SB Nation - Login for mobile commenting

Lookout Landing

The Best AL Pitcher in 2008 was...

Well, I'm not actually totally sure.

First of all, where does CC Sabathia rank? Do you discount him entirely because he changed leagues? Count him for the NL because that's where he (was best / finished the year)? Count him for the AL because that's where he threw the most pitches?

If you do count CC in the AL, does he run away with the award? No. In fact, he just further complicates the situation by joining two other people at the top, Cliff Lee and Roy Halladay.

Cliff Lee would seem to be the clear cut favorite even by our lovable new metrics. After all, he's the only AL SP with a tRA under 3 and his pRAA blow away anyone except for Sabathia's combined total. But there's an important factor first brought up by Royals Review and echoed by Geoff Baker recently, that is the strength of competition. Royals Review went ahead and grabbed the actual numbers for us, so I'll just quote him:

Lee's average challenger was a .262/.330/.405 hitter.
Halladay's average hitter was a .266/.342/.425 guy.

Well, how big of a different is that actually? It doesn't seem like much, but remember this is the average difference, manifested over nearly 900 batters faced. Small margins add up over a sample that big. I wanted to put an actual number on that and since I only had triple slash lines, I went to GPA since I knew that could handle those inputs and give me a decently good figure for the difference in runs. (See THT Glossary for formula).

Taking the difference in runs per plate appearance for those two hitters, I multiplied that out by the 987 batters that Halladay faced and ended up with a difference of about nine runs. That is, if Roy Halladay had faced the same average hitter that Cliff Lee had faced, you would expect Roy Halladay to surrender nine fewer runs.

Nine runs chopped off Roy Halladay's line lowers his tRA to 3.04, which is still slightly higher than Lee's, but Halladay faced more batters. It also likely vaults him ahead of CC Sabathia (I don't have CC's average hitter line, but given that half his season was in the NL and the other half was spent facing the same batters as Lee, I would be surprised if it was anywhere near high enough).

Basically you have:
CC Sabathia at about 46 pRAA, but half accumulated in the NL
Cliff Lee with 43 pRAA and the lowest tRA
Roy Halladay with 44 pRAA and the most batters faced

Uhhh....tie?

Poll
If there was an award intelligently given to the best pitcher and I were allowed to vote, my vote would go to:
Roy Halladay
262 votes
Cliff Lee
181 votes
CC Sabathia
29 votes
Other (please explain)
4 votes

476 votes | Poll has closed

0 recs  |  56 comments

Comments

How much of those numbers is a function of those batters facing more Cliff Lee?

I have to imagine some of it is explainable that way, even if it’s not a very large portion

I would say insignificantly small.
I fail to understand how those numbers

make Halladay look better than Lee. Batters faired worse against Lee than they did Halladay, of course the numbers are going to be lower.

Perhaps you are not understanding.

Those lines are not what the average hitter did against Lee/Halladay.

Those lines are the average lines of the the hitters that Lee and Halladay faced. Yes, a small small small fraction of those lines include their PAs against Lee and Roy, but that’s an insignificant part.

Roy Halladay faced a much tougher slate of hitters.

I remember Neyer trying to run a study like this for an ESPN column

and concluding that the variance in competition faces is negligible.

Come again?
It was several years ago

…but unless I’m becoming more senile, using AL v. AL players, that is what he concluded .

No, I mean, I need you to re-word your second sentence.

I want to be clear and the typo(s) is preventing me from that because if Neyer is saying what I think he’s saying (or said), then while I would agree for most cases, it’s clearly not a universal truth and this would be such a year that it’s not valid.

Change 'faces' to 'faced' or...

and concluding [he concluded] that the variance in [quality of] competition two players in the same league faces is negligible.

Oka,y that's what I thought, but I wanted to be sure.

Yeah, I have to disagree with him in this case. The AL East this season was just absurdly strong offensively this year. And while over the entire population of pitchers it might even out, regarding two specific pitchers, it clearly didn’t.

Baseball research has come a long ways in ~6 years

Plus, I should I do better.

Eh. I think it's one of those applies in 98% of cases thing.

This year was just insane. The five best teams in baseball were the Red Sox, Rays, Cubs, Jays and Yanks and the Yanks, Red Sox, Rays and Orioles all had well above average offenses. That Toronto pitching did what they did against playing those teams 72 times is remarkable.

Avg OPS+

AL EAST (sans toronto) = 104
AL CEN (sans cleveland) = 99.5

Toronto’s defense is better than Cleveland. Yeah I know tRA might neutralize for defense but I’m not sure I’m totally sold on that yet.

I say give it to Lee cause I want to say that.

Which would I rather have out of the three next year? Either but Lee.

"Yeah I know tRA might neutralize for defense but I’m not sure I’m totally sold on that yet."

What.

Being completely serious for a change:

How far has tRA caught on in the sabermetric world as a whole and and has it made wide spread penetration into the other, non-Seattle markets? If so how did you manage to do it in such a short time? Was it mainly name recognition?

Isn't tRA measured by batted ball types and their typical run values?

Wouldn’t that make them defense-neutral?

The only way for tRA to not be defence neutral is if people were robbing home runs left and right.
I know just the man for such a job

Goddamn Tori.

Breaking Graham’s statistic.

I hope he doesn't show up at my birthday.
I don't know if I'm totally convinced 100% that batted ball characteristics of a pitcher

are not also affected by the park and defense he plays in front of. I do think that at worst tRA (or any other “defense neutral stat”) probably is almost defense neutral It very well could be defense neutral and I just worry about small things. The only reason I brought this up was that Toronto was so good at defense this year and so far from typical that it may cause us to rethink how we do things a little. I need to think about this a little more before I put a statement like that out there.

For example

Jamie Moyer has a higher GB/FB ratio with the Phillies than he did with us. Its a small change but I bet significant. Has he become a better pitcher because he gets more groundballs? I would say no. In Safeco the run value for a flyball is less than in Phily and he has changed his pitching to suit his enviroment. I could also see a pitcher who has a good defense attempting to pitch to contact more.

In Halladay’s case, if he has a worse defense behind him do you think he throws as many IP (negating his main advantage over Lee)?

That is why we calculate batted ball park factors and apply those.
And I refer to expected outs, not actual outs

so again, defense doesn’t play nearly as big a role there.

Expected out and expected batters faced are not the same thing right?

If TOR defense sucks then Halladay faces more batters per inning which will reduce the number of innings he can pitch.

Yes, but it wouldn't reduce the number of batters he faced.
Of which he faced nearly 100 more than Cliff Lee, about 11% greater.
Can you point me in the direction of where I can find how expected outs are calculated

I couldn’t find it in the tRA post or on statcorner. I feel like my level of ignorance might be making me be wrong.

`

xOuts = kL + kS + number of groundballs (#GB) * E(outs) on groundballs (gbOV) + (#FB – # HRs on FB) * fbOV + #IF * ifOV + (#LD – # HRs on LD) * ldOV + #Bunts * buntOV + #nonOuts * E(outs) on basepaths

Sabathia's opponents:

0.260/0.330/0.406, AL, 507 PA
0.254/0.322/0.388, NL, 516 PA

Look who has sexy intuition.
Finally that free BP subscription comes in handy for something
Also holy crap Sabathia's NL opponents were pathetic
You're telling me

Maybe it has to do

With facing a pitcher 2-3 times/start

Tough choice.

I like Halladay because he pitched more innings and had more strikeouts. I also like the fact Rincewind sponsored Halladay’s BR page.

Halladay was a durable hoss, and against slightly tougher competition

He’s my pick.

Felix?

I kid, I kid…

Yeah, I’m with Baker for once. Halladay. Tougher competition and more innings. Halladay doesn’t quite get the recognition he deserves either. Consistency and saving bullpens really helps a team.

Felix?

Only if you count grand slams

Best pitcher: Halladay

Most Valuable: Lee

I'm the other way around.

Lee was the better pitcher (tRA) but Halladay was more valuable (pRAA).

So, Lee would outrank Halladay on my Cy Young ballot, but Halladay would outrank Lee on my MVP ballot.

For value I use WPA
I voted for Lee because of the tRA difference, even though a lot of that seems to be from an unsustainable HR/BIA rate

however, being the best over the course of a season occasionally involves a little luck, whether it be via BABIP, HR/BIA or you just faced suckier batters

There is no other place in the world where you could find a comment like this.
Amen.
You must feel like a proud parent.
I'll probably get stoned (with rocks) for saying this...

But part of “best” and “valuable” for me is bang for the buck; that is, how much each gets paid. Both earned their money, but Lee cost 40% what Halladay did. Also, while wins and losses are often dismissed by stats gurus, Lee did have 2 more wins and 8 fewer losses on an overall lesser team. Pitching is often times situational.

Cost matters in terms of value, but I hate the idea of it being used as part of performance evaluation.

The question here isn’t “who was the more valuable asset?” it’s “who performed better?” Those two things are quite different.

Also, pitching wins/losses are often dismissed by stats gurus because they don’t matter. They tell you nothing that other stats don’t, and if you rely upon them for evaluative purposes they might lead you to believe things that aren’t at all true.

Speaking of, why isn't Matsuzaka in this race?

18 wins can’t be wrong.

MIKE MUSSINA WON 20 GAMES

CY YOUNG OMGWTFBBQ

Good for him.
In this case...

I was merely using them as tiebreakers; it’s fairly even.

Halladay

20 more innings pitched, 3 more starts, etc.
Workload is a factor, and in a case where two pitchers are nearly tied in rate stats, I think it pushes Halladay over the top.

The gap in quality of hitter faced seals it.

I voted for Halladay too

Using my own nonsensical tiebreaker: I asked myself who I’d want next year, and it was Halladay… due to his consistency over the years, and his greater number of innings pitched. After all, wouldn’t you have to multiply their effectiveness by their innings pitched to see how “valuable” they were over a season? (Which also eliminates Harden from the discussion.)

But then someone pointed out above that Halladay’s edge in innings pitched may have something to do with the sick defense playing behind him. Now I’m wondering if I should have voted Lee. (But I also have a kneejerk reaction against voting for him because he was the comeback player of the year, which is always a feelgood story and people jump on board.)

Halladay's DER was indeed a lot better

but hey, he misses more bats. You can go back and forth on these all day, and I think we all agree that there’s no clear-cut wrong answer (Francisco Rodriguez WOULD be the wrong answer, however).

Yes, I think you do have to multiply effectiveness by playing time – not just for pitchers but hitters as well. That’s why I didn’t think it was such a travesty that Chase Utley didn’t win MVP last year, for example.

You must Login with your SB Nation account and be a member of Lookout Landing to post a comment.