Fluke watch – Ubaldo Jimenez

Going into 2010, Ubaldo Jimenez was known as a potential big-time ace. He had just finished the 2009 season with a 3.47 ERA and had a 3.63 xFIP to make it clear that this improvement was, in fact, real. In 2009, Jimenez had two fastballs that both averaged over 96 MPH, and he touched triple digits in velocity 10 times.

Meanwhile, he had a groundball percentage over 50 percent (52.5 percent according to fangraphs). Thus, Jimenez looked to be the rare pitcher with the ability to be both a power-strikeout pitcher and a groundball pitcher, a combination even more appreciated in Coors Field. Needless to say, his potential seemed quite great.

Then in April and May of 2010, Jimenez appeared to break out, with a sub-1.00 ERA through April and May. Of course, if you’re reading The Hardball Times, you were probably warned to be wary of Jimenez’s .251 and .202 BABIP in those months, which indicated that a good bit of regression was coming.

And boy did it come, with a 4.41 ERA in June and a 6.04 ERA in July to drive fantasy owners crazy. Still, if you kept Jimenez all year, you were probably happy with the overall 19-8, 2.88 ERA line that he put up.

Still, there are some odd things about Jimenez’s year:

First of all, he increased his strikeouts, but this increase mainly came against left-hand batters (Ks have actually decrease against righties).

Second, Jimenez’s walk rate remained lower than it was in 2008 for the second straight year. However, this change is also odd, as the BB rate against righties increased, while the BB rate against lefties dropped a good bit.

Finally, Jimenez’s ground ball percentage dropped to 48.8 percent, the lowest over the last three seasons. This appears to have been caused by the increasingly-extreme groundball splits of his pitches: against left-handed batters, Jimenez’s is a terrible groundball pitcher (39.4 percent GB rate), while against righties he’s a clear groundball pitcher (57.8 percent). This split was not as extreme in 2009 or 2008 at all. So what caused this?

These oddities raise the question: what can we expect from Jimenez next year? Let’s look at the pitches:

What Jimenez throws

Jimenez historically has thrown six pitches—a four-seam fastball, two-seam fastball, change-up, splitter, slider, and curveball—during his career. However, he appears to have broken out a seventh pitch in 2010: a cutter. That said, the slider was barely used in 2010 (it was thrown 37 times), so its impact on Jimenez’s performance is negligible and not worth any further discussion.

Other than the cutter, the movements of all six of Jimenez’s other pitches remained the same in 2010. The same was true of the velocity of these pitches. There also isn’t a large difference in the frequency of use of any of these pitches. No, whatever caused the odd results of Jimenez in 2010 appears to be in the location of where Jimenez threw his pitches.

Jimenez’ fastball and the location changes in 2010

The key pitches, of course, when discussing Jimenez are his two fastballs, which he throws a combined 60 percent of the time. The two-seamer is thrown twice as frequently as the four-seam fastball against both left- and right-handed batters, but both pitches are thrown fairly frequently. The four-seam fastball is slightly faster than the two-seamer (by about 0.5 mph), has about two inches more hop (“rises” two inches more), but has around three inches less tail.

As two-seam fastballs go, Jimenez’s pitch does not have a great amount of sink. However, this lack of sink is countered by the pitch’s great velocity, which increases the pitch’s ability to get ground balls (for more on this phenomenon, see here).

That said, it’s this pitch that seems to be the greatest cause in the emergence of Jimenez’s odd groundball splits and his improved ability to get strikeouts against left-handed batters in 2010. Take a look at the following table, which shows the results of these two fastballs against left-handed batters each of the last three years:

A Hardball Times Update
Goodbye for now.

Pitch Type Year Swinging Strike % Ground Ball %
Four-Seam 2008 8.71% 50.00%
Four-Seam 2009 9.06% 46.15%
Four-Seam 2010 12.71% 41.51%
Two-Seam 2008 4.90% 48.13%
Two-Seam 2009 6.76% 48.43%
Two-Seam 2010 8.20% 33.04%

Now, of course, the question is what caused this dramatic change in the results? The answer is where Jimenez has located his fastballs against left-handed batters. In 2009, Jimenez threw both fastballs at a higher location in the strike zone than he did in 2008, but he kept the horizontal location in the strike zone around the same location.

Higher sinkers will still get ground balls against same-handed batters, but will be less effective against opposite-handed batters: in this case, left-handed batters. Of course, high fastballs are more likely to result in swinging strikes, as well, which explains the increase in strikeouts.

In 2010, Jimenez kept his aim high, but altered his horizontal aim as well: he tried to throw both fastballs against right- and left-hand batters more away from hitters than he did previously. Against right-hand batters, this put the ball more in the middle of the zone (Jimenez had previously aimed inside on these batters). Against lefties, however, this resulted in these pitches being in the outside part of the plate more often than before.

The end result was an increase in swinging strikes (Jimenez’s best spot for getting swinging strikes the last two years against LHBs has indeed been away and high). However, as you could guess from the table above, the result was a drop in the groundball rate (the high-and-in pitches did get ground balls more frequently than high-and-away pitches.)

All in all, these appear to be real changes in Jimenez’s pitching habits, and we should expect the increasingly extreme strikeout and groundball splits to continue in the future.

But what about the decreased walks?

You may have noticed that the oddity regarding the reduced walks of Jimenez over the last two years is not accounted for yet in this analysis. This is because it’s hard to understand why Jimenez is getting a better BB Rate. Jimenez did do a better job in 2010 throwing strikes on three-ball counts. Of course, he threw more pitches in three-ball counts in 2010 than in 2009 (and in a roughly equivalent amount to 2008), negating the effect of this improved accuracy.

Moreover, Jimenez’s ability to hit the strike zone didn’t seem to change against either hand too much in 2010 as compared to 2009 or 2008. One theoretical explanation you might think of is that batters were swinging more in 2010 and 2009 at pitches out of the strike zone. Indeed, Fangraphs’ BIS data would tell you that was the case in 2010.

Unfortunately, this is one of those cases where the BIS data is incorrect….according to PitchF/X data, batters did not swing at pitches out of the strike zone significantly more frequently in 2010 or 2009 as compared to 2008 (For an explanation as to why BIS data is inaccurate in this area, see Colin Wyers’ article here.

The end result is that, without a discernible reason why Jimenez’s BB rate has dropped, don’t bet on it staying as low as it was in 2009 or 2010, and don’t be surprised if it returns to being above 4.0 BB/9.

Conclusion:

Ubaldo Jimenez has tantalizing stuff. And at times he looks like an ace. But as aces go, he’s a riskier proposition than other No. 1 starters for your fantasy team. This is the case for two reasons:

1. His walk rate is likely to increase in the future, resulting in his numbers not being as impressive as they were in 2009 or 2010.

2. While his changes in where he locates his fastballs have resulted in increased Ks, they also have increased the volatility of his pitch results. Because he no longer will get as many ground balls against LHBs, he’s more likely to give up home runs, which will ruin quite a few fantasy starts. Jimenez had a HR/FB of 3.3 percent (according to Fangraphs) against LHBs in 2010. That’s due for a good bit of regression, especially considering he plays in Coors Field, humidor notwithstanding.

The end result is that against teams with a large number of left-handed batters, Jimenez is more likely to be a feast-or-famine pitcher than he was before 2010: he could give you a start with a lot of Ks, or he could end up giving up a few gopher balls.

Thus, I probably wouldn’t have Jimenez too high on your list of choices to be your No. 1 starter in your 2011 fantasy league.


20 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
buck turgidson
13 years ago

How is 214 K’s in 222 innings and a 3.91 xfip in Colorado flukey?

It’s fairly arrogant for you to assume that he is not going to continue to improve just because he changed his approach for leftys and you can not point to a peripheral to support his lower walk rates.  Silly really.  Maybe he’s entering his 27 season and he’s figuring it out.

His season was not a fluke.  He was an immortal beast for two months, average and mediocre for two months and then fairly beastly.  His BABIP on the season was a little low.

Lucas Apostoleris
13 years ago

Josh, how did you determine the splitters from changeups?  I’ve had trouble with that whenever I look at Jimenez’s pitch data.

garik16
13 years ago

I didn’t say that his season was a total fluke.  Nor did I say that the Ks were a likely fluke (there’s a good explanation for them, that’s likely to continue). 

But when a player improves say his walk rate, usually there’s an explanation for it.  For example, we often say a player has better control….well that doesn’t seem to be it, as Jimenez was missing the zone at the same rate as the previous two years.  We might say that the pitches’ speed/movement changed, but that didn’t happen either.

Moreover, the improvement in walk rate the last two years has not been constant in its nature.  As I mentioned above, his walk rate stayed low due a drop in walks against lefties, but there was a rise against righties (despite there being a drop against righties the previous years).  With no obvious explanation (control, pitch usage, movement/speed, batter swings), this makes it especially odd and something I am skeptical of. 

It is certainly possible that I’m missing something, and that I’m incorrect.  Jimenez could also actually improve his control in a more noticeable way next year to make a better walk rate stick (or even get better).  But from what we’ve seen over the last three years, it doesn’t look like that’s the case.

garik16
13 years ago

Lucas, I’m not particularly confident in that classification difference myself, but I basically classified off-speed pitches that were more two-seam like as CH and those that were closer to FF like in movement as FS. I classified by going through each start individually.

That said, you’re right, it’s incredibly close, and I’m not really comfortable making a judgment about either the change-up or splitter as compared to the other.

garik16
13 years ago

Lucas: 
I have the Change-up (282 Thrown) in 2010 at:
87.9MPH, -6.46 pfx_x, +3.98 pfx_z

I have the split (254 Thrown) in 2010 at:
87.7MPH, -3.64 pfx_x, +2.05 pfx_z

Lucas Apostoleris
13 years ago

Okay; thank you.  I would have assumed that the split has more drop, less tail, and more velocity (though this may not be the case) than the change.  For the time being, I’ve lumped them all in my data as splitters, but I’ll go through game-by-game over the next few days and clear up the FF/FT and FS/CH distinctions to the best of my ability.  By the way, how many cutters did you find?

Lucas Apostoleris
13 years ago

Oh, you said 37 times.  My bad.

Kevin L. Wiley
13 years ago

Your logic in expecting a regression in Ubaldo’s walk rate toward 2008 levels seems very faulty to me.

I understand that you cannot pinpoint the reason for his improvement in 2009 and 2010 in his walk rate.  But because you can’t pin point the reason then to conclude that it is likely to return to 2008 levels lacks anything but speculation on your part.

You might just need to realize that two years worth and almost 450 IP means something and that Jimenez has established through his experience and maturity better command.  The numbers that we crunch may not be able to tell us why but the performance against major league hitters does.

Kevin L. Wiley
13 years ago

Also a general point I love all the pitch data that we now have access to and the breakdowns by people far more educated in it than me, but also one observation:  is there anyway to quantify what the pitcher is throwing from the hitters perspective?  I’ve heard that Ubaldo (and others in the past) throw a very heavy ball, and that different body types especially long armed long legged type present a different challenge to hitters.  How does or does it at all play into pitch data as we now know it?

garik16
13 years ago

Jimenez through a half season in 07, then a full season in 08.  His walk rates were a problem then, and his Ks weren’t what they became. 

Pitchers grow.  They develop.  Jimenez HAS.  Since 08, as I detail above, he’s changed how he’s pitched, and the result is extreme GB splits and increased Ks. 

Your new question (first one) is why couldn’t 08 be the fluke.  Well first, I’d point to 07 as the first reason, which while not a whole season, points to that being a real result and not a fluke. 

Second, there’s the fact that the improvements in the walk rate haven’t been constant in 09 or 10 (that’s where the splits-stuff comes in).  When combined with the fact that we see little change in his accuracy in 09-10 from 08, I suspect that regression TOWARD 08 is most likely to be the outcome.

Now, this doesn’t mean that his walk rate will go all the way back to 07-08 levels.  Likely, if we have 3 years in which his accuracy is the same, but 3 different results, the true walk rate will be somewhere in between.  Regression won’t be 100%, but it should happen if he doesn’t change anything.  Note the key here is that THERE HAVE BEEN THREE DIFFERENT RESULTS: Despite the Overall BB Rate being down in 09-10, that’s due to different odd split phenomenona, so it DOESNT lend weight to the idea we should weight those two years far more strongly than the 3rd year. 

In other words, the word fluke is a misnomer: no years are complete flukes, they all show some element of the real thing. 
==========================================
“very heavy ball, and that different body types especially long armed long legged type present a different challenge to hitters”

Very Heavy Ball shows up as the sinking motion of Jimenez, which I discuss above.  Pitchf/x can handle that.

Different body types are a problem that’s more of an issue….pitchf/x can extrapolate release points somewhat, but overall motions less so.  It’s possible that a pitcher has gotten better at not tipping his pitches or something of the like, and we couldn’t detect it.  FWIW, Jimenez’ release points haven’t changed much from 08-10 (he’s slightly seeming to slide toward the center of the rubber, but that’s about it and it’s’ small), and that doesn’t tell us anything. 

So yes, things like that are reasons why my cautions could be wrong, and he could theoretically keep it up.  But the point of this column isn’t to denounce Jimenez, but to throw a word of caution to fantasy players….Jimenez may be a #1 starter, but there are potential red flags, thus I was advising players to be cautious when picking him. 

I try not to give precise reasons for a reason (also there’s the fact pitchers can change their repertoires so you have to be careful)

Kevin L. Wiley
13 years ago

Bringing his rookie season into the equation (4.06 BB/IP) over half a season of starts is actually closer to 09 (-.55) than to 08 (+.61) and the difference was only (-.32) so I actually believe bringing his rookie numbers into the equation definitely supports a sub 4 walk rate more than a plus 4 walk rate, now that he is 700 innings into his ML career.

Maybe we need to have some sort of side bet, I’ll take the under you can take the over on a BB/IP of 4.

garik16
13 years ago

Sure, in fact, I think maybe I ought to make an offer in my next column.  (BB/9, not BB/IP though, obviously)

Of Note:  In addition to his 07 half season, Jimenez’s 06 and 07 minor league seasons had BB/9s of around 5.  Of course, I have no pitch data to s how on those years,

buck turgidson
13 years ago

I think putting so much of your hypothesis’ weight in split stats from single seasons is a mistake. I’m not buying this.

garik16
13 years ago

Buck, as I’ve said above, I’m NOT making these statements solely based upon one thing.  Looking at the pitches, they appear to have the same accuracy rates each of the three years. 

And yet the results (BB-wise), are different all three years, as becomes apparent by the splits. 

As I say above and have said to you already, this implies to me that the changes in the results are simply the results of randomness…of simple variation. 

The splits alone show that the improvement on a seasonal-level may not be stable or real.  The fact that the pitch data shows around the same amount of zone-accuracy further points in that direction, and thus I make the prediction.  I’m not sure what’s hard to see about that.

Chris
13 years ago

Is it at all possible that his command hasn’t gotten better over the span of a season, but he’s pitching better when it counts? I mean, his BB rate will drop if he tries to pitch more to contact against lefties while it should rise if he’s trying to get a right handed hitter to swing and miss at a pitch out of those zone. Perhaps this is what’s causing the problem? This article should probably be looking at pitchf/x splits for different counts to get a better grasp of what’s going on.

Having watched Jimenez extensively over the year, I can tell you that the eye test said he was exhausted. He was leaned on too heavily in the first half, his fastball seemed to be a tick slower and he wasn’t locating as well as he had in the first half. I think that if the bullpen and offense start working more in his favor so that he can leave games earlier he’ll perform much more evenly over the course of the 2011 season.

Personally I’m expecting a 3.21 ERA 20 W, 7 L, 215.1 IP 208 K, and a 1.13 WHIP

garik16
13 years ago

Chris, as I said above, Jimenez was slightly more accurate in 2010 on 3 ball counts.  Of course, he got into MORE three ball counts in 2010 than in 2009 (but on par with 2008), which should have negated part of this impact (Mind you, you don’t want to be in 3 ball counts to begin with).

“I mean, his BB rate will drop if he tries to pitch more to contact against lefties while it should rise if he’s trying to get a right handed hitter to swing and miss at a pitch out of those zone. “

This is the opposite of what has happened.  His BB rate against lefties has dropped this year as he pitched to try and get more swings and misses, not to get contact. 

I think you’re quite a bit optimistic regarding his numbers, really.

Derek Ambrosino
13 years ago

Just to set aside the numbers for a second and talk investment stategy –

Isn’t Jiminez 2011 kind of like Greinke 2010? A legitimate breakout candidate who broke out but even more so than his advocates could have realistically expected. Do we think he’s actually going to better in 2011 than 2010?

Of course, we need a price at which we we’re willing to buy each and every player, but given the overall context doesn’t it seem more likely that the general, public market will overvalue Jiminez vs. undervaluing him – meaning that somebody will jump for him before your measured self enters the bidding, even if you’re on the realistically optimistic side?

This is the year to buy Greinke again instead. no?

Basically, what I’m saying is that isn’t it generally a bad idea to buy a player coming off a career year unless you are especially confident that he’s going to repeat or best it? It’s one thing if the underlying data points toward – you know what, he should have been even better! But, if not, I’d say its better to avoid taunting the gods of regression.

That said, I plan to write a column tonight that might somewhat contradict this post.

garik16
13 years ago

Derek, I’m not sure the two are comparable, though I’ve done zero work on Greinke, so I’m not great at comparing, but just a quick look at his peripherals shows that from 2007-2009, there was steady improvement in his peripherals that seems normal for an improving pitcher (rather than 1 thing improving one year then getting worse in another year, while another thing compensates by getting better). 

And I’m not the best at investment strategies, but yes generally the idea would be to buy low, though you might still buy high on guys you think can repeat it.

Greinke’s an odd case (I should do a piece on him) in that his Ks regressed greatly (back to 07 levels) but in that he also had an increase in ground balls…suggesting there’s been some change in his regression.

That said, would love to see a column on the subject.

garik16
13 years ago

I admit, I don’t like making exact predictions because I do not have a projection system.  But I’m fairly confident on this one.  And you’re not actually making an argument based upon by facts. 

“two years worth and almost 450 IP means something and that Jimenez has established through his experience and maturity better command”

Except, the two years aren’t really similar.  The walk rate improvement in 2009 was caused by a big drop in the walk rate against RHB.  In 2010, it stayed down due to a drop against LHB (a good sign), but it ROSE against RHB.  If this was a constant steady improvement, we’d expect both to remain similar, but this is just inconsistent. 

Meanwhile “better command” is proven not to be true as he’s NOT being more accurate around the strike zone.  This shows that the odd, potentially flukish BB rates, are in fact more likely to be flukes than we would think otherwise.  It’s not that in 09 he was more accurate against RHBs than in 2008 or 2010, but just seemingly an odd result. 

The idea of this series is to identify pitchers whose basic peripheral statistics may lead us to believe that improvements are real, but who in fact are deceiving us with improved-peripherals that are caused by luck.  Jimenez’ strikeouts DO NOT fit this description…but his walks do.

Kevin L. Wiley
13 years ago

Here is my question, why are you pointing to 2008 as a baseline or some starting point to base the next two years and even 2011 off of?

2008 was his first year as a full time starter and he was 24 years old.

Now regardless of how the #‘s have been put together the last two years, to come to some conclusion that the 2011 version will closer reflect 2008 version is a strange analysis to me.

Just because you are unable to quantify why his BB rate has been down for what in essence is 66% of his career (and the most recent) to then want to say that the 1st 33% of his career is a truer reflection of the pitcher, once again is strange logic.

But that is nothing other than my lay opinion.

(No offense intended I appreciate all of you guys work here it, I just happen to strongly disagree with your conclusion on this one.)