Re: What would you deal Garza for?

[SouthSideRyan]

I thought I was correcting your error, because I knew Corey came up around 2001, but thinking back, he debuted in 2000, and probably straight from AA. So you're randomly like 6 years off in your previous post, which is weird.

[jumbo]

You don't really believe what you just posted, do you? ML teams call up players who have lousy numbers in the minors because they think those lousy minor league numbers will improve with age. :lol: :lol: :lol: :lol: :lol:

 

 

Yeah, okay, I was trying to play along with this like without treating you like you're dumb, but you are making that more and more difficult with each post.

 

A player's minor-league performance, especially high-minors, can be reliably translated into a Major League Equivalent. That MLE is just as predictive of that player's future stats as a major league performance would be. A .260/320/400 MLE is predictively equivalent to a 260/320/400 MLB stat line.

 

Where most prospects flame out is that their MLEs are close, but not quite, up to par. The teams hope they will continue to improve and they don't.

 

Take Felix Pie. At age 21 in 2001, he put up a .283/.341/.451 line at Iowa.

 

Plugging that into the handy online MLE calculator, that is the equivalent of a 230/274/352 line in the majors.

 

Despite that, the Cubs called him up in 2002, hoping he'd continue to improve because of his age. He didn't, posting a 215/271/333 line, eerily similar to his MLE the year before. He improved a small amount but mostly stagnated and became a .249/.298/.374 career hitter.

 

Again: It's a long-proven sabermetric tenant that properly-adjusted MiLB stats are just as predictive as MLB stats. The problem is that most people overestimate how predictive MLB stats are (even established players are pretty volatile) and they underestimate how harsh MLE adjustments can be. The idea that we have to just throw up our hands and pretend like we have no idea what to expect from these guys is lazy and wrong.

 

Very nice description of MLE, including an example. A+

[Rob]

Yeah, Felix Pie came up in 2007. Aside from that, nice post.

[Hairyducked Idiot]

Yeah, sorry, the stats were for Pie. Just got the wrong year.

[Little Slide Rooter]

Tigers Say Reported Interest In Garza Is Overstated

By Zach Links [January 14 at 2:11pm CST]

The Tigers say that their interest in Matt Garza is grossly exaggerated, tweets Bob Nightengale of USA Today.  Furthermore, the club won't part with its prized prospects to land the Cubs hurler.

 

Earlier this week, it was reported that the Cubs and Tigers were discussing a trade involving the right-hander.  The Tigers, according to the report, were open to dealing top prospects and it has been said that the club would part with pitcher Jacob Turner in the right trade.

 

Yesterday, ESPN.com's Jayson Stark wrote that rival teams don't believe the Cubs will trade Garza until July, when there are fewer alternatives on the open market.  Meanwhile, president of baseball operations Theo Epstein contends that he thinks "very highly" of the 28-year-old.

 

 

Considering that they've already stated that they'll move Turner for the right guy, if they want Garza, we want Turner. If not, by all means find the right guy. Now if they offered Castellanos, Smyly, and Porcello or another prospect, would that be worth it? What about Smyly, Crosby, and someone from the back end of their top 10?

[biittner77]

This thread got me thinking about star for multiple prospect trades. What was the last trade where the team getting the prospects won? I'm not suggesting that trading Garza is a bad adea by any means. I'm having trouble thinking of a deal that really backfired on the team that was giving the prospects and getting the established player. It just seems like teams really value their prospects more than ever before.

 

I'm sure I'm overlooking some obvious trades so fire away.

[Little Slide Rooter]

This thread got me thinking about star for multiple prospect trades. What was the last trade where the team getting the prospects won? I'm not suggesting that trading Garza is a bad adea by any means. I'm having trouble thinking of a deal that really backfired on the team that was giving the prospects and getting the established player. It just seems like teams really value their prospects more than ever before.

 

I'm sure I'm overlooking some obvious trades so fire away.

 

Holliday for Car Go, Bradley for Ethier, Garcia for Floyd, Mulder for are a few I can think of off the top of my head.

[TruffleShuffle]

This thread got me thinking about star for multiple prospect trades. What was the last trade where the team getting the prospects won? I'm not suggesting that trading Garza is a bad adea by any means. I'm having trouble thinking of a deal that really backfired on the team that was giving the prospects and getting the established player. It just seems like teams really value their prospects more than ever before.

 

I'm sure I'm overlooking some obvious trades so fire away.

 

http://www.baseballamerica.com/today/majors/trade-central/2010/2610442.html

 

http://www.baseballamerica.com/today/majors/trade-central/2009/269313.html

 

http://www.baseballamerica.com/today/majors/trade-central/2009/268800.html

 

http://www.baseballamerica.com/today/majors/trade-central/2009/268653.html

 

http://www.baseballamerica.com/today/majors/trade-central/2008/267390.html

 

http://www.baseballamerica.com/today/majors/trade-central/2008/267167.html

[davearm2]

Erik Bedard

 

Not quite the same formula, but the Red Sox did pretty well on the Nomar trade.

[seanimal]

I'm assuming Bill James' conclusion involved minor league stats were predictive of major league production. We're talking about the Cubs and major league baseball and not predicting what a player will do in his 3rd year of rookie ball.

 

cleanup on aisle 3

 

lol 3rd year of rookie ball. that one's going into my greatest hits folder

[Northside Blues]

This thread got me thinking about star for multiple prospect trades. What was the last trade where the team getting the prospects won? I'm not suggesting that trading Garza is a bad adea by any means. I'm having trouble thinking of a deal that really backfired on the team that was giving the prospects and getting the established player. It just seems like teams really value their prospects more than ever before.

 

I'm sure I'm overlooking some obvious trades so fire away.

 

Holliday for Car Go, Bradley for Ethier, Garcia for Floyd, Mulder for are a few I can think of off the top of my head.

 

Carlos Guillen and Freddy Garcia for two months of Randy Johnson wasn't all that bad considering the M's traded quite possibly the greatest lefty of all time.

[Careless]

Every player has significant error bars on their projection.

 

The difference is that when a Pujols underperforms by a few WAR, he's still a great player. When most players underperform by a few WAR, they fall into the mass of replacement-level fungibility.

I'm going to guess without knowledge that high-WAR players have a much larger standard deviation in WAR than more average players

[Northside Blues]

Every player has significant error bars on their projection.

 

The difference is that when a Pujols underperforms by a few WAR, he's still a great player. When most players underperform by a few WAR, they fall into the mass of replacement-level fungibility.

I'm going to guess without knowledge that high-WAR players have a much larger standard deviation in WAR than more average players

 

Yes, but if Pujols' distribution is something like

 

 

0.0-1.0 .000

1.0-2.0 .000

2.0-3.0 .000

3.0-4.0 .002

4.0-5.0 .015

5.0-6.0 .062

6.0-7.0 .161

7.0-8.0 .260

8.0-9.0 .260

9.0-10.0 .161

10.0-11.0 .062

11.0-12.0 .015

12.0+ .002

 

There's just a small likelihood of him being a sub 6.0 WAR player (1.7%). For comparison, a 6.0 WAR would have been 24th in MLB last year. None of these numbers were created out of thin air. I found the stdev for Pujols' career WARs and built the distribution on that. All WAR comes from Fan Graphs.

 

And FWIW, Albert Pujols' standard deviation for his career WAR is 1.41 and his average WAR is 8.0.

[Ski]

Every player has significant error bars on their projection.

 

The difference is that when a Pujols underperforms by a few WAR, he's still a great player. When most players underperform by a few WAR, they fall into the mass of replacement-level fungibility.

I'm going to guess without knowledge that high-WAR players have a much larger standard deviation in WAR than more average players

 

Yes, but if Pujols' distribution is something like

 

 

0.0-1.0 .000

1.0-2.0 .000

2.0-3.0 .000

3.0-4.0 .002

4.0-5.0 .015

5.0-6.0 .062

6.0-7.0 .161

7.0-8.0 .260

8.0-9.0 .260

9.0-10.0 .161

10.0-11.0 .062

11.0-12.0 .015

12.0+ .002

 

There's just a small likelihood of him being a sub 6.0 WAR player (1.7%). For comparison, a 6.0 WAR would have been 24th in MLB last year. None of these numbers were created out of thin air. I found the stdev for Pujols' career WARs and built the distribution on that. All WAR comes from Fan Graphs.

 

And FWIW, Albert Pujols' standard deviation for his career WAR is 1.41 and his average WAR is 8.0.

That's assuming his WAR follows a normal distribution, though. You really can't assume that with only a sample size of 11 seasons (Central Limit Theorem requires n is at least 30).

[Northside Blues]

Eh, I was making the point that for a star like Pujols a 5.0 WAR season or less is fairly improbable, as long as he stays healthy. And I did say Yes, but if Pujols' distribution is something like, not that it's that. It's certainly not going to be a normal distribution.

[biittner77]

I'm assuming Bill James' conclusion involved minor league stats were predictive of major league production. We're talking about the Cubs and major league baseball and not predicting what a player will do in his 3rd year of rookie ball.

 

cleanup on aisle 3

 

lol 3rd year of rookie ball. that one's going into my greatest hits folder

 

Didn't Earl Cunningham spend 3 years in rookie ball? It certainly seemed like it at the time.

[Guest]

That's assuming his WAR follows a normal distribution, though. You really can't assume that with only a sample size of 11 seasons (Central Limit Theorem requires n is at least 30).

That's a really good point and one that will be lost on most. Much of the statistical information that is currently in use and frequently mentioned here requires sample sizes that make their predictive utility of little practical value.

[mul21]

This thread got me thinking about star for multiple prospect trades. What was the last trade where the team getting the prospects won? I'm not suggesting that trading Garza is a bad adea by any means. I'm having trouble thinking of a deal that really backfired on the team that was giving the prospects and getting the established player. It just seems like teams really value their prospects more than ever before.

 

I'm sure I'm overlooking some obvious trades so fire away.

 

The first one that popped into my head was Beckett for Hanley and Anibal Sanchez.

[d_money]

This thread got me thinking about star for multiple prospect trades. What was the last trade where the team getting the prospects won? I'm not suggesting that trading Garza is a bad adea by any means. I'm having trouble thinking of a deal that really backfired on the team that was giving the prospects and getting the established player. It just seems like teams really value their prospects more than ever before.

 

I'm sure I'm overlooking some obvious trades so fire away.

 

The first one that popped into my head was Beckett for Hanley and Anibal Sanchez.

 

Brock for Broglio

[Little Slide Rooter]

This thread got me thinking about star for multiple prospect trades. What was the last trade where the team getting the prospects won? I'm not suggesting that trading Garza is a bad adea by any means. I'm having trouble thinking of a deal that really backfired on the team that was giving the prospects and getting the established player. It just seems like teams really value their prospects more than ever before.

 

I'm sure I'm overlooking some obvious trades so fire away.

 

The first one that popped into my head was Beckett for Hanley and Anibal Sanchez.

 

That was really a win for both sides, which is probably more rare. Beckett has been great for Boston, and while they'd surely love to have had Hanley in their lineup instead of Alex Gonzalez, Marco Scutoro, and whoever else has cycled through short stop for the Red Sox since the trade, they havnt exactly been having difficulty with offensive production since.

[Brian]

I'd say the Cards lost when they gave up Haren for Mulder

[bukie]

Sergio Mitre/Renyel Pinto/Ricky Nolasco for Juan Pierre...

[82TonyM]

Bonser Liriano Nathan for AJ Pierzynski

[Post Count Padder]

Cliff Lee, Brandon Phillips, Grady Sizemore for Bartolo Colon

[squally1313]

That's assuming his WAR follows a normal distribution, though. You really can't assume that with only a sample size of 11 seasons (Central Limit Theorem requires n is at least 30).

That's a really good point and one that will be lost on most. Much of the statistical information that is currently in use and frequently mentioned here requires sample sizes that make their predictive utility of little practical value.

Doesn't the fact that a single season WAR number is made up of a lot of information somewhat negate this though? For example, couldn't you just run the same analysis for his WAR in half seasons, or even months? Just splitting it up into 3 month periods would have you 22 numbers to run, which seems more significant, and 3 months of data is still a pretty good amount.

 

Just throwing it out there. Obviously there are many outside factors that would also play into any sort of performance prediction.