Starting pitchers and First Year in Boston Syndrome

The Gray Eagle

Member
SoSH Member
Aug 1, 2001
16,687
Alex Speier's 108 Stitches newsletter had an interesting look at veteran starting pitchers in their first season in Boston.

"Since the 2004 season, the Red Sox have added nine starting pitchers with at least 100 starts in the big leagues who went on to make at least 20 starts in their first season in Boston – most recently, David Price in 2016 and Rick Porcello as well as Wade Miley in 2015. The First Year In Boston dropoff from prior career standards is eye-opening.

Every one of the nine starters in question – Price, Porcello, Miley, Ryan Dempster, John Lackey, Brad Penny, Josh Beckett, Matt Clement, and David Wells – posted an ERA in their first season with the Red Sox that was higher than their pre-Boston career norm. On average, the group’s ERA was 0.75 runs higher in their first season with the Red Sox than they’d produced in their careers to that point; the rises ranged from 0.23 runs (Clement) to 1.55 runs (Penny)."

"Some of the increase would be expected given the impact of pitching in Fenway Park and the smaller ballparks of the American League East. Nonetheless, on average, the group saw its ERA+ -- which accounts for park effects – fall by an average of 12 points.

It’s interesting to note that Price’s FYIB experience was very much in line with those standards – his ERA was 0.90 runs above his pre-Boston norm, while his ERA+ went from 126 (meaning 26 percent better than league average) to 114. In other words, Price essentially experienced a “standard” departure from the elite levels he’d established in his career, part of the reason why the Red Sox tend to consider his 3.99 ERA of 2016 a solid foundation for what they hope will be an even better second season."

"There will be an expected transition for Sale, who arrives in Boston with a career 3.00 ERA and 135 ERA+; a normal FYIB adjustment might see him with a 3.75 ERA and 123 ERA+."


Speier doesn't really offer an explanation as to what might be the cause of this FYIB Syndrome. If it is real and not just a statistical fluke, then nobody really knows what might cause it. It might be adjustment to the ballpark and division, or it might be something more nebulous, like dealing with more scrutiny and a different atmosphere in Boston. Nobody knows if it's a real thing, and if it is, nobody really knows what causes it.

But 9 out of the last 9 have taken a hit to their numbers in their first year, so it seems to me that numbers like 3.75, 123 ERA+ should be a reasonable expectation for Sale in 2017. If he does that and throws his usual 200+ innings, that would make a huge positive impact, despite those numbers being well below average for him in his career.

Or maybe he will break the streak and be the first vet starter to come here since Schilling and be better than before. I just hope the fanbase keeps expectations lower and everyone doesn't freak out if Sale has the same FYIB type of season that all the others had.
 

dbn

Member
SoSH Member
Feb 10, 2007
7,785
La Mancha.
It would be interesting to see a box-and-whisker plot of ERA+ versus year (starting one or two years prior to 1st season in Boston and extending past the first year) for veteran Boston FA SPs, and the same for other teams or league average.

Maybe I'll find time to do so (but if someone else wants to please do because I'll probably not get around to it).

edit: or if someone with SAS-ability wants to grab the data for me we could collaborate.
 

NoXInNixon

Member
SoSH Member
Mar 24, 2008
5,282
Generally speaking, if you're starting a season with a new team, you're much more likely to be on the bad side of the aging curve than the good one. On average, every year after age 26 you can expect a pitcher to gain about .15 in ERA. So that explains 20% of the observed effect. Some of the rest is likely to be contract year ERA+ inflation, which is something every team is going to have to deal with. Sale wasn't pitching for a new contract last year, so that part of it shouldn't affect him.
 

dbn

Member
SoSH Member
Feb 10, 2007
7,785
La Mancha.
Generally speaking, if you're starting a season with a new team, you're much more likely to be on the bad side of the aging curve than the good one. On average, every year after age 26 you can expect a pitcher to gain about .15 in ERA. So that explains 20% of the observed effect. Some of the rest is likely to be contract year ERA+ inflation, which is something every team is going to have to deal with. Sale wasn't pitching for a new contract last year, so that part of it shouldn't affect him.
Anecdotally that sounds reasonable, but it'd still be fun and probably interesting to look at some data. For example, if year-2 ERA+s go back down, then there would seem to be something to the adjustment theory.
 

StupendousMan

Member
SoSH Member
Jul 20, 2005
1,897
If one does choose to test this idea, a good check would be to create a control sample of pitchers with properties similar to those who changed teams, but who did _not_ change teams ... and then compare the control sample's statistics in some given year to the pitchers'-who-changed-teams statistics.
 

O Captain! My Captain!

Member
SoSH Member
Mar 3, 2009
3,532
Generally speaking, if you're starting a season with a new team, you're much more likely to be on the bad side of the aging curve than the good one. On average, every year after age 26 you can expect a pitcher to gain about .15 in ERA. So that explains 20% of the observed effect. Some of the rest is likely to be contract year ERA+ inflation, which is something every team is going to have to deal with. Sale wasn't pitching for a new contract last year, so that part of it shouldn't affect him.
You could probably compare actual outcomes to projections which bake in age curves to account for this.
 

JimBoSox9

will you be my friend?
SoSH Member
Nov 1, 2005
16,664
Mid-surburbia
There are only maybe four names out of those nine who would would have expected at the time to post a Boston ERA within a half run of their career average overall, so linking it to a "first year" effect seems strange. As a temporary phenomenon that is then recovered from in years 2+, I'm pretty skeptical.
 

uk_sox_fan

Member
SoSH Member
Nov 11, 2006
1,273
London, England
Don't forget that another part of the perceived effect is the selector bias of choosing the last 12 seasons. Why 12? Because 13 would include Schilling?

If, in fact, you chose the past 20 years rather than the past 12 you would have discovered quite a few starters whose ERA+ in their first year in Boston was better than their career ERA+ to that point.

I've come up with 10 more starters in '97 - '04 that meet the criteria of 100+ starts before coming to Boston and 20+ starts in their Boston debut:
Schilling, Burkett, Castillo, Cone, Nomo, Fassero, Schourek, Rapp, Portugal and Pete somebody.

(Bret Saberhagen didn't meet the GS criteria as he started only 6 games in the 2nd half of '97 following his 18-month recovery from shoulder surgery, but his 124 ERA+ in 4 years with Boston held up well against his 128 career ERA+)

Code:
Year   P Age  GS  ERA+ prGS pcERA+ psERA+
 04   CS  37  32  148   338   130   159
 02   JB  37  29  101   364   100   147
 01   FC  32  26  107   218    94   142
 01   DC  38  25  104   390   122    70
 01   HN  32  33  100   181   104    99
 00   JF  37  23  106   194   115    69
 00   PS  31  21   99   155    90    87
 99   PR  31  26  122   155    95    90
 99   MP  36  27   91   256   101    99
 98   PM  26  33  163   120   140   219
The average ERA+ for these 10 veterans in their debut season in Boston was 114.1 vs a career ERA+ (using bbref's selection tool) of 109.1 and an average mark of 118.1 in the season previous to arriving. This performance is fairly remarkable given the age of some of these starters -- Mark Portugal was 36 when he arrived, G38, the Napkin and Jeff Fassero were 37 and David Cone was 38. Only Portugal failed to either match his career average or better his previous season and Rapp and Schourek managed to do both.

Taking that 26 year old out of the sample doesn't change much (except bring everything down): 108.7 average ERA+ vs 105.7 career mark and 106.9 previous season performance.

So unless you're arguing that winning the World Series in '04 created a veteran starting pitcher curse or something, I think this particular mediocrity streak is a clear case of selection bias.


edit: used the bbref season selector tool rather than calculating average career ERA+ weighted by IP and added back Jeff Fassero who somehow dropped out of my table

pcERA+ = ERA+ for career previous to coming to Boston
psERA+ = ERA+ for season immediately before arriving in Boston
 
Last edited:

uk_sox_fan

Member
SoSH Member
Nov 11, 2006
1,273
London, England
For comparison here are the same stats for the 9 most recent starters mentioned by Gray Eagle:

Code:
Year   P Age  GS  ERA+ prGS pcERA+ psERA+
 16   DP  30  35  114   213   126   164
 15   RP  26  28   87   180    97   113
 15   WM  28  32   96   102   103    86
 13   RD  36  29   90   322    99    86
 10   JL  31  33   99   233   116   115
 09   BP  31  24   83   245   105    67
 06   JB  26  33   95   103   118   118
 05   MC  30  32   99   192    98   120
 05   DW  42  30  102   417   111   104
I have to admit it is pretty remarkable how much this group underperformed their averages relative to the previous 10.
 

uk_sox_fan

Member
SoSH Member
Nov 11, 2006
1,273
London, England
And it would look worse if we had ERA+ marks for Daisuke's years with the Seibu Lions...

Code:
Year Age Team  W   L   ERA  G  IP   SO
1999 18 Seibu  16  5  2.60 25 180.0 151
2000 19 Seibu  14  7  3.97 27 167.6 144
2001 20 Seibu  15 15  3.60 33 240.3 214
2002 21 Seibu   6  2  3.68 14  73.3  78
2003 22 Seibu  16  7  2.83 29 194.0 215
2004 23 Seibu  10  6  2.90 23 146.0 127
2005 24 Seibu  14 13  2.30 28 215.0 226
2006 25 Seibu  17  5  2.13 25 186.1 200
2007 26 Boston 15 12  4.40 32 204.2 201
2008 27 Boston 18  3  2.90 29 167.2 154
2009 28 Boston  4  6  5.76 12  59.1  54
2010 29 Boston  9  6  4.69 25 153.2 133
2011 30 Boston  3  3  5.30  8  37.1  26
2012 31 Boston  1  7  8.28 11  45.2  41
2013 32 NY Mets 3  3  4.42  7  38.2  33
 

Reverend

for king and country
Lifetime Member
SoSH Member
Jan 20, 2007
63,822
It seems to me there is the possibility of a major selection bias with regard to these guys being targeted by the Red Sox and why.

Like, playing in Boston is not the only characteristic they share; they also share the characteristics that are the reasons that they were targeted by a perennial contender, here: Boston.

Is some of those characteristics are associated with having just put up a really strong season or two, then we should expect some regression to the mean.

How much of the observed effect can be explained by this kind of selection effect? I dunno. But it seems like there are strong reasons to expect all the guys in this class since Henry bought the team to be subject to regression to the mean following the kind of season they'd need to have had to make them an attractive target for acquisition.
 

lexrageorge

Member
SoSH Member
Jul 31, 2007
18,078
I hate case studies where all the players are bunched into one "bucket" without any regard for the fact that there were definitely several players with a unique set of circumstances.

Ryan Dempster was signed as a flyer by the Sox to fill the 5th starter slot. He was 36, and never pitched again after his 1 inning cameo in the World Series. And he wasn't really that good with Texas the prior half season.

John Lackey's first year likely coincided with the start of his elbow problems that eventually required TJ surgery.

Brad Penny was worse in his final year with the Dodgers before coming to Boston, and his stint with the Giants was helped tremendously by a 0.206 BABIP in September. His St Louis stats are helped by a tiny sample size (9 games)

Matt Clement's FIP in 2005 was essentially identical to his FIP the prior 2 seasons. This is also where I need to point out for the 900th time that he was a very different pitcher after he got hit in the head with a line drive late that season; his ERA would have shown significant improvement had he not had those 6 September starts post-beaning. Whether that falloff was due to the injury or the shoulder problems that ultimately ended his career is an unknown, but his "Boston effect' was indeed the reverse of what Speier claims.

This is where I also need to point out that David Wells' FIP of 3.83 was lower than his FIP of 3.88 and 3.94 from the prior 2 seasons. The big difference between 2004 David Wells and 2005 David Wells was a BABIP spike from 0.280 to 0.320. And Wells was 42 years old, and started to decline the following season.

Similarly, Wade Miley's FIP of 3.81 was lower than his prior season's FIP of 3.98, and is less than his career average of 3.99. Cafardo's periodic bleatings aside, he hasn't really been any better than average since he left Boston.

Beckett, Price, and Porcello are the only 3 pitchers that truly showed a "first year in Boston" effect. And as Rev correctly notes above, that "effect" could just be the result of basic statistics. Something I would have expected Speier to address, given that he's usually better than this.