Do I have to say it like an acronym or can I call it "See-air-ah"? You know, for when I'm talking baseball with the chicks.

VoxOrion wrote:Do I have to say it like an acronym or can I call it "See-air-ah"? You know, for when I'm talking baseball with the chicks.

I've been pronouncing it See-air-ah, so I think you're good

Yea, when I clicked on this I was definitely thinking that you had a baby and that there was going to be pictures.

MattS wrote:

bleh wrote:SIERA = 6.262 – 18.055*(SO/PA) + 11.292*(BB/PA) – 1.721*((GB-FB-PU)/PA) +10.169*((SO/PA)^2) – 7.069*(((GB-FB-PU)/PA)^2) + 9.561*(SO/PA)*((GB-FB-PU)/PA) – 4.027*(BB/PA)*((GB-FB-PU)/PA)

The beauty of it is its simplicity

I mean, it's something that you need to put into Excel or go to BP to get like everything else. How many people calculate xFIP in their head or by hand? The best thing about it, in reality, is that it gets rid of a lot of the nonsense simplification in other estimators. What good is an ERA estimator that misses badly for anything but reasonably average or somewhat above average pitchers? I think it's more useful to be able to look at elite pitchers and figure out how good they really might be.

It'll all be on the stats page very soon, so it's not like anyone will need to calculate by hand.

It sorta looks like a regression equation, but it really isn't, is it? Cause you've got interaction terms and squared terms and a bunch of other fun stuff.

TenuredVulture wrote:

MattS wrote:

bleh wrote:SIERA = 6.262 – 18.055*(SO/PA) + 11.292*(BB/PA) – 1.721*((GB-FB-PU)/PA) +10.169*((SO/PA)^2) – 7.069*(((GB-FB-PU)/PA)^2) + 9.561*(SO/PA)*((GB-FB-PU)/PA) – 4.027*(BB/PA)*((GB-FB-PU)/PA)

The beauty of it is its simplicity

I mean, it's something that you need to put into Excel or go to BP to get like everything else. How many people calculate xFIP in their head or by hand? The best thing about it, in reality, is that it gets rid of a lot of the nonsense simplification in other estimators. What good is an ERA estimator that misses badly for anything but reasonably average or somewhat above average pitchers? I think it's more useful to be able to look at elite pitchers and figure out how good they really might be.

It'll all be on the stats page very soon, so it's not like anyone will need to calculate by hand.

It sorta looks like a regression equation, but it really isn't, is it? Cause you've got interaction terms and squared terms and a bunch of other fun stuff.

Yes, it is a regression equation. The benefit of that is that you can actually test for these effects rather than guessing, as well as pick up any pitcher effects on BABIP (which do correlate a little bit with K%, BB%, and GB%, so it's not ignoring them like other metrics) while getting rid of the BABIP luck and defense effects.

Wheels Tupay wrote:Yea, when I clicked on this I was definitely thinking that you had a baby and that there was going to be pictures.

I was kind of imagining him emceeing at a really awful strip club.

Conclusion: Getting strikeouts is very, very good.

TenuredVulture wrote:Conclusion: Getting strikeouts is very, very good.

Haha, yeah, that's a pretty important conclusion. I think it gives a little more credit to strikeouts than other estimators because strikeout pitchers tend to have low BABIPs but it's impossible to see through all the noise (i.e. great strikeout pitchers will be at .290-ish instead of .300, but that really means the window is .260-.320 instead of .270-.330 so it's tough to say). This picks up all that stuff since it's a regression on real live pitchers!

I called Dan Shaughnessy and gave him the heads up about this, he'll be sending a goon squad to your office ASAP

I haven't had a math class since sophomore year of college. Looking at that equation makes my head spin.

fungible

has joboggi chimed in on this yet

CalvinBall wrote:I haven't had a math class since sophomore year of college. Looking at that equation makes my head spin.

you don't need to carry the ones yourself. you can plug it into excel or get it from the statistics reports once it's up (which should be shortly I'm told).

CalvinBall wrote:I haven't had a math class since sophomore year of college. Looking at that equation makes my head spin.

Really? It's pretty basic. I mean, it's derived from running a regression which involves matrices, but result is really just multiplication, addition, subtraction, and division.

i was being a bit sarcastic. just all the PU-GB-FB+B-SEXYIME stuff makes it crazy looking. and the constants are crazy that they are actually derived from something.

This is some good stuff. Someone could probably make some cash betting MLB using this.

HillMD wrote:This is some good stuff. Someone could probably make some cash betting MLB using this.

Sadly, I was thinking the same thing. Time for MattS to put his powers to good use and start posting in the Aces Up thread.

MattS wrote:For 2009:

Pitcher ERA SIERA FIP xFIP
Halladay 2.79 3.09 3.06 3.05
Hamels 4.32 3.55 3.72 3.69
Blanton 4.05 3.92 4.45 4.07
Happ 2.93 4.37 4.33 4.49
Moyer 4.89 4.68 4.94 4.74
Lidge 7.21 4.20 5.45 4.76
Madson 3.26 3.18 3.23 3.25

lee? kendrick? bastardo?

also, the nature of the test of siera as a predictor was a little unclear.
my guess is that "using 2003-08 data to generate a formula and then testing it on 2009 pitchers, SIERA emerged as the best estimator of park-adjusted ERA in the following year, and..." means that (a) 2003-8 data was used in the regression to determine the right formula for combining the components of pitching to predict era; then (b) that formula was used to calculate siera for pitcher each season, including 2008 and 2009; the reliability/accuracy of siera as an era predictor relative to other measures was tested BOTH by seeing how well 2008 siera predicted 2009 era AND by seeing how well 2009 siera matched up with 2009 era. do i have that right?

finally, it looks like your discussion of lidge elsewhere in this thread suggests that one reason siera beats fip in the next season despite losing out in-season is that siera treats injuries as luck. an injury is likely to produce a consistent streak of hr/fb bad luck during one season that doesn't persist into the next. is that a fair interpretation?

danrosz wrote:

HillMD wrote:This is some good stuff. Someone could probably make some cash betting MLB using this.

Sadly, I was thinking the same thing. Time for MattS to put his powers to good use and start posting in the Aces Up thread.

As long as Matt and Eric are not bent on world domination, I am OK with all this.

Hey did everybody hear about Cliff Lee's foot surgery yesterday?

If I can come up with an intelligent question or comment on all this, I will. Promise.

phorever wrote:

MattS wrote:For 2009:

Pitcher ERA SIERA FIP xFIP
Halladay 2.79 3.09 3.06 3.05
Hamels 4.32 3.55 3.72 3.69
Blanton 4.05 3.92 4.45 4.07
Happ 2.93 4.37 4.33 4.49
Moyer 4.89 4.68 4.94 4.74
Lidge 7.21 4.20 5.45 4.76
Madson 3.26 3.18 3.23 3.25

lee? kendrick? bastardo?

also, the nature of the test of siera as a predictor was a little unclear.
my guess is that "using 2003-08 data to generate a formula and then testing it on 2009 pitchers, SIERA emerged as the best estimator of park-adjusted ERA in the following year, and..." means that (a) 2003-8 data was used in the regression to determine the right formula for combining the components of pitching to predict era; then (b) that formula was used to calculate siera for pitcher each season, including 2008 and 2009; the reliability/accuracy of siera as an era predictor relative to other measures was tested BOTH by seeing how well 2008 siera predicted 2009 era AND by seeing how well 2009 siera matched up with 2009 era. do i have that right?

finally, it looks like your discussion of lidge elsewhere in this thread suggests that one reason siera beats fip in the next season despite losing out in-season is that siera treats injuries as luck. an injury is likely to produce a consistent streak of hr/fb bad luck during one season that doesn't persist into the next. is that a fair interpretation?

Lee: 3.73 combined (4.07 CLE, 3.05 PHI)

Bastardo: 4.69 (but that's a pretty small sample size...but I guess a benefit of SIERA is that it will stabilize quickly like QERA)

Kendrick: 4.92, 5.30, 4.26 in 07, 08, 09 (though the sample size plus pitching in relief makes the 09 number not really all that great)

You have some of the tests we ran right, though we ran tons and it pretty much won every way. It was the best at predicting next year ERA in every year, and the best among estimators that treat HR/FB as luck in estimating same year ERA in every year, I think. It was pretty consistently ahead in all of the tests. The primary ones, though, consisted of predicting same-year ERA better than QERA and xFIP, and next-year ERA better than QERA, xFIP, FIP, and tRA. We also tested data using coefficients generated from 2003-08 to test prediction in 09 just to make sure there were no biases. And using slightly different coefficients that were generated without knowing 2009 happened (in fact, generated last year), the 2008 SIERAs predicted 2009 ERAs best too.

I wouldn't say that SIERA treats injuries as luck or anything, just that injuries can clearly make a pitcher pitch way behind his peripherals as I learned in my more novice days where I declared Freddy Garcia fine, heheh FIP happens to pick up on HR/FB luck and credit it to the pitcher when the reality is that HR/FB is probably correlated with pitching while injured, and it's just that so many people with high HR/FB drop out of the sample that we don't pick up this injury effect in testing.