Yesterday was the first day of the college lacrosse season (D1 Men’s that is).  It was also the re-launch of my LacrosseReference website that I wrote about previously.

It’s been an exciting couple of days for a few reasons.

LaxRefBot

The primary one is that got to see my first auto-tweets going out from the bot account that I created.  I am still trying to get a hang of the balance between stats and analysis.  Go to heavy on the stats site and people won’t really be engaging with the site.  Go to heavy on the analysis, and you lose a great way to pique people’s attention.

LaxRefBot is a sort of compromise.  I have scripts that create statistical game previews, calculate and visualize the win odds for each game, and do other things automatically.  I added a feature to these scripts to schedule a tweet with the relevant data viz for LaxRefBot to send out automatically.

For those in my audience who want to see the steady stream of advanced stats nuggets, they can follow LaxRefBot in addition to my main account.  I can also retweet anything that I think the larger audience would want to see.

This way, I am still in control of managing the balance of analysis and stats, but I get the benefit of automation.  The real question is how the follower populations between the main account and the bot account will end up overlapping.

Year on Year Progress

This is a sort of obvious one, but I am on track to have more people visit the site in the first 3 days than I had in the entire month of February last year.

Back then, I was just starting and no one knew what LacrosseReference was.  The numbers make sense

This year, we have a more built in audience.  It’s not as if this is a surprise, but it’s going to be cool to see how much growth there has been in the audience in 12 months.  I have absolutely zero goals about traffic growth, so it’s not as if this is something I’m optimizing for.  It’s just that warm fuzzy feeling.

Flying in Daylight

I have already noticed that my efficiency in all things LacrosseReference is way higher than last year.  When the first game finished on Friday, my script dutifully grabbed all the play by play data and stored it in the database.  I probably spent 200 hundred hours last year getting this capability right.  And this year, it’s looking like it’ll be zero.

A few of my script had hard-coded year values in them, which accounted for a few hiccups in running the actual stat-tabulating scripts.  But it took no more than 90 minutes to get these sorted out.  And since everything else pretty much flows from these two set of scripts, the rest of the content automation is running as expected.

I am excited by the fact that to spin up a new analysis or a new post, I’ll be able to quickly call up the relevant data in no time.  Less time fiddling with data means more time writing and doing actual analysis.  More writing generally means more ideas of things to analyze.  It’s a virtuous circle.

So rather than the fly-by-night operation of last year, I’ll be able to focus on the more valuable things.  And that should make LacrosseReference year 2 way more enjoyable than the last.