How about this guy for SS?

[RShack]

First bullet - Agree

Second bullet - Agree

Third bullet - Agree

Fourth bullet - Of course

Last bullet - Sounds fair - I'd like to see you put it into action. I'll take it a step further and say it sounds like a good idea to do it the way I think you're going to do it.

Well, I need something to use for whatever is average. Dunno what's best, so lemme just make something up. This year, everybody kept saying how Miggi was at least average, so I'll pretend he was average. In 124 games he had 522 chances... but I wanna use 150 games, which translates into 631 chances. Then, I'll scale up his E's to 150 games and subtract them. I'll also scale up his DP's to 150 games and subtract half the DP's (good range has zip to do with relaying flips from the 2Bman to 1B). Which drops his extrapolated 631 chances down to 567 extrapolated outs. So, I'll use 567 outs as the baseline, just because it's the only number I've got.

As for the numbers you gave about range involving 350 balls, I don't know what's average. So I'll just take the average of the numbers for the guys you mentioned (just because that's all I've got), which turns out to be 338 (combining both numbers you gave). Comparing each guy you mentioned to that average, I get:

• LH: 113.3%
• Ryan: 101.8%
• Cedeno: 100.3%
• Hu: 93.5%
• Aybar: 91.1%

Using those, that tells us that each guy's D would be worth the following number of outs (compared to the 567 I'm using for average):

• LH: 642
• Ryan: 577
• Cedeno: 569
• Hu: 530
• Aybar: 517

If we use those, then LH at SS is worth the following extra outs compared to the these guys:

• Ryan: 65
• Cedeno: 74
• Hu: 112
• Aybar: 126

So, how to factor those outs into each guy's O? The idea is to pretend all these guys are the same D-wise, and instead give them credit for their D-outs by crediting them appropriately as times they got on base as hitters instead of the times they prevented the Bad Guys from getting on base. I'd prefer to divvy them up between BB's and hits, just because that seems right, and because that would impact both OBP and SLG, but I don't have enough info to do that. So let's just pretend they're all BB's (God knows LH needs help with his BB's anyway ;-), and then let's look only at OBP. If we leave everybody else's projected OBP alone, and use LH's extra D-outs to adjust his OBP upward by the appropriate amount, what do we get? Using 600 ABs as the basis for the OBP projections, and factoring LH's extra D-outs into his OBP calc as BB's, then LH's New Adjusted OBP vs. the projected OBP's you gave for the other guys is:

• Ryan=.319, LH=.390
• Cedeno=.341, LH=.405
• Hu=.306, LH-.469
• Aybar=.293, LH=.492

The OBP numbers for LH, above, are different compared to each other guy because his OBP is getting adjusted differently, based on how many more outs LH is expected to get than the guy who we're comparing him to. Another way to do it would be, rather than adjust LH's projected-OBP to each of the other guys' projected-OBP, we could adjust everybody's OBP (either upward or downward) based on the the number of D-outs each guy's range translates to vs. the average. If we do that (assuming 600 AB's once again), then we get the following "D-range-adjusted OBP" for each guy:

• LH: .418 (up from .282)
• Ryan: .337 (up from .319)
• Cedeno: .334 (down from .341)
• Hu: .239 (down from .306)
• Aybar: .202 (down from .293)

I had no idea how this was gonna turn out. I'm sure I made some bad assumption somewhere, but it wasn't intentional. (One obvious flaw is that I basically applied Miggi's fielding pct to everybody.) Regardless, LH comes out of this looking better than I thought he was gonna look. Now, before some people who don't bother to actually read things go completely berserk, this is just a back-of-the-envelope guess, and all the assumptions I used are explicitly mentioned above. I'm not saying it's right. I bet it's not right. But at this point we need constructive suggestions about how to do it better, not just people mindlessly screaming that LH is the worst player in baseball. (Just according to the back of my envelope, it looks like that title might belong to Aybar :-)

[El Gordo]

Well, I need something to use for whatever is average. Dunno what's best, so lemme just make something up. This year, everybody kept saying how Miggi was at least average, so I'll pretend he was average. In 124 games he had 522 chances... but I wanna use 150 games, which translates into 631 chances. Then, I'll scale up his E's to 150 games and subtract them. I'll also scale up his DP's to 150 games and subtract half the DP's (good range has zip to do with relaying flips from the 2Bman to 1B). Which drops his extrapolated 631 chances down to 567 extrapolated outs. So, I'll use 567 outs as the baseline, just because it's the only number I've got.

As for the numbers you gave about range involving 350 balls, I don't know what's average. So I'll just take the average of the numbers for the guys you mentioned (just because that's all I've got), which turns out to be 338. Comparing each guy you mentioned to that average, I get:

• LH: 113.3%
• Ryan: 101.8%
• Cedeno: 100.3%
• Hu: 93.5%
• Aybar: 91.1%

Using those, that tells us that each guy's D would be worth the following number of outs (compared to the 567 I'm using for average):

• LH: 642
• Ryan: 577
• Cedeno: 569
• Hu: 530
• Aybar: 517

If we use those, then LH at SS is worth the following extra outs compared to the these guys:

• Ryan: 65
• Cedeno: 74
• Hu: 112
• Aybar: 126

So, how to factor those outs into each guy's O? The idea is to pretend all these guys are the same D-wise, and instead give them credit for their D-outs by crediting them appropriately as times they got on base as hitters instead of the times they each prevented the Bad Guys from getting on base. I'd prefer to divvy them up between BB's and hits, just because that seems right, and because that would impact both OBP and SLG, but I don't have enough info to do that. So let's just pretend they're all BB's (God knows LH needs help with his BB's anyway ;-), and then let's look only at OBP. If we leave everybody else's projected OBP alone, and use LH's extra D-outs to adjust his OBP upward by the appropriate amount, what do we get? Using 600 ABs as the basis for the OBP projections, and factoring LH's extra D-outs into his OBP calc as BB's, then LH's New Adjusted OBP vs. each one of the OBP's you gave for the other guys would be:

• Ryan=.319, LH=.390
• Cedeno=.341, LH=.405
• Hu=.306, LH-.469
• Aybar=.293, LH=.492

The numbers for LH are different above because his OBP is getting adjusted differently compared to each other guy, based on how many more outs LH is expected to get than each of those guys. Another way to do it would be, instead of adjusting LH's OBP to the other guys' projected OBP, we could adjust everybody's OBP (either upward or downward) based on the outs their range numbers provide. If we do that, then we get the following "D-range-adjusted OBP" for each guy:

• LH: .418 (up from .282)
• Ryan: .337 (up from .319)
• Cedeno: .334 (down from .341)
• Hu: .239 (down from .306)
• Aybar: .202 (down from .293)

I had no idea how this was gonna turn out. I'm sure I made some bad assumption somewhere, but it wasn't intentional. (One obvious flaw is that I basically applied Miggi's fielding pct to everybody.) Regardless, LH comes out of this looking better than I thought he was gonna look. Now, before some people who don't bother to actually read things go completely berserk, this is just a back-of-the-envelope guess, and all the assumptions I used are explicitly mentioned above. I'm not saying it's right. In fact, I'll bet it's not right. But at this point we need constructive suggestions about how to do it better, not just people mindlessly screaming that LH is the worst player in baseball.

The first thing that comes to mind is that on defense most of those outs will actually be assists involving at least one other player and on offense the walks willl be the sole responsibility of the batter.

[RShack]

The first thing that comes to mind is that on defense most of those outs will actually be assists involving at least one other player and on offense the walks willl be the sole responsibility of the batter.

I agree that what you said is true. However, I don't see how it matters. If a SS has the range to go grab a ball up the middle that another SS would have just waved at, do you think the SS who got the ball doesn't deserve credit for the out just because the 1B-man caught his throw to nail the batter? I'm not trying to be difficult, I just don't see how it matters much.

[El Gordo]

I agree that what you said is true. However, I don't see how it matters. If a SS has the range to go grab a ball up the middle that another SS would have just waved at, do you think the SS who got the ball doesn't deserve credit for the out just because the 1B-man caught his throw to nail the batter? I'm not trying to be difficult, I just don't see how it matters much.

I don't think it matters all that much in terms of your experiment. I guess it just underscores the idea that defense is interdependent. Does the range of the 3B also contribute to the ability the of the SS to make plays ? How does the team's skill in positioning a player contribute to his defense? How many times does a good 1B convert a poor throw into a PO? Etc.

[RShack]

How good are the best computer simulations? Can you turn off all the video junk, punch in the different numbers we're using for SS hitting and for SS range, but keep everything else the same, and then tell it to play 162 games each way (or maybe 1,620,000 games, just to get sample size ;-) while you go to bed, and then see what it says in the morning?

[clarence]

Regardless of how all these stats turn out, I think it is awesome to read and I appreciate you guys for doing it. It is fun to look at numbers like these and wonder how they matter over the course of a season.

I wonder how many runs a season a shortstop with range saves just by stopping a ball from going to the outfield with a runner on second base, even if they don't get the out at first? I think there are way too many variables to consider when playing with defensive stats.

[El Gordo]

Okay. It may be time for me to admit that I may owe an apology to Luis Hernandez.

I did some research, and found this article which tells us how to turn the RZR data into +/- Plays for Shortstops. The article points us to another article which tells us how to convert those plays into defensive runs (assuming, of course, that you don't have a problem doing so).

I ran the numbers to determine how a group of shortstops compare to each other over the course of 400 balls hit into their zone (BIZ). Let me tell you how I came up with 400. MLB shortstops played 43,398 defensive innings and had 13,019 balls hit into the SS zone during the 2007 season. That works out to 432 BIZ per 1400 innings (roughly a full season for a team). I rounded down to 400 because players not named Cal Ripken don't typically play every inning.

The 2007 MLB SS leader based on this method was Troy Tulowitzki with 30 "defensive runs." Tulowitzki had 467 balls hit into his zone. The worst? Derek Jeter at -28 runs in 386 BIZ. Miguel Tejada finished with 4 runs in 317 BIZ. Our boy Luis Hernandez finished with 4 runs in 43 BIZ.

Next, I took every SS who played at least 125 innings, and extrapolated the numbers for each to 400 BIZ. Ramon Santiago led all SS with 43 runs, followed by Adam Everett with 40, and Mr. Hernandez with 39.

Based on these numbers, Luis Hernandez is a pretty darn good shorstop. Before we get too crazy, there's something we need to consider - sample size. We have to remember that Luis Hernandez only played 139 defensive innings in 2007. During my research, I found a comment that said something like this: "just as we can't tell how good someone is offensively based on 100 plate appearances, we also can't tell how good someone is defensively based on only 100 innings." By the way, the aforementioned Santiago only played 186 innings last year.

Next, I looked at offensive runs created. To do this I put together a list of shorstops who may or may not be available this offseason. I did primarily for the sake of comparisons, not because I want the Orioles to go after one or more of these players. I also determined RC by the very basic formula I mentioned earlier (AB*OBP*SLG). I gave everyone 500 ABs and used their Bill James projection where available. If a BJ projection was not available, I used the BBHQ projection, and in one case I used the player's MLE from last season. (NOTE: the reason for using the BJ projections and not the BBHQ projections is that the BJ projections are free.)

This chart shows the actual SS innings played in 2007, the offensive runs created, the defensive runs, and the total runs (RC+DR).

	Inn	RC	DR	Total RRouse		65	47	66	113Tejada		1068	86	5	91Gomez		71	57	34	91Hernandez	139	42	39	81	Wilson		1142	64	13	77Keppinger	390	80	-7	73	Bynum		66	58	11	69Cedeno		1397	66	2	68Fahey		94	42	24	66Uribe		4015	63	3	66Ryan		163	56	6	62Santiago	426	43	14	57	Hu		70	66	-19	47Aybar		154	51	-25	26

I cannot emphasize enough the important role that sample size plays in this data. It would be foolish for a team to make a decision based on such small sample sizes - fortunately, we don't have that responsibility. Bottom line - read as much or as little into this as you want.

For instance, I would just throw out the numbers for Rouse, Gomez, Bynum, Fahey, and Hu.

One item of clarification. For anyone on the above list who played SS in more than one of the last four seasons, I ran the defensive numbers for all affected seasons, and for just 2007 - I then went with the final tally that gave him the highest defensive run total (hey, it's my game).

The bottom line for me. If (and I think this is a big if) Luis Hernandez is anywhere near as good as his 2007 defensive numbers suggest, then he actually may not be a bad option at SS for the 2008 Orioles.

I think most of us agree that someone like Hu or even Cedeno would be a better long term option, but Hernandez might not be so bad after all for 2008. If I were being paid to make the decision I would rely heavily on the feedback of my scouts, manager, and coaching staff as to whether 2007 was a true indication of his defensive ability.

Thank you for your work. This was something I wanted to do but had no idea where to begin. Facinating stuff. So maybe DT might know what he's doing just a little?

[RShack]

Hey 1970... Just for grins, I compared the R numbers your method provided with the D-adjusted-OPB my guestimethod came up with. Your numbers are in the first column. My numbers are in the second column. The third column is what you get when you multiply my number by a constant value, which happens to be 197. (There is no big secret to 197, it's just that's the number that works best.) As you can see, our two results show not only the same ranking, but for most of these guys it also gives pretty much the same numerical result. My guestimethod gives Aybar way more value than yours does, but the other numbers are either pretty close or else exactly the same. So, it looks like we're using different means of finding more-or-less the same thing, whatever it is ;-)

Hernandez	81    .418    82Cedeno		68    .344    68Ryan            62    .337    66Hu		47    .239    47Aybar	        26    .202    40

[Sports Guy]

The problem is, as 1970 said, you are using a small sample size in this data and extrapolating it out.

It is like putting a lot of weight in his BA even though he had a 328 BABIP.

[RShack]

The problem is, as 1970 said, you are using a small sample size in this data and extrapolating it out.

It is like putting a lot of weight in his BA even though he had a 328 BABIP.

I'm not. I'm using Bill James' 2008 projections, and THT's Revised Zone Rating, (both as provided by 1970). Now, I suppose you can try to lecture Bill James and whoever the THT-boss is about sample size if you really want to. Whatever beef you have about the numbers is with them, not me.

And if you're not gonna use projections from those sources here, then I'm sure that you'll agree to not use them for anything else, right?

[Sports Guy]

I'm not. I'm using Bill James' 2008 projections, and THT's Revised Zone Rating, (both as provided by 1970). Now, I suppose you can try to lecture Bill James and whoever the THT-boss is about sample size if you really want to. Whatever beef you have about the numbers is with them, not me.

And if you're not gonna use projections from those sources here, then I'm sure that you'll agree to not use them for anything else, right?

The projections are based on a small sample size.

If he had a shaky 2 weeks of fielding, it hurts these stats...That is my point.

BTW, let's not pretend that people on here don't agree with the idea that LH is an above average defender...Most agree with that.

[RShack]

The projections are based on a small sample size.

If he had a shaky 2 weeks of fielding, it hurts these stats...That is my point.

So, did Bill James attach a special note to his projection, telling everybody it's not trustworthy? Did the THT guys? Was there an asterisk where they said, "Don't trust this projection! We published it, our own name is on it, but trusting it would be just like putting a lot of weight in LH's BA even though he had a 328 BABIP." Did they say that? Or is this just another case of you back-peddling again?

BTW, let's not pretend that people on here don't agree with the idea that LH is an above average defender...Most agree with that.

Never mind, you just answered that question. Let's not pretend the brouhaha has been about that, shall we? That's BS and you know it. The brouhaha has been about people yelling that LH is a terrible ballplayer and has no value that justifies him actually playing ML games at SS.

[TommyD4207]

Many on here know that my biggest (most likely only beef) I have with some of the people on here is how they treat many things as if it's a complete and total foregone conclusion and the file is closed with no chance of ever being re-opened.

Such has been the case with Luis Hernandez, in my opinion.

I am thankful that 1970 went through the time to explore the numbers as deeply as he did. He is a very well researched and intelligent poster and he brings a ton to this forum, in my opinion.

What many people on this board seem to forget is that the Orioles FO more than likely has access to numbers that the general public does not. I would be willing to bet a very, very large sum of money on that. I'm pretty darn sure that they have scouts record everything that happens in the minors, and that they have nerds (not an offensive term, being that I use BP numbers, as you'll see if you look at my posts) that crunch the numbers by hand.

Furthermore, like 1970 said, I wouldn't be surprised if the THT guys used MLE's in their analysis of Luis Hernandez.

The conclusion? Luis Hernandez is a helluva defensive SS in my opinion. His defense makes up for much, much more than was believed on here in the past. Much. Based on 1970's numbers, Hernandez is a mere 10 runs below Tejada's production. If Eli Eon had claimed that during his huge debates, he would've gotten slaughtered.

Note: I am not saying that 1970 is in any way promoting that Hernandez is only 10 runs below Tejada's production or that I believe that at all. I realize that the numbers are extrapolated. I am not sure, however, if the numbers include MLE's.

As a side note, the conclusion, once again, is that the general public does not have the info that the front office has. This is not a "these guys have jobs so therefore anything that is said on this forum is null and void" argument at all. This is a "they have more info on the situation" argument. Unless there's anyone on here that went to all of Hernandez's Bowie games and charted what he did and has access to advanced defensive formulas, it's very difficult to gauge how good he is, for the general public. I would not be surprised, however, if the front office has already crunched the numbers with the information they have, that we don't, and came to a conclusion.

Because I'm sort of cranky, and can't resist ( ) I will throw in there that the numbers do reveal that Hernandez played very, very well in his short time here. Now yes, yes, that doesn't mean that he's necessarily a great SS (which his numbers extrapolated would state) yeah, I know, got it. But it does show that in the limited time, his defense WAS great. It was. This does give some credibility to those that said "His defense was very good in his limited time" and takes away from the "It was above average I guess, nothing special" crowd.

Conclusion? As far as I can tell, I agree with 1970 for the most part. As far as the opposition to him starting, I really don't want us to give up anything that could be future value. When Hernandez is quite possibly better than many of the names thrown around here, it seems irresponsible, to me, to give up anything in the form of a prospect that could amount to anything later.

I like the "70% opposed" belief. My first and foremost hope is that we get a young SS back in a deal which make all this discussion a moot point.

[Pickles]

This is a really, really great thread. It should be required reading and it reminds me a lot of the thread tommy started this summer regarding Bako vs. House. Nobody here, with very few, if any, exceptions has more insight into the game than a Trembley or a Macphail. The simple fact is we don't have thier expertise nor thier resources. I've long felt LH starting next year was a non-issue, imo. If we obtained a true ss prospect in another trade, great. If we didn't then he was clearly the best of the in-house canidates and giving up ANYTHING in order to bring in a stopgag soloution, simply to keep LH from starting next year, was a foolish waste of resources.

The main objections seem to be that 1) There are literally "dozens" of players we can get for nothing who are better options than LH. I've never felt that was remotely true and this thread has only strengthed that opinion. And 2) A better team than the 08 O's could afford to carry a no-hit good-glove SS but because the rest of the team isn't as strong offensively as an 70 O's team or a 04 Red Sox team we can't afford that. That argument is the defintion of ingenous imo. So if the O's were good enough to compete in 08 at the other 8 positions they could carry a LH, or like player, but since it's a rebuilding year with the likely destination of 70 wins and a 4th or 5th place finish they can't? I fail to see the logic there.

Anyway, great thread. I almost chimed in on another thread that had deginerated to discussing the merits of rshack as a poster but, imo, this thread serves to show exactly why he is so valuable to this board. He doesn't simply swallow the "company line" and let it go at that. He wants to explore issues more deeply and if that ruffles some peoples feathers, particularly long-tenured posters who don't like their opinions to be questioned, the more power to him.

[Sports Guy]

Never mind, you just answered that question. Let's not pretend the brouhaha has been about that, shall we? That's BS and you know it. The brouhaha has been about people yelling that LH is a terrible ballplayer and has no value that justifies him actually playing ML games at SS.

I have always said he is an above average SS....I just don't pretend that he is Jesus Christ with a glove.

And stop putting words into my mouth.