THE BOOK--Playing The Percentages In Baseball

I don’t know what i’m more excited about ... getting linked to by Neyer or by Tango…

Why 50-55 games?  Why 3 seasons?

My equivalency in terms of reliability is 200 PA (50 games) tells you as much as 400 BIP (100 games for a neutral position).

So, if you’ve got 150 games for a hitter (one season), the uncertainty level of his stats would be equivalent to the uncertainty level of 300 games (two seasons) of UZR.

’My equivalency in terms of reliability is 200 PA (50 games) tells you as much as 400 BIP (100 games for a neutral position).”

Based on what?  And 400 BIP or UZR chances?  WHich you don’t use the same IIRC.

BIP = BIP.

And the reason I say 200 PA and 400 BIP is because at those levels the year-to-year r is .50 for Linear Weights and UZR.

But many players play more than 700 IP which is obviously a good deal more than 400 BIP.  I mean, starters get 400+ chances, much less on the field for 400 BIP.

WHat am I misunderstanding here?

Would that mean for players that are on the field for more than 1400 BIP, their UZR is “accurate”?

Also, what are the implications for a single season WAR?  If you are using numbers that aren’t “projectable” to determine a projected contract value - isn’t that an issue?

Ahhhhh… my poor explaining.  The “BIP” is “assigned BIP”, where I assign the SS, 2B 5 BIP per game, 4 for 3B, CF, and 3 for LF, RF, 1B.

So, that’s 27 BIP for 7 positions, or an average of almost 4 BIP per position.  So when I say 400 BIP, I mean 100 games for 3B, CF.

Sooooooooo.... 200 PA for an individual batter tells you as much as 2800 TEAM BIP for an individual fielder (excluding bunts).  (You need fewer team BIP for SS, 2B, and more for LF, RF, 1B.)

***

For forecasting WAR, I regress.  For evaluating the past season, I don’t regress.  Is this what you are talking about?

So you do mean chances (average chances).  Why do you assign them those #s?  So 100 games, but teams play significantly more than that, right?

For instance, in a season, a SS gets 500-520 ZR chances, and 500 PAs.

In a game, a SS gets 5 “assigned” BIP and 4.25 PA per game.  So, if a SS plays 120 games, he has 600 BIP and 510 PA.

But, I don’t understand your point.  I said that when a fielder gets 400 assigned BIP, that tells you as much as his 200 PA.  So, the fielding can NEVER catch up to the hitting.

If you can make a determination on a shortstop’s 510 PA as a hitter, in order to be able to have that same level of uncertainty, you need 1020 assigned BIP.  And that means you’ll need 204 games.

I don’t understand what you are trying to get at.

"If you can make a determination on a shortstop’s 510 PA as a hitter, in order to be able to have that same level of uncertainty, you need 1020 assigned BIP.  And that means you’ll need 204 games.”

Because I don’t understand that.  How is that level of uncertainty being determined?

Answered in post 5.