Tangotiger Blog

Friday, March 29, 2013

Felix Verlander

And the pairing just keeps on pairing.  Three years ago, they signed nearly identical 5/80 deals.  And two years later, they got nearly identical 5/135 extensions.  (Verlander got a bit more, in return for a team option beyond that.) 

But their career stats are ridiculously similar.?  Both have been in the league 2005-2012.  They are six starts apart, Felix has a slightly better ERA, but when adjusted, Verlander is a bit ahead.  Felix is ahead by 33 strikeouts, but with ten more walks.  Any fan, any scout, and player, they'll think the same thing, when they look at their overall careers.  And while Verlander may be ahead today, he's older, and therefore more likely to lose value than Felix by the end.  So that, overall, they may indeed be equals over the five years of the extension.

Where they do differ is the W/L record, with Verlander at 124-65, and Felix at only 98-76.  But all of us, the fans, the scouts, and the players, we've all decided to give the middle finger to the W/L record.  It's a relic that adds nothing, once you consider everything else.  The GWRBI died a quick and painless death.  Such is the power of inertia, that it can disguise itself as a longstanding tradition, and fool those who don't want to think too much about it.   The W/L record had zero impact on Verlander/Felix contracts.  It has zero impact if you watch baseball.  So, stop looking at it, just because someone has decided to keep counting it.  It's dead already.

I leave you with this list, pitchers with at least 3700 batters faced (half of CC's league-leading total), 2005-2012, of WAR per 1000 batters faced, according to Baseball Reference:

WAR/1000PA 
 6.7  Josh Johnson
 6.6  Roy Halladay
 6.3  Clayton Kershaw
 6.1  Johan Santana
 5.6  CC Sabathia
 5.6  Justin Verlander
 5.5  Jered Weaver
 5.3  Roy Oswalt
 5.2  Cole Hamels
 5.2  Chris Carpenter
 5.1  Jon Lester
 5.1  Cliff Lee
 5.0  Matt Cain
 5.0  Felix Hernandez
 4.9  Zack Greinke

Say hello to Clayton Kershaw, the first MLB pitcher that will be guaranteed 200MM$ when he signs his next contract.  And Josh Johnson has a chance to be the best pitcher that time forgot in twenty years.

Thursday, March 28, 2013

Astros going to tandem starters, including the upper minor league levels

?Here's the plan:

There will be four groups of two starting pitchers at each of the full season affiliates. That gives each affiliate eight starters to work through. The first pitcher will go five innings or seventy-five pitches, whichever comes first, and will be piggy-backed by another starter who will go the remaining four innings or sixty pitches. They'll go through those four groups, and flip-flop the starter-reliever combo and go through again. This is something Jeff Luhnow has experience with since he implemented this in the lower levels of the minors. However, this is the first time it's been done in the upper levels.

Eventually, the hope is that five starters will ultimately separate themselves from the pack and the affiliate will transition back into the traditional five-man rotation. Although, there is no set timetable for that transition.

Relievers are unfortunately not guaranteed much work. The managers are at their own discretion to use the relievers for game situations closing out games, but the starters are guaranteed their work.

This is a fascinating development.

Monday, March 25, 2013

Tandem Starters

Pizza proposes that the whole starting group becomes tandem starters.

Under his proposal, you have six starters paired in groups of three.  That would mean they'd each pitch 54 games, and they'd each throw 50 pitches.  That's 2700 per pitcher.  Basically, the 100 pitches per game times 162 games would be spread over six pitchers instead of five.

A top starter (Verlander, Felix, Kershaw) would presently average over 100 pitches per start and some 33 starts each.  In his last 4 years, Verlander has averaged 3850 pitches per year, Felix 3600.  Trivia: Kershaw ?has thrown 3469 pitches in each of the last two years.  Exact same number!

Anyway, Pizza proposal could work, but obviously, it can't be tested with a team that has an ace starting pitcher.  You can't test a theory if you already are one of the best pitchers in the league, on the idea that it MIGHT help the whole team for you to pitch less often overall.

So, we'd need a team where there's no obvious ace starting pitcher, a team that likely has pitchers who would throw 90-100 per game, and start 29-32.  A team that where the top starter would likely get to 3000 pitches in a normal usage pattern, and therefore could go for 2700 pitchers in a tandem-rotation.

And pitchers have (sort of) pitched like this in the past.  I looked for pitchers with at least 50 games and at least 140 IP, and at most 5 starts.  John Hiller in 1974 threw 150 IP in 59 relief games.  Willie Hernandez in his 1984 MVP season.  Mike Marshall (of course) did it three times.  There was Bob Stanley twice, Bill Campbell twice, Clay Carroll three times.  And the last time it happened was with rookie Mark Eichorn (who I'd have awarded the ERA title).

I think it's worthy of a tryout, and naturally, you'd have to try this out in the lower-levels first. 

Friday, March 22, 2013

Rivera or Pettitte?

A couple of years back, I posted where I thought Pettitte and Rivera ranked among their peers:

#1. Pedro #2. Mussina #3. Mariano Rivera #4. Andy Pettitte #5. Tim Hudson

I also compared them last year:

From 1995-2010, Pettitte and Rivera’s careers overlapped.  According to Baseball-Reference, they both had close to 50 wins above replacement in that time period.  And they both made close to 125 million dollars.  However you ultimately value Andy Pettitte, that’s right where you should peg Mariano Rivera.  They made the same kind of impact (in the regular season anyway). When it comes to the playoffs, Rivera’s impact is something else to behold. 

Ben brought up Rivera and where he ranks.

But when the two teammates hit the Hall of Fame ballot, perhaps simultaneously, Rivera will almost certainly be inducted immediately, while Pettitte’s candidacy could linger until his years of eligibility are up. Both pitchers, despite the different ways that they’re used, have essentially the same job—to get batters out—and Pettitte has retired many more of them. It’s not necessarily fair that Rivera will waltz into the Hall while Pettitte watches and waits, but no one will object. Rivera, after all, is the one with the case for the all-time team.

I'm not sure how small you  have to make the Hall of Fame that you would exclude Pedro, but I'd say it would have to be exceptionally puny.  How small do you make the Hall of Fame to exclude Mariano Rivera?  And how big to include Andy Pettitte.

I think on a gut level, the perception is as Ben is suggesting, that there is a chasm between Mo and Pettitte, even though they've been paid about the same, and our best guess is that they've had similar impact.  Even if you want to argue for Mo over Pettitte, it's not an easy call like Pedro over Pettitte.

But, Mo I think will be more highly considered than even Mussina.  And possibly more than Glavine and Smoltz.  And Smoltz is a good comp I think that shows that Mo should fall below Smoltz.

This is one of those where A > B > C > A.  That is, you can argue for Mo being better than Glavine, and you can argue for Glavine being better than Smoltz, and then Smoltz being better than Mo.

By "better" I mean "more accomplished career".

?

Thursday, March 21, 2013

Are pitchers conducive to a certain role (closer, non-closer reliever, starter)?

Someone I worked with talked to a big league manager several years back, and ?that manager told my buddy that a certain reliever specifically told him that he did NOT want to pitch in the ninth.

Eric Gagne in his book said that once he got the taste of closing the game, he knew that's the only thing he wanted to do.

Adam Wainwright started his MLB career fairly late for such a talent, and spent the whole season in the bullpen in a non-closing role, but being awesome.  He became a closer in the playoffs (2006 World Series winner), but then immediately became a starter.

Daniel Bard was never a closer, but wasn't able to make the transition to starter.  The very young Chris Sale was also not a closer, but he dominated as a starter.  Gossage dominated as a reliever, failed as a starter, and went back to dominating as a reliever.

We can come up with a whole bunch of examples with a whole bunch of different combinations of background.  Is it about personality, talent, repertoire?  What do you think?

Monday, March 18, 2013

Mythbuster: day-to-day same kind of pitcher is no disadvantage to pitching team

?Pizza looks at whether it matters that you follow a power starter with a power starter, and he finds: no.  If you are looking for this effect, it would be more magnified among same-game pitchers, so a power RH starter followed by a power RH reliever, etc.  Pizza, can you look into this?

He also notes that there is something going on here:

I coded each game for whether the batting team was facing the same team from

It didn't matter who the pitcher was or what sort of approach he generally used. Teams facing off for the second game in a row were less likely to hit home runs, more likely to strike out, and more likely to be hit by a pitch. The effect sizes were small (on the order of a tenth of a percentage point), but significant, and they generally favored the pitcher.

We've talked about getaway days, day games following night games, etc.  So, I'd like to see some more controls here.  Looks interesting.

Tuesday, March 12, 2013

Strasburg v the field: look at the baseline

I was looking at the forecasts for Strasburg on Fangraphs.  Fangraphs collects them from Steamer, Oliver, Bill James, ZiPS, and the community.  But, it's hard to compare, because they each have their own scale.

Take for example Oliver, which forecasts him for 2.46 ERA, which seems absurdly low for a mean forecast.  But the top 3 other starting pitchers (Kershaw, Verlander, Weaver)? average a forecast of 2.49. 

The Fans are similar: forecasting 2.76 ERA, but the other top 3 SP (Kershaw, Verlander, Felix) average 2.79.

ZiPS seems more optimistic than the Fans with a 2.69 ERA forecast, but his other top 3 (Kershaw, Cain, Weaver) average 2.85.

And Steamer, seemingly the most pessimistic and most out of line with the other guys, with an ERA forecast of 2.91, shows that they are the most optimistic compared to the other top 3 starters (Kershaw, Greinke, Lee), who average 3.25!

(Bill James forecasts are not provided for everyone on one page, so if someone wants to pick out all the pitchers I named above, take the top 3, and do that, feel free.)

UPDATE: thanks to Steve, the other top 3 for Bill James averaged 2.87 (Kershaw, Cain, Verlander)

PECOTA has Stras with 2.61, compared to the other top 3 (Kershaw, Felix, Doc) of 2.83.

Clay Davenport also has the forecasts, but he has it team-by-team.  Seeing how Strasburg doesn't compare favorably to Gio, Clay might not be so hot on Stras.

UPDATE: Thanks to Sky, Clay's other top 3 (Kershaw, Cain, Weaver) are at 3.03.

UPDATE: Craig at Bloomberg tweeted that Stras was at 3.02, and the other top 3 (Kershaw, Verlander, Lee) are at 3.07.

UPDATE: Shandler has Stras at 2.93, and the other top 3 (Kershaw, Price, Hamels) at 2.77.

UPDATE: MGL has Stras at 2.67 (league of 4.00), and the other top 3 at 2.64.

To summarize, this is Stras, relative to the other top 3 SP:

90% Steamer

92% PECOTA

93% Bill James

94% ZiPS

98% Bloomberg

99% Oliver

99% Fans

101% MGL

106% Shandler

115% Clay

I would LOVE if everyone provided their forecasts relative to a COMMON baseline.  I mean, really, no one out there is actually forecasting the league ERA are they?  Aren't they just using something based on the last couple of years?  Can't we all just say "hey guys, in 2013, let's agree to an MLB ERA of 4.20", or whatever.

Thursday, March 07, 2013

Pitchers without pitchers-as-batters

?Great stuff from Jeff, noting that if you remove pitchers-as-batters, Gio Gonzalez the NL pitcher looks like Gio the AL pitcher.  He also notes that RA Dickey strikes out nonpitchers and pitchers at the same rate, so, he should not be as affected in going to the better league as other pitchers would be. 

This is funny, because this is actually one of the subtopics we focused on in my class last semester.  If you want to have fun, check out Clayton Richard.

Wednesday, March 06, 2013

FIP 2.0?

Dave makes an appealing case:

Maybe we should consider that the IFFB is essentially not that different from a HR, in that it measures the results of specific batted balls that have a distinct run value and that aren’t influenced by the defenders behind any given pitcher. Just like we penalize pitchers for giving up home runs, logic would suggest that we should be giving them credit for infield flies.

I suppose that the error rate on IFFB is pretty close to the error rate on K (PB, WP, that allows batter to reach base). I also suppose that the hit rate on IFFB is going to be virtually zero (and naturally, the hits on K is zero by definition). So, for all intents and purposes a PROPERLY CLASSIFIED IFFB is the same thing as a K, which is why I have it as I do in bbFIP. The question therefore is if we can treat the IFFB distinct from the other three batted ball events, in light of Dan's findings. It has a certain appeal.  I mean, the fielder for the IFFB is as inconsequential as the catcher is for the K.  Indeed, even more inconsequential, since the catcher can influence a K.

The two main issues is lack of historical data, and any potential bias in misclassifying such plays.

So, does it make sense to have something in-between FIP and BattedBall FIP (bbFIP)?  I'm intrigued anyway.?

Measuring BABIP skill

Dan notes:

I studied a bunch of variables I thought might have something to do with hit suppression on balls in play. I came up with two — both FanGraphs stats — that seem to have significant predictive power. The first is pop up rate. The second is z-contact, which is when batters swing at a strike — balls in the strike zone — thrown by a pitcher. What percent of those times does the batter make contact? It turns out that, just like inducing pop ups, it reduces BABIP and correlates consistently year to year. Getting batters to swing and miss at your strikes has strong predictive power on hit suppression.

***

I should correct this:

Tom Tango’s FIP assumes that all pitchers have exactly league-average BABIP ability.

Would someone suggest that OBP "assumes" that all times on base are created equally, that OBP "assumes" that a walk is worth the same as a HR?  No.  OBP is very clear: it is only interested in the NUMBER of times a batter reaches base and is AGNOSTIC as to how far he got.  It takes no position on the QUALITY of reaching on base.  OBP looks at the SUBSET of a hitter's performance.

Similarly, FIP? is also looking at the subset of a pitcher's performance (those that do not require his fielders be involved) and is agnostic on everything else about his performance.  The difference is that FIP explains a pitcher's skill far more than OBP explains a hitter's skill, and so, we then get fooled into thinking that FIP is saying that a pitcher has no skill on balls in play.  But, FIP is not saying that at all, and neither am I.

Saturday, March 02, 2013

GB v FB pitchers

?Bill has an article about the two, and how much he (initially) hated GB pitchers, until he decided to study the issue.

Well. ..I still don’t really like Ground Ball pitchers, but I do concede that I was, in the past, too radical in my distrust of them.    I think that the advantages of throwing ground balls have been horribly overstated, and that the best way to get batters out is to find pitchers who can throw high fastballs.   However, throwing ground balls does appear to be a small advantage for the pitcher, if you control for the number of strikeouts.    Given two pitchers with the same strikeout rate and the same walk rate, it does appear that we should favor the one who gets more ground balls, although this advantage is probably no larger than the advantage of being a good fielder or the advantage of having a good move to first.

If you exclude homeruns, the run value of groundballs (hits, outs, DP, errors, etc) is virtually identical to the run value of flyballs. If you include HR, then it's no contest, and GB pitchers are better. That is, Bill's quote here:

"Given two pitchers with the same strikeout rate and the same walk rate, it does appear that we should favor the one who gets more ground balls"

That advantage is totally because of the HR. So, if you extend Bill's quote to pitchers with the same K, BB, and HR rates (or, in other words, the same FIP), then there is no advantage to GB or FB pitchers. Which is basically the point of DIPS.

***

Related thread on old blog.

Tuesday, February 26, 2013

Twins change term of “pitching to contact” to now become “attack the zone”

It's a semantical shift, but, when words can be used against you and taken out of context, it makes good sense to me.

"Pitching to contact" would have the implication of throwing the ball close to the middle, and if you miss your spot, it'll still be in the strike zone. 

"It does NOT mean, here, hit this ball," Anderson said emphatically. But that's what it looked like.

"Attack the zone" would have the implication of throwing in the strike zone (but not down the middle), and if you miss your spot, you'll still land predominantly in the strike zone.

"It's about attacking the strike zone," Anderson explained. "It means don't go out and start picking and trying to be too fine. Then you'll wind up pitching behind in the count. And if you pitch behind in the count, you'll get your ass kicked. It's about trusting your stuff."

"Nibbling" would have the implication of throwing at the edges of the strike zone, and if  you miss your spot, you'll land predominantly outside the strike zone.

That's why I had been insisting they should try "pitching to no contact."

"Think about that," Anderson said. "Pitching to no contact would be ball one and ball two. Throw it over. Throw it over. How many times have you heard that?"?

Saturday, February 23, 2013

(Almost) Complete History of DIPS

?A fantastic historical account of DIPS.  A required reading if ever there was one.

The one part I would add is that in the 1980's Bill James had DER, which is in essence 1 minus BABIP.  Bill did it at the team level, and never really explored it at the pitcher level.  And, I think, Eric Walker also had DER a few years before Bill (but I'm not sure).

Monday, January 28, 2013

Verducci Effect article for 2013

The only thing I've learned about the Verducci Effect is that there's alot of bright young guys out there willing to tackle a problem with eyes wide open.  This year, it's Pizza Cutter that gives it a go in this excellent piece.

To say that pitchers on the Verducci list are more likely to experience an injury is correct. However, it's actually the control group that is more likely to land on the DL. In fact, 29.2 percent of controls spent time on the DL compared to 24.0 percent of the Verducci group (not significant, for the record). Perhaps we might interpret this in context. Teams probably treat their young pitchers gingerly, especially in the context of having pushed them a bit in the previous year. Maybe they are a little more careful with minor injuries and push back a start here and there. But there's no evidence that Verducci Effect pitchers are more likely to sustain serious injuries. In fact, what evidence is there runs in the opposite direction.

...

Verducci status was not associated with differences year over year in strikeout rate (per batter faced), walk rate, home run rate, or ERA (i.e., his strikeout rate went from 12.1 percent to 13.2 percent for a difference of 1.1). This held after controlling for the previous season's number to account for possible ceiling or floor effects.

...

What should be apparent is that Verducci Effect pitchers are a very mixed bunch. There are plenty of ways to get on the list, and different ones might bode well, poorly, or neutrally for a pitcher's future. The problem of the Verducci Effect formulation is that the sample is far too heterogeneous to expect coherent effects out of it. Maybe the real frontier here is in breaking players down into sub-groups based on how they got onto that list to begin with. It's much more complex, doesn't fit nicely onto the page of a magazine, and it's the way that real research is done.

So here's to hoping that I don't have to resurrect this column a year from now. It's time to just admit that the Verducci Effect doesn't hold water and move on.

The one thing that Pizza noted is about looking at batters faced and not innings, because a guy can pile up innings by getting lucky outs from his fielders. So, why not just presume that Verducci meant to say 130 batters faced and not 30 innings? Or that he meant 500 pitches and not 30 innings? We talked about this last year in regards to Strasburg, where the game-by-game (presumed) control was based on pitch counts, but then, when we talked about the aggregate, it was always about innings.

And since Verducci threw out 30 innings as a proxy for 500 pitches, he was obviously not in a position to say 500 pitches any more than he could say 250 pitches or 750 pitches of increased workload. So, I don't think we necessarily want to hold him down to the 500 pitches.

There's plenty of great young saberists out there, including Pizza and all the other guys he linked to in his article (David, Jeremy, Derek). And Pizza noted a long list of potential biases, all of which are valid.

Really, the legacy of the Verducci Effect is how creative the saberists can be in trying to find something to the notion of forecasting the injury probability of a pitcher. For the aspiring among you, this is a great place to cut your teeth.