Is this the most WAR lost in a season?

[TuckerDuckson]

Uh, Freddie Freeman had a heck of a retirement tour I guess.



It was probably his defense, he had a -26.3 ZR, but had a .991 FP%, plus the worst bat in the league. He also got paid $12,100,000 to lose almost six and a half WAR.

Brought his career total from 72.7 to 66.3

[CBeisbol]

Yikes

[GTAltman]

Originally Posted by TuckerDuckson

Brought his career total from 72.7 to 66.3

Working on delegitimizing a possible HOF Career....
Needs just one more season like that

[Dave Stieb II]

I love Freddie Freeman, so to see him outdo Chris Davis by a substantial margin in terms of rapid declines makes me sad.

[rudel.dietrich]

That is very on brand for the Rangers

[Leo_The_Lip]

The actual lowest WAR in the MLB ever calculated was -4.0 by Jim Levey in 1933.

https://www.theringer.com/mlb/2020/9...ll-war-history

[BBGiovanni]

Did the AI play him in CF?

[rudel.dietrich]

Originally Posted by Leo_The_Lip

The actual lowest WAR in the MLB ever calculated was -4.0 by Jim Levey in 1933.

https://www.theringer.com/mlb/2020/9...ll-war-history

That is one AI thing that I wish would be looked at is that the AI sometimes will roll a guy out there even if they are historically terrible.

In real life he would have been benched of DFA'd regardless of that contract.

[TuckerDuckson]

Originally Posted by BBGiovanni

Did the AI play him in CF?

Nope, all at 1st base. Here are his ratings at his retirement:



And his fielding stats:

[Leo_The_Lip]

Originally Posted by rudel.dietrich

That is one AI thing that I wish would be looked at is that the AI sometimes will roll a guy out there even if they are historically terrible.

In real life he would have been benched of DFA'd regardless of that contract.

Yes, it is strange. But looking at the ratings posted above, it does seem the AI based its decision on ratings and "regression to the mean".

Every day.

And yet the mean was never regressed to.

Sometimes the penny lands tails 1000 times in a row. Doesn't mean the odds on the next flip aren't 50/50 and that's how the AI program is written.

[CBeisbol]

Originally Posted by Leo_The_Lip

Yes, it is strange. But looking at the ratings posted above, it does seem the AI based its decision on ratings and "regression to the mean".

Every day.

And yet the mean was never regressed to.

Sometimes the penny lands tails 1000 times in a row. Doesn't mean the odds on the next flip aren't 50/50 and that's how the AI program is written.

Yup
Those ratings aren't historically bad by any stretch

[Syd Thrift]

Originally Posted by rudel.dietrich

That is one AI thing that I wish would be looked at is that the AI sometimes will roll a guy out there even if they are historically terrible.

In real life he would have been benched of DFA'd regardless of that contract.

Khris Davis just had a somewhat similarly awful season in 2018, so I’m not so sure of that. I guess that if a guy is hitting below .200 with no power, he does probably lose his job regardless (unless he’s a catcher) but it was kind of a flukey bad season, it looks like.

Incidentally I looked up the worst single season WARs of all time and...

- the guy with the worst ever actually also had the 3rd worst ever as well. I guess the early 30s St Louis Browns were hard pressed for talent...

- A lot of the players are guys putting up deadball era looking numbers in high offense eras. In other words, it’s almost explainable...

- A big chunk of them have come in the last 30 years. Neifi Perez’s 2002 is up there, as is Jose Guillen’s 1997 (which doesn’t even look that bad at a glance), Davis of course... but also guys like Luis Polonia and Ruben Sierra.