I've been working on forecasting NHL games at nhlforecasts.com.
It's been a fun challenge as the games are pretty clustered in terms of scoring, and the games themselves are random with minimal points scored. I'm also not the biggest fan of hockey, so it's been fun for me to see which teams are ranked high.
I've been leaning on AI for the first time which has been interesting; I see a ton of content with AI around web dev, but less around more data science. It's interesting how quickly AI will break a common sense rule, like data leakage. Really fun learning experience!
In terms of platform, I've been having a ton of fun with static sites. Cheaper to host and more secure, all I need is a domain name to get it accessible on the web.
For a more data science-y project, I'm having a blast building https://nhlforecasts.com. I live in a town with a hockey team, and figured the best way to become a fan would be to understand the sport from a data driven approach.
I did a ton of work on building an Elo model first, but was getting very compressed results in terms of postseason predictions. I swapped to a Bayesian approach which has really taken off. Not sure how it's going to handle the second round of games approaching, but that's a problem for the future!
I’ve been writing on and off for a few years now! I’ve mostly been using this to note any particularly troublesome bugs/commands as well as use writing to think through some analysis throughout my career.
I really made it for myself, but I did set up a website last year (pugilis.com) and there is an older version there that I've made available for free. That version does not have the "MatchUps" feature though.
I'm really not a "marketing" guy but this week I started reaching out to some of the YouTube channels via email that cover boxing to see if any would want to partner with me to promote and sell the version with the analytics.
So far no one has got back to me. I think it's fair to say that boxing fans, as a group, may not be very tech savvy, and maybe even a bit intimidated by the idea of using an "app". And they may think my email is a scam. I put a link to the app in the email and they may be reluctant to click on it. Can't say I blame them if that's the case.
I think my next step is to make a video and put it on YouTube so they can see how it works and confirm it really is an app.
This is an awesome idea! I’ve been thinking about doing something similar in the hot sauce area. Any lessons/takeaways? Would love to know how you got your first customers.
1/ Have a flavor profile that's targeted towards a demographic you can reach. It's a much easier sell if you have access to an initial distribution channel and that demographic understands immediately what it might taste like.
2/ First customers: I started off with a couple friends, and then guessed where else I might find more like them. Kept trying until I found more. Don't start with paying for ads unless you're doing MVP testing pre-product.
3/ Have a unique flavor that they can't get anywhere else, but is legible (by reading the description, they can kinda imagine it). This may be tough, and your best bet is probably testimonies from your friends.
4/ Don't worry about scaling up initially. Take the time to make sure the product is something people actually want. Be skeptical and look for indicators, cuz your friends will lie to your face. Look for what they do, not what they say: spreading by word of mouth, repeat orders, unboxing on social media without you asking. etc. Work out the product kinks while you're small, because small problems also become big problems at scale.
5/ Work out the order fulfillment process and keep iterating on it. I find it really eats up time when I get an order that I'm not expecting to handle. Be organized and keep track of things.
“Why, of all his creatures, did He choose this loud, dirty, unkempt, obnoxious, uncontrollable, megalomaniacal madman to be His personal bread baker? How was it that this disgrace as an employee, as a citizen, as a human being-this undocumented, untrained, uneducated and unwashed mental case who's been employed (for about ten minutes) by every kitchen in New York-could throw together a little flour and water and make magic happen? And I'm talking real magic here, people. I may have wanted Adam dead a thousand times over. I may have imagined, even planned his demise-torn apart by rabid dogs, his entrails snapped at by ravenous dachshunds, chained to a pillory post and flogged with chains and barbed wire before being drawn and quartered-but his bread and his pizza crust are simply divine. To see his bread coming out of the oven, to smell it, that deeply satisfying, spiritually comforting waft of yeasty goodness, to tear into it, breaking apart that floury, dusty crust and into the ethereally textured interior . . . to taste it is to experience real genius. His peasant-style boules are the perfect objects, an arrangement of atoms unimprovable by God or man, pleasing to all the senses at once. Cezanne would have wanted to paint them-but might not have considered himself up to the job. Adam Real-Last-Name-Unknown may be the enemy of polite society, a menace to any happy kitchen, a security risk and a potential serial killer, but the man can bake. He's an idiot-savant with whom God has serious, frequent and intimate conversations. I just can't imagine what He's telling him-or whether the message is getting garbled during transmission.”
It's been a fun challenge as the games are pretty clustered in terms of scoring, and the games themselves are random with minimal points scored. I'm also not the biggest fan of hockey, so it's been fun for me to see which teams are ranked high.
I've been leaning on AI for the first time which has been interesting; I see a ton of content with AI around web dev, but less around more data science. It's interesting how quickly AI will break a common sense rule, like data leakage. Really fun learning experience!
In terms of platform, I've been having a ton of fun with static sites. Cheaper to host and more secure, all I need is a domain name to get it accessible on the web.