Overrated = Brian Roberts?

[Baltimoron]

Migrant Redbird,

Read this.

[Pruke]

You didn't answer my question.

If Roberts had been snubbed, which guy would the Cubs get to keep: Gallagher, Marshall, or Cedeno?

None.

Why would they?

If the Cubs don't need him, they should make a low ball offer and everyone can move on.

I guess your point is that because he was an all-star - but not necessarily because he was a pretty darn good 2B - the Cubs have extended themselves with that offer? I don't see that.

Perhaps this will clarify my position

- Remove the all-star tag from Roberts

- The O's can keep his statisical production and glove - not to mention the positive marketing and goodwill be brings or look to another team that values him beyond his all-star tag.

- The Cubs can keep all of them, hope Gallagher becomes something special, suffer, hope Marshall improves and Cedeno becomes something period and we can all move on.

[Frobby]

I'm not sure UZR is right or wrong on Roberts, I'm saying I think you don't know.

When has that ever stopped me?

Actually, the idea that Roberts is 3 runs below average, or .3, doesn't really bug me. The idea that he is 3rd worst in the league just seems very wrong on a visceral level.

[davearm]

None.

Why would they?

If the Cubs don't need him, they should make a low ball offer and everyone can move on.

I guess your point is that because he was an all-star - but not necessarily because he was a pretty darn good 2B - the Cubs have extended themselves with that offer? I don't see that.

Perhaps this will clarify my position

- Remove the all-star tag from Roberts

- The O's can keep his statisical production and glove - not to mention the positive marketing and goodwill be brings or look to another team that values him beyond his all-star tag.

- The Cubs can keep all of them, hope Gallagher becomes something special, suffer, hope Marshall improves and Cedeno becomes something period and we can all move on.

Thank you.

By (correctly) answering "none," you support my position that having been named to the All-Star team doesn't enhance Roberts' (or anyone's) trade value.

[davearm]

Yeh, I agree. I definitely believe there are people on this board who would assign some amount of value to things like:

- being selected previously to an All Star team

- being an MVP, Cy Young, Gold Glove, etc... winner previously

- meeting certain milestones like winning 20 games, 200 hits in a season, hitting 300, etc...

- being a "type" of hitter like lead off, clean up, etc....

I realize it seem to be in vogue in this thread to rip Dave about this issue, but there are pretty clearly two different trains of thought on the other side of this debate. Those of you who don't believe that a classification in the categories I mentioned above increases today's value and those of you who do. Those of you who don't are getting pretty upset with him and insinuating he is being a butthead about this argument, but I can definitely see why he believes that some of the early posts in this argument were arguing that being selected as an All Star increases value today. And for the record, I don't believe he is painting those of you who do feel that way and those of you who don't with the same paintbrush.

I actually think we're all just a bit testy because of the current state of affairs. Lets all sing "Cum by Yah", have a Bud, and say "I Love You, Man"!!!

Well said.

Throughout this thread I keep having flashbacks to Moneyball, and how Billy Beane and Paul DePodesta made such a killing taking advantage of other teams' willingness to value things that just don't matter, simply because those other teams didn't realize what they thought mattered, really doesn't.

[Moose Milligan]

I find it odd that blueberryale hasn't shown up yet, I have the feeling she's off somewhere fighting urban sprawl.

So on her behalf, here's an obligatory post about how great Brian Roberts is because of all the community service he does and how that shouldn't go overlooked:

Brian Roberts is a community leader and that shouldn't go overlooked when it comes to talking about him being overvalued. If anything, he's undervalued because he's such a good guy.

[Lucky Jim]

I find it odd that blueberryale hasn't shown up yet, I have the feeling she's off somewhere fighting urban sprawl.

So on her behalf, here's an obligatory post about how great Brian Roberts is because of all the community service he does and how that shouldn't go overlooked:

Brian Roberts is a community leader and that shouldn't go overlooked when it comes to talking about him being overvalued. If anything, he's undervalued because he's such a good guy.

With an aesthetically pleasing fanny.

[Moose Milligan]

With an aesthetically pleasing fanny.

And cute dimples.

[Baltimoron]

That's only because you're right ;-)

Here's one way to think about it:

• OBP is like RZR (Revised Zone Rating = how did he do at balls he was s'posed to get, aka how did he do with his opportunities to make a play).
• SLG is like OOZ (Out of Zone = how many extra balls did he get that he wasn't supposed to, but he made the play anyway.)
• OPS is like the Mystery Number we don't have for D.
• OPS+ is like "+/- Plays" (kinda-sorta)
• We have a rough formula for RC (Runs Created) based on multiplying AB and OBP and SLG.
• We have a rough formula for DR (Defensive Runs) that's based on +/- Plays.
• That's like calc'ing RC based on OPS+ without even knowing what OPS is *and* not even caring about AB's. (How crazy is that?)

Lots of people think that RZR and OOZ are the right inputs into the Mystery Number we want that's analogous to OPS. I don't know enough to know if they really are the right inputs, but it sounds plausible at first blush. If they aren't the right numbers, I figure that they each are trying to get at something that would be the right numbers if we were doing it right. But let's not worry about that. Let's just assume for the moment that they are the right numbers. If RZR and OOZ are the right numbers, the problem is that we don't know what to do with them to get the Mystery Number we want. What we need is some easy way to take RZR and OOZ and come up with the equivalent of OPS (maybe "ZPO" for "Zone plus Out-of-zone"). Then, we need a simple formula for turning that into DR. Right now, we don't have either one of those things.

Part of the problem is because it's hard. Another big part of the problem is that D-stats involve secret obscure information. You don't need to wait for somebody to tell you what OBP and SLG are. They're directly extracted from any simple log of simple game events. There's no interpretation required. Everything you need is crystal clear. In contrast, neither RZR nor OOZ are like that. You can't get them from a simple log of game events. They require somebody to watch what happens to every ball-in-play and make notations about how it's hit and where it goes. They require some interpretation of game events, not just a simple log of game events. So, we're dependent on middlemen making judgments.

It's fine with me that we're dependent on human judgment, simply because people can be trained to reliably make correct judgments, we know that's true. The problem is that that the observations/notations about each play that are the basis for RZR and OOZ are not things that show up in Box Scores. So, nobody can just calculate RZR and OOZ. Instead, you have to wait for somebody to tell you. The people who know only tell you whenever they feel like it. And they don't even tell you the data points they're based on. So, it's like being told what OBP and SLG numbers are for a given guy without knowing how it happened. It's not like the announcer says, "That shoulda been a double into the gap, but Whatshisname grabbed it, and mark that as an OOZ play for him!" With RZR and OOZ, we don't even know what they are when we see them. To make things worse, AFAIK "+/- Plays" is propriety information that stays secret except for the Top-10 lists they throw to us.

Bottom line: None of the D-stats can be extracted from simple game events. So, we've got what amounts to be the D-equivalents of OBP and SLG as mystery numbers based on events that we can't just see evolve during the game like we can with OBP and SLG... plus, we've got DR being calc'd as a secret score based on secret data, so we have no way to construct it. This means that, whatever fragments of truth we have in D-stats are either inaccessible, mysterious and dispensed only periodically, or else they are outright trade secrets. Meanwhile, the basic numbers and formula we need just aren't there. Maybe somebody knows what they are, but if they do, they ain't telling us. What they're doing is treating it like the Coca Cola formula and selling the resulting information. They're not giving it away. So, we may have some actual stat-truth happening here, but if we do, it's not in the public domain. So we don't really know. (BTW, this is *not* how science is supposed to work.)

ps: Why isn't OPS really OTS? It's not OBP *plus* SLG, it's OBP *times* SLG.

There is so much wrong with this post I'm not sure where to begin, so I won't.

Mr science and fair and reliable research,

Instead of defending an agenda, why don't you try to educate yourself and present what you have learned, while being careful not to tread into the "i have no idea what I'm talking about but maybe someone won't call me on it"

[RShack]

There is so much wrong with this post I'm not sure where to begin, so I won't.

Mr science and fair and reliable research,

Instead of defending an agenda, why don't you try to educate yourself and present what you have learned, while being careful not to tread into the "i have no idea what I'm talking about but maybe someone won't call me on it"

Great, another quality post that does *absolutely nothing* to help everybody understand D-stats better. Look, I took the trouble to explain how I look at it. It's fine if you disagree, but where's the point and effort in cheap sniping? Why don't you actually *add something constructive*, other than just links to articles that often include stuff some folks don't understand. If you're so up on things, and if I'm so wrong about them (which I could be), then howsabout if you help boil-it-down a little bit, so that people can understand the important parts. People have a hard time making sense of D-stats, and I don't think it's their fault either. I think it's completely normal to be confused by this stuff.

I explained what I think is one good way to help make sense of the frequently-referenced parts of confusing and if-fy D-stats. If you disagree, then why don't you provide a better one instead of empty sniping. In particular, if I'm so wrong, then howsabout if you:

• Tell us how to take a simple log of game events and calc the RZR and OOZ values (or the comparable numbers of your choice) for each player without us waiting for middlemen to look at how everybody and his brother does, and then giving us the resulting numbers whenever they feel like it. Tell us how we can see how these thing evolve during a game, or from looking at a box score, just like we can do with OBP and SLG. I say you can't, am I wrong?
• Tell us how to take RZR and OOZ (or the comparable numbers of your choice) and give us a simple quick-and-dirty formula for turning that into something that's meaningful and more-or-less accurate for each individual guy, providing a number that's analogous to OPS but for D. I say you can't, am I wrong?
• Tell us how to turn RZR and OOZ, or that OPS-like D-number we don't even have, into a DR number that is meaningful and more-or-less accurate for each individual guy *based on what he actually did* in absolute and not relative terms. I say you can't, am I wrong?
• Tell us *any way at all* to take publicly available data and easily figure out how much value an individual guy's D has, so we can compare it to his offense. I'm not asking for some "less-worse than it used to be" way to do it for ballplayers-as-a-whole. I'm asking for some more-or-less fair and accurate way to do it for each individual guy. I say you can't, am I wrong?

If you can do those things, then I will happily say I'm wrong. In addition, I will be thankful for your gift of providing us with something that we can actually use. I've asked you this before, and I'm still waiting. I know you can post links, take cheap shots, and use big fonts. But instead of doing just those things, why don't you provide meaningful answers to these basic questions to help people understand better? All we have so far is you claiming I'm wrong, but you can't be troubled to say how and why. Exactly what did I say that is so wrong? Or are we supposed to just believe it's wrong just because you made an empty claim?

As for "defending an agenda", exactly what does that mean? Here's my agenda: I think things like RZR, OOZ, and +/- Plays are on the right track, but they're not nearly good enough if nobody knows how to use them to figure out the DR value of each individual guy with some accuracy. At least some part of the problem is that things like +/- Plays and other proprietary things (like WARP's) are figured out in secret, so nobody can test to see how right they are or aren't, i.e., there is no opportunity for the peer-review process to help things improve. AFAIK, they won't tell us how reliable they are, and neither will they provide a transparent process that lets other people figure out how reliable they are. So, a lot of it boils down to "just trust us about something we're selling for a profit". That's not how science is supposed to work. All I want is to be able to see how good these various things actually are without all the veils of secrecy and hidden magic. Since when is it a bad idea to want to know how trustworthy stat-claims are? Since when is it a bad idea to want to know how many runs a guy saves with his glove that's based on stuff we can trust? Exactly what about this "agenda" do you think is so wrong?

[Lucky Jim]

And cute dimples.

The problem with this is that we haven't developed a equation/system through which we can quantify these attributes and so I think we tend to discount them. I blame it all on the sabermaticians who obviously need to look at more fannies and fewer numbers. Or at least come up with an equation that factors in a tight ass and white teeth to a player's overall value.

[cindyluvsbrady]

I find it odd that blueberryale hasn't shown up yet, I have the feeling she's off somewhere fighting urban sprawl.

So on her behalf, here's an obligatory post about how great Brian Roberts is because of all the community service he does and how that shouldn't go overlooked:

Brian Roberts is a community leader and that shouldn't go overlooked when it comes to talking about him being overvalued. If anything, he's undervalued because he's such a good guy.

We agree:) He is priceless!

[Sports Guy]

Are we still discussing defensive stats with Rshack?

Rshack, i have some advice for you(not trying to be smart ass either).....Write to the guys at BP or Tom Tango or Bill James or whatever.

See if they can give you the answers you are looking for.

[rammyfanny]

Anyone know why my "Rumor with the Cubs" thread was deleted? There's rumors all over this site, and the guy who wrote it had reported that we signed Chad Fox a few hours before it was reported first by Levine, so I think he knows someone. I know the Fox thing wasn't anything big, but still.

[Pruke]

Thank you.

By (correctly) answering "none," you support my position that having been named to the All-Star team doesn't enhance Roberts' (or anyone's) trade value.

Great - trades should be based on merits. I would happily trade for Albert Pujols even if he never made it to an all-star game. I don't believe St Louis would expect anything less in a trade.