mgl

Yes, I don’t have nix projected all that badly. Around 15 runs worse than Jeter I think. Maybe less. I’m not sure off the top if my head. And of course my projection for Jeter is far worse than his 2012 season.

At some point, when the season is over, don't the actual results matter more than the projection?

A couple years ago this was something I would post all over the place. I'm not touching this at all.

At some point, when the season is over, don't the actual results matter more than the projection?

No. What happened in the past doesn't matter; what matters is what we expect him to do in the future. To expect him to simply play the same as his 2012 season is as foolish as expecting him to play the same as his 2011 season, or his 2012 August or whatever. It's as foolish as assuming Cano would be the same in the O's series as he was in his 13 games prior where he hit .500 (or whatever.) None of it matters, it all matters.

Again, this mostly comes down to defense anyway. Nix's defense at SS is largely unknown.

Not reading the article but it depends on what he means by his projection. If he means a pre-season 2012 projection then, yes, he needs to incorporate actual 2012 performance. But if he means a projection of Jeter's current talent level which includes his 2012, then he's already done that. ZiPS provides a "rest-of-season" projection, no reason MGL can't do the same. A simple 3-year Marcel puts him at a 105 OPS+. Ding him for being a year older and that's probably 100. Using bWAR, Marcel would put him about 15 runs below average on defense or at about 1.6 WAR overall, both before applying the age penalty. An up-to-the-minute assessment of Jeter's true talent might well put him well below average because, even this year, he's barely above it.

But that's using bWAR. Using fWAR, in part thanks to their much lower replacement level, Jeter's compiled 8.3 WAR over the last 3 years compared to just 4.6 under bWAR. Since MGL is a fangraphs boy, it's hard to see an argument for a team being better without their above-average SS.

In his career, Nix went from being a 2B to being a 3B. Why would we expect him to be even adequate at SS?

Well in terms of actual player evaluation, projections (small 'p') are technically all that ever matters. The past is always the past. Obviously any projection of playing ability for right now should include 2012 and include it to a considerable extent.

That said I can't imagine that Nix is a better player at short than Jeter. Sure Jeter as a shortstop defensively has all sorts of problems, and even offensively Jeter's big year with the bat is tempered by what looks to be the disappearance of his once large baserunning value. But there's no real evidence Nix is much of a shortstop either (and some that suggests he's a bad one), and while Nix is likely a little better on the bases with at least as much power, his on-base skills are pretty terrible.

I'd agree that on paper the dropoff between Jeter and Nix over a seven game series (or even two of them) has a negligible difference on the Yankees' chances, but I would say the loss of Jeter and the resulting sudden increase in the spotlight for Nix could have some more serious negative consequences as well.

#5,

His first year out of high school in rookie ball he played short, so you'd assume he played there in high school. He played it so well he was a second baseman from then until now, with only occasional stops at short (with varying levels of success).

In his career, Nix went from being a 2B to being a 3B. Why would we expect him to be even adequate at SS?

Because he only has to be as good as Jeter?

At some point, when the season is over, don't the actual results matter more than the projection?

Not for the likes of MGL, Cameron and Ray among others.

If we're talking about evaluating the 2012 season, of course it's the results that "matter." If we're talking about how a player is going to perform over the next X games, they only "matter" in so much as they are important inputs for our estimation...

Because he only has to be as good as Jeter?

I think the ratio of the numbers (.03/.07) makes it look weird. MGL first posted that Jeter was 15 runs better with the bat; that ratio would make Nix 35 runs better with the glove. The actual numbers are about 5 and 11 which is more reasonable on defense, but doesn't look right on offense.

MGL clarified that:

First of all, I said that the 15 runs was a guess off the top of my head. I was not in front of my computer.

Secondly, the .03 runs came from the sim. I said, one the lineup in the sim is completely different - I didn’t just replace Jeter in the 2 hole with Nix in the two hole. The sim also is a sim which means that it interacts all the players and includes base running, base stealing, etc. And finally, because the sim is a sim, the number of runs scored has sample error. The .03 could be .08 or .01 or just about anything.

In other words, he wasn't comparing Jeter to Nix -- he was comparing the Yankees lineup with Jeter vs. the Yankees lineup with Nix. A slightly different (but more important) measurement.

That makes sense. Cano, Tex, A-Rod, Swisher and Granderson all probably project better than Jeter and they all move up in the lineup now. Plus, you get some benefit in not having your best hitter 3rd.

It does (implicitly) underrate Jeter because they'd probably gain from just dropping him (or Ichiro) to 8th in the lineup.

If we're talking about evaluating the 2012 season, of course it's the results that "matter." If we're talking about how a player is going to perform over the next X games, they only "matter" in so much as they are important inputs for our estimation..

The important thing there is how good mgl's estimation is. I think there's too much credit given to the projections when we get to outliers like 38 year old shortstops. I would be curious to see how many times mgl has actually done a good job projecting Jeter. My memory is that he hasn't done well at all.

I think there's too much credit given to the projections when we get to outliers like 38 year old shortstops.

The problem is that ultimately you have to produce some sort of projection. Projections aren't just something math geeks like me do, everybody who does player evaluation has to do it, it's just the geeks do it mathematically. So does a mathematical projection do better or worse in such situations than other methods? Don't know without knowing the other methods and a way to compare the two.

Don't forget, the 2013 projection is for a 38 year old ss coming off a significant ankle injury. Of course, since its Jeter, he'll probably have a 5 war season and make all the doubters look foolish once again.

I would rather have Jeter.

MGL Out

MGL Out

It's like for a robot he actually has emotions.

MGL Out

Did something prompt that?

MGL Out

That is one of the all-time great blog posts.

Did something prompt that?

An argument with David Cameron followed by a post by Cameron here:

http://www.insidethebook.com/ee/index.php/site/comments/so_how_many_times_does_leyland_have_to_get_burned/#comments

That is no longer there because it got binned.

Geez, if there's one saber-guy that could benefit from seroquel, it's gotta be MGL.

{deleted} :D

Yeah, yeah--and Rany's going to disengage from the Royals. Promises, promises. :-)

I missed Cameron's comment in the Leyland thread. What did Cameron say?

I was not in front of my computer.

Secondly, the .03 runs came from the sim.

Now wait a second ...

MGL will no longer be contributing to the book blog? It's his own blog (shared with Tango). It's pretty bad when you get run off your own blog. Just kill the comments section or require that commenters make a $10,000 bet first or something.

MGL will no longer be contributing to the book blog? It's his own blog (shared with Tango). It's pretty bad when you get run off your own blog. Just kill the comments section or require that commenters make a $10,000 bet first or something.

Tres bizarre! Hey, does anyone have an idea how to push Chass over the edge?

EDIT: Stranger still is that MGL disabled the comments section for his good-bye post.