Well, Lopez’s 2008 was well within what Dave HIMSELF thought was Lopez’s skill set, about 9 months ago.

Also, I don’t think Lopez is going to be a truly “premiere” offensive weapon (like, say, Vladimir Guererro)- the best case is Lopez turns into something like Carlos Lee, maybe. That’s only a “premiere” offensive weapon if he stays at 2B, because 2B as a class hit so poorly- if he moves to a corner position, that’s a good player but not “premiere”, and Carlos Lee kinds of players are the ones who are most often overpaid for on the FA market (as such, you really want to grow your own if you can, and have them during the “team control” years).

I dunno- the guy I saw through Lopez’s career in the minors is a guy who, when he finally shows up hits .290-.300, has decent doubles/HR power, doesn’t walk a ton but has decent plate coverage. That’s the guy who seemed to show up in 2008.

The other thing I noticed, while idly going through Lopez’s hit charts for 2008, is his power is still pull/LF- even though his home/road splits were heavily biased towards Safeco, which might make you think “oh, he’s learned to hit the other way”. I suspect that were Lopez to go somewhere where LF wasn’t a complete bastard to a RHB, he’d have numbers that would look even better (not that it would change his skills and actual worth- he’d just be in an environment where they would shine more).

TIF,
It seems like the heart of this argument is the ‘ignoring data’ issue. I agree that you can’t discount what happened in the second half of 2007, no matter what it’s cause, because we just don’t know the reasons and probably never will. It happened, so consider it.

Precisely.

But it seems like some people knocking Lopez are doing the same thing in disregarding his production in 2008. Being third among all second basemen in both doubles and RBIs at his age is worthy of considering. It has to cut both ways, doesn’t it?

The “knocking” of Lopez is less a knocking of Lopez and more of an attempt to curb the enthusiasm some folks are having for him. Kind of a “Lopez had a good season, but let’s not go bonkers for the guy just yet” kind of approach. The supporters seem to think Lopez is ready to turn the corner and become a premiere offensive weapon and they are actually the ones taking 2008’s production to lengths that are strained.

Before the defensive metrics provided evidence that Lopez, at worst, is an average defender, his offensive production was not seen as anything but a mitigation of loss. Now he’s seen as less of a concern and with more value, a potential trading piece. So far that’s what I’m gathering from Dave’s comments, and my own stance is fairly in line with his.

Also, I can’t see how using an example like Brady Anderson’s 50HR year is honestly relevant. He was 32 years old that season, and nothing previously indicated that could happen.

Exactly… results can’t provide a good indicator of future performance. Brady before that season wasn’t even close to thought of as a power hitter and Brady afterward shouldn’t have been so overly hyped as he was. Folks were taking results and running with them, ignoring all other facts.

On the other hand, I think a more compelling case could be made that Lopez, over the course of his career, is improving as a hitter.
And of course, all of this is contextual. Whatever his value, if indeed we see the departures of Beltre and Raul, do we know if there’s anyone else better suited right now to hit in the middle of the order? Which was my original point…we have a lot more serious problems to worry about.

I don’t think anybody’s arguing against that last point, especially not the authors here at USSM.

So far, CMC and eponymous are winning me over, even though I usually take Dave’s analysis as gospel truth.

Dave makes great arguments and is extremely knowledgeable, but nobody’s analysis should be taken as the gospel truth. I would think that Dave appreciates people liking and understanding his work, but even he has to admit he’s not always right. But, remember, accuracy and being “right” are two completely different things. Like “fact” and “truth”, they get exchanged a lot as if they mean the same things.

I think it’s unfair for any of us to say “you’re just ignoring data to support your premise,” because that’s presuming motive and inferring willful ignorance.

Well, if you are on a long city street and pass through 5 signals, and afterward someone points out that one of them was a red light and you say “meh”, what would you infer from that? I’m grossly oversimplifying (another fallacy of logic), but I’m trying to make something clear. If someone is going to ignore data, especially intentionally because they consider it irrelevant, isn’t that willful ignorance?

CMC’s argument is not ignoring 2007, it’s discounting it by examining quantifiable data and the personal story we know. Sure, we can’t be certain that Lopez’s personal tragedy had anything to do with his collapse (can we agree that 2007 2nd half was a collapse?), but neither can we categorically declare that deaths in the family do not cause offensive slumps.

This is the argumentum ad ingorantiam fallacy.

“You can’t prove God doesn’t exist, so God exists.”

and / or

“You can’t prove God does exist, so God doesn’t exist.”

No, we have no evidence of it one way or the other. Unless you want to go start randomly causing the deaths of baseball player’s family members at various intervals and test their results afterward, then we’ll never know. If you don’t want a criminal record, we can go check the 100+ years of baseball history to see if anybody else had this problem. I’m not sure there will be enough evidence, but it’s worth a shot.

CMC is not saying they always do, therefore we must discount 2007; he’s saying the family death might have been a non-skill-related reason for the 2007 collapse.

All we have to go on is the information provided to us. Raul’s 2007, for example, was apparently heavily influenced by a back issue that he didn’t tell anybody about. Before we knew that, the speculation was that, at his age, he’d finally hit the decline phase and was plummeting like Mo Vaughn after ingesting a six-pack of Vidro. I do not recall if Lopez publicly complained about his family loss having an impact on his season. If he did, that might be something.

That’s relevant, and I don’t see anyone arguing the contrary, that Lopez messed with his swing, got eye surgery, returned to his minor league hacking, or any other such explanation that would point to why his 2007 collapse was skill-related. Essentially, the counterargument has been “deaths in the family don’t cause hitting slumps.” Is that so? Then prove it.

Already mentioned this above, but you can’t take this illogical stance.

TIF
I don’t know that I’m arguing with you specifically. I’m just trying to figure out, along with a couple others on this thread, why it’s a no-brainer to get rid of Lopez.

Sorry, I guess I got confused since you quoted something I said and kept going. I figured the whole post was directed at me.

Regarding the second point, Dave mentioned it was “smart” to trade Jose Lopez and there’s a lot of reasons why. He’s not advocating dumping him for nothing, but to listen to offers and if someone makes you a good one, you shouldn’t shy away from it. Lopez is signed to a manageable contract, is apparently a near average defender, and has the skill sets to hover around the league average mark for hitting. There is some value to him and we have a replacement part in Luis Valbuena. But, Dave mentions in this article that if Lopez’s defensive woes have been greatly exaggerated, then he should be a lower priority to replace then, say, shortstop.

There may be reasons that escape some of us, but this response, I admit, is baffling…

you’re using past results to project future performance, and that’s also bad analysis

Isn’t that what this site is all about? Looking at a player’s history to try to determine future performance, rather that ‘trusting your gut’ or deferring to the assertion that someone ‘is a good clubhouse leader’?

Not quite. We can use results in a couple of productive ways, especially if the results someone put up don’t match his skill sets, but when looking to future performance projections using results based analysis doesn’t work. If it were that simple, Brady Anderson would have always been projected to hit 50+ home runs after one monster season of results. That’s an extreme case, but you get the point. What has happened doesn’t always mean it will happen again.

The basic point I was trying to make in the last post was to say that essentially I agree with you and Dave in the sense that you can’t throw out data that doesn’t agree with your premise…but that seems to be what the Lopez detractors are doing. In other words, looking at his entire career path, with last year logically the most relevant, where does that leave us? More specifically, given his age, why is there not reason to assume further progression?

Why is last year logically the most relevant? It’s the most recent, but it should not be weighed more heavily into your calculations then 2004, 2005, 2006, or 2007. What Dave is trying to say is that only once has Lopez shown himself to be an above league average hitter (2008), and while that’s encouraging, there is nothing set in stone that provides evidence that the improvement trend will continue. Lopez has the skills to repeat 2008’s performance just like he has the skills to repeat 2007’s performance. Neither one, and both must be considered outliers at this point since one is the extreme bottom and the other is the current extreme top, is more important then the others and the remainder of the data we have on him.

I didn’t pull this out of the air…I was referring to eponymous’ statement earlier in the thread.

And I think he answered this one below, so I wont reply on this subject.

The thought’s the same (that Lopez’s skill set as a hitter has a bit more room to grow), but Guillen is a more extreme case of that.

I remember Jefferies as a small cult hero in my youth. Looking back, he was never as good as the hype.

Interestingly, Yuni is his #9 comp for ‘most similar batters’.

Looking at his comps we can maybe expect him to be a league average hitter (OPS+ wise) for his career. Jose’s apparent dislike of the walk coupled with the tendancy of batting average to vary wildly from one year to the next will probably make for a rollercoaster career.

That said, I’m not opposed to having him play another season with the M’s. But, with the season he just had, his youth and and relatively cheap contract, Lopez could be moved for more useful pieces to the team for a 2010 or so.

I think it’s unfair for any of us to say “you’re just ignoring data to support your premise,” because that’s presuming motive and inferring willful ignorance.

CMC’s argument is not ignoring 2007, it’s discounting it by examining quantifiable data and the personal story we know. Sure, we can’t be certain that Lopez’s personal tragedy had anything to do with his collapse (can we agree that 2007 2nd half was a collapse?), but neither can we categorically declare that deaths in the family do not cause offensive slumps. CMC is not saying they always do, therefore we must discount 2007; he’s saying the family death might have been a non-skill-related reason for the 2007 collapse.

That’s relevant, and I don’t see anyone arguing the contrary, that Lopez messed with his swing, got eye surgery, returned to his minor league hacking, or any other such explanation that would point to why his 2007 collapse was skill-related. Essentially, the counterargument has been “deaths in the family don’t cause hitting slumps.” Is that so? Then prove it.

you’re using past results to project future performance, and that’s also bad analysis

Isn’t that what this site is all about? Looking at a player’s history to try to determine future performance, rather that ‘trusting your gut’ or deferring to the assertion that someone ‘is a good clubhouse leader’?
The basic point I was trying to make in the last post was to say that essentially I agree with you and Dave in the sense that you can’t throw out data that doesn’t agree with your premise…but that seems to be what the Lopez detractors are doing. In other words, looking at his entire career path, with last year logically the most relevant, where does that leave us? More specifically, given his age, why is there not reason to assume further progression?

And at his age, I don’t know how you can conclude that he can’t be the next Carlos Guillen.

I didn’t pull this out of the air…I was referring to eponymous’ statement earlier in the thread.

Cut CMC and eponymous some slack. You can’t count the second half of 2007 as a data point…but not the rest of his career, too.

What I said is that excluding the latter half of 2007 as an outlier is bad analysis. I didn’t say anything about ignoring the rest of his career, so I don’t really understand what you mean.

Take the entire sample and do your analysis, not part of the sample and do your analysis. That’s it.

So isn’t the following ‘pure stat analysis’?: in the four years since his baseline -14, he’s up a net 18. That averages to a yearly increase of 4.5. Thus, why should we not predict him at +8.5 next year…and then +13 the year after that?

While that’s theoretically possible, you’re using past results to project future performance, and that’s also bad analysis. Trend analysis doesn’t really fly in baseball unless you have a much larger sample to draw comparisons from. IE: the trend of age curves is based on historical data from thousands of players over the last 100 years. Lopez improved this season, for sure, and he improved within the realm of possibility based on his skill sets. ZIPS, PECOTA, Marcel, CHONE, and other methods use all kinds of factors to try to predict future performance, with varying and surprisingly accurate results. Their methods are not so simplistic as simple trend analysis of a player’s annual stat improvements.

I assume the central argument against this is that given 2007, it’s illogical to assume a steady rise. But if so, isn’t that considering 2007 the outlier…just as you said is not acceptable?

Are you arguing with me or with Stags? I was deliberately saying you HAVE to include the outliers are part of the whole. If the outlier season, which is still part of Lopez’s career stats, knocks someone’s theory for a loop, you can’t just say “well that sucks, let’s not include it and my theory will be fine”. That’s just horrible analysis.

And at his age, I don’t know how you can conclude that he can’t be the next Carlos Guillen.

Well, putting aside that this is a fine example of an irrelevant conclusion fallacy (because nothing you’ve said in this entire post has anything to do with Carlos Guillen until now), what evidence does anybody have that he will be the next Carlos Guillen?

I believe what we’re seeing here is “gunshy paranoia”. We’ve seen, or think we’ve seen, our players leave us as “bad” players and become “good” players elsewhere, so we believe we should just hang onto whatever we’ve got regardless of what the stat analysis or scouting reports tell us. Nevermind the fact that there’s also a lot of our players who left bad and stayed bad…

Well, that’s why we do these things. We analyze and find out as much fact (not truth) as we can so we can make the best decisions. Or, at least, we would HOPE that’s what the people in charge do. It’s understandable that after a string of bad decisions made by bad people in charge, we’re cynical about any decision, but that doesn’t mean we have to start gnawing our own legs off when all we’re doing is DISCUSSING the possibility of moving someone around who may be a weak link to the team.

This is supposed to be fun, guys. =)