When I first started this site, I was 26 years old. The first year I ran the site, I was driving 45 minutes back and forth from Lynchburg every night. The second year on the site, I finished the season with a driving tour that took me from Charlottesville, VA to Richmond, Bowie, State College, Williamsport, Altoona, West Virginia, and back to Charlottesville, all in the span of two weeks. I don’t consider my current age of 31 to be old. But it is old enough to make me wonder how I ever did all of that driving, with very few breaks.
I spent most of today in the car, driving the near 700 miles from Florida to Raleigh, which is my official “I can’t make the drive to Virginia in one day, so this gives me an excuse to stop in Raleigh and see my best friend” trip. I’ll be heading up to Virginia tomorrow, which will hopefully be my travel hub for a long playoff run. Then I’ll be going up to Pittsburgh on Tuesday to start my Wild Card game coverage.
Tonight’s article is going to be short, because I’m exhausted. But I did get to spend the entire day in the car, which consisted of about three turns through the Guardians of the Galaxy soundtrack, listening to the Pirates game, hearing about the Steelers loss in a call from my Dad shortly after it happened, and a lot of focus on who will start the Wild Card game in Pittsburgh. For the purposes of this article, we’ll focus on the latter.
There were rumors today, along with a lot of speculation, that Edinson Volquez would be the starter in the Wild Card game for the Pirates on Wednesday. Volquez is currently listed as the starter on the Pirates’ main page as of this writing. It appears that he will be the guy, which suggests that the Pirates don’t want to go with Liriano on short rest.
I wrote over the weekend that I think Liriano is the clear option for the Wild Card start, even on short rest. If he’s healthy, that’s who I’d start. Volquez has been great this year from a results standpoint, but the advanced metrics say that he won’t keep this up. That’s true even in his final two months, which have produced an impressive 1.63 ERA, but a 3.87 xFIP. The Pirates are taking a gamble here that Volquez can hold on for at least one more game without the regression kicking in.
You might be able to go back and second guess the decision to go for the division, which made Liriano and Gerrit Cole unavailable for the Wild Card start. I think it was a good move to go for the division. It could have been won with a few key hits by Arizona, and a few key hits by the Pirates.
So now the Pirates are left with Volquez, unless they eventually decide to go with Liriano on short rest. The flip side to this is that if they win, they will have Liriano and Cole for the first two games against the Nationals, along with one of those starters for a possible game five. That’s a big if, though, as it would require a win from Volquez.
We’ll probably find out more on the Pirates’ strategy on Tuesday. For now, the matchup looks to carry some risk. If they can get the “results” version of Volquez from the last two months, then they will be in great shape. If they get the “xFIP” version, then the Pirates had better hope that their top offense shows up against Madison Bumgarner, because they’ll need to provide some support.
Links and Notes
**Improved Fastball Command Propels Buddy Borden in 2014
**2014 GCL Pirates Season Recap and Top 10 Prospects
**Morning Report: West Virginia Could Go From First to Worst
Tim started Pirates Prospects in 2009 from his home in Virginia, which was 40 minutes from where Pedro Alvarez made his pro debut in Lynchburg. That year, the Lynchburg Hillcats won the Carolina League championship, and Pirates Prospects was born from Tim's reporting along the way. The site has grown over the years to include many more writers, and Tim has gone on to become a credentialed MLB reporter, producing Pirates Prospects each year, and will publish his 11th Prospect Guide this offseason. He has also served as the Pittsburgh Pirates correspondent for Baseball America since 2019. Behind the scenes, Tim is an avid music lover, and most of the money he gets paid to run this site goes to vinyl records.
The fact that Liriano sucked in his last start has no bearing? Volquez is the better choice, although Cole should start the NLDS assuming Volquez wins.
Volquez is the better choice?? I disagree completely!!
You are more concerned with make believe results than actual results. That’s why I can’t stand guys who base decisions solely on sabermetrics. So if Eddie goes 7 shutout in wildcard game, you will probably say the Pirates should have lost because of a stat based on an estimate and an assumption, not my words, but the actual definition of the stat according to FanGraphs. You know, in a make believe baseball game where there are no fielders. You are basing your decision making on this formula FIP = ((13*HR)+(3*(BB+HBP))-(2*K))/IP + constant – rather than actual results.
Explain that formula to me. I’ll go with what my eyes and actual stats that you don’t need a physicist to try and explain, thanks.
I am not opposed to considering xFIP, FIP, etc., as additional sources of information and perspective, but it does seem inconsistent that saber metric fans on one hand rave about Pirate defense and shifts, etc., but then want to predict the success of a Pirate pitcher by divorcing his performance from that same defense. What is logical about that? That is especially ironic because saber metrics, as I understand it, is suppose to be logic driven. Without having thought about this a whole lot, it strikes me that one of the most important and distinctive characteristics of a good pitcher is his ability to strand base runners, particularly those in scoring position – FIP and xFIP largely ignore that characteristic.
According to FanGraphs, “FIP is a measurement of a pitcher’s performance that strips out the role of defense, luck, and sequencing. . . .” In trying to predict how well the Pirates and Volquez are going to do against the Giants on Wednesday, i.e., how well they are going to defend on Wednesday, why would it be helpful to eliminate consideration of the defense, and look only at Volquez’s past statistics independent of the defense that is playing behind him on Wednesday?? I also do not understand the assumption that “sequencing” is beyond the pitcher’s control. To the contrary, it seems to me that good pitchers are all about “sequencing” and are masters of that characteristic – when there is a runner on and particularly when there is a runner in scoring position the good ones get better and throw tougher pitches, i.e., they pitch better in the clutch. Why is it logical to eliminate consideration of that factor – a factor that many experts consider the essence of good pitching – when predicting the likelihood of success by a pitcher? Why would we track a batter’s average w/risp and then seek to ignore that same consideration when evaluating a pitcher? I also have some thoughts about “luck” – but enough already!
No statistics are going to very helpful in predicting what will occur in a single game, luck, timing, and other factors predominate, this is why Baseball Prospectus scrapped its’ playoff prediction model years ago. In isolation assuming equal rest, Cole or Liriano compared to Volquez on average is going to improve the Pirates chances of winning a single game with the probability shifting from 52% to something like 58%.
Average with RISP is not what one would consider a saber-metric statistic, it is important for cataloging what happen in a prior game, but isn’t of any predictive value. In addition there isn’t a sustainable skill to strand runners by pitchers LOB% has one of the lowest correlations of any pitching metric.
You mention the Pirates defense and shifting, it has an impact, but it is hard to demonstrate because we don’t know what would have happen in the absence of that shifting. Looking at the difference between FIP WAR and RA-9 WAR (runs against) there really is a pattern.
Also, good shifting teams reduce batted ball average by about 10%. EV is 40 points of batting average better than he should be, shifting or not.
10% I have not seen any like that at all?
I know that with three successive players on one side of the bag BABIP drops to .267, and with three players one side, but SS or 2B more up the middle BABIP drops to .296. But those account for about 6,000 of 100,000 balls in play.
Very interesting, thanks for the reply! I’m out of time right now, but maybe we can continue the discussion later – hopefully as it relates to the Pirates being in the NLDS, for example!
FIP is looking for what components of past performance are the best indicators of future performance, and it requires 4th grade level math skills to calculate.
The human mind is not a great objective elevator of information and the eye test, whatever it means, is extremely limited in that what sample size of actual games can one person watch? Then there are the issues of what events are remembered and then recalled.
You are making an argument against a position that no one advocates for.
I completely agree that calculation of the FIP number using the formula is elemental, but I think the underlying assumptions about the construction of the formula, the value of the output from the formula, and the use which should be made of the output are much more complicated.
What that stat basically says is that Volquez walks 3x more people than Bumgarner but only gives up 2/3rd the # of HR based on % of fly balls. So if Eddie gets a couple of DPs (or a TP!) behind him, it’s about even.
Volquez allowed 9 homers in his first 9 starts.
Volquez allowed 8 homers in his last 22 starts.
His K%-BB& stinks, but he’s likely to keep them in the game by keeping the ball in the park.
His BB are exactly what I am worried about. Putting men on base is a bad deal especially in a one game playoff against a high caliber team. The Giants have hitters who will make EV pay for inefficiency and BB.
“Should” Bumgarner beat Volquez based on projectable statistics? Yes, if the Bucs were an MLB-average offense. But they’re not.
If you want to get into one game situations against a specific pitcher, keep in mind Bumgarner lives high in the zone, not on the low outside corners. He is the type of pitcher I’d expect the Pirates to have success against. Now, if Bumgarner increases the # of pitches he throws inside, especially under the hands of RHB, then he’ll find more success. But going from his standard heat maps, I’d expect guys like Harrison, Marte, Gaby, Mercer and Martin to have good contact (and in SSS, they have).
The numbers bear it out – Career numbers for Pirates against Bumgarner:
Harrison – 4 for 5
Marte – Never faced him
Gaby – 3 for 9
Mercer – 3 for 7
Martin – 3 for 7
That’s a combined 13 for 28 (.464 AVE). Even Neil Walker batting right handed is 3 for 9 career against him.
Lets hope it works out that way. For some reason I was much, much more confident last season going into the WC game vs. Cincy.
I would have had Cole take Sunday off, but that’s out the window now.
I’m not a fan of short rest, if only because they haven’t asked him to do that all year. Volquez it is, and I’m OK with that.
I know Liriano is on short rest, but the Pirates have only two guys who can absolutely dominate another team’s lineup, and Volquez ain’t one of them.
The Braves, Brewers, Cubs, Cardinals and Reds all called to say “Eddie can absolutely dominate a lineup”. And that’s just been in his last 6 starts.
I’d rather have Volquez than Liriano, and the way Liriano has been pitching lately, I don’t trust him for the wildcard game anyway. I wish it could have been Cole, he would be the clear go-to guy. But I’m not worried about Volquez. I believe he will hold the Giants to 3 runs or less in seven innings. It’s really up to the Pirates bats, that is what got them this far.
I have no clue how you or anyone else can expect EV to go 7 with 3-runs or less. His peripherals say that just isn’t to be expected. Maybe he can keep the lightning he’s caught in the bottle and go with it…but to expect anything more than what the numbers say to expect is wishful thinking.
“I have no clue how you or anyone else can expect EV to go 7 with 3-runs or less.”
Because he’s somehow managed it in 1/3 of his starts this season (11 of 32) including the last three?
I’d even say with the extra bullpen for the WC game, all he needs is 6 IP with three runs or less. Which he’s done in over half of his starts (17 of 32) including his last six.
For comparison, Liriano has gone 7IP with 3 runs or less in 6 of 29 starts and 6IP with 3 runs or less in 14 of 29; Cole has gone 7IP with 3 runs or less in 8 of 22 starts and 6IP with 3 runs or less in 13 of 22.
The next time this year that the Pirates start a pitcher on short rest will be the first.
Literally. The only pitcher to start a game after less than 4 days of rest was Volquez, who started on April 6 after pitching two innings in relief on April 3 in what was essentially a rehab outing.
Volquez is starting Wednesday. Bank on it.
To be fair though, the next game the Pirates play will be their first playoff game of the year. It’s definitely different than regular season.
But I agree completely that it will be Volquez, especially since they announced him. If there was any indecision on their part, they could have waited until today (or maybe tomorrow) to announce it. And it’s not like Frankie has been lights out his last 3 outings. Oh the walks!
The only real decision to make is who starts game 1 in the next series, and thus gets 2 starts if necessary. I think I’d go Cole, but I’m fine with Liriano
And here I was hoping for a First Pitch review of the Guardians of the Galaxy soundtrack.
Oh, well. Maybe tomorrow.
OTOH, Bumgarner’s xFIP for September is 3.89 and his K/9 dropped below 7.
I don’t think I’d want Liriano on regular rest let alone short rest based on his performance on Saturday.
Martin is a big key as Bumgarner lives high in the strike zone, an area Martin loves to attack. Although Stewie has been a huge surprise and is the new designated LHP killer. Not sure if Volquez has pitched to anyone but Martin this year?
xFIP is no longer predictive over such a small sample. I think you need like 50 innings just to have walk rate stabilize. Here’s the interesting part..
Jeff Locke’s first half of 2013: 6.03 K/9, 3.88 BB/9, 52.3 GB%, 2.15 ERA, 4.21 xFIP
Edinson Volquez’s second half of 2014: 7.38 K/9, 3.75 BB/9, 52.8 GB%, 2.20 ERA, 4.13 xFIP
They are almost identical in the “true talent” area. We should be no more confident in Volquez as we would be in Jeff Locke circa 2013 for this game.
Save me some time doing research, when you say “xFIP is no longer predictive over such a small sample” are you referring to the WC game as the “small sample” – in other words, are you saying xFIP is not a very valuable predictor of how Volquez is going to pitch on Wednesday? I’m also curious and interested in the second point – except maybe for the psychological aspects of dealing with the pressure of a big game, why is it a problem if we were to get pitching performance from EV on Wednesday along the lines of Locke’s first half from last year when he was an All-Star? (Hopefully, EV’s being a veteran and Locke’s being a rookie last year will allow EV to handle the pressure of the WC game). Maybe if we should lack confidence in Volquez that is a suggestion that we should be concerned about xFIP rather than his actual performance; however, that suggests that xFIP is predictive over a sample size of one game (Wednesday’s game)?? Maybe you’re simply agreeing with Tim that we need to hope for the “results” EV rather than the “xFIP” EV?
Well there’s quite a few things here..
First, I was referring to Bumgarner’s September numbers (xFIP) being higher than we’d expect. It’s only natural for that number to change over a small sample and it’s not very predictive. For example, his season statistics would be a much better indication of how good you’d expect him to be.
That ties into the Locke/Volquez situation. Just because their ERAs are remarkably low, it doesn’t mean they are pitching well. They both had pretty low strikeout numbers paired with pretty high walk numbers. That’s not a good way to keep your ERA down. FIP takes in strikeouts, walks, and groundball rate to predict ERA. Essentially, there’s pretty good evidence that a pitcher cannot control other factors like batted balls in play average.
Volquez has such a low ERA because his batted ball average is extremely low (.263 this year). It should be about 40 points higher as guys will trend toward .300. xFIP gives you the true talent of a pitcher by taking out “luck factors” like batted ball average. In the grand scheme of things, Locke pitched over his head last year before the break and xFIP would have told you that he was extremely lucky when his ERA was two runs better than it should have been. It caught up with him big time in the second half, as we know. He didn’t really pitch any different, just regressed towards his true talent of a 4.20ish ERA.
The same exact thing has been happening to Volquez the second half of this season. His balls in play are finding gloves, a lot more gloves than they should which has produced an ERA much lower than it should be. Now, he could go pitch a shutout in this game (hopefully!). But his true talent would indicate he’s been lucky and there will be regression at some point soon. That doesn’t mean he can’t pitch well either; just to this date he isn’t as good as his ERA suggests. Sorry, I know this is a lot to take in.
Hey, thanks for taking your time to do a great explanation! Sorry, I missed that you were referring to MadBum in your first two sentences – I thought you were referring to the predictive value of Volquez’ xFIP concerning his likely performance on Wednesday. Looks like Volquez has the biggest disparity between ERA and xFIP on the staff – adds even more interest to the game. Presumably, given the couple data points that I am aware of (EV’s stats for (1) the whole year and (2) for the last two months (the latter provided by Tim in his article); however, I understand that only two data points could be deceptive) that EV’s disparity between ERA and xFIP generally has continued to increase all year – I would think that is even more unusual. Given the numbers provided by Tim, EV has a variance of well over 200% over the last two months between his ERA and xFIP – for the season, that same variance is approximately 38%. Crazy
It is quite the disparity. If we look at this another way, though, we might be able to explain it a bit. Throughout his entire career, he’s pretty much been “unlucky” or worse than his metrics would suggest. Even with this year, his career ERA is 4.44 with a 4.17 xFIP. Maybe it’s not out of the realm to think he’s having his positive regression in one huge “lucky” year to catch up with his ERA. There’s probably better evidence (larger sample) that this may be the case compared to just looking at one season of data.
It’s so funny that with the MVP on our team and with Harrison and last years Comeback Player of the Year…we’re talking about Russ Martin. I LOVE it. I loved it when we signed him bc I loved his makeup…and heading into the huge game Wed it’s awesome that we’re talking about it being Martin as the likely difference maker. I love it.
Btw, no one seems to be talking about Bum vs. Cutch…and Cutches “struggles” this year against LH pitching.
Got to win the whole thing. With that in mind I go with Volquez and no one on short rest, especially in a play-in game. That allows you to go Liriano/Cole in 4 of the potential 5 games against the Nats and Volquez in Game 3. That is our best shot to get to the NLCS.
Oops, my bad. With games on Friday, Saturday, Monday, Tuesday, Thursday you will need 4 starters if no one goes on short rest. I still go Volquez in the WC Game. I then go Cole/Liriano/Volquez/?/Cole. I haven’t taken the time to see who has the best career splits against the Nats, Locke or Worley. If Locke not dominant against them I’d go with Worley.
Why would you switch Cole and Liriano? Liriano pitches before Cole in the order but you have it reversed.
I think the more you can pitch Cole the better. He’s pretty dominant right now. Remember last year in the final game against St Louis, Hurdle choose Cole over AJ even though it was AJ’s turn.
Worley is looking fantastic all of a sudden.
Have all the faith in the world in Volquez. Liriano hasn’t looked all that sharp to me these last three games despite what the numbers might say. It has taken him damn near 100 pitches to get through five innings. No way I’m starting him on short rest. Volquez has looked good. Plus this game’s at PNC. I think we win this game.
Agree on all points. Hot hand is what matters right now. This staff hasn’t been consistent enough to have a true “ace” this year.
Whomever toes the rubber Wednesday may pale in comparison to which player is behind the dish. I certainly hope Martin is healthy enough to play. If not, the Pirates may be relegated to have Sanchez back there, which is a much bigger concern to me.
Unless Martin is chained to a hospital bed with armed security watching him, he’s definitely going to play.
Even then I think it depends on how many armed security guards there are. 4 or less, he still plays.
Haven’t heard anything about Martin except day-to-day. We will know when the rosters are released. Martin playing is the key in my opinion.
Stewart’s wrist was not broken on x-ray and he’s listed as day-to-day with soreness according to MLB.com.
Go with the hot hand – Volquez. And get Worley involved early in the NLDS.
I would go with Liriano on short rest. If for no other reason, he has the experience of pitching this game last year. Volquez has been great for us but I am very concerned about this game.