Alright, so today I’m gonna yap about something I was messing around with yesterday: the whole “garbage time” meaning thing. Yeah, you know, that part of the game where it’s basically decided, and everyone’s just kinda going through the motions. I wanted to see if I could actually, like, quantify it, you know? Turn it into something measurable.
So, first thing I did was, I started digging around. I mean, what IS garbage time, really? Everyone knows it when they see it, but how do you define it? I started looking at game scores, time left on the clock, and how much the score changed. My initial thought was: if the score difference is X and there’s only Y minutes left, then it’s garbage time. Simple, right?
Turns out, not so simple. Because X and Y are, like, totally dependent on the sport. A 20-point lead with 5 minutes left in basketball? Definitely not garbage time. A 3-0 lead with 5 minutes left in soccer? Pretty much over. So, I needed something more flexible.
I fiddled around with different formulas. Tried to find some ratio of score difference to remaining time that would work across different sports. I even looked at win probability calculators – those things that tell you the percentage chance a team has of winning at any given point. My thought was, if the win probability is above a certain threshold (like 99%) for one team, then it’s probably garbage time.
I pulled some data from a bunch of different games – football, basketball, even some baseball. Scraped it all off some sports stats websites (don’t ask me which ones, I don’t wanna get sued!). Then I started plugging it into my little Python script. The script basically calculated these win probabilities and checked if they crossed my threshold.
It was a mess, honestly. Some games, the script would flag garbage time way too early. Other games, it would miss it completely. I realized the problem was that win probability isn’t the only factor. You also need to consider things like momentum, team strategy, and even just random luck.
So, I started adding more variables to my script. Things like the rate of scoring in the game, the number of turnovers, the number of timeouts remaining. I even tried to factor in things like whether the leading team was playing at home or away.
The results got a little better, but still not great. The biggest issue was that I was trying to automate something that’s inherently subjective. Garbage time is, at the end of the day, a feeling. It’s the feeling that the game is over, even if the numbers don’t quite say so.
Eventually, I gave up on trying to create some perfect, all-encompassing algorithm. Instead, I decided to focus on identifying potential garbage time situations. My script now flags periods of the game where the win probability is very high, the score difference is significant, and the scoring rate has slowed down. Then, it’s up to the user (me!) to decide whether it’s really garbage time.
It’s not perfect, but it’s a start. And honestly, it was a fun little project. I learned a lot about data analysis, win probabilities, and the limitations of trying to quantify something that’s inherently human.
- Pulled data from various sports
- Tried different formulas based on score difference and time left
- Incorporated win probability calculators
- Added factors like momentum and team strategy
- Realized the subjective nature of garbage time
- Shifted focus to identifying potential garbage time situations
What’s next? Maybe I’ll try to train a machine learning model to identify garbage time based on video footage. But that’s a project for another day.
