THE BOOK--Playing The Percentages In Baseball

Gee, you guys really are picking on the Mariners today.
I agree that it was dumb to release Sexson. If only because some his at bats will now go to Miguel Cairo.
Concerning Vidro, someone like Dave Cameron might tell you that in this case, your projections may be off due to individual circumstances. For instance, last season, Vidro had lots of flukey infield singles although he is not known as a fast runner.

My current projections of the 3 are:

A.J. 261/310/393 (If he were traded to Seattle, I wanted to be ballpark neutral about this, his Chicago projection is better)

Vidro: 265/339/367 - close but 29 OBP beats 26 SLG

Sexson: 232/332/437 - Looks like the Mariners released the wrong washed up, defensively challenge stiff.

This is as good a place as any to post this. Deciding playing time and moves based only on an up to date projection is not a good way to manage a real team, IMO.

For example, what if ARod decided at the start of the season not to run hard on routine ground balls, because he might only lose 4 singles over the season (or whatever), and his projection will still be much better then any other Yankee option?

The real world aspect of baseball as an entertainment, and the necessity to keep players (who are already rich) motivated, seems to me to demand a greater focus on shorter-term results than our formulas would suggest is optimal.

David S., I don’t get your argument.  If A-Rod decides not to run out ground balls this year, then we take his projection and dock him however many singles or errors that is likely to cost him.  Just like if a player is playing hurt we take his projection and reduce it accordingly.  Etc.

VJ, everyone has a reason, after the fact, for why a player is doing well or poorly and why his projection is wrong - just for this guy and this one time, of course.  Yeah, right!  Just like everyone is in the best shape of their lives in spring training.  Just like Chipper was not really a .310 or .320 hitter a few weeks ago (he was a .340 or .350 hitter), even though he hasn’t gotten a hit or driven in a run since since the Momma’s and Papa’s had a hit.

If our projections never apply to anyone (they don’t apply to Vidro, to Sexson, to Edmonds, to Bonds, to Cristian Guzman, etc., etc.), who do they apply to?

Marcel, CHONE, and ZIPS all had Vidro projected around a .740 OPS heading into this year, thanks to a pretty significant regression from his flukey BABIP last year.  250 miserable at-bats at this age have to have knocked him down to a true talent .700 to .720 OPS. 

He can’t play the field.  Literally, his knees are in such bad shape that any time they tried to give him even occasional playing time as anything besides a DH, he’d come up lame. 

He can’t run at all.  He’s remarkably slow. 

So, we have a guy who can only play DH, is probably a true talent .720 OPS guy, and can’t run.  What possible use could he be to a major league team?

He’s not a good enough hitter to start for anyone at DH.  He’s not a good enough hitter that you’d ever want to pinch hit him for anyone besides a pitcher, but since he has to play in a league with a DH, that’s not an option.  He can’t pinch run.  He can’t play the field. 

What can he do? What is the point of any major league team carrying Jose Vidro?

This is why we mock the M’s for keeping him around.  At least replacement level players like Bloomquist can do some things to help a team win.  There’s literally no role Vidro can fill. 

None of this has anything to do with your post, of course.  This is just why we ride him so hard.  He’s totally useless.

The replacement-level for a DH is around a league average hitter.  If you are an average hitter (with no glove and no legs), you are a fill-in at DH and worth nothing.

While that doesn’t make Vidro one of the worst HITTERS in baseball, it does make him one of the most useless.  People I think are trying to say one thing, but are saying the other.  (That is, they are including replacement level, without actually saying it.)

I’ll take AJ. He can catch.

He can also kick your team’s trainer in the balls.

I mean, he’s not a useless player, but I get the idea that next to him, Barry Bonds and Roger Clemens are Mother Theresa and Mahatma Gandhi.

I agree that he is a useless player (as are several other MLB players).

My quibble is that he is not nearly the worst hitter in baseball.  Not nearly.  As many people make him out to be.

My principal point, which I have made 100 times already on this blog, is that I am sick of people quoting this year’s stats to make a (usually incorrect) point about what we should or should not do with a player.

I don’t expect most people to understand the concept behind a Marcel or other projection, but…

If I were King of the World, I would make an edict that everyone has to cite a player’s last 3 years’ stats when making any point about their talent or their future performance.

Now that Fangraphs has rolled out “last x Calendar years”, I think we should always quote at least last 2 Calendar years.

Tennis and golf are always working on last calendar year, and it works fine there.  No need to start everybody at zero on Apr 1.

Also, in my experience, well more often than not the person making the comments like “is he the worst dh in the last 20 years” isn’t smart enough to be talking about what the team should do in the future.  Rather, they are merely saying, in a round about way, over the last 3 months the player is the worst.

I think you need to be careful about sticking too dogmatically to a mechanical projection for a player with significant medical issues. As noted earlier, Vidro’s knees are degenerative, and absolutely shot. It’s legitimately possible that they’ve crossed a tipping point that would significantly lower his true talent level.

There’s no question that you need to consider all information.  If a guy has a known health issue, that should get included.  If the guy takes 5 seconds (as an illustration) to run to 1B, then we know his BABIP will be much lower than someone else.

Where do they think that we get our projection algorithms from?  From real life!  When REAL LIVE PLAYERS who are 33 years old hit .780 in their careers and then .577 in 3 months, they hit .740 (or whatever the number is) from then on in!  We don’t just make these projections up.  They are based on what players actually do, given their histories.

This is a point that doesn’t get made enough.  Regression isn’t used because math geeks are inherently pessimistic, regression is used because that makes you right more often.  Thanks, MGL.

I agree with Sky, that the point isn’t being made enough, strong enough, or well enough.

All we are doing is taking the thousands of statistical lines, empirical actual real-life data, and constructing some simple rules of thumb.  They are not out of the blue here.  We are simply looking at similar situations and saying: “This is what usually happens, so this is what we should prepare ourselves to happen.”

This is just like weather forecasting.

That first question from BTF is mine.

Your argument that there are 20 full-time players who are worse hitters than Vidro is fallacious, since I would assume they almost all play premium defensive positions.

I asked about DH.  He has zero defensive value, and zero baserunning ability.

Also, isn’t there a significant chance he’s just done?  I think projecting significant regression for guys as beat up and seemingly physically cooked as he is, is a poor bet.  I’ll also think his career numbers are meaningless, since all his good performance was 5+ years ago.

Thanks

Just to add, Marcel forecasts for 2008 are based on 2005-07 only, with 2005 representing about 20% of the weight (once you include regression toward the mean).

Your argument that there are 20 full-time players who are worse hitters than Vidro is fallacious, since I would assume they almost all play premium defensive positions.

I’m sorry, but that makes no sense at all.

The thread starts out with this:

Whom would you rather have in your lineup tomorrow?

not counting defense.

Vidro may be beat up and physically cooked, whereas the average player we base these projections on is not.  However…

It has already been mentioned several times (and another 3,748) times in various blogs, articles, etc., that a projection includes all pretinent information, including the current and past health of a player.

I am afraid that everyone who does poorly and is somewhat old is “cooked” and “done” and “beat up”.  Unfortunately, that is always after the fact.

I am not saying that Vidro is not “done” but I am skeptical.  Wouldn’t the Mariners organization be 1000 times better at assessing something like that than we are?  I realize that they are a bad organization, but let’s be realistic.  Would they be playing a player that they thought was truly a .600 OPS (or whatever) player? I don’t think so.  I think they think he is a .700 (or so) OPS player who has had bad luck this season (and they are probably right).  Not that a .700 OPS player should be playing DH, of course, but that is a far cry from a .600 player playing DH.

Arthur, you give me a reasonable over/under for what Vidro is going to hit for the remainder of the season (assuming he lasts for a while), as a “cooked” and “done” player, and then offer me a bet whereby I can take the over or under for an amount that would hurt a little if you lost.  Then we’ll see what you really think his true talent is!

I just started two polls to forecast Vidro over the remained of the season: one is for Mariner fans, and one is for non-Mariner fans.

Ideally, one would suspect that the Mariner fan is all in-tune with all things Vidro, and should come up with the better forecast.  Let’s see…

MGL,

I was strictly defending my observation about Vidro being one of the worst players to get regular time as a DH.  That was the crux of the BTF thread.  If Vidro was a decent defensive 2B, nobody would have been mocking the M’s for sticking with him.  So, your reply was not on point to the discussion you were criticizing.

If Vidro hit 270/330/400 from here on out, a bit better than your 700 OPS player, he’d still have no business as a starting DH.

There’s no reason to have a declining 700 OPS player with zero defensive or baserunning value on an MLB team.  At least TRY someone younger who might have a chance to improve.  Or a good defender/baserunner who gives you more roster flexibility.

Add on top of that a POSSIBILITY he may have deteriorated so much physically that he IS a 600 OPS player, it is unjustifiable to have him as a starting DH.

They just CUT an actual 700 OPS player.

Cheers

MGL said

“I am not saying that Vidro is not “done” but I am skeptical.  Wouldn’t the Mariners organization be 1000 times better at assessing something like that than we are?  I realize that they are a bad organization, but let’s be realistic.  Would they be playing a player that they thought was truly a .600 OPS (or whatever) player? I don’t think so.  I think they think he is a .700 (or so) OPS player who has had bad luck this season (and they are probably right).  Not that a .700 OPS player should be playing DH, of course, but that is a far cry from a .600 player playing DH.”

While I agree with your original point - saying that the Mariners organization can eyeball a player better than the fans is sort of having your cake and eating it too. The idea that the Mariner’s “know” that Vidro is a 750 OPS hitter and the peanut gallery “knows” he’s a 650 OPS hitter is equally falacious. The projection is the projection.

I realize you said they “know” he’s a 700 OPS hitter, but if that were true then they must be incompetent. There must be at least one or two guys on their AAA team that can hit better than that. They must think he’s better than 700 OPS.

I would say that many MLB teams are just as bad about ignoring projection as the average fan is. I suspect that they even include even more suspect subjective information. If Sexson pouts and Vidro sucks, but with quiet nobility, guess who gets the ax? (just a ‘for instance’ scenario - I have no idea who the better clubhouse guy is).

Regards

Arthur of course I have no quarrel with the fact that he is worthless (as a player with no defensive or baserunning value) and that he has no business playing on an MLB team, or basically anything else you said in the last post.  But please don’t tell me what argument I was making or what argument I was arguing against.

My argument was simply about using current season stats to argue a player’s true talent. That is it.  It had nothing to do with Vidro as a DH.  I used him as an example because lots of folks, even ones who consider themselves serious baseball fans, were using his .577 OPS so far this year, as evidence that he is one of the worst hitters of all time (yeah, I know someone said the worst DH of all time), the worst hitter in baseball this year, etc., not withstanding his positions as DH, although obviously that is an intrinsic part of the argument.  I could have used any one of 100’s of examples of players and comments about those players, regardless of their defensive positions.

There is a big difference between what a player is or what he is likely going to be, and what he has been, a (simple) concept which seems to be quite illusive to the average or even serious baseball fan and certainly almost all of the baseball writers and commentators.

Sexon, is more like a .750 or better player (again, projection-wise, and I have not heard anyone say that he is beat up or injured).  So sure, if nothing else, they would have gained (a lot) by releasing Vidro and keeping Sexson as DH.

MGL’s rule # 1,273:

When someone (or some organization) does something that seems inexplicably bad or wrong, you cannot fully lambaste it (well, you can if you want) until you hear their explanation.  Some (non-insignificant) percentage of the time, you will say, “Yeah, you’re right (or at least somewhat right).  I never thought of that.”

Funny, how now that you put it to a vote and made people put their money where there mouths are, he is suddenly a .700 hitters, which is a far cry from what people were saying before, notwithstanding his position as DH.

It usually works that way when you force people to think rather than just shoot their mouths off.  Having to put your “money with your mouth is” (I do not mean betting money literally) is the great equalizer.

One more thing:  READ the title of the thread and the first sentence in the text. The premise of the thread was very simple and had NOTHING to do with what position Vidro, AJ, or anyone else, plays!

Talk about red herrings!

MGL,

Yes, I read the title, and you are correct there.

BUT, you were explicitly replying to comments on a BTF thread that is all about Vidro as a DH.

Look at the stats quoted, worst performance at DH.

Your’re just making a different argument than what we were discussing, so quoting us was not the best way to frame your argument.  You’re right that people overweight recent performance, but we (the BTFers who slammed him) were right that he is a historically bad choice for DH.

That’s all I wanted to point out

Arthur, no problem.  We are on the same page, I think.  Thanks for the input.

Hey I’m as down with not using current season stats alone to argue a player’s worth as the next guy.

That said, using Vidro as the poster boy to prove the point to the masses may not be the best way to market the concept....

Basically picking Vidro to illustrate the concept is saying, “I know you’re saying he sucks based upon his first half, but hah! You’re using junk analysis. The reality is that he does in fact suck but just not as much as you think!

Terry, that is a good point, but on the other hand, people have an intuitive sense that players like Cabrera (and others like him that I can’t think of off the top of my head) will rebound.  I am not sure who the best examples are.  I do know, however, that when I see someone quoting someone’s season-to-date stats to make a point about their true talent, which is the case 99% of the time, I stop reading and throw the paper or the web page in the trash, because that is where it belongs.