Gallardo is Better Than Bush/Vargas.

[BraunsBoyfriend]

I think I like Vargas more than Bush this year so far. Vargas doesn't go very long and his WHIP is rather high...but he gets it done. He seems to become a different pitcher when runners get on...

 

Bush has had a rough going at it so far. Couple good games could change some things, and maybe his second half will be better.

 

But right now I think I keep Vargas as the #5 when Cappuano comes back:

 

Sheets

Cappuano

Suppan

Yo

Vargas

 

Or something to that effect.

[mothershipconnection]

But I am not sure that GMs only look at this year's stats when evaluating players

 

I agree that even the worst GM's are not the pus-brains we view them as. Still, lots of times they're forced to make a move, looking at only the immediate future. Vargas is the kind of guy who fits right into a desperate GM's mold.

[DougJones43]

I think I like Vargas more than Bush this year so far.

 

Yes, but I'll take Bush in the long run anyday.

 

 

...but he gets it done.

 

Recently he has. Throughout most of his career, he hasn't. That's probably a more reliable source for what to expect from him.

 

 

He seems to become a different pitcher when runners get on...

 

That most likely can be chalked up to a lot of luck. It's also unlikely that he can sustain that ability.

[berniebrewer4life]

Quote:

---

Well, it's specific to tonights game because we actually saw Yo go against big league hitting.

---

 

Let's remember that there is yet to be a book out on Yo(i would assume anyways). If he is still putting up these kinds of numbers 2 months from now, sure. But Let's not get carried away.

[1992casey]

Let's see if I can summarize the stats presented last week?

 

Vargas uses up a lot of pitches and carries a LOB percentage that he isn't likely to sustain.

 

Bush's BBs, Ks, etc. are fine. His BABIP indicates he's been unlucky. Dave's SLG pct. against sticks out as a yellow flag, but if you regress his BABIP to the mean, SLG 'normalizes', too.

[foamboy1]

He seems to become a different pitcher when runners get on...

 

That most likely can be chalked up to a lot of luck. It's also unlikely that he can sustain that ability.

 

I heard an interview with Yost a few weeks ago when the announcer asked about moving Bush to the pen and bringing up Gallardo or moving Villy to the rotation.

 

His answer was very interesting and insightful. He said that moving Bush to the pen would only be considered as a final option because it plays to Bush's weakness, which is pitching from the stretch.

 

Yost mentioned that Bush struggles to throw strikes from the stretch. This explains why he gives up the big innings. When he gets runners on base, his control is not as good and he is susceptible to the big inning.

 

The way Yost was talking about it, they have no doubt quantified statistically the difference in his control from the stretch, which is why he would only move him to the pen if it was a last resort.

 

I found it quite interesting....everything seemed to make sense then.

 

I guess my point is....there is an explanation....or partial explanation beyond luck.

[rluzinski]

Bush actually walks less guys with men on, so I'm not sure what that means. Batterers' isolated power against Bush also increases when runners are on. The only way that would jive with the "less control from the stretch" theory is that Bush is so against walking guys with runners on, that he grooves one down the middle instead of throwing ball 4? The data suggests something is going on, but there are sample size issues.

 

When do pitchers typically work from the stretch? Only when a base runner could steal a base, right? For example, if the bases are loaded, they don't even bother, correct? If that's true, there are probably less than 450 AB worth of "from the stretch" data for me to look at. Pretty hard to extract much from that.

 

Perhaps Yost is simply aware of some mechanical changes to Bush's delivery that he feels hurts Bush's accuracy when he pitches from the stretch. Otherwise, there probably isn't enough data to draw any firm conclusions from.

[JohnBriggs12]

Nobody's suggested it so for but eventually someone will so I'll throw this out there for discussion:

 

How about a 6 man rotation?

 

I'm not necessarily advocating it. If I did it, I'd have Sheets every 5th day, and the other 5 in rotation around that. That would keep innings down and arms fresh.

[hauser42]

Yo pitched great last night very few can argue that. But it was against the weak hitting Giants. FSN showed last night how bad they are in a bunch of hitting categories. Yes they have Bonds, but he is not Barry of 5 years ago, and who after that in the batting order scares you??

 

Yes wait and see when the NL has a book out on Yo and see what he does. Vargas has earned his spot in this rotation and should not be traded. If any thing I would send Bush down to the minors to work on his pitches as a starter if Yo is to stay in the rotation.

[bcf5575]

I think that Bush and Vargas are basically pitching for their starting jobs over their next 2 or 3 starts. Lets see who wants it more.

 

Whatever happens its good to have "problems" like this. At least it gives the Brewers some options:

 

1. Any of the 3 in the pen makes the pen better.

2. Using Bush or Vargas as trade bait could bring in another quality player to fill a need.

3. Having Bush and Vargas pitch so good that Gallardo has to go back to the minors is not a bad thing.

[Ennder]

Bush goes almost an inning deeper into games on average and has a more proven track record than Vargas, I know which one I'd send to the pen if one ends up there.

[GenoSeligPrieb]

"...there are probably less than 450 AB worth of "from the stretch" data for me to look at. Pretty hard to extract much from that.

 

Perhaps Yost is simply aware of some mechanical changes to Bush's delivery that he feels hurts Bush's accuracy when he pitches from the stretch. Otherwise, there probably isn't enough data to draw any firm conclusions from. "

 

A question, Russ.

 

Correct me if I'm off track here, but you dismissed (Sabremetrically? http://forum.brewerfan.net/images/smilies/wink.gif ) those <450 at bats as "pretty hard to extract much" from, rendering them, in your opinion at least, not statistically meaningful enough, right?

 

Even if there were 80 at bats to draw from, for a sample size, what's Ned supposed to do? Ignore what he sees, and wait until August to react to things, because then, the sample size becomes sizable enough to draw conclusions from?

 

Being a manager means you have to react to mistakes and bad trends you see. I understand I'm drawing this example to a logical extreme, but stay with me here, alright RL?

 

What if YoGa goes out there last night and gets torched in the first inning for hits (even, godforbid, unlucky ones!) to the first 8 batters? Do you do nothing, because YoGa's MLB sample size isn't big enough yet? Say, he ends Game One by giving up 10 earnies, and loses. Then, in his second start, you're Ned or Maddux and you see YoGa get b-slapped again. Game 3, same thing. Game 4? Bad! And over these 4 starts, he pitches 8 innings, with an ERA in double digits. Do you demote him back to AAA, or is that considered overreacting? The sample size sure as hell isn't large enough for the Sabremetric community (I'm starting to like that term) to embrace as a sure trend, so what would you do?

 

My rhetorical answer is, you've GOT to do something, and not just for Gallardo. If you're a manager or coach, and you see enough to convince you that Boosh struggles in the stretch position, you try to limit the exposure to that damage, right?

 

Say what you will about Ned or any other skipper who manages with some instinct, but all season long we fans who engage in debates with those of you card-carrying members of the...say it with me...Sabremetric Community...are sample-sized to within an inch of our lives, because "it's too early." Well, there has to come a time, be it April, May or June, when adjustments have to be made based on how a guy is performing in 2007. Last year, he may have been fine, 2005, too. But that was then, this is now.

 

Otherwise, it's paralysis by analysis as you sit and wait for your sample size to grow large enough to be considered significant and actionable. But by then, some bad juju can happen, and you can start falling in the standings, maybe even too far back to recover.

[recte44]

"He's just like you guys, he's a stud."

 

-Ned Yost to Fielder, Weeks, Hardy and Braun as Gallardo came off the field last night.

 

He isn't leaving the rotation anytime soon.

[hauser42]

He isn't leaving the rotation anytime soon.

 

If he gets rocked the next couple times he will be back in the minors.

[Fatter than Joey]

He isn't leaving the rotation anytime soon

 

The difference between Yo and Fielder/Braun/Weeks/Hardy -- Is that the Brewers currently have 6 guys that can do Yo's job. If Vargas and Bush start to improve, and Cappy gets healthy, it will be pretty hard to take these guys out of the rotation.

[endaround]

He either leaves the rotation soon or he leaves it in July. Again, he only has so many innings to pitch.

[rluzinski]

If you're a manager or coach, and you see enough to convince you that Boosh struggles in the stretch position, you try to limit the exposure to that damage, right?

 

The point I was driving across was that, if you ONLY have the raw data, making decisions based on a small sample is foolish, especially when a larger one is available. It's like flipping a coin 10 times, getting 7 heads and deciding that the coin flips heads more than tails. As I said, if Yost has other information (like scouting) at his fingertips that I'm not privy to (which is what I assume), fine. If he looked at the coin and saw something that might cause the odd results, that's additional information and it needs to be considered. But 7 heads in 10 flips in itself isn't enough evidence to conclude the unexpected.

 

Quote:

---

Say what you will about Ned or any other skipper who manages with some instinct, but all season long we fans who engage in debates with those of you card-carrying members of the...say it with me...Sabremetric Community...are sample-sized to within an inch of our lives, because "it's too early."

---

 

And how many times are those silly saber guys correct when they suggested that you are better off ignoring those early numbers? This year alone has countless examples. I read from a guy that I respect that Sheets could be replaced by a pitcher with a 4.5 ERA because that's what his current ERA was. We had someone suggest Suppan and Capuano were better pitchers than Sheets!

 

Tony Gwynn Jr. was supposed to be the everyday lead off hitter based almost exclusively on the results of 36 AB. Hart was over rated and shouldn't start every day, if at all. So many examples just from this year. It's no wonder you hear "small sample" until you want to explode.

 

What a typical baseball fan wants to be able do is skim the sports page in his Sunday newspaper in May and start making generalizations based on 50 IP or 150 AB. There's so much more information available... use it. If a person wants to formulate an educated opinion on a player, they should consider all the information available, whether it be stats from last year or current scouting information. They should also appreciate the inherent uncertainty of the stats they are looking at.

 

And you're always rhetorically asking, "when is a sample big enough?" but I don't know if you've ever really wanted to know the answer. I've shown many times the uncertainty associated with small samples but I don't think you aren't very interested in that kind of stuff. You don't have to be but then I think it's unfair to still question the validity of the concept. I didn't make those darn, pesky statistical laws, so that I could win arguments over people on the internet. They exist because they exist.

 

If you or anyone else wants to conclude a coin is flawed because it flipped 7 heads out of 10, don't be surprised when someone tells you that you are being hasty.

[Never Outhustled]

I think dealing Capuano has to be part of this brainstorming session. He has far more value to most teams, but he's only slightly more valuable to the Brewers. He is the one guy in this equation that has the value to bring an impact player in return.

 

Vargas has quality stuff, so its believable that he has improved. Bush is a quality pitcher despite his present 5.48 ERA. His career ERA is still just 4.47 in 530 innings, almost half of those in the AL. He's improved each month, and looks like he could be the guy we expect for the rest of the season.

 

Capuano has also been significantly worse in the 2nd half in each of his 3 Brewer seasons.Dealing him after his return has to be part of the discussion.

[GenoSeligPrieb]

"And you're always rhetorically asking, "when is a sample big enough?" but I don't know if you've ever really wanted to know the answer. I've shown many times the uncertainty associated with small samples but I don't think you aren't very interested in that kind of stuff."

 

Define "small" then. In April, we're told it's too early, and that's fine. But what if it's May and a heretofore effective pitcher is getting torched for Month #2?

 

"And how many times are those silly saber guys correct when they suggested that you are better off ignoring those early numbers? This year alone has countless examples."

 

So do I.

 

Roy Oswalt has been amazing for years now. But in 2007 so far, his walk rate is nearly double, and his ERA is up half a run. Do you just sit there and let it continue? Past stats say to ignore it, but he hasn't been getting past his surprising wild streak. I'm of course not sayiing Oswalt's got to be shipped to the pen or traded. But I'm sure Ol' Scrap Iron at least talked with him about 2007, and he and his pitching coach are monitoring his bullpen sessions between starts.

 

OK, I wish Oswalt's 2007 "struggles" were something we had to deal with. I chose him because he's a very soft example of what I'm trying to convey here.

 

But what if we're looking at hitters who had been terrific in the last few years, and who now stink in 2007 (Jermaine Dye, Nomar, JD Drew) and successful pitchers who suddenly find themselves struggling this year (Bronson Arroyo, Dontrelle Willis, Mariano Rivera)? Do you do nothing, out of fear you'll overreact?

 

I'd lay cash that their managers and coaches are working more closely with them...and reacting (hey, it could be they discover an injury the pitcher had been either downplaying or hiding, once they examine him, or as in Brian McCann's case, yesterday on BBTN, a flaw was discovered in Brian McCann's swing this year, and he started hitting last night (4 RBIs). What if Cox had reacted Sabremetrically? McCann's 2007 season would keep slipping into failure.

 

May and June are not too early to turn around someone who's struggling and salvage their season. Especially someone who hasn't established longer term success in the majors. You've GOT to at least react.

[Ennder]

Bush hasn't pitched poorly this year though, not only is it reacting to a small sample its reacting to a poor choice in metrics. He very well may be struggling in the stretch and of course they should do what they can to get his mechanics fixed, but I'd say his previous 500+ IP sample shows that he is capable of pitching from the stretch at the major league level.

[BillScottCanRake]

Roy Oswalt has been amazing for years now. But in 2007 so far, his walk rate is nearly double, and his ERA is up half a run. Do you just sit there and let it continue? Past stats say to ignore it, but he hasn't been getting past his surprising wild streak.

 

Past statistics don't say you should ignore anything. There is always a reason for a given performance producing stats. Where care needs to be applied is in determining why that's the case and what needs to be done about it.

 

In the case of Oswalt, simply knowing his walk rate has doubled is important but the "why" is just as (or more) important. Are hitters taking more pitches? Is he tipping his curve ball so hitters aren't swinging at it? Have his mechanics gotten out of whack? Statistical variation? Bad luck?

 

If something is broken and can be fixed, chances are decent that his walk rate will trend towards career norms. If he's simply strung together a series of bad outings, and there isn't anything to "fix", it seems probable that his outings will improve and his stats will move towards his career numbers. If there is a problem, either that is unidentified, or can't be fixed, he could well not revert towards career norms.

 

With someone like Corey Hart, his early season lack of success can be attributed largely to a sore wrist and a small sample. Getting healthy was as important as getting a larger sample to "prove" he wasn't a .220 hitter.

 

The thing is now one really "knows" what is going to happen in the future. The best one can do is use the data on hand to try to predict what will happen. The best data at hand is the data with the largest sample. It's not perfect, but it's the best there is with regard to trying to predict future performance.

 

But what if we're looking at hitters who had been terrific in the last few years, and who now stink in 2007 (Jermaine Dye, Nomar, JD Drew) and successful pitchers who suddenly find themselves struggling this year (Bronson Arroyo, Dontrelle Willis, Mariano Rivera)? Do you do nothing, out of fear you'll overreact?

 

The answer, as with almost all things, is "it depends". If they are pretty much the same player as they were in the past, it's entirely possible they are "slumping" or experiencing "bad luck" and with additional time their numbers will rebound towards their career norms. In that case, doing nothing could well be the right move.

 

However, if something is "wrong" and that something isn't fixed, doing nothing is stupid.

 

May and June are not too early to turn around someone who's struggling and salvage their season. Especially someone who hasn't established longer term success in the majors. You've GOT to at least react.

 

I don't think that's in dispute. It's the degree of the reaction that is.

[GenoSeligPrieb]

So, you'll allow this:

 

"He very well may be struggling in the stretch and of course they should do what they can to get his mechanics fixed, "

 

but you then say his 500+ IP overrides that?

 

"but I'd say his previous 500+ IP sample shows that he is capable of pitching from the stretch at the major league level. "

 

Is Ned wrong to react to this, or not?

 

He should, especially because this is only Dave Bush and not Roy Oswalt or Dontrelle Willis, who are much more likely to right the ship on their own.

 

Your mechanics can change a bit, whether you plan for them to, or not. You could have been hit in that balky left shin on a come-backer last week and you're not planting correctly. Maybe your contract's up this season and you want to try to reach 95 on the gun when more scouts are in the crowd. Maybe Adam "Pac-Man" Jones moved in, next door and you can't sleep nights and it's effecting your stamina. Maybe in 2007, your wife left you, to take up with the dashing Bobby Jenks, and you're pitching angry now? Past stats tell a lot, but perhaps not about what's going down these past few months.

 

Do you have stats, over those 500+ IP, which suggest Boosh CAN be effective under potential SB situations, i.e. from the stretch? I'd be interested to see them, because maybe he's rushing his delivery in 2007, and perhaps throwing too many straight fastballs, where he hadn't been, in years past. Could he have issues with Johnny Estrada's ability to gun down runners this year? Maybe he's not as confident or comfortable pitching to JE, and it's harming him. Either way, Boosh has been inconsistent in 2007, and I'm glad Ned's not passively waiting for the sampling size to get large enough before exploring what's been wrong.

[rluzinski]

Quote:

---

I'd lay cash that their managers and coaches are working more closely with them...and reacting (hey, it could be they discover an injury the pitcher had been either downplaying or hiding, once they examine him, or as in Brian McCann's case, yesterday on BBTN, a flaw was discovered in Brian McCann's swing this year, and he started hitting last night (4 RBIs). What if Cox had reacted Sabremetrically? McCann's 2007 season would keep slipping into failure.

---

 

I keep saying, "consider all the information available", whether it be scouting and statistical. If Yost knows that Bush is doing something mechanically that might cause him trouble in the stretch, that's additional information that needs to be considered. Is Coco fatigued or hurt? Additional information. Mangers should always be on the look out for a shift in performance, so they can investigate further. That doesn't cost the team anything. What might cost the team something is making a hasty personnel decision based only on a small sample. They need to weigh the cost of making the wrong decision with the probability of being correct.

 

And the idea isn't that you pass some magical threshold that makes an otherwise statistically insignificant sample, 100% rock on solid. The uncertainty of the data is proportional to the sample size. The larger the sample, the more accurate it is. Using the coin flip example, it's not surprising if we flip 6 out of 10 heads, right? It would only mildly surprising if we flipped 30 out of 50 heads. Now, flip 1,000 times and you are going to be pretty darn close to 500 heads, because you flipped it enough times to let the odds even themselves out. Here are some odds odds:

 

Flip 6 or more heads out of 10: 38%

Flip 30 or more out of 50: 10%

Flip 60 or more out of 100: 3%

Flip 600 or more out of 1000: ~0%

 

LINK

 

But for real baseball players, we have to contend with the uncertainty of their true, current skill level plus the above "coin flip" uncertainty. Even if your player were a robot that had a .300 BA skill level, you wouldn't really see that solely based on the results of 10 AB.

 

Of course, if your robot batter was also hacking at every first pitch, that's more information...

[homer]

Quote:

---

Bush actually walks less guys with men on, so I'm not sure what that means.

---

 

But couldn't it also mean that he's missing over the plate? So instead of inside corner on the black, he throws a meatball that gets crushed. I don't think the fact that his walk total is less with guys on tells the whole story about his control in that situation.

[GenoSeligPrieb]

I know WHY you're employing that coin toss paragon, Russ. I just don't agree that it applies to human beings pitching baseball games.

 

The coin doesn't get tired after being flipped 100 times. There's no stress on the coin's elbow. The coin can't be going through a divorce, a steroids investigation, nor can the coin be worried about possibly being outed as gay (unless it's one of those flamboyant ones from South America).

 

Under the statistical mindset, you seem (again, correct me if I'm wrong) to hold that a given player's career has 2,000 innings pitched in it, or 7,000 plate appearances, and that over time, it all evens out in the end, right?

 

But what I'm saying is that, yes, a player in his first 2-3 years is Version A of himself...the young, inexperienced pre-arby guy who gets hazed and is looking t stick in the bigs. That guy is a .260 hitter who attempts SBs but his power ain't there yet. Version B kicks in, and he's maturing. Version C has inked that first set-for-life contract, and is still at/near his peak. Version D is hurt. Maybe out for the year. His stats are skewed. Version E is a fat millionaire who's looking to land that one last big deal so he can buy off his mistress and relax when this gig is over. Version F is usually a Yankee, or a late-90's Diamondback, and is only in it for the cash.

 

I think every player's 1-2-3 year set of stats defines him in most situations for a while, but then he evolves, and you can't readily apply those same templates from his past, to his current probabilities, indefinitely.