Steve Trachsel Traded

[Eight]

I'll be the first to admit that he pitched better than I expected. And yes, he did earn some of my respect. (I personally stayed away from calling him names such as Trashel [it irritates me when people do that], and the like).

Yes, he did do exactly what was asked of him - and his 4.48 ERA is an indicator that he did it well.

The question is do you believe 4.48 is a true indicator of what's left in his tank? Personally, I don't. Some people like to make fun of the peripherals, and roll their eyes at stats such as FIP, xFIP, and DIPS, but I will rely on them over flawed stats such as ERA and W/L record any day of the week (and be right more often than I'm wrong).

I'm a big believer in the DIPS family of metrics. I don't think it's reasonable to expect him to put up a sub 4.50 ERA going forward and have long been dreading seeing his performance fall of a cliff.

That being said, Trachsel has outperformed by at least .50 runs in each of the past four seasons. He has outperformed his xFIP doing so my an average of almost 1.00 runs per game.

To try to find some reason for this, I looked at his splits and found that his OPS against was over 60 points better with men on than with the bases empty over his career (.797 in 6103 PA to .735 in 4423 PA).

The only other pitcher in the top 15 in innings among active pitchers to have a lower OPS against with men on than with the bases empty is Livan Hernandez, and even his differential (.766 to .765) is almost nonexistent. Pedro Martinez was the closest otherwise with a .004 difference between his OPS with the bases empty and with men on base.

Trachsel w/bases empty: .274/.334/.463 -14.3% K% - 9.20% XBH% - 7.75% UIBB% - BABIP .290

Trachsel w/men on base: .257/.331/.404 -15.6% K% - 7.20% XBH% - 8.45% UIBB% - BABIP .282

So he strikes out more and gives up fewer hits, particularly extra base hits while walking more batters. My questions are:

1) Is this likely to be a meaningful skill - either that he pitches particularly well from the stretch or that he pitches to limit big hits by working the corners or both. I would think it is meaningful given the .060 OPS difference over 4500+ PA in both categories, but if it is I wonder why no other pitchers that I looked at show any sort of a difference. Many of them had a similar difference in the other direction.

2) How much could this explain the differences between his ERA and his theoretical/peripheral ERAs over the course of his career?

[scOtt]

I don't know if this has been posted anywhere... Roch says Moore is starting tonight at 3B. We'll get a chance to see him right away.

[scOtt]

I don't know if this has been posted anywhere... Roch says Moore is starting tonight at 3B. We'll get a chance to see him right away.

Actually it was Zrebiec.

Moore and Cherry are expected to be in uniform when rosters are eligible to expand tonight. Orioles manager Dave Trembley said he has told Moore that he'll start at third base tonight in place of Melvin Mora.

[mweb]

This should be required re-reading for a bunch of you guys. Hopefully, you know who you are.

The part that begins with "They didn't just give him away...", yeah that is required reading for many including me, the part before that is called luck.

[davearm]

The question is do you believe 4.48 is a true indicator of what's left in his tank? Personally, I don't. Some people like to make fun of the peripherals, and roll their eyes at stats such as FIP, xFIP, and DIPS, but I will rely on them over flawed stats such as ERA and W/L record any day of the week (and be right more often than I'm wrong).

Just speaking for myself, FIP, xFIP, DIPS, BABIP, K/9, K/BB and all the rest are great to look at when projecting future production, because they have better predictive value than ERA and the like.

However I have a much harder time using those peripherals in backward-looking analyses. Past production is measured by how many runs you allowed to score, above all else.

The bottom line is that the statistics of choice depend entirely on the form of the question -- are you asking who was better last year, or who will be better next year?

[Ohfan67]

Just speaking for myself, FIP, xFIP, DIPS, BABIP, K/9, K/BB and all the rest are great to look at when projecting future production, because they have better predictive value than ERA and the like.

However I have a much harder time using those peripherals in backward-looking analyses. Past production is measured by how many runs you allowed to score, above all else.

The bottom line is that the statistics of choice depend entirely on the form of the question -- are you asking who was better last year, or who will be better next year?

Nicely put.

[davearm]

The problem is that the pitcher isn't alone in determining how many runs are charged to his record. Trachsel and his defense teamed together with other factors (a.k.a. luck) to post a 4.48 ERA with the Orioles. Wandy Rodriguez of Houston has a 4.49 ERA. Do you really think Trachsel is as good a pitcher as Rodriguez? I know I certainly don't. Trachsel largely had two things in his favor over Rodriguez. His defense apparently played better behind him (.725 DER vs .704), and he had a better LOB% (74.6% vs 66.0%). [Keeping in mind that LOB% doesn't appear to be a stat that pitchers have a lot of control over.]

People really need to get away from this idea that DIPS/FIP etc are useful strictly for predicting future performance - they are more valuable than that. Make no mistake, stats such as K-Rate are real time stats.

I never said that.

[davearm]

You sure as heck keep implying it.

By the way, if you want to play that game, I never said that you said it.

You're inferring something different than what I'm implying.

DIPS/FIP and the rest are fine as secondary indicators of past performance, but they take a backseat to the bottom line -- runs allowed.

FIP would go bonkers for a guy with a line like...

5 IP, 5 ER, 8 H, 0 HR, 8 K, 1 BB.

Meanwhile FIP would snicker at a guy with a line like...

7 IP, 2 ER, 4 H, 1 HR, 2 K, 2 BB.

The reason? FIP doesn't care about hits or runs. Ks, BBs, and HRs are all that FIP cares about.

I'm sorry, but defense, luck, variance etc. etc. just doesn't come anywhere close to tilting the scales enough for me to view the former outing as superior to the latter. It's borderline ludicrous, actually.

Peripherals: great tools for predicting the future, but severely lacking as tools for assessing the past (IMO).

[mweb]

You're inferring something different than what I'm implying.

DIPS/FIP and the rest are fine as secondary indicators of past performance, but they take a backseat to the bottom line -- runs allowed.

FIP would go bonkers for a guy with a line like...

5 IP, 5 ER, 8 H, 0 HR, 8 K, 1 BB.

Meanwhile FIP would snicker at a guy with a line like...

7 IP, 2 ER, 4 H, 1 HR, 2 K, 2 BB.

The reason? FIP doesn't care about hits or runs. Ks, BBs, and HRs are all that FIP cares about.

I'm sorry, but defense, luck, variance etc. etc. just doesn't come anywhere close to tilting the scales enough for me to view the former outing as superior to the latter. It's borderline ludicrous, actually.

Peripherals: great tools for predicting the future, but severely lacking as tools for assessing the past (IMO).

For a game or two, sure, stats like FIP aren't very meaningful, but over a course of an entire season, they should be strongly considered imo.

No one can convince me that Trax hasn't been very lucky this year, and that's based on stats like FIP, command rate, k rate, etc.

[davearm]

No offense, but anyone who would try to use FIP for one or two appearances doesn't get it to begin with.

By the way, what does ERA does us about either of those guys?

As for your first part. The pitcher isn't solely responsible for the number of runs he allows, so how do you personally account for that?

Then multiply all of those numbers by 30 and call it a season. The point still holds -- FIP still loves the first guy and scoffs at the second guy.

The pitcher isn't solely responsible for the number of runs he allows, making ERA flawed. I get that. But FIP disavows the pitcher of ALL responsibility for runs allowed. That sure as heck can't be right, either. It's asinine, actually.

As I've demonstrated, FIP way overcorrects by completely ignoring runs allowed, which in reality is the only thing that does matter, thus giving us something that's even less valuable than ERA for backward-looking purposes. What FIP does is tantamount to throwing the baby out with the bath water, IMO.

What I'd imagine doing is starting out with a guy's actual ER, and making some adjustments to tweak ER up or down to address things like defense and luck, then multiplying the normalized ER by 9 and dividing by IP to get a normalized ERA. On an intuitive level, that's what makes the most sense.

Absent those adjustment mechanisms though, ERA is the lesser of two evils.

[Sports Guy]

If a pitcher K's 8, walks none and gives up no homers, i like his long term chances better than someone who is giving up more homers, striking out next to no one and walking more than his strikes out.

FIP has no use for 1 game...The use for it is over the course of a season and predicting the future based off of that.

[davearm]

I think the flaw in your thinking is that FIP disavows the pitcher of all responsibility for runs allowed. That absolutely isn't true.

That absolutely *is* true.

Runs allowed is nowhere to be found in the calculation of FIP.

Therefore FIP believes that runs allowed has absolutely nothing to do with a pitcher's effectiveness. Zero. Nada. Nein.

It's that simple. And that ludicrous (IMO).

[davearm]

If a pitcher K's 8, walks none and gives up no homers, i like his long term chances better than someone who is giving up more homers, striking out next to no one and walking more than his strikes out.

FIP has no use for 1 game...The use for it is over the course of a season and predicting the future based off of that.

We're talking about how best to determine who was more effective in the past, not who we expect to be more effective in the future.

[Lucky Jim]

You're right, walks and home runs aren't related to runs allowed at all.

Sounds to me like policy debates involving statistical lives. lol.

You're not going to convince Dave to change his position. He's pretty much decided on this point. For him...

...a walk that does not lead to a run is a fact that is incontrovertible. In other words, a walk may be related to a run hypothetically, but until that run crosses the plate it is irrelevant. In his eye, I think, a pitcher can control the number of walks (or other base-runners) who cross the plate.

Am I wrong, Dave?

You're just against constructing hypothetical runs when you have actual runs at your disposal?

I don't agree with him. But I can't see you changing his mind.

[davearm]

You're right, walks and home runs aren't related to runs allowed at all.

Just give it up already.

All else being equal, a guy's FIP number is going to be the same regardless of whether his statline shows 1 earned run or 1,000.

That's undeniable and incontrovertable fact.

This discussion is completely beyond hope if you cannot even concede that which is undeniable and incontrovertable.