Predictions vs. Outcome

I agree with your grade of the Giants offseason.

I think he was grading his predictions from before the Hot Stove league, could be wrong.

If we could add the Sanchez trade to this list of moves, then it’s an F. Total F.

I’d say the Huff prediction is the more likely by far. Wrist surgery plus mid-30s is a very worrisome combination.

You’re forgetting that though I’m taking the Huff+DeRosa+Molina money, the impact of each player is relative to the player they replace.

You don’t sum up Nick Johnson’s contributions and compare them to the contributions of those three.

Take Molina: he’s projected to hit .263/.292/.424, far worse than Posey’s projection of .265/.341/.402. The net gain of a year of Molina relative to a year of Posey is negative. In fact, every at-bat that Molina takes from Posey is a net negative for the team.

Next, look at DeRosa: he’s got a .262/.343/.415 projection, as compared to a .261/.344/.403 one for Lewis. For this signing, one gets 12 points of slugging.

Lastly, Huff: he’s projected to hit .267/.332/.437, only marginally better than Ishikawa’s projected .261/.329/.433. When you compare Ishikawa’s close to Gold-Glove-level defense to Huff’s DH defense at 1B, Ishikawa is actually expected to be a better player next year. Johnson is projected to go .268/.392/.427, much better than either line. With Johnson, you get 60 more points of OBP, which are much more valuable than the 10 points less of SLG, especially seeing as 1 point of OBP is about 3 times as valuable as one point of SLG for scoring runs.

Just as every at-bat Molina takes away from Posey is detrimental, every at-bat Johnson would take away from Ishikawa would be beneficial. Thus, even if Johnson played 81 games, one of the worst-case scenarios, 81 games of Johnson and 81 games of Ishikawa would be much better than 162 games of Huff and Ishikawa.

Summing up the net impact of the three: Molina is a big negative, Huff is a negative, and DeRosa is a tiny positive.
Johnson: a huge positive, even if he plays half of the games! The rest is icing on the cake.

Much better.

That Scutaro is any better than FSanchez, let alone FSanchez and Uribe, is highly dubious. 34-year-old Scutaro had a good year at the plate last year, but at his age one can hardly project that for two years into the future. He has a lower career BA and OPS+, and about the same career OBP as Sanchez. Last year, he was better; in his mid-’30s, no one has any idea that this will continue.

Neither of us can purport to know which years should be weighted most strongly when comparing players.

There are those who have that knowledge- the creators of the highly reputed projection systems. Their projections of players reflect their past years. So let’s look at the conclusions of one of these people, who is more knowledgeable on the subject than you or I.

Freddy Sanchez: .285/.323/.400
Juan Uribe: .257/.304/.421
Marco Scutaro: .268/.360/.367

You can see that Scutaro’s projected line is worse than he hit last year, so it does reflect the years of mediocrity that he had before his two seasons of excellent production.

Even so, Scutaro is a far better hitter relative to Sanchez, as we see the 37 points of OBP gained and 33 points of SLG lost (remember than 1 point of OBP is 3x more valuable than 1 point of SLG when it comes to scoring runs). Scutaro is also a better hitter than Uribe, gaining 56 points of OBP and losing 54 points of SLG.

And not even mentioned in your post is Scutaro’s strong defense, which he is known for, rather than for his hitting which you focus on.

Thanks for the response!

Oversimplification. This post reeks of it.

Here’s the first example: DeRosa got lots of interest in free agency, and the Giants haven’t gotten a decent offer for Lewis, therefore DeRosa>Lewis. Well, the free agent market is not the trade market. Imagine that! What does this mean for the particular example?

-DeRosa is available when the GMs have a large pool of cash to draw from, and a large number of holes to fill
-the cost of making a phone call to DeRosa’s agent is nothing; teams can express interest, see what the asking figure is, and back away later if need be
-DeRosa is a recognizable player to a prospective fan base; GMs consider popularity of moves in addition to their effect on the field, and even if Lewis and DeRosa may be equivalent, many GMs would be willing to pay the extra price to acquire the “cagey vet” rather than the young player that none of his fans knows who didn’t start on the worst offense in baseball (even though he was benched wrongly, ultimately)

The post decides that receiving interest from teams in free agency is strong evidence of a player’s value, yet for the other player to be deemed worthy, he must be traded! One of these occurences is much more common than the other.

The second example, Scutaro vs. FSanchez:

-Sanchez signed all of his contracts in the boomtime market (the current extension is based off of the $8 million option, which was part of a previously written contract), while Scutaro signed the current contract after the market slowdown
-Freddy Sanchez signed contracts with the Pirates pre-Huntington era and the Giants in the Sabean era, two organizations known more for their ineptitude in signings, free agent and otherwise, during that time period, and thus very much inclined to overpay for a players services, while Scutaro signed his first free agent contract this year, with the Theo Epstein Red Sox, one of the highly esteemed front offices.

Comparing one signing by a proficient front office to a signing by a front office with a bad reputation, and then calling it the “projection of the market”, is a very large oversight.

I think DeRosa has more value than FLew because he can play all over the field. FLew is pretty much a left fielder only which is one of the easiest positions to fill. But I agree with the concept that some GMs don’t pay much attention to the projection systems.

I think a lot of the background assumptions here are incorrect.

1) GMs are accountable for their results. Not necessarily. Lots of really bad GMs persist for many years and fill multiple vacancies without producing any results. Ed Wade, Bill Bavasi, Allan Baird, Cam Bonifay. Even good GMs are not held accountable everytime they misjudge the value of a player. If that were true, Sabean would have been gone after Zito and Rowand and Billy Beane’s career would have been ended by Eric Chavez.

2) Baseball players are valued approximately according to their contribution on the field. If the baseball labor market were perfectly competitive, this would probably be true. But it is not. Players are not payed precisely what they are worth in FA because of A) limited numbers of perspective employers B) imperfect information about what they contributed have and will contribute.It’s been well documented that high OBP players were undervalued for many years and defense is probably still not correctly compensated.

That being said I don’t think that the market for Fred Lewis is all that mysterious. He’s not a particularly good player, and really only useful on as terrible an offensive team as the Giants. He’s really not that young and only a good fielder for a position so bereft of athleticism as left field (where even bad teams usually have no trouble finding a real hitter.

1) + reply. You should use it.

2) Your responses are not coherent, flitting about from topic to topic and indicative of misreading of my own comment (see use of “simplistic”), making it nigh impossible to make any sense out of it.

You haven’t replied to a response (by ResDog) to your previous comment, which highlighted a serious error in your logic.

DeRosa gets a large amount of his market value from his positional versatility. Lewis can play one position: corner outfield. For the Giants, they are effectively equal positionally, but for a whole slew of other teams, the demand for DeRosa was to fill needs at other positions. Faulty comparison: DeRosa has very little value as a corner outfielder only, which is how the Giants are playing him, but you are taking the demand for him as an infielder who can play at three spots on the diamond and pretending that that is a function of his hitting and fielding ability in left field, which it is not.

We have an explanation of why I found it prudent to use projections here:

Not trying to start a huge argument here, but can you guys talk about why you rely so heavily on projections?

No problem. Basically, we’re trying to figure out what a guy will do in the future. It’s essential to the decision making process. We can all go back and look through their numbers to try to figure it out ourselves, but we don’t have a database in our heads of every player and their trends. We don’t have the math figured out for how much you should properly regress certain variables and how heavily to weight them. The projections systems use all this data that we are simply incapable of using – databases of players, doing the math on trends and how heavily certain variables influence performance and future performance, figuring out how players age, what minor league/major league performance tells us – basically, they take an objective look at all the data to tell us what a player will do in the future based on what all players historically have done. We cannot do this as objectively or accurately as projections systems, which is why we use them. Simply put, they’re better at figuring it out then we are, because they know a hell of a lot more than we do.
by Missing Barry on Feb 14, 2010 6:56 PM PST

Perhaps you thought if you made a blanket denial of all projections’ usefulness, that would leave me with no evidence to prove that Scutaro>Sanchez and Lewis=DeRosa in LF. Well, that’s not so. As you can see from the explanation, the projections are just a condensation of past performance with a weighting formula to decide which years in the past are most relevant for the player’s actual ability this year.

Well, anyways, here’s where I show you without projections that Lewis and DeRosa are roughly equivalent as hitters.

Fred Lewis

Mark DeRosa

Background info: Lewis’s 2006 and 2007 fall under extreme small sample sizes, and his 2008 is his only real year of consistent starting.

Not a substantive difference between the two. DeRosa’s average for the last four years is right with Lewis’. Trends indicate that DeRosa will drop below Lewis, though both could level out to about the same.

Let’s see: fundamental fallacy in the market value theory identified, and explained and justification of DeRosa and Lewis’ equivalence as DeRosa is being used by the Giants currently accomplished! My feel-good juices are flowing.

Lo and behold! I can do the same non-projection-based justification for Scutaro being better than Sanchez.

1) Contracts: Scutaro 2 yr $12.5 mil, Sanchez 2 yr $12 mil
Right off the bat, the market says: Scutaro’s better. Then consider that Sanchez got his contract with the negotiating power of an $8 mil option for 2010 behind him, which Scutaro didn’t have. The Giants went in with Alderson, a buyout was not an option, so their contract was actually a restructuring of an $8 mil, 1 year contract that was the only possible alternative.
Players of similar value to Sanchez like Orlando Hudson got no more than 1 year, $5 mil in free agency. So now the market says: Scutaro’s even better than these contracts indicate.

2) Past stats
Scutaro

FSanchez

Background info: Scutaro gets fewer than 100 PA in 02 and 03. Same goes for Sanchez 02-04.

First, look at the years other than the two peaks for each graph. They average out to be roughly equivalent. Sanchez had his peak 4 years ago, and hasn’t been above average since. In Scutaro’s 2nd year of playing more than 140 games, he had his career year, and it was last year. Doesn’t take a rocket scientist to see which one is the better hitter at this point in their careers.

And of course, no appraisal of FSanchez’s ability is complete without reminding everybody that injuries to both shoulders and one knee in the last season are the kinds of ailments that can cause a drop in productivity on the offensive and defensive sides of the ball.