We all know there’s a Perfect Manhattan—that is, the kind that’s splits the vermouth quotient between sweet and dry varieties. But is there a perfect Manhattan?

The PUNCH staff decided to find out, culling recipes for the classic cocktail from 17 leading bartenders across America and sampling them in a blind tasting. Joining PUNCH in the tasting were bartenders Meaghan Dorman (Raines Law Room, Dear Irving, The Bennett), Joaquín Simó (Pouring Ribbons), Sother Teague (Amor y Amargo) and this reporter.

If ever a cocktail was worth the bother of such an evaluation, it’s the Manhattan. The whiskey drink’s place in the cocktail pantheon has never been questioned since it emerged from its namesake borough in the 1870s to become a national and international favorite. The Manhattan was the first of the great modern cocktails of the late-19th-century golden age of mixology to make use of vermouth as an ingredient. It predated even the mighty Martini.

Some early recipes called for equal-parts whiskey and sweet vermouth, while others asked for dry vermouth. The variety of bitters used, too, ranged wildly. But by the 1930s, the cocktail had settled down to the now familiar two-parts whiskey, one-part sweet vermouth, dash of bitters formula we recognize today as a Manhattan.

Unlike other classic cocktails, the Manhattan was never completely forgotten by the drinking public during the dark ages at the end of the last century. But it did frequently suffer from the indignities of bottom-shelf liquor, spoiled vermouth, absent bitters and maraschino cherries that had never seen a tree branch. Perhaps because of this, modern bartenders have made it a point of pride to serve a superior version of the drink.

The first decision when crafting a Manhattan is, of course, whether to use rye or bourbon. The bias of young bartenders toward rye was evident; 12 or the 17 submitted recipes were made with the spicier spirit. The majority of panelists, too, admitted to preferring a rye Manhattan. “This is the problem with bourbon Manhattans: not enough spice notes,” declared Teague.

Still, Simó, waxing philosophical, allowed that, “It’s less about the rye. All of the ingredients in the cocktail are great on their own. But a great Manhattan is two plus two equals five. For me, a Manhattan is an idea. And there are so many roads to go there.”

The roads taken by the bartenders who were summoned, however, were relatively narrow. There was one recipe that used dry vermouth instead of sweet. Another threw in a couple dashes of Herbsaint. A third spec asked for a Tennessee whiskey. But, by and large, we were dealing with drinks made of whiskey (rye or bourbon; sometimes a mix of two bourbons, or two ryes, or a bourbon and a rye), sweet vermouth and Angostura bitters, though a few entries did opt for other varieties, such as Peychaud’s or orange bitters.

“I use free license to use whatever bitters I want based on the whiskey and vermouth,” said Teague. Asked about bartender mistakes they had encountered when ordering Manhattans, the panelists all cited shaken Manhattans as the most common sin, but a few also pointed to the omission of bitters (“Bitters is that central spoke” to the drink, Simó said) and the tendency of some bars mix them with too little vermouth, thinking they’re doing the customer a favor with a heavier pour of whiskey.

Like many classic cocktails, the Manhattan looks simple on paper: three ingredients, all easy to acquire. But the flip side on that perceived ease is that the fewer the ingredients there are, the greater the need is for skill and discipline in getting them in alliance. The Daiquiri is, perhaps, the more notable case of a simple, classic cocktail that can easily fall apart if the proportions and ingredients aren’t just so. And I have chronicled my struggle to find a balanced Sidecar—another three-ingredient classic—in these pages. The Manhattan is arguably a more forgiving model; you don’t need a flashy whiskey or vermouth to make a good Manhattan. But the whiskey and vermouth must get along, and the bitters bottle must be wielded judiciously.

After the first round of tastings, only seven of the 17 Manhattans made the cut. The other ten were thought imbalanced in one way or another. Some were considered too thin, lacking body. Others were deemed flat and flabby, the victim of too much or too heavy a vermouth, or a whiskey with insufficient authority. Overall, the rye Manhattans were preferred over the bourbon entries, though two bourbon Manhattans did make the finals.

In the final estimation, the panelists—who registered only their first, second and third favorites—veered toward the Manhattans that clung most closely to the drink’s classic profile.

The top vote-getter, by New York bartender Jeremy Oertel (Death & Co.), couldn’t have been more classic: two-and-a-half ounces of Rittenhouse rye, one ounce of Carpano Antica Formula sweet vermouth and two dashes of Angostura bitters. The panel found the cocktail rich and full-bodied, with notes of cocoa nibs on the finish.

Close behind in the vote tally was a rendition by one of the judges himself, Joaquín Simó. Simó split the spirit base with one ounce each of Rittenhouse rye and Russell’s Reserve 10-year-old bourbon. The sweet vermouth, too, was divided: one half-ounce of Cinzano and one half-ounce of Martelletti Classico. Even the bitters were a double act, with one dash of Angostura and one dash of Dale DeGroff’s Pimento Bitters.

A bourbon Manhattan by Caitlin Laman (formerly of Trick Dog) came in third. Her two-parts whiskey to one-part vermouth, plus Angostura, formula was classic, but her liquor choices unusual. The bourbon was the oft-overlooked Johnny Drum 101, and the vermouth was the California-made Tempus Fugit Alessio Vermouth di Torino.

What can be learned from the fate of the other 14 drinks? Older whiskeys fared better with the panel then younger ones did. And rye was vastly more popular than bourbon. (Among the general public, the opposite is true.) Perhaps the biggest lesson was that a little cleverness goes only a short way. It’s fun to play around with classic cocktail specs. But the Manhattan model has probably stood the test of time for so long because it works.

It also became clear that that model, or what we expect it to be, is strikingly consistent amongst drinkers: Given the many choices, there was remarkable consensus among the panelists. Oertel’s and Simó’s Manhattans were chosen by every single judge. If the Manhattan is an idea, as Simó said, we all seemed to share the same one.