PECOTA Predicts Last Place Finish For Fish

[CrimsonCane]

Shit happens.

Or, it doesn't.

[Bobbob1313]

Or, it doesn't.

Exactly.

[Namaste]

PECOTA models are Dafts best friend.

I can't remember the last time the Marlins did not win more than the total set by offshore oddsmakers. Maybe 2004. Maybe.
--------------------
What I don't get is.......

They released a projection and then a few hours later released an updated projection?

What's that all about.

[Bobbob1313]

The first was released a couple of days ago, and there were some bugs and various other problems with them, since they put an emphasis on getting the info out soon this year with the caveat that it's "Beta" until they announce it official. It was my mistake on not realizing that they were working on updating.

The stuff I posted at the top is the old stuff, but the offensive numbers don't really change with the updated numbers. Has us scoring marginally fewer runs (like, 12 or something) while allowing 60 less runs, putting us around an even run differential. That seems about right.

[THE_REAL_MIBS]

I just don't like computer models trying to forecast what humans will do. I never have. We've already been over this but I think that the players stats should just be used for a summary of what they have done not what they will do. Human beings are subject to thoughts and emotions and other events that happen in their lives that computers cannot predict. That's why I don't like them.

I also don't put much stock in what people say when they predict how many games teams will win or lose or how a player will do. I'm not going to give you a prediction on how many games we'll win but I believe that this team is better than that 76 game revord the computer originally predicted just as I don't believe that the Nats are a .500 team like it said. I'm not physic but that's just what I believe. I'm just not going to try to make these projections of team records and individual player stats because they can change wildly.

[Bobbob1313]

So we shouldn't try to predict anything ever.

[CrimsonCane]

I think people would be able to respect the role statistics plays if they understood something about statistics. People assume that these things are being put up on some pedestal by the people devising them.

They are subject to variance, just like any projection. But, they are more accurate, on average, than human observers making similar predictions. That's their value. Nothing more, nothing less.

[Ramp]

I wonder what PECOTA would say if we had Josh Willingham in the lineup

[Beef]

I don't really put any stock into what any predictions like this say, but I don't think it isn't meant to be something that tells us what will happen. What will happen is unknown. This is approximately what they expect to happen based on the past. The past also includes human emotions and the like, so those are used in the predictions in some sense and then, again, there is a variance. Yeah, some people look at these and act as if we are being slighted or getting no respect and what not, but to that I say who the fuck cares. But I imagine it is those types of responses and a need to feel respected by the predictions that create the need for some to say these predictions have no validity.

[CrimsonCane]

Their primary use is either for (1) running a baseball team or (2) gambling. Since I do neither, I simply look at these projections and say, "OK. That's interesting."

[Beef]

Yep. They certainly have validity. To me, they just do not matter.
--------------------
Certainly more validity than some douche on tv saying Team A gon lose because they are young and inexperienced and Team B gon win because they are young and talented

[Bobbob1313]

I wonder what PECOTA would say if we had Josh Willingham in the lineup

506 PA 60 R 18 HR 61 RBI 5 SB .254/.359/.450

Basically what CHONE and Marcel have him for this season as well. Bill James, as always, is slightly more optimistic.

They have him to be slightly better than Coghlan (.010 OPS points), significantly worse than Cody (.848).

So basically, they think we'd probably be worse if we had him in our outfield. I guess he could be an upgrade at 1B, but considering what it would cost to get Hammer at this point, wouldn't it be in our best interests to let the kids play?

I was thinking about Hammer being available earlier today, but it doesn't make any sense.

Or are you doing the "We never should have traded Hammer" thing?
--------------------

Their primary use is either for (1) running a baseball team or (2) gambling. Since I do neither, I simply look at these projections and say, "OK. That's interesting."

I just like arguing about baseball, and these lead to sometimes interesting discussions.

[Ramp]

Not to totally change the topic of this, but interesting that Hammer is projected to sport a .809 OPS when he's been well over that the last four seasons.

[nny]

but I believe that this team is better than that 76 game revord the computer originally predicted just as I don't believe that the Nats are a .500 team like it said.

You do realize for the upteenth time that those predictions were from a borked system and it's now fixed, yes? Why do you keep harping on it?

[Beef]

bork and bindy