tiles to come to me, though; other players got hot while my tiles got cold, and I moved down to 6^th place by Saturday evening. Sunday started slowly, too, but I eked out two quick wins in the last round on Sunday. It was VERY close: my 520 points meant that I ended in a tie for 8^th place.

What about you? What are your tips for winning in a tournament? You can always contact me at bubbefischer@gmail.com ; I love hearing from you!

Talk to you soon!

Bubbe Fischer

Boston Bubbe--Part ONE

Greetings and salutations from Bubbe, just back from a Destination Mah Jongg weekend tournament in Waltham (just outside Boston), Massachusetts!

Who doesn’t love summers in New England? Apparently no one—that is to say, EVERYONE loves summer in New England, as evidenced by the tremendous traffic we encountered on the Mass Pike, Route 84, and the Merritt Parkway, but it was worth every minute to spend time with a lovely group of 83 women and one man, all fixated on mah jongg.

By the way, Destination Mah Jongg is one of many excellent organizations that run three-day tournaments. They are (at minimum) 12 rounds, starting with three the first evening, three (or more) round sessions the next day both in the morning and afternoon, and a final morning session on the third day. Sometimes participants also join in three-round evening “mini-tournaments.” Bubbe has been fortunate to participate in similar events with Mah Jongg Madness, Mah Jongg at the Kalahari Resort, and Four Seasons MahJongg, as well.

If you want to play mah jongg to your heart’s content, tournaments provide the perfect opportunity. Over the years, your Bubbe has used the 48+ games as a way of testing hypotheses about her own play. In fact, at every table I sat at, I whipped out my little hotel note pad and asked if the other players minded if I kept notes.

This weekend, I was a “West,” so I had a lot more liberty to take notes. It’s very hard to do so as an East, because East’s responsibility includes keeping track of everyone’s score sheet. As a West, I moved from table to table for all twelve rounds. Everyone was very friendly and some were downright hilarious; I made many new friends from New York, New Jersey, New Hampshire, Massachusetts, Vermont—and some who even travelled from as far as California, Florida, and Washington State!

Anyway…Bubbe took copious notes and will get into some heavy details in later posts, but here are some very brief examples of what I tested.

JOKERS, FLOWERS, AND THE DEAL: 

Everybody whines that they “never get any jokers,” but is this really true? Sometimes you will get none, but other times you will get one, two, or even three—based on probability and the total number of tiles available, I estimate that a player will average 2/3 of a Joker and Flower per deal.

How did this bear out in reality? In each of the 48 games, I wrote down how many Jokers and Flowers I was dealt. In total, I received 32 Jokers and 36 Flowers. Quick, get out your calculators: that’s 32/48 or 2/3 of a Joker and 36/48 or 3/4 of a Flower. Pretty spot on, I’d say!! The most Jokers I was ever dealt was THREE; the most Flowers I received was FOUR. And guess what? In each of those situations, I won the hand!!

CHANGING HANDS:

It always drives me crazy when a bettor, in making her selection, wants to know what each player is “going for.” Because I pick and throw so many tiles, and other tiles go dead because of other players’ discards, my plan after the Charleston often has little bearing on what I end up playing. I estimate that I change my hand after the Charleston about half the time (50%).

How did this bear out in reality? After the Charleston I would look at my hand and think about which hand made sense for THOSE tiles.  I also recorded where my hand actually ended up. Strictly looking at the hands I won, 46% were exactly the hand I expected after the Charleston; 54% of my winning hands were NOT what I expected when I started.

JOKERS, JOKERS, JOKERS:

This particular tournament awards a 20-point bonus for every winning hand that is “jokerless”: I don’t mean a Singles and Pairs hand, I mean a hand that includes Pungs and Kongs yet doesn’t use any Jokers (obviously a Quint needs Jokers).  That bonus is hard to get! I estimate that most winning (non-Singles and Pairs) hands include at least one joker. 

How did this bear out in reality? Of MY winning (non-Singles and Pairs) hands, 92% used Jokers!! I only got the 20-point bonus once! Of all the “regular” (non-Wall, non-Singles and Pairs) games at my tables, only two were jokerless. In other words, fully 95% of all winning non-Singles and Pairs hands included Jokers!!

WHAT WAS THE MOST SUCCESSFUL HAND?

I know everyone has inherent biases when they look at their tiles. Some of us, like Bubbe, are suckers for a Soap—if we see a White Dragon, we’re off to the Year section of the card! Others really love to play Winds, or 369, or Like Numbers. I often tell about a former student who would only play “Dots” because she didn’t understand the Bams or Craks.

In my book, Searching for Bubbe Fischer, I encourage you to know the WHOLE card, not just any one section. Don’t shy away from Quints or Singles and Pairs. Don’t let tri-color (three suit) hands intimidate you. 

So, did I stay in one section of the card? Was any one type of hand more attainable than another? I will look at this issue more closely in another article.

For now, I will talk about one specific hand that has received a lot of “buzz” this year: “CR5,” otherwise known as FFF 1111 2222 DDD with any two consecutive numbers and matching Dragons. It doesn’t require any pairs (as opposed to previous years when either the Dragons or the Flowers were a pair); it’s easy to collect. Some say that CR5 is the most popular, winnable hand.

How did this bear out in reality? Of MY winning hands, 31 percent were CR5! Among my opponents, 20 percent of the wins used CR5, so a total of 24 percent of all wins I saw were CR5. That is ONE hand of over 60 on the card, and yet it was the winning hand in almost a quarter of all games!!!

I think this kind of information is fascinating. If you do, too, please drop me a line at bubbefischer@gmail.com and ask me what else to look into. I love hearing from you!

Talk to you soon.

Bubbe Fischer