So, now I am at my laptop, so it's easier to write something than with my mobile. The thing is that past performance is more or less an indicator for future performance. But ERA (along other oldschool statistics) is influenced by way too many pitcher-independent things. I don't want to write a big article about sabers and how they are calculated. The internet is full with blogs and comments about that and in my experience it's rather true, that xFIP and especially SIERA beat the conventional stats by a wide, wide margin regarding the quality of a pitcher. It began with FIP, then came xFIP and SIERA is such a strong stat that everyone should look into it. If you ask LTProfits, he will confirm you that his very strong model takes sabers into account. And here comes why I love sabermetrics.
All in all we can say if a pitcher has an ERA which is way above his sabermetric numbers, he was rather unlucky, which means that he will get better raw numbers in the future. If a pitcher has an ERA which is way below his sabermetric numbers, he was rather lucky, which means that he will get worse raw numbers in the future. The first pitcher who came to my mind is Chris Young, now with the Royals. I said it time and time again over the last few days that he will get knocked around sooner or later because his sabers tell us that he was one of the biggest lucksters ever. And then came the Red Sox and boom he got crushed. It was the same thing with Tim Lincecum this season. Same story. I said it time and time again, that although his raw (oldschool) numbers were rather good at the beginning of the season, that he is still the same bad pitcher as in the previous seasons. Why? Because his ERA was much better than in the past few years, but his xFIP and SIERA were even worse (at least in parts). And what happened? From 4/10 to 5/20 he allowed 11 earned runs in in 8 starts (47.2 IP, so a 2.08 ERA). Then in his next six starts he got worse and worse, last time he lasted only 1.1 IP. So in his next 6 starts he allowed 21 earned runs (27.0 IP, so a 7.00 ERA). Was this a surprise? No. Why? Sabermetrics. You won't have luck forever. Your BABIP won't stay below .250 forever. Your LOB% will come down sooner or later. And talking about Chi Chi he owns a BABIP of .183 (unsustainable, league average is about .300) and a LOB% of 94.6% (league average is about 70-71%, so unsustainable, too).
But to keep things easier let's only look at SIERA and xFIP. Yes, there are some guys who outplay their peripherals, which means they are lucky over the course of a whole season (but HOW lucky? as lucky as Chi Chi?) or unlucky over the course of a whole season. More or less the sabers include the BABIP and LOB%, so we can forget them and simply look at the difference between ERA and xFIP or ERA and SIERA. If the sabers are higher than his ERA, he was lucky, if it's the opposite he was unlucky. (If there's a difference of -0.5 or +0.5 I would not consider it big enough.)
Over the last 5 seasons (2010-2014, taking all qualified starting pitchers into account), the average difference of ERA and xFIP was -0.09 and of ERA and SIERA it was -0.14. So you see, that those sabers are a good predictor - at least on average.
If you take a further look, you will see that those guys were the luckiest pitchers regarding the difference of ERA and xFIP:
Season
Name
Team
ERA
xFIP
SIERA
ERA-xFIP
2011
Jeremy Hellickson
Rays
2,95
4,72
4,77
-1,77
2010
Clay Buchholz
Red Sox
2,33
4,07
4,27
-1,74
2014
Chris Young
Mariners
3,65
5,19
5,24
-1,54
2014
Doug Fister
Nationals
2,41
3,85
3,93
-1,44
2013
Travis Wood
Cubs
3,11
4,50
4,43
-1,39
2011
Jered Weaver
Angels
2,41
3,80
3,66
-1,39
2012
Jered Weaver
Angels
2,81
4,18
4,09
-1,37
2012
Jeremy Hellickson
Rays
3,10
4,44
4,51
-1,34
2012
Aaron Harang
Dodgers
3,61
4,95
4,94
-1,34
2013
Bartolo Colon
Athletics
2,65
3,95
4,10
-1,30
2014
Edinson Volquez
Pirates
3,04
4,20
4,20
-1,16
2010
Johan Santana
Mets
2,98
4,13
4,19
-1,15
2011
Ryan Vogelsong
Giants
2,71
3,85
3,96
-1,14
2010
David Price
Rays
2,72
3,83
3,83
-1,11
2012
Kyle Lohse
Cardinals
2,86
3,96
4,13
-1,10
2014
Lance Lynn
Cardinals
2,74
3,81
3,84
-1,07
2013
Clayton Kershaw
Dodgers
1,83
2,88
2,99
-1,05
2012
Matt Cain
Giants
2,79
3,82
3,69
-1,03
2010
Trevor Cahill
Athletics
2,97
3,99
4,13
-1,02
(I could show you more tables, bigger tables, blablabla, but there's not enough space for it lol.)
First: Look at the names. Do you see some familiar names, familiar in the meaning of 'boy did they regress'?
Second: Now look at the differences. Not one single qualified starter had a higher difference than -1.77. As I said, a lot of them regressed. Surprise? No!
And...They are only 19 guys. Regarding SIERA there are a few more guys (27) who have been luckier than 1.00 run, but it's stilly only 27 of 443 starters! As I said, Chi Chi has about a 5 run difference between his sabers and his ERA. This - can't - be - kept - up. Never has, never will.
When you turn the page around you will see that from 2010 to 2014 there have been 25 starters who have been rather unlucky regarding the difference between ERA and xFIP (+1.00 or higher) and about the same number (21 starters) regarding SIERA (difference between ERA and SIERA +1.00 or higher). Regarding both calculations, the highest difference was a +1.64 (Edinson Volquez in 2013, there he had an ERA of 5.71 and a xFIP of 4.07 along a SIERA of 4.26, in 2014 he had an ERA of 3.03...surprise? No!) and a +1.63 (James Shields in 2010, he had an ERA of 5.18, a 3.55 xFIP and a 3.55 SIERA, and what did he have in 2011? A 2.82 ERA...surprise? No!).
I hope now you can better understand, why I 'hate' (at least regarding betting) Chi Chi Gonzalez...