Long post (you’ll need a drink after this):

My previous post about the illogical math for the results of Asia’s 50 Best Bars 2024 (you can read it here) prompted conversations and, needless to say, a bit of consternation.

While most people in the industry (sponsors, bar owners, distributors) were glad I raised the issue regarding the math and number of voters, some feathers were obviously ruffled. (I may have lost two friends already. Oh well.)

The people from A50BB wrote me an official response to my post, which raised more questions than before. So, I sent them (through the PR company) follow-up questions and some voting scenarios to consider. They replied last night. There are some areas not cleared up, and many conclusions to be drawn. I’ll post all their responses in full so nothing is taken out of context.

Their First Response to My Post:

“We appreciate how you have taken time to review and understand the 50 Best voting system. Please see below responses from Emma Sleight, Head of Content at 50 Best. We hope these can provide greater clarity and accuracy on how the 50 Best voting system works. 

How does voting work? 
The voting academy for Asia’s 50 Best Bars consists of 265 independent, anonymous voters that are spread across the continent. We do not disclose how many voters there are in each region as each region’s votes are allocated based on the size and relative development of that region’s bar scene. Hence, the ranking cannot be mathematically (or accurately) assumed.

Voters must vote for 7 bars in order of preference, and can vote for up to 5 within their territory. They can vote for as many bars as they like (up to the maximum of 7) that are located outside of their territory.

What happens in the event of a tie?
Voters record their votes based on their personal preference of what they consider their best experiences, from 1-7 as part of their 7 allocated votes. In the event that a bar has received the same number of votes, we assess in what order of preference they were voted for to discern their final ranking. All voting for Asia’s 50 Best Bars undergoes authentication and adjudication by independent body Deloitte.

Is there a higher committee, whose votes carry more weight? 
There is no higher panel committee at 50 Best. The list is the result of a simple computation of votes from across the 265-strong academy of voters in Asia. Members of the 50 Best team do not vote, nor do they impact or affect the vote or the resultant list. The voting academy is made up of bartenders, bar owners, drinks writers and cocktail aficionados. 50 Best also works with professional services consultancy Deloitte as its official independent adjudication partner to help protect the integrity and authenticity of the voting process and the resulting list of Asia’s 50 Best Bars.”

<snip>

So, the bars are ranked 1-7. Can 265 voters with a maximum of 7 votes do justice to hundreds of bars across Asia? Are there enough votes to go around? Yes, we can conclude that Asia has at least 102 bars. Remember MO Bar and Tesouro? No?

Spare a thought for MO Bar (Singapore), placed #8 in 2022. Not in the top 100 in 2023. Not in the top 100 in 2024. They didn’t even warrant one vote after being in the top 10. In the same year, Tesouro (Goa, India) placed #4! In 2023, guess what – not in the top 100. 2024? Nada. Both disappeared like wine coolers. And these are your Top Ten bars.

Is ranking and order important? How do you tie-break? How is the distribution of voters decided? Inquiring minds want to know!

I emailed these questions:

Regarding ranking:

“Voters record their votes based on their personal preference of what they consider their best experiences, from 1-7 as part of their 7 allocated votes.”

Q: Is the ranking and order important? Consider this scenario:

Bar A beats B and C because it received more votes.
B and C are tied (same number of votes).
B fairly beats C.
But both B and C should clearly beat A.

Q: So then, what happens?

Regarding tie breaks:

“In the event that a bar has received the same number of votes, we assess in what order of preference they were voted for to discern their final ranking.”

Assuming a “1” is a ‘most favourite bar’, and “7” is for ‘slightly less favourite bar’. (Yes, I mean, all bars are favourites and beloved.) Three scenarios:

• If Bar A receives one “1” placing vote, and Bar B receives seven “7” point votes, how would that tie break work?  
• More simply, if Bar C, D, E, F and G each receive only a ‘4’ placing, all things being equal, how would that tie-break work? Someone would have to decide, no?

A real-world example would be:

If Bar Leone (Hong Kong), Zest (Seoul), Jigger & Pony (Singapore), Coa (Hong Kong), Bar Benfiddich (Tokyo) ALL RECEIVED the same number of ‘1’ votes, same number of ‘2’ votes, ‘3’ votes, ‘4’ votes and ‘5’ votes, how do you break that tie?

• If Bar H receives a 3 vote, a 5 vote and a 6 vote (three votes), and say Bar I receives a 2 vote, a 5 vote and a 7 vote. Also three votes. How would this tie-break work?

Regarding the Academy:

“The Academy of voters is spread across Asia with the distribution of voters designed to reflect the relative development and sophistication of the drinks sector, and the concentration of quality bars, in each country.”

Q: So there ABSOLUTELY IS a higher panel which decides how good the bars are in each country to decide how many local judges that country gets, right?

<><><>

They responded last night. Their response doesn’t address my points one by one but defends their position as a bastion of integrity and impartiality.

“The 50 Best list is the result of a simple computation of votes from across the 265-strong academy of voters in Asia. Voters cast 7 votes in order of preference or their ‘best’ experience. Their preference order will have an impact on how that bar ranks in the voting. If bars are tied, the splitting of those bars is a two-stage process. Stage one is to add up points for positioning within the 7 votes. The second stage is how many no.1 positions vs other tied bars, then the process is repeated for no.2 position etc. to determine overall position. There is simply and categorically no way of knowing how voters vote and in what sequence they cast their preferences.

We reiterate there is, categorically, no higher panel or committee, whose “votes carry a lot more weight”, or “decide on placings and positions”. As mentioned above, the 50 Best list is the result of a simple computation of votes from across the 265-strong academy of voters in Asia. Academy Chairs do not determine the amount of voters in each region. The role of the Academy Chair is simply to source credible, confidential voters. Academy Chairs votes are worth the same weight as any other voter and they are unable to control or influence the composition of the list and have no access to or knowledge of any other voters’ selections. 50 Best does not impact or control the resultant list. 50 Best will only assign a volume of voters based on the sophistication and development of the cocktail industry in that relative region.”

<><><>

Where do we start, dear readers? They did not address the tie-break situation. This scenario that I mentioned is highly probable and mathematically possible:

If Bar Leone (Hong Kong), Zest (Seoul), Jigger & Pony (Singapore), Coa (Hong Kong), and Bar Benfiddich (Tokyo) ALL RECEIVED the same number of ‘1’ votes, the same number of ‘2’ votes, ‘3’ votes, ‘4’ votes and ‘5’ votes, how do you break that tie?

They chose not to answer.

If the top 5 bars all had the same number of votes and were ranked in the same way, then how oh how oh how do you decide who is number one if “50 Best does not impact or control the resultant list”? Okay, is it the guy in Deloitte who can do the most tequila shots? Because by simple logic, if there’s no human interference, then we’d have 5 bars at number one, no #2-5, and the list continues with #6. The more they try to be transparent, the more confusing it gets.

I asked about the composition of voters in Asia and how it is determined. Remember, they said this the first time: “The Academy of voters is spread across Asia with the distribution of voters designed to reflect the relative development and sophistication of the drinks sector, and the concentration of quality bars, in each country.”

I asked then: So there ABSOLUTELY IS a higher panel which decides how good the bars are in each country to decide how many local judges that country gets, right?

Readers, a committee has to decide how many voters a country gets, right? The Academy Chair picks the voters yes (“The role of the Academy Chair is simply to source credible, confidential voters.”.) Confidential is laughable, but that’s another issue. The Chair is told they have 25 voters to pick, so they only pick 25 voters (“Academy Chairs do not determine the amount of voters in each region.”), BUT A COMMITTEE has to decide how many voters each country’s Chair gets to select, and that, readers, is how votes are skewed.

[Simple math: If Hong Kong has 200 voters, and the remaining 65 voters are spread across Asia, guess who wins come voting time? This scenario would never happen, of course, nudge, nudge.]

An industry insider puts it more eloquently: “There is no panel above the judge’s decisions, but there is a panel over the judges themselves that can arbitrarily decide the number of judges from each country, potentially giving one country an advantage over others. And then these results can be seen as the measure of ‘development and sophistication’ in a country, giving them even more judges next year. This would have been particularly influential during Covid when no one could travel. In essence, the committee decided which country’s bars would get the most votes.”

More conclusions:

1  The process is fair, equitable and beyond reproach, in which case the data would not be controversial or

2  There is something questionable or at least unexpected in the process.

Which is it?

In 1980, the Oscars awarded Best Picture to ‘Ordinary People’ over ‘Ragingbull’. ‘Ragingbull’, my peeps, is in the film canon of the greatest movie ever made.

“Hey guys, let’s get together this weekend and watch ‘Ordinary People’, okay?” SAID NO ONE EVER.

In 1990, the Oscars awarded Best Picture to ‘Dances With Wolves’ instead of ‘Goodfellas’. ‘Goodfellas’! It’s one of the most eminently rewatchable and astonishing movies by one of film’s greatest directors on a career peak. Did you even finish watching Costner dancing with lupine pups the first time around?

With any voting process, there will be anomalies, banalities and controversies. Whether Oscars or bars, if the voting process is straightforward and transparent, people will accept the results, personal preferences aside. (Democracy, yeah!).

We’ve been trying to make sense of the results of the A50BB for years (this year in particular was suss as hell), and the votes have often been clickbaity and inadequate. The math does not add up, the placings do not make sense, and the system is silly at best and skewed at worst. (We haven’t even broached the campaigning issue. Shush, it doesn’t happen.)

Ultimately, do what you want with the results. If you feel strongly about it, speak up. If not, hope you enjoyed the show. Either way, use the list as a bookmark and go visit the bars — they could all enjoy our support, now more than ever.

Your barfly friend.

8