{
"path": "/beer-off",
"site": "at://did:plc:gfrmhdmjvxn2sjedzboeudef/site.standard.publication/3md7ylshxzk2y",
"$type": "site.standard.document",
"title": "Shitty beer statistics",
"content": {
"$type": "site.standard.document#markdown",
"value": "My girlfriend Grace threw a \"Shitty beer off\" party, where participants were tasked with drinking a shitty beer and then rating it on a scale of 1 to 10. We provided 5 beers and double-blindly asked participants to rate them. \n\nBeers available were \n\n- PBR\n- Coors\n- Miller\n- Hamms\n- Bud\n\nWe asked participants to rate each beer on the following dimensions:\n\n- Drinkability\n- Bubble\n- America Fuck Yeah-ness\n- Mouthfeel\n- Overall\n\nHere's the data on the [actual beers](/assets/data/beer/actual.csv) and the participant's ratings on the [shitty beers](/assets/data/beer/shitty-beer.csv).\n\n```\nusing CSV, DataFrames, Plots\n\n\n```\n\n## Average ratings\n\n```\n Row │ beer_name overall drinkability bubble america mouthfeel \n │ String15? Float64 Float64 Float64 Float64 Float64 \n─────┼───────────────────────────────────────────────────────────────\n 1 │ PBR 6.83333 7.46667 6.33333 5.66667 6.28571\n 2 │ Budweiser 6.53846 7.15385 5.26923 7.76923 6.38462\n 3 │ Hamms 6.375 6.75 5.4375 6.75 6.375\n 4 │ Coors 4.43333 5.66667 3.93333 6.8 4.2\n 5 │ Miller 4.33333 5.33333 4.33333 6.0 3.6\n```\n\nTakeaways:\n\n- Budweiser is a drinkable beer, and also has the most America fuck-yeah-ness. *eagle noise*\n- PBR is the best beer, but it is also has the least America fuck-yeah-ness. It's the shitty beer of the landed gentry.\n- Coors and Miller are tied for the worst beers overall. Awful.\n\n## Variance in ratings\n\n```\n Row │ beer_name overall_var drinkability_var bubble_var america_var mouthfeel_var \n │ String15? Float64 Float64 Float64 Float64 Float64 \n─────┼──────────────────────────────────────────────────────────────────────────────────\n 1 │ PBR 2.84524 3.98095 3.38095 3.95238 3.91209\n 2 │ Budweiser 5.9359 5.47436 5.19231 2.69231 3.25641\n 3 │ Hamms 3.58333 3.26667 4.39583 5.4 4.11667\n 4 │ Coors 4.3881 9.52381 7.78095 2.74286 7.31429\n 5 │ Miller 2.66667 7.66667 3.95238 4.61538 3.97143\n```\n\nTakeaways:\n\n- Most people agree that PBR is the best beer.\n- Most people agree that Miller is the worst beer.\n- Bud is pretty controversial, with some people liking it and some people hating it, but overall it had a very high score.\n- Coors has a lot of variance in ratings, as well as a low average rating.\n\n## Correlation between attributes\n\n\n\nTakeaways:\n\n- Overall and drinkability are the most correlated, perhaps this is obvious.\n- Bubble also seemed to be an important attribute for the overall score, but to a lesser extent than drinkability.\n- America fuck-yeah-ness seems to be correlated with _worse_ beers -- this is unsurprising. America-fuck-yeah-ness is about being on a speedboat on a lake, holding a rifle, waving a flag, and shooting at things with a tepid Coors Light in your hand.\n\n## Guessing statistics\n\nParticipants were not good at guessing the correct beer. Total accuracy was 20.37%.\n\nThis did however vary by beer. Some beers were easier to guess than others.\n\n```\n Row │ beer_name accuracy \n │ String15 Float64 \n─────┼─────────────────────\n 1 │ PBR 0.636364\n 2 │ Miller 0.363636\n 3 │ Budweiser 0.0\n 4 │ Hamms 0.0\n 5 │ Coors 0.0\n```\n\nMost participants identified PBR quickly, which was surprising. Miller was the only other beer that was guessed correctly. Coors, Bud, and Hamms were all guessed incorrectly by everyone, though this may have been due to the fact that the participants slowly became drunker the longer they were at the party.\n\nWe can also break this down by participant. We have several observations where no name was available, so this is restricted to the participants who provided a name.\n\n```\n Row │ name accuracy \n │ String7 Float64 \n─────┼───────────────────\n 1 │ anna 0.5\n 2 │ emily 0.4\n 3 │ david 0.333333\n 4 │ conner 0.25\n 5 │ ashley 0.25\n 6 │ andrew 0.2\n 7 │ mark 0.2\n 8 │ summer 0.2\n 9 │ max 0.0\n 10 │ pranay 0.0\n```\n\nTakeaways:\n\n- Pranay and Max are idiots.\n- Anna and Emily have classy, refined palates.\n- David, Conner, and Ashley are not very good at guessing.\n- Mark and Summer are somewhere in the middle.\n\n## Conclusion\n\nThis was a fun party! Really great idea. I strongly recommend hosting a shitty-beer off.\n\nGo buy some PBR. \n\nOr, well, basically any of the American craft beers, which are excellent. \n\n-- Cameron"
},
"publishedAt": "2024-10-20T07:00:00.000Z",
"textContent": "My girlfriend Grace threw a \"Shitty beer off\" party, where participants were tasked with drinking a shitty beer and then rating it on a scale of 1 to 10. We provided 5 beers and double-blindly asked participants to rate them. \n\nBeers available were \nPBR\nCoors\nMiller\nHamms\nBud\n\nWe asked participants to rate each beer on the following dimensions:\nDrinkability\nBubble\nAmerica Fuck Yeah-ness\nMouthfeel\nOverall\n\nHere's the data on the actual beers and the participant's ratings on the shitty beers.\n\nAverage ratings\n\nTakeaways:\nBudweiser is a drinkable beer, and also has the most America fuck-yeah-ness. eagle noise\nPBR is the best beer, but it is also has the least America fuck-yeah-ness. It's the shitty beer of the landed gentry.\nCoors and Miller are tied for the worst beers overall. Awful.\n\nVariance in ratings\n\nTakeaways:\nMost people agree that PBR is the best beer.\nMost people agree that Miller is the worst beer.\nBud is pretty controversial, with some people liking it and some people hating it, but overall it had a very high score.\nCoors has a lot of variance in ratings, as well as a low average rating.\n\nCorrelation between attributes\n\n!Correlation between attributes\n\nTakeaways:\nOverall and drinkability are the most correlated, perhaps this is obvious.\nBubble also seemed to be an important attribute for the overall score, but to a lesser extent than drinkability.\nAmerica fuck-yeah-ness seems to be correlated with _worse_ beers -- this is unsurprising. America-fuck-yeah-ness is about being on a speedboat on a lake, holding a rifle, waving a flag, and shooting at things with a tepid Coors Light in your hand.\n\nGuessing statistics\n\nParticipants were not good at guessing the correct beer. Total accuracy was 20.37%.\n\nThis did however vary by beer. Some beers were easier to guess than others.\n\nMost participants identified PBR quickly, which was surprising. Miller was the only other beer that was guessed correctly. Coors, Bud, and Hamms were all guessed incorrectly by everyone, though this may have been due to the fact that the participants slowly became drunker the longer they were at the party.\n\nWe can also break this down by participant. We have several observations where no name was available, so this is restricted to the participants who provided a name.\n\nTakeaways:\nPranay and Max are idiots.\nAnna and Emily have classy, refined palates.\nDavid, Conner, and Ashley are not very good at guessing.\nMark and Summer are somewhere in the middle.\n\nConclusion\n\nThis was a fun party! Really great idea. I strongly recommend hosting a shitty-beer off.\n\nGo buy some PBR. \n\nOr, well, basically any of the American craft beers, which are excellent. \n\n-- Cameron"
}