RSS Twitter Facebook YouTube
Expand Menu

Second Thoughts Game #110: Texas 0, Cleveland 2

Second Thoughts Game #110: Texas 0, Cleveland 2
Terry Francona & T.J. House (Photo: AP)
August 3, 2014
Share via: Share: Facebook Share: Twitter Share: Google Share: Pinterest Share: Print Share: Email

T.J. House as Dallas Keuchel

Before the 2014 season, every comparison to Houston Astros players – without exception – was either in positive reference to former players like Roger Clemens or was a pejorative. In 2014, Houston still has the third-worst record in the league, but at the same time, the Astros team has surprising bright spots – foremost among them the sinister specter, Dallas Keuchel.

Dallas Keuchel is perhaps the most favorable comparison for Cleveland Indians left-hander T.J. House, the sole southpaw in the Indians’ rotation. T.J. House was a largely-unheralded prospect; Keuchel even less so. Yet both of them share one similarity: an incredible propensity for groundballs. In Keuchel’s breakout 2014 campaign, his walk rate (6.0%) has been good and his strikeout rate (18.7%) sufficient, but his truly distinguishing characteristic is his extreme reluctance to surrender home runs. Fueling this is Keuchel’s truly extreme groundball tendency; at 61.8% of all batted balls, Keuchel leads all ERA-qualified pitchers in the majors in ground ball rate. Consequently, Keuchel has the ninth-lowest home runs per nine innings (0.46) of all ERA-qualified pitchers, a rate even better than Corey Kluber (0.57). Overall, at an ERA of 2.79, a Fielding Independent Pitching (FIP) of 3.11, and an Expected Fielding Independent Pitching (xFIP) of 3.19, Dallas Keuchel would be a strong #2 pitcher on a playoff team – an extremely valuable asset indeed.

Yet for Keuchel’s 2014 campaign to be a ‘breakout,’ it requires a substantially less-successful past. In 2012 and 2013, Keuchel had run an ERA over 5, fueled in large part by his home run rate, which remained above 1.00 HR/9 IP in spite of a ground ball rate that remained near the top of the league. What this meant, in turn, was that Keuchel’s initial Home Run to Fly Ball ratio (HR/FB) was well above league average, at an anomalously high 15.9% in 2012 and at a yet higher 17.4% in 2013. Given that HR/FB hovers around 10% for the majority of the league with very deviations of even 2% being dramatic, Keuchel’s HR/FB ratio remaining consistently above 15% was either exceedingly unlikely or indicative of the fact that Keuchel was not a major-league caliber pitcher. Most would have assumed the latter; Keuchel’s 2014 HR/FB ratio, however, at a better-than-average 8.6%, has put the assumption to lie.

Because pitchers are snowflakes, because House has velocity that Keuchel cannot match, and because Keuchel’s career and minor-league command numbers are far better than House, the comparison is not one-to-one. If House could replicate Keuchel’s 2014, moreover, Indians fans should be ecstatic, as that would make him a solid #2 on a playoff team rather than the fringe #5 starter he has been to date. The comparison is imperfect for a number of reasons. That caveat granted, however, it’s an intriguing line of comparison to pursue.

Like Keuchel, House’s groundball rate is extreme – in fact, at 61.7%, House’s GB% is effectively identical to Keuchel’s 61.8%, with Saturday’s game serving as an extreme manifestation of this theme – of the 12 balls House allowed in play on Saturday, 11 were groundballs, for a game GB% of 91.7%. In turn, this would imply that House is quite capable of repressing home runs, yet fueled by an astronomical HR/FB rate, T.J. House’s home run rate remains at a far higher-than-average 1.15 HR/9. Entering Tuesday’s contest, House’s HR/FB ratio was an astronomical 21.9%. After a homerless game on Saturday, House’s HR/FB declined imperceptibly to 21.2%, a decline that was so amusingly low because House allowed only one fly ball on the night.

The conclusion one might draw from the elevated HR/FB ratio might at face value indicate that House is not a major league pitcher, but one must recall that HR/FB is extremely volatile and that, even in the bandbox of Columbus’s Huntington Park, House posted quite passable home run rates. If, soured by the performance of Salazar, one regards the gap between major- and minor-league competition as too pronounced for there to exist any correlation between the two, one need only look to Keuchel to recall that extreme groundball pitchers can run better-than-average HR/FB rates.

T.J. House’s outing was imperfect but left very little room for serious complaint. With an improved line score of 5.0 IP, 0 R/ER, 3 H, 2 BB, 7 K, 97 pitches, 67 strikes, and 9 swinging strikes, House had a scoreless outing, pounded the strike zone, and induced whiffs at an average-to-above-average rate. What he failed to accomplish in terms of length largely resulted from his 24 fouled pitches, twelve of which came in two-strike counts. Ideally, the orthodoxy proscribes, starters should give length. Much of House’s failure to do so, however, stemmed from factors that do not reflect negatively on House; foul balls, because they have the possibility for a foul fly out with no possibility for a hit, are very much beneficial. That House was prevented from going deep into the game as a result is not a condemnation of this philosophy; after all, House did pitch five shutout innings in a game wherein one Texas run might have changed the entire tenor of the game.

House is an exceedingly intriguing pitcher moving forward. For House to experience a home run repression breakthrough is neither likely nor impossible. Given his ground ball rate, the bat-missing and control displayed on Saturday would be more than adequate given a HR/FB ratio that was even slightly worse than league average. In spite of the substantial uncertainty surrounding the back end of the rotation, House has the potential to recreate himself anew, no longer disadvantaged by Asdrubal Cabrera at shortstop; going into 2015, a strong argument could be made that House’s groundball skills would play exceedingly­ well in front of an elite defensive shortstop prospect such as Lindor, as well as a third-base situation no longer plagued by doubt.

The following numbers are not guarantees of future success, but Expected Fielding-Independent Pitching (xFIP) is a better predictive mechanism than ERA or FIP, and are scaled to ERA. The HR/FB ratio remains, but what follows is what would amount to T.J. House’s company were his HR/FB to decline to league average levels.

Season xFIP Numbers

T.J. House: 3.51
Jake Odorizzi: 3.55
Anibal Sanchez: 3.72
James Shields: 3.59
Adam Wainwright: 3.43
Scott Kazmir: 3.40

John can be reached on Twitter at @JHGrimmHe can also be reached by e-mail at

User Comments

August 4, 2014 - 1:51 PM EDT
I don't see House developing into a Kuechel (although I would love to be wrong), but he could turn into a good 3-4 starter with an elite ground ball ability. It all kind of depends on the defense behind him. If we had Lindor at short, Jose Ramirez at 2nd, Gio Urshela at 3b and Santana at 1st, then House could be a Kuechel like pitcher. But if we have Lonnie at 3B, Aviles at SS, Kip at 2B, and Swish at 1B, then he's a fringe BOR guy like he is currently. With pitchers like him, it all depends on who is behind him, and thats one of the primary reasons House had frankly, not been great this season.
August 4, 2014 - 4:41 AM EDT
As most everyone knows, I am not a great believer in WAR as an evaluation tool which I believe is roughly the equivalent of BABIP and fortunetelling. When it comes to assessing ML players, I do see some value in those assessments. I think John's piece is good and inclined to agree with Tony's post.

Few on this site remember but Tony and I worked with Art and two others in evaluating the 2008 draft. I had seen him in HS and thought he had two quality pitches but little projection since he was a big lad even then. As I recall, we evaluated him in the top 20 of Indian prospects. I am not sure that Tony ever waivered but I did not really jump back on the wagon until House was at Akron.

Never underestimate a LH who can locate his fastball and can make hitters miss with his breaking ball If the change is working to RH batters, then you likely have someone wo will find a home in a ML rotation. I still don't see a lot of projection but I see a LHP I am pleased is in Cleveland's organization and believe John expressed it well.
August 3, 2014 - 10:30 PM EDT
Nice piece by John and I like the analytical aspect he brings. House is an interesting pitcher and someone who may be of more use than we think, and this piece kind of explains why he is so intriguing.
Billy Lybarger
August 3, 2014 - 10:28 PM EDT
My comments have not been an indictment of advanced statistics or sabremetrics at all. I have read for several weeks John Grrimm's sophmoric misapplication, or misunderstanding of this obvious recent exposure of his to data, and have several issues with his analysis. He has also wontonly thrown out phrases about pitcher's careers that are much too incendiary for someone with such a blatant and clumsy misunderstanding of the facts.

Mr. Grimm stated Dallas Keuchel's "career and minor league command numbers are far better than (TJ) House." Now, unless Mr. Grimm is in fact, a unicorn in human form, I seriously doubt he possesses the two pitcher's command numbers. As recently as this Spring Training Dave Cameron of Fangraphs wrote of pitcher's command being of particular interest to him because no one keeps public data of such stats. I am going to make the assumption Mr. Grimm being the neophite he is, meant control numbers.

However, a more than just passing examination of the two pitcher's Fielding Independent Pitching (FIP) numbers show Keuchel is not the far better control guy. Keuchel owns a 4.18 major league and what I calculated to be a 3.49 minor league FIP, and House has posted a 4.41 and 3.76 (calculated) respectively. The innings pitched comparison gives a decided advantage (375 to 55) to Keuchel at the big league level, where TJ's bit of bad luck for his higher than career norm HR rate has inflated his FIP. The regression is expected as shown in TJ's xFIP being better at 3.51 vs. 3.79 for Dallas. The picture for TJ may actually be a little rosier for House, as his career K/9 and BB/9 are both slightly better than Keuchel.Comparing the two minor careers shows similar numbers for HR rates, with Keuchel surrendering less walks (near great) and House much better at strike outs.

This backhanded complimentary piece by Grimm is in stark contrast to one he wrote just weeks ago, where he predicted doom and gloom for TJ. Still, after House had arguably his best MLB start to his very young career, he dismisses the nice jump in strikeouts as average to above average. A 12.6 K/9 is great, but acknowledging it would further show how out of touch his post of July 6th (where he dismisses House with an "extremely unimpressive strikeout rate,").

Mr. Grimm is exactly what is wrong with young guys joining the punditry with a rudimentary understanding of advanced statistics, and no appreciation of scouting reports that help paint a more complete picture of a pitcher's value. He dismisses House as "largely unheralded" when in fact TJ was a top 20 talent when he was signed away from Tulane University. I guess only the Indians were stupid enough to throw $1,000,000 after an unheralded kid. House struggled in the lower minors, before changing his off season workout routine (actually documented on this site a few years ago) and reinvigorating his minor league career. Since then House has been able to be promoted and figure out how to be successful at every MiLB stop he's made. Most likely only those idiot Indian FO guys have been able to appreciate any of that as well.

I hope this retort shows I was not being snotty nor dismissive. I think it is plain to see which writer is much more resembling those accusations. I would advise Mr. Grimm to keep gaining an understanding of the data, but not be hyperbolic when analyzing said data. He also should know his audience and realize most people here have been following these players for a long time, so we might have a bit of a defensive attitude stand when some new guy wants to take a limited amount of advanced stats and start taking pot
shots at our guys. pot shots at the
August 3, 2014 - 8:53 PM EDT
wasn't talking about you being snooty, your tribe. just missing a chance to better understand why things happen.
August 3, 2014 - 6:09 PM EDT
yourtribe, we're not talking about analyzing high school or college players here, where anyone who's major league caliber should be dominating the competition, these are big league pitchers. Pitching in front of Cleveland's defense and pitching in front of, say, the Royals defense at Koffman Stadium, are very different things, so I don't know why you'd only look at those 3 stats that a.) aren't real predictive of future success and b.) are dependent on variables like defense and park that are outside of the pitcher's control. Stats aren't the end-all or the only thing you should look at, but it doesn't make sense to ignore some and just look at others. If you were looking at ERA, whip and BA, then I'm guessing you probably didn't think it was a bad idea when the Indians passed on Kazmir, Had you looked at xFIP, his k/BB rate, the velocity of his pitches, well, you'd have probably thought differently. I remember saying before 2013 how it was a horrible idea to sign Brett Myers, because of his velocity, low strikeout numbers and his poor performance even as a reliever, and said that Kluber would likely be better. And pretty much no one agreed, because they were looking at Kluber's 5+ ERA the previous year. Just curious, what predictions are you making based on pitchers' ERA, baa and whip? You think the indians should maybe target Josh Beckett for next year? He's got some good numbers in those categories. Or do you think someone should pay Chris Young more than Max Scherzer? The obsession with WHIP is funny to me, it's the sabermetric stat that people who say they don't use sabermetrics like to use all the time. Even though there are better things to look at.
August 3, 2014 - 3:42 PM EDT
not being snooty at all. just find all the stats overkill. which they are. dont' over complicate this, it's baseball. not building a nuclear reactor. (just like the nfl combine. biggest waste ever. just watch game tape.) you cannot predict talent by plugging in stats into a computer. and for every kluber, there are 25 guys we took in the first few rounds in the last 10 years or so who flat stunk. luck of the draw.

starters for example. era, batting avg. against and whip. those three are about all you need to know. era can be a bit misleading, but no one is taking a college pitcher in the first round if his era is 5, his batting avg against is .285 and his whip is 1.8!!! don't care what his other ratios and stuff is. fact.

hitters, batting avg, hitting with risp and k/bb ratio.

the team we put together in the 90s had nothing to do with computers. they were just good ball players that were scouted and panned out. we got lucky. may not happen again in 50 years. it's been 20 and counting and we are not close. our drafting over the last 15 years has been awful as is well documented. computers didn't help squat.
August 3, 2014 - 2:07 PM EDT
no offense taken, though i think you're missing some important insights. this same story could have been - and probably was - written about kluber a couple of years ago (how production had yet to catch up to his peripherals).

but someone's reasoned analysis deserves better than a dismissive, snotty shot.
August 3, 2014 - 1:05 PM EDT
No knock on Jim but all this stat overkill these days is borderline useless. I can watch a player over the course of the year and know if he's any good or not without any of these stats, ratios and advanced stat analysis.

I don't need any of that to know masty stunk and Brantley is really good. Or to know Lonnie has had two good months in over 3 years etc.
August 3, 2014 - 12:06 PM EDT
feel free to do some research and offer analysis that might provide something to think about.
August 3, 2014 - 12:01 PM EDT
Wow, why the hate Billy boy? Seems rather unsolicited, gratuitously mean spirited and ungrounded in any supportive explanation.
Wtf dude?
Billy Lybarger
August 3, 2014 - 11:56 AM EDT
So because TJ has an xFIP lower than James Shield, we now should expect him to be a #2? I think Jim doesn't understand how statistics work as predictors of future outcomes either.

I put my name by my comment because I stand by it. The author adds little analysis, and critizices based solely on data he obtained from a dashboard. Maybe if he moved down the Fangraphs page a bit he might find new data to spout, but he might have to read the glossary first to see if TJ's O-Swing% is indicative of his high GB%. I'm sorry, Jim, but your defense of the author is about as weak as his analysis.
August 3, 2014 - 11:17 AM EDT
that's a harsh assessment, billy, and more likely reflects your inability to digest complicated material. this is a very strong piece and a welcome suggestion that house may be a better than a #5-6 guy
Billy Lybarger
August 3, 2014 - 11:10 AM EDT
John Grimm is a perfect example of a writer than can read and regurgitate Fangraphs data, but add nothing to the conversation, most likely because he has no clue what he is looking at on the field.
August 3, 2014 - 10:20 AM EDT
House had a much better game than Masterson last night - but Masterson got the win. I doubt his performance would have yielded a win for the Tribe.

Your Name:
Leave a Comment:
Security Code: