How about this guy for SS?

[Frobby]

Oh, I don't think he is either. Having said that, let me show you something.

Sticking with the RZR and OOZ metrics, LH had a RZR of .930 and a OOZ/BIZ ratio of .163. The other Orioles shortstops combined for .835 and .157. Overall, AL shortstops were .807 and .147. The first thing to note is how well Orioles shortstops performed according to this metric.

Even if you regress LH's stats 50% towards the AL averages, he'd finish at .869 and .155, respectively. Both still very good rates. Plug in the +/- and defensive runs equations and that works out to +7 and 5 defensive runs per 100 BIZ for LH compared to 0 and 0 for the league average SS.

As a comparison, the consensus best defensive SS in the game (Everett) was worth a +14 and 11 per 100 BIZ last year. The MLB leader amongst SS last year in +/- and DR was Troy Tulowitzki. He averaged a +9 and 7 DR per 100 BIZ.

So even regressed 50% toward the league average LH is pretty good.

That is still some pretty good defense.

Agreed, but I'm stuck on the very small sample size here. Compare Tejada's defensive numbers for April vs. May -- one set of numbers would make you think he was the worst SS ever, the other set would make you think he's a gold glover. I just don't trust 139 innings of defensive data.

[El Gordo]

Coming up with defensive numbers based on what he did in his limited time in the majors is a spointless exercise IMO.

1970 did the work(not rshack) but it is still based on a small sample size...He just said that himself.

At this point, relying on what you see and what scouting reports tell you is much more important than projecting out defensive stats(which are shaky stats to begin with) based off of a very small sample size.

LH is an above average defensive SS...How far above average is up for debate but not one single person on this board knows this for sure.

What we do know is that the guy can't hit, doesn't walk and has no power.

Even Adam Everett only had a WARP3 of 1.6 last year.

Everett's best year, according to WARP3, was last year when it was 5.7. But most years, he averages around a WARP3 of 3.I think that is LH's ceiling...A WARP3 of 3.

I think mostly he will be somewhere between a 1-2 win player.

You've lost me here, not that you ever had me, but how could Everett have a WARP 3 last year of 1.6, as well as a WARP 3 last year of 5.7 Some difference there; which was it? If LH is a Warp 3 player, that would mean he is worth 3 wins, right? As a stop gap solution for a year, what's wrong with that? And RShack did perform his own back of the envelop experiment, which is there for you to read should you care to do so, and came up with similar numbers. I've yet to see anything other than repetitous bluster from you.

[Sports Guy]

Agreed, but I'm stuck on the very small sample size here. Compare Tejada's defensive numbers for April vs. May -- one set of numbers would make you think he was the worst SS ever, the other set would make you think he's a gold glover. I just don't trust 139 innings of defensive data.

Exactly...1970's work is good and all and it is always appreciated but, at the end of the day, it is meaningless because of the small sample size.

Here is what I want to know....What does LH have to do, defensively, for him to carry a 550 OPS over the course of 450+ at bats??

Whatever that answer is, if he can do that and still be worth 2 or more wins a year, I would be OK with him being our starting SS for 2008. My guess is, he can't do that though.

[Sports Guy]

You've lost me here, not that you ever had me, but how could Everett have a WARP 3 last year of 1.6, as well as a WARP 3 last year of 5.7 Some difference there; which was it? If LH is a Warp 3 player, that would mean he is worth 3 wins, right? As a stop gap solution for a year, what's wrong with that? And RShack did perform his own back of the envelop experiment, which is there for you to read should you care to do so, and came up with similar numbers. I've yet to see anything other than repetitous bluster from you.

Sorry, i meant he had a 1.6 WARP3 this past season...5.7 the year before.

And I said LH's CEILING is 3 wins...Not that I expected that.

And what is wrong with that? Simple...We should be doing better. That has been the constant message since all of this started.

[El Gordo]

Sorry, i meant he had a 1.6 WARP3 this past season...5.7 the year before.

And I said LH's CEILING is 3 wins...Not that I expected that.

And what is wrong with that? Simple...We should be doing better. That has been the constant message since all of this started.

My constant message has been, if we can't get the long term solution to SS this year, why not put up with a good defensive 1-2 win stop gap, in house, and not waste a prospect?

[Sports Guy]

My constant message has been, if we can't get the long term solution to SS this year, why not put up with a good defensive 1-2 win stop gap, in house, and not waste a prospect?

Who says you have to waste a prospect though? And who says it would be a prospect that we would miss??

We have seen some MiL FAs shown on here that would be better than LH in all likelihood.

And besides, we have to get that long term solution first...I would prefer to just go get that ML ready guy now and not even worry about any of this.

[El Gordo]

Who says you have to waste a prospect though? And who says it would be a prospect that we would miss??

We have seen some MiL FAs shown on here that would be better than LH in all likelihood.

And besides, we have to get that long term solution first...I would prefer to just go get that ML ready guy now and not even worry about any of this.

I haven't. The only one who looked to be possible was Rouse. Please show me some of the others and give me the numbers to demonstrate they would be better than WARP3 1-2.

[Sports Guy]

I haven't. The only one who looked to be possible was Rouse. Please show me some of the others and give me the numbers to demonstrate they would be better than WARP3 1-2.

http://www.thebaseballcube.com/players/P/Antonio-Perez.shtml

Someone mentioned he was picked up by the Nats I believe...I would make a deal for him...I am sure something small could be worked out.

Perez was a top prospect before and was part of the Sweet Lou compensation.

I would look for more players like him...The next Brandon Phillips type guy. You could probably get guys like this for prospects like JJ Johnson or guys like Rocky Cherry...We need to thin out the 40 man roster somewhat if we trade BRob and Bedard, so we need to trade some players...Do it for a better SS stop gap than LH.

As for the MiL FAs, that research has already been done...I am not going to go back over it.

[RShack]

...for the guys being discussed, there is between 5 and 7 years of more-or-less full-time experience to go by. So why wouldn't you trust that data?

Because I don't regard defensive stats as being particularly reliable measurands of defensive ability. I'd take a solid scouting report from an experienced infield scout over 5 to 7 years of minor league defensive data any time, simply because I don't know how to interpret the defensive data. It's difficult enough for major league defensive stats, on which a world of analysis has been lavished by the best statistical minds in baseball.

Let's pause here. I agree that one should refer to high-quality scouting reports, and I routinely point out that we need to be mindful of what stats are and are not good for. So, it's not like I'm a believer in the idea that you should blow off the scouts and just mindlessly trust stats. I think we can all agree that my view of this has been consistent. But that's not the same thing as saying we should do the reverse. I don't beleive we should blindly trust scouting reports and dismiss any stats that might contradict those reports either. I think it's important to look at both, and to consider what we know about the strengths and weakness of each. My guess is that you agree with this basic idea.

Of course, the problem with this idea is that it is not convenient. What people want is some single number they can refer to in order to get a good basic idea of a player's value. Having such a value is a mixed blessing, witness the dumb things some folks conclude simply because they trust OPS to the exclusion of other stuff. But I think we can all agree that this is one place where D-stats lack something that hitting and pitching stats have: a single number that gives you a basic idea, plus a short-list of additional numbers that, when taken-in-total, can indeed paint a pretty definitive picture of a guy's value. You can do that with hitting and pitching stats, and you can't do that nearly as well with D-stats. But that doesn't mean that D-stats are neither reliable nor useful. Rather, it's just that D-stats are incomplete. The better D-stats we do have each tell us something useful, but they can also be misleading if you don't realize the important D-things that they don't measure. For example, consider range data. I think we can all agree that range data is importantly informative, especially when talking about a SS. While we can argue about which version of range-data is best, I think we can all agree that range-data matters. The problem is that it doesn't tell you everything you need to know. So, for example, we can all agree that if a given SS has great range, but also has clank-hands and a lousy arm, then that's a guy who we would not consider to be a D-asset. In that example, it's not that the range-data doesn't tell you something useful, but rather that it doesn't tell you everything you need to know to have a useful picture. What we don't have is some D-stat that gives us an integrated view of a SS's range-plus-hands-plus-arm. On that basis, then D-stats are a failure.

But that's not the basis we're stuck with. The better approach is to look at the range data, and then factor in what scouts say (or what your eyes see on your TV) about the guy's hands and his arm. IMO, if you use reasonable judgment about considering all of those things, then the D-stats do indeed help a great deal in permitting us to form a solid and accurate picture of the SS's D-contribution. Thus, if I see a guy who by all accounts has a strong arm and good hands, and if the range-stats show him to be way better than other guys, then I'm quite happy to trust methods of assigning D-runs to the guy's defensive contribution. I don't think there is anything wrong with trusting the idea of D-runs, provided that you have adequate data about range, combined with appropriate reports about hands and arm. AFAIK, that's what's happening here. Given the available data about LH's range (which evidently is based on a large sample of combined ML-and-MiL play at SS), combined with what I see (and what scouting reports say) about his hands and his arm, then I see no reason to distrust the idea that his D-runs go a huge way towards compensating for his weak bat. This is 100% consistent with the demonstrated value of guys like Belanger. While I am not suggesting that the 2 stat-exercises that 1970 and I have done are "the last word", I do think it's very interesting that we went about our exercises in very different ways but nonetheless wound up with guestimates of Total Value that were eerily, scarily similar. To me, it's flat-out dumb to arbitrarily dismiss that in favor of a knee-jerk reliance on OPS which arbitrarily dismisses D-value as trivial when there are clear indications that D-value can actually be quite substantial, and that stats can permit us to weight it appropriately with offensive contribution. For anyone who claims to support a SABRE view of baseball and who understands what stats can be useful for, but who also values scouting reports, the idea of arbitrarily dismissing an integrated Total Runs view while blindly trusting just OPS is IMO simply indefensible.

In the case at hand, the only big question I have is whether the BBHQ range data is really based on both ML and MiL innings. Based on what 1970 said somewhere in here, it's my understanding that it's based on both. If that's true, then I am comfortable saying that LH is the better option than the other guys we were comparing him to. However, I wonder if it's true, simply because he and I are getting eerily similar findings and he's using ML-only data. That makes me wonder if either I misunderstood what 1970 was saying about that, or if he accidentally said something that's not true. If the range-data for all those guys was based solely on the small ML sample, then I agree it's not trustworthy. However, the way it's not trustworthy has very little to do with the value assigned to LH. If you have range-glove-arm, you have range-glove-arm. That's not where the problem would be. If the range-data is based on a small sample size, then it could easily be misleading about Hu and Aybar. The danger of small sample size in this case would be that it is undervaluing their D-value because the small sample size fails to let them demonstrate that they have better range than the data shows. So, I am quite comfortable with the value it assigns to LH. My only big concern would be that the small sample size unfairly undervalues Hu and Aybar. Notice however that this is different from what some of the foot-stompers are saying: some are saying that a small sample means that LH is being way, way over-valued in his Total-Runs. That's just silly. It could be that it's way-undervaluing Hu and Aybar, but that says nothing about the Total-Runs value of LH, which is the part that some guys have their heads in the sand about.

[El Gordo]

http://www.thebaseballcube.com/players/P/Antonio-Perez.shtml

Someone mentioned he was picked up by the Nats I believe...I would make a deal for him...I am sure something small could be worked out.

Perez was a top prospect before and was part of the Sweet Lou compensation.

I would look for more players like him...The next Brandon Phillips type guy. You could probably get guys like this for prospects like JJ Johnson or guys like Rocky Cherry...We need to thin out the 40 man roster somewhat if we trade BRob and Bedard, so we need to trade some players...Do it for a better SS stop gap than LH.

As for the MiL FAs, that research has already been done...I am not going to go back over it.

The point of this thread is the value of defense vs offense and the relative trade off. So your suggestion is a guy who has played 16 innings at SS in the past 3 years? If you look at his stats you will see he has mostly been a 2B,3B since 2000. Do you really think his glove at SS is as good as a guy who's career up to this point is based on it. Or that if it was, he would be wasted at the other two positions all that time? The idea is to get a good defender who's bat won't kill us not an average or less defender who's bat is maybe ok.

[Sports Guy]

The pont of this thread is the value of defense vs offense and the relative trade off. So your suggestion is a guy who has played 16 innings at SS in the past 3 years? If you look at his stats you will see he has mostly been a 2B,3B since 2000. Do you really think his glove at SS is as good as a guy who's career up to this point is based on it. Or that if it was, he would be wasted at the other two positions all that time? The idea is to get a good defender who's bat won't kill us not an average or less defender who's bat is maybe ok.

Actual, the point of this thread was finding another SS...It has turned into what you said.

And that may be what you want but neccassarily what I think we should go for.

I would rather have a 700 OPS SS with an average glove than a 580 OPS SS with an above average glove.

[Roll Tide]

Hopefully we can obtain a SS in a deal & dont have to worry about combing the scrap heap.

[El Gordo]

Actual, the point of this thread was finding another SS...It has turned into what you said.

And that may be what you want but neccassarily what I think we should go for.

I would rather have a 700 OPS SS with an average glove than a 580 OPS SS with an above average glove.

Don't see how the defense/offense trade off isn't part of a discussion about what to look for in another SS. Uribe is about a .700 OPS guy and slightly above average at SS(+9 to Miggi's +5) I believe earlier you said you preferred LH to him, "I would prefer LH to Uribe as well". Why the change of heart?

[Sports Guy]

Don't see how the defense/offense trade off isn't part of a discussion about what to look for in another SS. Uribe is about a .700 OPS guy and slightly above average at SS(+9 to Miggi's +5) I believe earlier you said you preferred LH to him, "I would prefer LH to Uribe as well". Why the change of heart?

The problem with Uribe is that he is no guarantee for a 700 OPS and his OBP is awful...And, of course, he comes with a big price tag.

If I thought we could get a 320ish OBP out of Uribe, I wouldn't mind him for the year.

EDIT: Actually, i change my mind....If we don't acquire a good ML ready SS, i would trade for Uribe assuming it doesn't cost as much:

http://www.baseballprospectus.com/dt/uribeju01.php

I had not realized his WARP3 was as strong as it was.

[RShack]

...Frankly, if you just stop and think for a sec about the fundamental difference between hitting and defensive performance, you can see that it makes sense to trust upside MiL defensive performance *more* than you trust upside MiL hitting performance, simply due to the different nature of the phenomena.

Frankly, I can't agree with you. While I haven't studied in detail the analysis of how well minor league hitting stats project to the major leagues, I'm aware that it's been done and it's reliable to a certain degree. Even if MiLB defensive stats did project as well into major league defensive stats, what would it actually gain us?

I agree that we can come up with proper functions for adjusting MiL hitting stats to ML hitting stats on a normative basis. The only issue I have is when people confuse the normative with the individual. For a concrete example of the difference, just look at Matt Holliday. He came "this close" to being NL MVP, but his MiL hitting stats say he was nothing special. Pretty much everybody who looked at him as a MiL guy put him as a grade-C guy, but that's not how it turned out. So, I'm not saying that MiL hitting stats aren't useful, I'm just saying you need to be careful about what you do with them, that's all. And I'm certainly not saying that LH is gonna turn out to be a good hitter. I'm very confident that he won't. When it comes to his bat, I see no reason to disagree with Bill James projections about the guy.

What I was referrring to here is something different. There is something very basic about how D-stats are different. Both Hitting and Pitching stats are based on what happens in direct head-to-head competition between 2 guys. In contrast, D-stats are not based on that same kind of direct head-to-head competition. This makes a big diff in how much we can trust them to project well. A guy can hit (or pitch, either way, doesn't matter) well in the MiL's up to a certain level, and then just hit the wall. This happens all the time. Why? Because a guy reaches a level of competition where the guys he is competing directly against head-to-head (for a hitter, the group of pitchers he faces) are just too good for him. At each level up the ladder, the quality of that kind of competition gets stiffer and stiffer. Everybody knows this. This is exactly why we expect that most guys' ML numbers will be worse than their MiL numbers, especially at first. It doesn't happen to everybody, but for most guys we fully expect it. That's why Bill James projected LH to have a 2008 OPS of only .578 when his actual ML OPS was .662 and his MiL OPS was .624. It's not mainly because of small ML sample size, it's mainly because of the large MiL sample size, combined with the expectation that he'll do worse in the ML's than he did in the MiL's. While some of the screamers have claimed that LH will be lucky to have a.550 ML OPS, I have yet to hear anybody make any kind of rational argument about why Bill James projection of .578 is out of line. Since I don't know anybody smarter about these things than Bill James, I'm perfectly happy to defer to him about it. Now, if most people are willing to more-or-less trust Bill James about this (and I think they are), but if they still don't like the Total-Runs projection, that means that the part they don't like is the D-part of the equation. That has to be it, there's nothing else to not like. Either you think you're smarter than Bill James about how much LH is gonna suck at the plate, or else you think the conclusions about LH's D-contribution are not only on the high-side but are crazy-high. It comes down to just those two possibilities, because there is no third factor.

Now, assuming that the D-projections are based on good data and aren't confounded by things that the stats are blind to, then there is a good reason why it is dumb to trust the D-projections less than the hitting projections. The reason is that Fielding is basically a non-competitive activity. When a guy goes from A to AA to AAA to ML, the ground balls don't get harder to field. If anything, they get easier to field, simply because the fields keep getting better, and the ML fields are the best ones of all. The real competition is between hitter-and-pitcher, that's what determines which balls get hit where and how hard. But once a ball gets hit, it's just a hit ball, that's all. Once the ball is hit, the competition between the hitter and the pitcher is over (for that one AB), and now it's just a matter of seeing who can go get the dang ball and throw it to the right place. This means that, with D-stats, you just don't normally see the phenemona where a guy suddenly hits the wall. To the contrary, many guys become better and better fielders, just because of practice and experience and proper coaching. They can do this because they're not competing against stiffer opposition. A sharp grounder in the hole is a sharp grounder in the hole, nomatter what league it's hit in. This means that the main reason we have for projecting ML hitting and pitching stats to be significantly worse than MiL stats does not apply to fielding stats. However much you trust good MiL hitting stats to project to good ML performance, you should trust good MiL D-stats more than that.

So, if we use range numbers that are based on both ML and MiL play, why would anybody think they are somehow crazily inflated? If anything, they're gonna be way more trustworthy than the projections based on MiL hitting stats. In general, a good fielder's defensive play is more likely to improve over what it had been, simply because it's non-competitive in nature. Defense is not about level-of-competition in the same way that hitting and pitching are. Defense is mostly about individual ability and skill when it comes to going after the ball, getting it, and throwing it to the right guy. Succeeding at this is fundamentally different from succeeding at hitting or pitching because it is much more determined by the individual all-by-himself, and not how he performs against others. The fielder is not competing against another guy as much as he's competing against a hit ball. Once we realize this, then common sense says that good-MiL-D-stats are more trustworthy than good-MiL-hitting-stats. Individual change is more likely to trend upward rather than downward for some D-stats, while the reverse is generally true for both hitting and pitching stats whenever a guy moves up to a higher level. AFAIK, there's nothing about playing at a higher-level that would make D-stats get worse (beyond a kid being nervous at first and trying too hard). In contrast, there's lots of reason to expect that moving up to a higher-level will tax a kid's hitting (or pitching) performance.