Gattis Traded To Astros (pg. 13)

[gilesfan]

I gave you ERA+ as well. I think a pitcher with 2 consecutive years with an ERA right around 3.00 and ERA+ around 121 is pretty likely to repeat those numbers. Especially a pitcher as talented as JT is.

You said they weren't going to finish top 20. Well they have certainly been top 20 over the last seasons. But I guess since FIP says so it must be true.....

I don't think you come close to understanding when you say things like "I gave you ERA+" as if it has any predictive measure to it.

[Heyward]

I don't think you come close to understanding when you say things like "I gave you ERA+" as if it has any predictive measure to it.

So what do you use to determine how good pitchers are then, SP?

[Carp]

That doesn't mean they will be top 20. They have solid peripherals, but accounting for the decline in defense in 2014 and the predictive nature of other statistics, it's likely wishful thinking that both will be able to perform as top 20 starters.

Here are the Steamer predictions for both; just as a guide:

Teheran 3.80 ERA, 4.10 FIP.
Wood 3.60 ERA, 3.50 FIP

Steamer projects him to have the highest HR rate of his career by a good margin and give up a career high BB%. How exactly is the defense going to affect that?

[thethe]

I'd love to see how accurate the steamer projections were last year. These predictive tools are close to useless ofr young players who haven't logged significant major league innings.

[Carp]

I don't think you come close to understanding when you say things like "I gave you ERA+" as if it has any predictive measure to it.

I'll take 400 actual innings pitched in the last 2 seasons rather than some arbitrary projection system.

He's easily been a top 20 pitcher over 2 the last years, and now because we lose Heyward, he's going to explode to 3.80 ERA..... during his prime.....

Yet somehow, to you, that is a MORE likely than him simply continuing to pitch as well as he has over the last 2 years.

[gilesfan]

Steamer projects him to have the highest HR rate of his career by a good margin and give up a career high BB%. How exactly is the defense going to affect that?

Defense won't affect that part. Highest HR rate because he's an extreme fly ball pitcher that had an unusually low HR/FB rate last year (8.1%) .

[gilesfan]

I'd love to see how accurate the steamer projections were last year. These predictive tools are close to useless ofr young players who haven't logged significant major league innings.

http://www.hardballtimes.com/evaluat...ction-systems/

Even though Steamer does not take age into account, their projections do a wonderful job with young players. Stating they are close to useless for young players is pretty ignorant (and dumb).

[gilesfan]

I'll take 400 actual innings pitched in the last 2 seasons rather than some arbitrary projection system.

He's easily been a top 20 pitcher over 2 the last years, and now because we lose Heyward, he's going to explode to 3.80 ERA..... during his prime.....

Yet somehow, to you, that is a MORE likely than him simply continuing to pitch as well as he has over the last 2 years.

While there is certainly a chance he continues to outperform his peripherals, to expect him to do so is wishful thinking. He's certainly not going to fall into Horacio Ramirez bad, but he has a career FIP of 3.66 and an xFIP of 3.81.

If he can continue to keep his HR/FB ratio under 10% while stranding 80% of base runners, then he's going to be really good; however how likely is that to happen? And those are areas that fluctuate with luck from year to year.

[gilesfan]

where did Chop2Chip's response go?

[chop2chip]

I take these projections with a grain of salt. pretty much you can get the exact same accuracy from a projection system by regressing every player to the league average player. Obviously they don't do that, but players that perform well in a given season seem to be regressed back towards average. It's not a bad methodology on the aggregate, but I don't believe projection systems to be all that good of predicting players to break out in individual cases.

[chop2chip]

where did Chop2Chip's response go?

I confused steamer with marcel. Hence me taking back what I said.

[chop2chip]

I tend to side with GF in that JT and Wood are prime regression candidates, but they still have projectable upside. Something that I feel projection systems are poor at capturing.

[gilesfan]

I tend to side with GF in that JT and Wood are prime regression candidates, but they still have projectable upside. Something that I feel projection systems are poor at capturing.

They certainly are which is why is tough to expect not just one, but both to be top 20 pitchers in the upcoming season. If they are, great, the baseball gods shined down on us, but it's highly unlikely.

The things that are skewed toward league average are luck driven statistics.

Just compare JTs first and second half. Slight difference in all of his statistics,yet he gave up more hits and less guys were left on base. His ERA jumped 41 points in the 2nd half.

And not expecting them to be top 20 starters doesn't mean they are bad pitchers or untalented.

[emk418]

I tend to side with GF in that JT and Wood are prime regression candidates, but they still have projectable upside. Something that I feel projection systems are poor at capturing.

Not saying you're right or wrong but why are they due for prime regression outside of Gattis being in LF?

[dak]

I'd love to see how accurate the steamer projections were last year. These predictive tools are close to useless ofr young players who haven't logged significant major league innings.

This study looks at the accuracy of the various projection systems in 2014 using wOBA as a metric. While it might not answer the core of your question, it's a good overview on the basics of how these systems work and their differences. Looks like ZIPS had the edge in this study.

To state the obvious, all these projection systems are quite imperfect . . . especially when the underlying data is limited. Still, these are the best predictive tools we have in the public domain. A hell of a lot better than looking at a players career average stats.

http://www.hardballtimes.com/evaluat...medium=twitter

[nsacpi]

Wood's stuff impresses me more than Teheran's. And so far he has the lower career FIP. And not surprisingly Steamer projects him to have a lower FIP.

The thing that Teheran has going for him is a deeper repertoire. So far in his brief major league career he already seems to have reinvented himself a couple times. That kind of adaptability is going to serve him well in the long run, assuming good health.

[chop2chip]

They certainly are which is why is tough to expect not just one, but both to be top 20 pitchers in the upcoming season. If they are, great, the baseball gods shined down on us, but it's highly unlikely.

The things that are skewed toward league average are luck driven statistics.

Just compare JTs first and second half. Slight difference in all of his statistics,yet he gave up more hits and less guys were left on base. His ERA jumped 41 points in the 2nd half.

And not expecting them to be top 20 starters doesn't mean they are bad pitchers or untalented.

Agreed on all those points.

I just think for those two young pitchers that there is still a bit of projectable upside. Upside that's not necessarily to their ability but simply from becoming smarter pitchers, working with a better catcher, becoming more confident in secondary pitches, etc. These are all factors that are impossible to capture in a projection system.

But at the same time where I agree with you, these improvements aren't necessarily a given. Most pitchers don't improve in those areas. And like you said losing Heyward should have a material impact on Teheran who is a rather extreme fly ball pitcher.

[Carp]

Defense won't affect that part. Highest HR rate because he's an extreme fly ball pitcher that had an unusually low HR/FB rate last year (8.1%) .

Career HR rate 1.00 in the majors. Career HR rate .7 in the minors.

But he's going to have the highest of his professional career just because.

Great logic.

[chop2chip]

Not saying you're right or wrong but why are they due for prime regression outside of Gattis being in LF?

Because they have been pitchers that have benefitted from good defense which is evident by pretty low BABIP's. That could also be a result of limiting hard contact, but it's been shown that pitchers struggle being able to do that consistently.

I do believe steamer is being rather pessimistic on Teheran and Wood.

[thethe]

This study looks at the accuracy of the various projection systems in 2014 using wOBA as a metric. While it might not answer the core of your question, it's a good overview on the basics of how these systems work and their differences. Looks like ZIPS had the edge in this study.

To state the obvious, all these projection systems are quite imperfect . . . especially when the underlying data is limited. Still, these are the best predictive tools we have in the public domain. A hell of a lot better than looking at a players career average stats.

http://www.hardballtimes.com/evaluat...medium=twitter

I think looking at its effectiveness using every player is not getting to my point though. A large enough sample size and almost anything is predictable.