Soup dumplings are one of Shanghai’s classic creations. Unlike hairy crabs, hongshao rou (red-braised pork) or sheng jian bao (pan-fried pork dumplings), the standards for what makes a good soup dumpling are easily measured with a few simple tools.
Zhi duo, pi bao, rou duo, rou xian (a lot of soup, a lot of meat, thin skin, savoury meat) – these are the standards any Shanghainese will tell you make a proper soup dumpling (though those in Suzhou or Wuxi will certainly disagree). A sensitive scale and a set of callipers are about all you need to measure the first three, and that’s exactly what I’ve been doing since December 2013. I’ve since visited more than 50 places serving the city’s iconic dumpling that either purport to be experts themselves, or are represented that way by others. I have dissembled hundreds of dumplings right from the steamer with a pair of hair-cutting scissors, a scale that measures to the hundredth of a gram and Japanese digital callipers sensitive enough to measure the thickness of a piece of paper. I won’t eat another for a long time.
The goal has been to eliminate all of the subjective factors that go into a decision about what is the ‘best’ soup dumpling. To some, a dirty hole-in-the-wall connotes authenticity. To others, it just means dirty. Some like strong pork flavour and others put the strong ones right back in the basket. Price is also held up as a common measure of goodness – ‘yes, they are very good, but they’re 10 times the price!’ For all the soup dumplings I’ve eaten, I’ve yet to find one that tastes like money. No, the goal has been an objective reckoning of the engineering challenge that is the xiaolongbao, Shanghai’s home-grown soup dumpling. It is a delicate balance. If the wrapper is too thin, the filling will break the dumpling. If the filling is limited, there won’t be enough soup.
Even if this balance is achieved, the steamer is another hurdle. Thin wrappers break easily after steaming, a major foul at the dumpling party. A delicate wrapper won’t hold indefinitely in a steamer, waiting for a customer. It will absorb water and become mushy, compromised. So, the perfect dumpling must not just meet the engineering standards but also be steamed fresh, ideally to order, to arrive at the table just so: looking exhausted, collapsed, wrapper folded over to one side, hot soup pressing at the insides, threatening to throw all the work into chaos. And that’s to say nothing of the thickness of the top, where the folds meet, how many folds there are, the level of fattiness and clarity in the soup, if there is sugar or spring onion or ginger in the filling...
The result is the Shanghai Soup Dumpling Index, a scientific survey, a printed guide and a map to 20 of the city’s best places for xiaolongbao. Notes about the methodology, ‘what happened to flavour?’ and how to get a copy are below. In the course of this incredibly serious study, the stories of several places stood out. And while I have tried as hard as I can to suppress the human element, the personality, and anything else that might compromise my rigid scrutiny, my absolute objectivity and my scientific posturing, the stories of the following five restaurants are just too human. Fortunately, they also happen to serve some of the best soup dumplings in a city famous for them.
The Shanghai Soup Dumpling Index is a wholly scientific solution to one of the most pressing problems in the world. If someone tells you a xiaolongbao at their favourite place is ‘good’, what exactly does that mean? What are their points of reference? Where does ‘good’ fall on their rating scale? What’s ‘good’ about it?
The Shanghai Soup Dumpling Index, a year in the making, addresses this vagueness with science. Starting in December 2013, I have visited 53 xiaolongbao shops and a few restaurants known for their dumplings. In the process, I have weighed a total of 7,244.76 grams of dumplings, or 1,190.13 grams of soup, 2,456.99 grams of pork, and 2,313.34 grams of wrapper. I have devised a simple formula that calculates the engineering of a soup dumpling by dividing the thickness of the skin by the total weight of the filling, and then categorised these restaurants in three classes, based on their results. To my everlasting regret, ‘flavour’ has had to be judged subjectively, given a lack of access to analysis by potentiometric solid-state electrodes or near-infrared spectroscopy.
The result is a 49cm x 59.4cm publication that lists the full results of the survey, and indexes all 53 restaurants accordingly. The front gives the most essential information in infographic form; the back gives detailed measurements from all restaurants. In addition, there is a discussion about regional variations, a map with precise coordinates for all Class A and Class B locations, a detailed methodology, and a greater volume of data.