www.beyondtheboxscore.com/2019/11/20/2...

I was thinking recently about how I spent two months during 2003 driving coast-to-coast and back again staying at multiple friends' houses on the way. I had no iPhone, no SatNav, no cell phone of any kind, and I'm pretty sure I didn't own a laptop.

I have no idea how I did that. Seems impossible.

@davidjroth.bsky.social Never felt more seen than when you referred to “people who are itching to name Morgan Burkhart” on Twitch. Femtoseconds between the time I saw the grid and thought of him

FWIW, Trea Turner does significantly better by fWAR vs. rWAR:
Turner - 46.7 fWAR w/ 4 seasons >=5.0 fWAR
Bogaerts - 43.7, 2 seasons
Seager - 40.8, 3 seasons
Correa - 40.0, 1 season

Boils down to which defensive metric you trust more. It's a 4 player race if we split the difference.

This is abjectly stupid reasoning. If shutting down the government is beneficial to Trump, then the GOP would simply not pass a CR or Trump would veto it! Do you think your supporters are idiots or are you simply tremendously dumb? Do us all a favor and resign!

First change is that players are ineligible for the next ballot if they receive fewer than 5 votes. I like that because it encourages fresh faces on the ballot.

But the second rule says that any player who receives fewer than 5 votes twice is permanently ineligible for future ballots. That stinks.

Punishing other players because Jane Forbes Clark really wants to ensure Barry Bonds is never elected to the Hall. I’d hate to see him and Clemens take up 25% of the ballot in perpetuity, but the first rule change helps with that.

I predict we’ll see fewer candidates elected going forward. Voters will be less willing to set personal favorites aside to push someone else over the line lest the former lose eligibility permanently. Like voting for a player to keep them from dropping below 5% but with far fewer votes to spare.

I love the first rule change as it forces the committee to consider different players instead of recycling the same names every year. But I hate the permaban. It’s transparently intended to keep Bonds permanently out of the Hall with no consideration for other consequences.

Tried statrifying in my Beltrán model based on his trends with ARod/Manny voters vs. antiPED voters while also controlling for the stinginess of the voter. I suspect Bonds/Clemens voting history would have been a better proxy but it can't easily be applied to voters who didn't have ballots then.

Not anymore. Think you passed mine! Used to be a lot easier to project flip rate trends when PED-tolerance and space on ballot were the main differences between public and post/private voters. But those are now less relevant and it's clear there are other biases that aren't captured in my model.

Congrats on your performance! Pretty sure you did better than my model. I assumed CC would drop a lot, and I tried several alternative approaches to stratify public voters to reflect post/private voter behavior, but I couldn't get <90% median no matter how hard I tried.

Interesting. Is it a flip rate model or one that tries to adjust the raw post-reveal percentages based on historical Tracker comps like @baseballot.bsky.social's old approach?

As always, massive thanks go to @notmrtibbs.com, @tonycal.bsky.social, and @shutthedore.bsky.social for all of the hard work that goes into tracking ballots. And thanks to all of the BBWAA members who are willing to share their ballots and the @baseballhall.org for letting us have our fun!

Your projections align better with my gut feelings than my own model, especally for CC, so will be interesting to see how close each of us gets. Definitely interested in hearing more about your methodology sometime.

Why is Beltrán only elected in 33% of simulations when his median is so close to 75%? Basically he still needs to flip a lot of remaining voters - roughly a third more than he's gotten with the public voters. So he needs a lot to go right but there are many ways that something could go wrong.