Posnanski’s Radical Idea
Pretty much everyone loves Joe Posnanski, including me. But while “Poz” has been perhaps the most prominent advocate of objective analysis among “mainstream” sportswriters, few would call him a “radical.” His blog post today, in the aftermath of a series of Tweets and counter-Tweets regarding Matt Thornton’s appearance in last night’s All-Star Game, however, paints a different picture. While he doesn’t explicitly say so, Posnanski implicitly has a pretty radical idea about what sort of pitcher belongs (and does not belong) in the All-Star Game.
While acknowledging that Matt Thornton is a good middle reliever, Posnanski does
not want to see a middle reliever, no matter how good, pitch with the game on the line in the All-Star Game. I just don’t. I’m not necessarily overjoyed by the way managers use their staffs, but I know the realities of the game — managers put their best pitchers in the rotation, and they put their best reliever in the closer role, and they take the best of the rest (or the young pitchers they are trying to break in) and put them in middle relief. Some thrive in that role. And their value tends to be undervalued by everyone … including me.* But that doesn’t make it interesting in any way.
This is set up by this perfectly reasonable point:
Middle relief is a specialty position. It’s extremely important in day-to-day baseball, but, to me at least, the All-Star Game is not day-to-day baseball. It’s a game for the best. And in a league that has Lee, Liriano, Lester, King Felix, Verlander, Price, Greinke, Weaver, Sabathia, Pettitte and many other great, great starting pitchers, I’m sorry, I don’t want Matt Thornton pitching in the most important moment of the game. I don’t buy that having that “experience” of pitching the seventh inning with runners on is of great value, not in this kind of game. Give me the best pitcher.
Poz is right: we want to see the best pitcher out there, and we know that relief pitchers are generally relievers because they aren’t good enough to start. Yes, most fans would rather see Zack Greinke or Jon Lester than Matt Thornton — they are better pitchers than Thornton.
The “experience” point is also excellent, and pushes this idea over into the “radical” category. After all, if the “‘experience’ of pitching the seventh inning with runners on” isn’t important in the All-Star Game, then really, why would “experience of pitching in the ninth” be any different? The “experience” issue is a major part of the basis for modern closer “position,” one of the myths that sabermetric analysis has been debunking for a while. Closers do what they do because, just like every other reliever, most of them couldn’t handle starting for some reason or another. This gets at the heart of Posnanski’s implicit radicality: if we want to see the best pitchers, why should any relief pitchers make the All-Star Game?
If we agree with Posnanski that Matt Thornton isn’t good enough to pitch in the All-Star Game, which relievers are left? Among qualified relievers 2008-present, only two had a better FIP than All-Star Scourge Matt Thornton (2.50): Jonathan Broxton (2.13) and Mariano Rivera (2.36). Has Thornton been lucky on home runs? Here is an exhaustive list of qualified relievers with a better xFIP than Thornton (2.84) 2008-present: Broxton (2.32) and Rivera (2.57).
Of course, Rivera and Broxton are both “proven closers,” but even if you buy that load of “clutch” nonsense (you shouldn’t), as Posnanski pointed out, that kind of “experience” shouldn’t matter in the All-Star Game. In any case, leverage accounts for just how important the situations in which the pitcher has been placed in have been. Wins Above Replacement (WAR) helps us properly value “mere specialists” by incorporating leverage into their valuation as well as how much they’ve pitched. Since 2008, Matt Thornton has accumulated 5.9 WAR. Which relievers are ahead of him? You guessed it: Jonathan Broxton (6.9) and Mariano Rivera (6.4).
Maybe guys like Rivera and Broxton could make an appearance in the All-Star Game, if all the starters were used up or something, but hey, if we don’t want to see Matt Thornton, one of the most valuable relievers (middle or otherwise) in the game over the last few seasons, what’s the point of having relievers on the roster at all? Heck yeah, I’d rather see King Felix pitch than Matt Thornton. I hope Posnanski uses his prominent position to advocate for the position logically implied by his arguments.
However, I suppose in one way Posnanski should probably be happy that Matt Thornton was the reliever inserted into the game. Can you imagine the backlash Posnanski would have endured from Royals fans in his hometown of Kansas City if he’d tweeted about the silliness of Joe Girardi putting in Joakim Soria, someone Thornton has clearly outpitched the last several years?
Matt Klaassen reads and writes obituaries in the Greater Toronto Area. If you can't get enough of him, follow him on Twitter.
Its possible that teams are valuing closers too highly and the Jonathon Broxton’s of the world would be more valuable to their teams if they started.
Possibly. You’d need to look back into the reasons why pitchers get moved to the bullpen. Do they fail as starters? Do they have trouble maintaining their velocity beyond a few innings? Do they have only have 1 plus pitch (or 2 good pitches) but lack a solid 3rd or 4th pitch to keep hitters off balance for multiple PA’s? Has injury / Have injuries forced them into a relief role?
There are multiple reasons why guys end up in relief. I would hypothesize that the majority are due to failing as starters or just not developing the arsenal needed to start, while a smaller % are moved to the bullpen prematurely (or held there after success at the MLB level).
Papelbon, Soria, Broxton, Rivera, Wagner, and I’m guessing a lot more were starters in the minors, with really good numbers, but the needs of their major league teams led them to where they are. I personally, wouldn’t punish a pitcher for not being a starter during the AS game since I believe that if you are going to use a closer, you better use a good one. Ask yourselves if you would want Ryan Franklin, Brad Lidge, or Byung-Hyun Kim closing those post-season games for you.
CesarV, with all due respect, Rivera posted very mediocre numbers in the minors:
Also, he was already 25 when he debuted in the majors. Very few players debut that late and go on to Hall of Fame careers (the only one I can think of is Sam Rice).
If you look at his age 25 season, and search for all Hall of Famers who threw 60+ innings, with an ERA+ between 80 and 90, you find two Hall of Famers of 42 who threw 60+ innings, vs. 34 who were over 110.
To Rivera’s credit, though, one of those pitchers was Christy Mathewson (ERA+ of 88 in his age 25 year).. But to not make the majors until age 25, and start off so slow, is really an uncommon path to such success.
Very possible. Let’s call Broxton a 3 WAR reliever (he’s on pace for more this year, but he’s never hit it before.) Here are some ways he could hit 3 WAR as a starter:
4.20 ERA in 180 IP
3.40 ERA in 120 IP
Could he out-do that? Good question.
Broxton was made into a reliever because he didn’t have as many pitches at the time and there were worries about his size affecting his longevity as a starter. That said, he has developed his command and control to the point that he could easily pass as a starter.
Hong-Chih Kuo is another example who had to convert to relief because of injury.
The ASG somehow keeps finding a way to get worse. Buck went on and on last night about how the presence of guys like Thornton (who is actually kinda good) and Infante and Kuo was a sign of how serious the teams are taking the game now. Which he also implied was of course a good thing.
I know ASG quibbling is an internet cliche, but when you think about all the ingredients of the recipe (fan voting for starters, the byzantine process that selects reserves, the one player from every team rule) coupled with the still curious way the game is managed, it’s really amazing that it would count for anything.
Then again, the NCAA is adopting a policy which will feature some higher seeded teams being forced to play more games than lower seeded ones, and the media absolutely LOVES IT. So it could be worse.
Kuo was selected, eventually, because he is one of the best relief pitchers in baseball, and has been since 2008. The reason he doesn’t either close or start is because of his injury history, not his effectiveness.
I’m sure Posnanski would be overwhelmed by the hate mail from all of the 57 Royals fans out there.
I hope he reads this. Good post Matt.
“…they put their best reliever in the closer role…”
Except that Matt Thornton has been the best reliever on the White Sox since 2008. What’s more, everybody knows it. Not just the stats guys but the casual fan. His manager seems to know it too. Guillen does want him “pitching in the most important moment of the game” and as a white sox fan, I do too.
I was shocked he was picked because I think most people feel as Posnanski does – if he’s not the closer, he’s not the best reliever on the team and so he can he be an all star? This site tends to be of the (correct) impression that having a designated ninth inning pitcher isn’t the best use of one of the best pitchers in the bullpen. Therefore, why should anybody be anything less than pleased when two major league managers -Guillen and Girardi – seem to agree?
I wholeheartedly agree with your point about relievers in general. If you look at the top 15 pitchers in each league, 14 of them are probably starters. If you look at relievers in general, virtually all of them are failed starters. Mariano Rivera started 10 games in 1995, Broxton started 50 games in the minors so even they weren’t acquired as relievers. Almost nobody gets drafted because he projects as a great relief pitcher.
come on, man. do you really think Guillen uses Thornton in the middle innings because he thinks he’s the best reliever on the team? there isn’t a single general manager in baseball today that constructs their team this way… and not a single manager in baseball today that uses their bullpen this way.
I think the point is less about who is the “best” reliever on the team, and more about who Ozzie is most comfortable with using. He goes to Thornton very frequently, and not just to get lefties out, because he has burned him the least amount of times.
Bobby Jenks has gotten his act together of late, but he can be a train wreck. Even when getting saves, he tends to make it interesting. Ozzie sticks with him in that defined closer’s role, no doubt. But having watched the majority of the Sox’ games over the past couple seasons, I definitely see a pattern where Ozzie tries to give Thornton some pretty important strings of outs to get.
Should Thornton have been on the All-Star team? No, I agree with that general argument. But the fact that Brian McCann beat him in a very important situation and won the game should not be an indictment of Thornton’s skill. It was one at-bat that he lost to a fellow All-Star, not necessarily something to be too ashamed of, in my opinion.
That’s exactly why he uses him in the “middle innings” and not as a closer. Sorry but you’re wrong in thinking that every team has their best reliever as their closer.
Thornton is the best White Sox reliever, and would be a waste if used as their closer. Saving your best reliever for only save situations is a waste.
For one inning at a time, Matt Thornton is one of the very best pitchers in baseball. ERA, FIP, and xFIP all below 3 the last couple of years. Strikes out 12 per 9 innings, lefty throwing 96 MPH.
He’s on the same level as the elite closers in the game. I don’t care if the White Sox use him for the 9th inning or not, the simple fact is Thornton is among the elite. He deserved to be there Tuesday night.
He got beat. Hindsight. Big deal. Unless it’s 2002 then somebody has to get beat in the allstar game.
Broxton started because they wanted him to get more comfortable with his off-speed pitches (which have become ridiculously good). They always knew they wanted to convert him.
Isn’t Posnanski only contending that he doesn’t want to see middle (not all) relievers pitch, based on the premise that teams “put their best reliever in the closer role?” It just so happens that the White Sox are one of the few teams that do not follow that premise. Though his argument still breaks down a bit if that’s correct.
I love this posts and this site, but almost every single piece that gets put up here has multiple spelling and/or grammatical errors.
Now that’s funny.
totally agree, there shouldn’t be any relievers on allstar game pitching staffs.
Isn’t there something to be said for the differences in skill-set between relievers and starters? I’m not talking about psychology, but differences in actual pitching ability.
Starters, by virtue of seeing the same batter a few times, generally have more pitches in their arsenal (e.g. 4 vs. 2), and have more endurance (let’s pretend to be scientific and say more slow-twitch muscles).
Relievers seem to be able to throw a bit harder and faster because they lay it all out there over the course of 20 pitches, leaving nothing in reserve (fast twitch).
Switching from one skill set to the other requires a period of “stretching out” over the course of a few games and in the other direction, perhaps learning to exert maximally (if you’ve ever gone from being a mid-distance runner to a sprinter, you know what I mean – even from 400m where you run as fast as you can, to 100m where you run even faster than that).
This suggests a difference in skill and training based on number of pitches you expect to throw. So if a pitcher is going to go just 1 inning, doesn’t it make sense to pick a reliever, as long as they’re (nearly) as good as whatever starter you have?
I was shocked when I saw Poz complain about Matt Thornton pitching in the game. Normally Poz makes some very good points, but this was ridiculous. Thornton has been very good for several years now. This felt like it was a complaint towards the role this pitcher normally has, rather than a complaint towards the pitcher’s talent level.
His point wasn’t whether Thornton has been very good or not, he’s saying that no matter how good the reliever is, he would rather see a very good starter instead, for instance F. Hernandez or Z. Greinke.
Even being a White Sox fan, i gotta agree with him. Although if i was trying to win the game, I would rather have Thornton pithching in that spot over most any other pitcher in baseball
I like how you conveniently choose the 2008 to present timeframe. That wouldn’t have anything to do with the fact that from 2007 to present, Soria has a WAR of 6.7 compared to Thornton’s 6.9 would it? It’s also worth noting that in the 2007 to present timeframe, Soria has pitched in front of what have been widely considered some of the worst defensive players in baseball. With the exception of DeJesus, he has had to rely on some terrible fielders. His WPA for the 2007 to present timeframe blows Thornton’s out of the water. Given that he has put up comparable numbers (peripheral as well, K/9, K/BB,etc.) while pitching in much higher leverage situations, isn’t it safe to assume that he is a much better pitcher than Thornton? Also, Thornton derives much of his value from his split versus lefties (career FIP of 4.00 against RHB, 2.93 against LHB). Soria on the other hand is equally dominant against either handed batter (career FIP of 2.98 vs LHB, 2.75 vs RHB). It seems clear to me at least that Soria is the more complete pitcher and clearly the better pitcher.
Actually, I used 2008-2010 because the FanGraphs sort function includes a “last three years” function, and I wanted to check all relievers. I guess I could have done 2007-2009, but oddly enough, I thought including 2010 might be more relevant to the 2010 All-Star game.
WPA is a good story stat, but as its creators will tell you, is a bad idea for valuing the contributions and skills of individual players.
*Jon Lester
So Ian Snell is a better pitcher than Mariano Rivera, and that’s why the logicians who run baseball included him in the elite group we dub “starting pitchers.”
I don’t think people are starters because they are too good to be relievers. Certainly elite starters are more valuable than elite relievers, because their performance stretches across so many more innings. But is there really evidence that one inning of Lester, Price or Verlander is better than one inning of Thornton, Rivera, or Broxton?
Why is POZ being so NEG?
One good thing about all star games is that starters don’t get used for as many innings as they normally would be used. A problem with a guy like Thornton is that this is likely how he would normally be used in a real game situation. I think we automatically assume that starters are at a disadvantage if they pitch in an all star game, but perhaps relievers are at more of a disadvantage because they are used as such. At least if you had 9 starters each pitch an inning, you’d be spreading the work load around and not hurting any one pitcher or team disproportionately (assuming similar pitch totals).
“Posnanski would have endured from Royals fans in his hometown of Kansas City if he’d tweeted about the silliness of Joe Girardi putting in Joakim Soria, someone Thornton has clearly outpitched the last several years?”
Why trash Soria to make your point ? and a poor example at that to back your point up.
Thronton has outpitched Soria last several years- in what universe -he’s never outpitched Soria.
Soria War 2010-2007 1.5 2.7 3.4 2.9
Thornton War 2010-7 1.3 2.7 2.3 0.5
if you play an analysts on fangraphs then get the numbers right
Your data is completely wrong about Soria. Get the numbers right. 🙂
He’s using BaseballProjection.com/Baseball-Reference’s numbers which uses actual runs allowed adjusted for quality of defense as opposed to FIP.
As noted above, I used FanGraphs numbers. One can make good arguments for using ERA/RA for pitchers (although I don’t agree with them) rather than FIP, but I think particularly in the case of relievers, were we get such a small sample size relative to starters, that FIP is the much better choice.
It kind of all depends on what “best” is.
If we’re talking about who has the most value over the course of a season, then the starters win. If we’re talking about who has the most value on a pitch per pitch basis then the relievers win.
To be honest, if you were trying to win a single game it makes more sense to load up on relief pitchers since they are going to better on a pitch by pitch basis.
If you are trying to showcase the most valuable players then you would pick all starters.
If you want to put on a show for the fans then you would pick the most well known players that the fans want to see play. Since there’s a lot of prestige in being a closer then, fairly or not, the fans want to see them pitch.
It seems like we end up with a mishmash of the previous criteria.
Thornton made a huge sustained leap in 2008. Since then he has a 2.71 ERA/2.45 FIP/2.70 xFIP to Soria’s career 2.13/2.86/3.27. I’m not arguing for one or the other, but Soria’s edge in WAR as calculated at Baseball-Reference.com and BaseballProjection.com is a reflection of a higher leverage index (1.9 to 1.4) and superior strand rate (84.6% to 78.1%).
I’m sorry, that was meant as a response to ribman. His was the last comment on the page at the time I posted mine so I’m not sure how it ended up as a response to you.
The minimum is three relievers. I want them to stop taking more than that.
Eh, as a random fan of baseball I love seeing some of the crazy good strikeout relievers. I mean, I love my Lincecum, King Felix, Verlander, Ubaldo, etc etc as much as the next guy. But the Broxtons, Marmols and Matt Thortons of the world are awesome to watch as well.
I totally agree – plenty of starters on the team already, I certainly think the very best relievers bring something to the table and deserve to be on the rosters, too. I mean, they’ve bloated them out to, what, 78 players per team now? How can you find no room there for a dominant, shut-down reliever?
(And I know, JoePoz’s point seems to be specifically about middle relievers; but Thornton has been good enough that his selection shouldn’t really strike anyone as ridiculous or out-of-place.)
Rather than get all bent out of shape about something small, I agree …
NO relievers on All-Star teams. There’s plenty of good starters that can pitch in relief.
Interesting that Strabwrg shouldn’t be in the All-Star team because he only has 30-some IP. How many IP do Broxton or Thornton have?
I love relievers, and as MGL routinely points out, an average reliever is superior to a tiring starter … But heck, none of that applies to an ASG … Where ALL of your delivers could be “Aces”.
I do believe, and have said it, that the “relievers are just failed starters” bit is said by morons. Elite relievers are more “specialist” than “failed starter”. It’s like saying Chase Utley is merely a “failed shortstop”. Dumb. Using a middle reliever or loogy as an example of reliever when we’re most commonly discussing closets is a subtle strawman.
Strasberg not Strawbrg. Stupid touch screen.
Overall good article. But no, Thorton has not clearly outpitched Soria over the last few years. That is wrong using any set of metrics.
MM
Meant Thornton sorry.
I used to like Posnanski, but last night on Twitter he was a yeast infected vagina about this.
I disagree that a reliever is to a starter what a pinch hitter or defensive replacement is to an everyday hitter/fielder. While both represent steps down in value, whether a reliever was always a reliever or a failed starter, it involves slightly different skill sets in a way that the batting/fielding equivalent doesn’t.
I agree with some of the points made above. Over multiple plate appearances and multiple innings and multiple games, of course the starters are more valuable, that’s why they’re starters. But if you had to get 1 out to win the world series, and all of your pitchers are rested, wouldn’t you send Broxton or Rivera or Marmol out? Not Lee or Strasburg or Halladay? On a per batter faced scale, don’t the best relievers look better than the best starters? Relievers focus on stuff and out pitches and throwing everything they have into an inning or two. Starters use more deception and subtlety and endurance.
I suggest a far more revolutionary idea: if it’s supposed to count, treat every inning like the 9th. If you’re the AL, roll out Rivera, Soriano, Thornton, Bard, Feliz, Valverde, and Soria. Give them all 3-5 outs. If you’ve got one game, and that’s all that matters, on a per batter faced scale, doesn’t that staff look untouchable? To throw out an ill-conceived metaphor, the best marathon runner in the world wouldn’t stand a chance against 100 good to great sprinters if they were allowed to function as one entrant in a race. Or something like that.
Some have argued that middle relievers are somehow better equipped to throw all-out for an inning or two, but the starters in the ASG were clearly able to do that. Price, Jiminez, Johnson,etc were all lighting up the gun at 3-4 mph higher than their averages.
You don’t need to convince me of the overrated-ness of relievers. But even if this knowledge were to pervade the mainstream, is it even possible at this point to get rid of the hallowed “closer” position, given the politics of the game? Unless every GM suddenly became more forward-thinking (and as a Mets fan, I have prayed one too many nights begging for this to come to fruition), as long as the players continue to have as much leverage as they do, it’s not happening.
If a closer by committee scenario were to be successfully implemented, one key cog in the machine would bolt to another team to become their full-time, overpaid “relief ace.” I feel as if relievers are also conditioned into having defined roles. Ballplayers egos are too big to not demand titles – closer, eighth inning guy, loogy, etc. The nebulous “reliever” won’t do. After all, to devalue the closer is to devalue any closer. Because of Mariano Rivera’s untouchable status, debunking the myth of the relief ace is a complicated task. If Posnanski takes up this endeavor, I wish him luck. No one is in a better position to do so.
“, most of them couldn’t handle starting for some reason or another. This gets at the heart of Posnanski’s implicit radicality: if we want to see the best pitchers, why should any relief pitchers make the All-Star Game?”
One thing does not follow from the other. The fact that a guy can’t handle being a starter doesn’t mean that he can’t give the single best inning performance. A guy may be no good throwing 93, but terrific throwing 97, and only throwing one inning allows him to throw harder, whereas some guys when they throw harder, things flatten out.
Baseball has been around for over 125 years. Sometimes the stats geeks forget they are not the first people to debate issues in the game.
Whatever the debate, whatever the discussion regarding the use of closers/middle relievers/starters, one rule of thumb should be not to mention WAR.
That being said, some elite relievers have a place in an All-Star game. Maybe I’d include a Matt Thornton among them. The fact is, relievers are used to pitching from the stretch and only working one-inning. When you’re talking about one-inning, the gap in value between an elite reliever and a starter is somewhat bridged – because in that one inning, the strengths of a relief pitcher are accentuated, and his weaknesses less manifest. For one inning, I’d rather have Mariano Rivera on the mound than the Cy Young winner in most given years.
Poz is wrong. Why isn’t the occasional, talented middle reliever who makes the all-star game a charming part of its story?
We can’t afford a couple of players on that big stage who we might not know much about, or might not be all that desperate to see in action but who, solely because of their manager’s use of them and not their lack of talent, consigns them to an underappreciated role and therefore anonymity? Please.