No-no, not so fast

Tuesday night, Francisco Liriano pitched the first no-hitter of the 2011 Major League Baseball season, and in doing so offered us an opportunity to marvel at the beauty and peculiarity that is the game of baseball, and thereby fantasy baseball. I’d like to explore Liriano’s no-no from a few different angles, some of which are more fantasy-relevant than others.

KISS strikes again

I trust many of you are familiar with the “keep it simple, stupid” philosophy. While, sometimes hard to employ and justifiable from which to deviate, KISS often comes back to bite those who exercise hubris and try to get too cute with their stratagems. I’m willing to bet Liriano’s performance occurred on many a fantasy team’s bench.

Let me say right here, all those who decided to bench Liriano on Tuesday night made an understandable choice. (His ownership is only 84 percent in Yahoo leagues, and I can’t offer the same consolation to those who chose to drop him outright). At the same time, Liriano’s performance goes to reinforce an important lesson as well. You never really know when a slumping player with a proven track record is going to turn it around, or show a glimpse of the talent that produced his resume. Let me pose a thought exercise here.

Placido Polanco is a player with a proven track record. We know he is a quality hitter with the potential of hitting well into the .300s and putting together some very useful stats for a middle infielder. He’s playing out of his mind right now, unsustainably so.

Imagine yourself a Polanco owner, would you try to jump the curve and guess when the 0-5 is coming? Would you bench him speculatively? Of course not. Most likely, the best strategy is just to ride it out day to day and bank on the right numbers being there at the end of the season, only making a move if you reach a point where you think you have a clearly better option at your disposal.

In one respect, this experiment could be argued to be making the opposite point of my own, that Liriano has stunk and that you shouldn’t try to predict when that streak will end, but rather ride it out—on your bench. But that’s not what I’m saying at all.

Quite simply, if you thought Liriano was going to stay that bad, then you should have dropped him. So, if you didn’t, you’re displaying some faith that the real Liriano will emerge. My point is, it’s a fool’s errand and an act of disillusionment to think you might know when that is going to happen.

Those of you who had him on your bench for that start are now, in one respect, in a worse position with him than you were before Tuesday’s start. Not only have you absorbed what are likely to be several of Liriano’s worst outings of the season, but you’ve also forfeited one of his best. If you’re dollar-cost-averaging, you’ve just increased your holdings of “the bad Liriano.”

The unpredictability of the game of baseball is a primary ingredient of its beauty. The juxtaposition of a game that can be so precisely measured and quantified, yet still manages to shock and awe every day, is part of what makes the sport sublime. Try to outsmart the game on a microscale at your own peril.

Before moving on, allow me to offer one more point here. Although I used Polanco as my counterexample, I do see a difference between hitters and pitchers. (Maybe I should have chosen James Shields for my example here instead.)

A pitcher makes only 30-odd starts a season, at best. It is dangerous to sit a pitcher as talented as Liriano because he has so few chances to help your team, and it really hurts if he comes through big time and it doesn’t count.

I struggled through a season of Aramis Ramirez in 2010, and I hit points where I just benched him. But, as a hitter, short of a Zobristian outburst, one day’s worth of at-bats is unlikely to produce as much positive (or negative, granted) value as an outing from a starter. In fact, I think Ramirez’s 2010 proves my point, too. Regardless of how painful and short of expectation that season was, Ramirez still put up 25 home runs and 83 RBI in 124 games.

Unless you really have a plan, like the top-notch daily fantasy game players do, and a team constructed around platooning and such, you’re playing with fire when you try to cherry pick the innings or at-bats you want from your high-level players.

On the surface, this may seem no different than sitting your borderline pitchers against the highest-octane offenses, but that’s not the case. Sitting a stud pitcher who had few bad outings is not nearly as calculated a move as it looks from the bird’s-eye view.

Unless you have hard evidence-based reason to believe he’s not actually a stud pitcher any longer, you’re basically just randomly choosing to bench a high-quality player because of the outcomes of largely independent trials that previously conducted.

The forty-footer

After indirectly extolling the virtues of Liriano (but largely to make a more general point), I’m going look at the underbelly of Tuesday’s start. Men are pigs, and we have several terms for women who look gorgeous from afar, but substantially less so up close. “Forty-footers,” for example. Liriano’s no-no was a forty-footer, and this is more apparent when viewed through the fantasy lens than the mainstream baseball coverage lens.

Though he gave up no hits, Liriano only struck out two, and walked six in the process. He turned in nine scoreless frames to the tune of a 0.67 WHIP, earned a win and chalked up two strikeouts.

Over in Beantown, Jon Lester also earned a win, while giving up one earned run over seven innings on his way to a 1.00 WHIP, and he struck out 11 in the process.

In terms of how well each pitcher pitched, was Liriano any better from either a fantasy or actual standpoint? Going back through Liriano’s career, Tuesday’s outings wouldn’t even rank on the short list of his best outings, as he certainly was better here, here, here, here, and here, for starters (no pun intended).

So, while fantasy baseball takes its share of ribbing for not always mirroring the way the game really works (relying on batting average, overvaluing stolen bases, etc.), sometimes sabermetric justice is served in the fantasy universe, as well. Liriano’s outing (or really, the outcome of his outing) was certainly very good, but it was not incredible.

Keeping it in perspective

A good friend of mine who is a Twins fan remarked to me today that he was taken aback by the tone of some of the coverage of Liriano’s performance. Many reports focused on his recent struggles (and rightfully so—juxtaposition is always a good literary or rhetorical tool), but seemed to define Liriano by them. From this premise, the improbability of this performance was elevated. I agree that this is quite silly, and for two reasons, or maybe even three.

First, Liriano is hardly a nobody, despite what the tiny sample size of his 2011 starts has to say. For the period of time he was a starter in 2006, he was basically the best pitcher in the sport. And last year, back to full health, he was quite good again.

However, and this is the second point, even if he was a nobody, so what? The record book is chock full of mediocre pitchers who have thrown no-hitters. Just last season, Edwin Jackson (Liriano’s opponent on Tuesday) tossed one, and Armando Galaragga pitched what should have been a perfect game.

And you know why else this wasn’t all that improbable? As I mentioned earlier, it wasn’t even that great a performance.

Liriano’s outing this past Tuesday is a perfect example of something many of my THT Fantasy colleagues have said in the past when discussing scoring systems for points-based leagues: Don’t give extra awards for no-hitters.

All the evidence we have at our disposal, and the anecdotal case of Liriano, tells us that these events are flukish. Liriano’s outing may have already earned more points in your points-based league than his actual performance was worth in pure merit. Don’t compound it by giving him bonus points, such that his performance outearns Lester’s by leaps and bounds.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: AL West: Texas isn’t finding it easy
Next: Roy Campanella: A study in contrasts »


  1. bob said...

    You’re arguing semantics. Getting through 9 innings at this level without surrendering a hit is an amazing feat. Nobody is saying that Liriano will post a 0.00 ERA the rest of the year.

    You want to make the point Lester is a better pitcher? That remains to be seen. I don’t think anyone would argue that Lester is a comparable talent.

  2. Andrew said...

    The idea that you should never play the matchups is wrong, wrong, wrong. Take a look at the work THT’s own Derek Carty has done over at FanDuel recently. The difference in facing, say, the Yankees in New York compared to the Padrs in San Diego cannot be overstated. Sure, anything can happen on a given day, but that doesn’t mean we shouldn’t make the best decisions possible.

  3. Kevin said...

    “…if you thought Liriano was going to stay that bad, then you should have dropped him. So, if you didn’t, you’re displaying some faith that the real Liriano will emerge.”
    Alternatively, many owners who keep him on their bench are displaying some faith that a trade-able Liriano may emerge.

  4. Derek Ambrosino said...


    Getting through 9 innings at this level without surrendering a hit is an amazing feat.

    I’m going to argue semantics again. I’m not sure I’d say it’s an amazing feat, it’s an incredibly improbable feat. A no-hitter is an inherently improbable proposition, as we saw with Liriano just two days ago, the performance that generates the improbable result is not necessarily “amazing.” A no-hitter is not the apotheosis of dominant pitching. There’s lots of luck involved in a no-hitter, and often lots of defense. Liriano clearly got a heavy helping of both.


    I’m not saying you shouldn’t play match-ups. I’m saying that sample size is small for pitchers and it’s very easy to get burnt by trying to outsmart yourself by overmanaging your stars. Playing match-ups is something different altogether. Sitting a pitcher because he’s playing the Yankees in the Bronx is a different proposition than sitting a high quality pitcher simply because it looks like he’s been “in a funk.” The former is a calculated decision based on risk assessment, the latter isn’t, but kind of looks like one.

    And, daily fantasy is its own animal. I’m not saying that if you were assembling a daily team, you should have bought Liriano over your other options. You’re making 1 play and have all alternatives theoretically open to you when you make any single decision, so you should play the odds and hope it turns out right for you. In traditional fantasy, the question is Liriano the best choice here (I probably wouldn’t have picked him either), but rather, Liriano – take it or leave it. My point is that in the absence of evidence that Liriano is not in fact what we know Liriano to be, I’ll take it.

    The operative question here is whether there was any reason to think Liriano was anything other than what he had proved himself to be over a much larger sample size, simply because he strung together a few poor outings. If that’s the reason you are sitting him – if you don’t have any fundamental evidence supporting a reason for those poor outings – then you aren’t playing match-ups at all, you’re just doing the equivalent of picking up the guy who want 4-4 yesterday. (If the “The Nena” is reading, he knows that’s called a “Nomad Special”).

    One shouldn’t expect a pitcher’s outcomes (good or bad) to continue as such, unless those outcomes are supported by fundamental, performance-based peripheral indicators. Now, honestly, I haven’t explored Liriano in-depth (I don’t own him at all), so maybe there was true reason for concern (his Ks are down, perhaps that’s a real issue and not just a sample size issue, we’d have to dig even deeper to get a beat on that). Now, if a decision to bench him was based on those kind of indicators, then I can get behind that because that’s saying these starts aren’t fully independent trials that happen to cluster as aberrations, from which we should se regression to the mean. If, however, the reason for benching him was, he’s got a 9.00 ERA, then that’s not all too strategic at all.


    The same central question applies to you as well. Holding on to him on the bench is basically a hedge, and ostensibly what you’re doing is making what would largely be considered the safe play. But, my point is, I’m not sure if it’s really so safe. Tuesday, you didn’t have the confidence to start Liriano, but that isn’t the same thing as there being increased odds he’d pitch poorly. Next time around the rotation, I assume you’ll start him, but does Tuesday’s outing actually make it any more likely that he’ll pitch great next time around? Why that game have any more bearing on the next than the start before Tuesday’s had on Tuesday’s?

    Now, I do believe that should you wind up going down, you should do so in they way that is true to you, such that regrets are minimal. Me, I’m an ennjoy the ride kind of guy. I’d rather go down because of the 3 innings 8 run stinker than not make it because I left the amazing outing on the bench. But, some people prefer the opposite philosophy. I understand that, and like Dres of Black Sheep told me back in 1991, the choice is yours.

  5. Jeffrey Gross said...

    Now, if Kerry Woods 20K game turned out to be a perfect game, I’d say it was talent wink

  6. bob said...

    Wrong again. To say this is largely luck based, is to say there is no such thing as a good investor, or poker player….it’s to say there is no value in managing risk. It’s to say that you don’t understand the nature of pitching at all.

    Pitching is not black and white. It’s not, if you do A, B will occur. You’re dealing with probability. You’re buying into the broadcaster’s notion that almost every ball that’s hit hard is a “mistake” pitch. It’s selection bias, and it’s total bullshit.

    Pitching is an art form. Success is dependent on a combination of a myriad of factors, most of which are not conducive to statistical analysis as predictive indicators.

    The exact same pitch offers different odds of success in different contexts. Your “analysis” relies on the notion that pitchers have no control over batted balls, and this is a vast oversimplification by oversimplified minds.

    Not all no-hitters are equally “impressive”. With that said, we celebrate far lesser achievements every single day. We are dealing with one of the top young talents in the world, who is making a conscious effort to addjust his approach with a very specific objective. In short, he is trying to garner weak contact to get deeper into games. To fail to mention this in defining the merits of this achievement is akin to evaluating a musician without accounting for their genre.

    He effectively mixed about 4 pitches with varying location in an effort to garner weak contact and produce groundballs when needed. Sure those groundballs could have gone for hits, but he conveyed the ability to increase the potential for a groundball with their timeliness. If you actually watch baseball, you would realize he had the Sox hitters on their heels for most of the evening. Konerko hit one that could have just as easily been a HR. Beckham JUST missed one. The last out is a hit if it’s 3 feet to the left. If you take away from the no-no for these events, then a miniscule # of games would qualify as “impressive”.

    Liriano greatly increased his odds for success by making timely pitches, and executing on his pitches. The White Sox have an extremely dangerous lineup. The fastball that Konerko hit to right may not have been all that great. An alternative narrative (and at least equally credible) would be to say that Liriano kept them off balance enough that this was the best Konerko could do with that pitch, because he was forced to respect 4 pitches on either side of the plate.

    Take away Liriano’s plus slider, change, consistent delivery across different pitches, and that same pitch is 15 rows deep for a HR.

  7. bob said...

    Disclaimer: My wording could be nicer. This is a reflection of my frustration with sports journalism, who report their biases as opposed to the events. There is a severe lack of objectivity in writing nowadays, and it’s getting worse with every new blog.

  8. Derek Ambrosino said...


    I think it is you who is actually quite blatantly revealing your bias. Suffice to say, I think just about every word of characterization of my post and beliefs are wholly incorrect. I’ll just call out a precious few.

    1. Of course I believe there’s such a thing as a good poker player; in fact I have personal friends who make (at least most of) their livings as such. The point is that poker players, like pitchers – even good ones – also benefit or suffer from luck as well. The following are all quotes from your post that are actually points that support my stance

    - You’re dealing with probability.
    – Success is dependent on a combination of a myriad of factors
    – Not all no-hitters are equally “impressive”.
    – Sure those groundballs could have gone for hits,
    – Konerko hit one that could have just as easily been a HR. Beckham JUST missed one. The last out is a hit if it’s 3 feet to the left.

    I’m not saying Liriano didn’t pitch well (though I notice you mention nothing about the 6 walks) just that this game was not one of the best pitched of Liriano’s career (or of the night, for that matter). There are 3 things entirely in a pitchers control – or independent of a team’s defense – Ks, BBs, and HRs. Liriano allowed no homers, which is great. He struck out 2 of the 33 men who faced him – that’s bad. He walked 6 of the 33 that faced him – also bad. He did a lot of good things on the mound that day, and they factor into the result, but he also had a whole lot of help. But, this is part of what makes baseball so wonderful, the fact that everything is so precise, but also so unpredictable. The stability of the big picture vs. the wild swings of the small picture.

    2. THT is not a “new blog”

    3. I don’t really think much about labeling, because philosophically I think it’s intellectually poisonous, but if pressed to label what I do here, I certainly wouldn’t go with “sports journalist.” First, I’m a columnist, not a reporter. Second, I write opinion and advice columns about fantasy baseball that sometimes use real life events at their impetus. …Again, this column wasn’t really about Liriano, or his no-hitter. I certainly wasn’t intending to “report events” – I’d have been wildly late for that, now.

    4. I watch as much baseball as is probably humanly possible for somebody who has the other responsibilities that I do. So, please don’t throw the – if you would watch baseball canard. If you watch more baseball than I do, I really hope you get paid to do so (as sometimes I actually do as well, btw), otherwise, while I’m sure it’s a badge of honor in the myopic context of this specific conversation, it’s probably not all that reflective of too much I’d be envious of otherwise.

    - No need for the disclaimer. I can take it, and I can dish it.

  9. Drew said...

    Regarding your last comment, Derek, you mention there are 3 things a pitcher can control: Ks, BBs, and HRs.  Don’t pitchers also possess some control over weakly hit balls?  I think it is slightly arbitrary to say that a ball that misses the bat by 1/16” is more in the pitcher’s control than one that only grazed the bat and was hardly put in play.  Don’t these weakly hit balls have a much greater probability of being fielded for an out than a line drive?  That is part of why LD%, etc, is a nice complement to other batted ball stats.

    That said, I’m a Twins fan who didn’t even watch most of the no-hitter.  As soon as Liriano walked Juan Pierre for the second time, I was so fed up that I turned it off until my friend called me after the 8th inning.  I was hoping he would have been traded before the season started.  He felt soreness after his first workout in spring, and when Rick Anderson asked him if he had done his offseason workouts, Liriano simply said “No.”  That is why I have no faith in him becoming a quality pitcher.  He is so lazy.

  10. Eddie said...

    “Liriano’s outing may have already earned more points in your points-based league than his actual performance was worth in pure merit. Don’t compound it by giving him bonus points, such that his performance outearns Lester’s by leaps and bounds.”

    Different strokes for different folks, sir. My league does award a bonus for the NH. It also deemphasizes stolen bases, penalizes caught stealings, and uses OB% rather than BA. Would you condemn all 5 X 5 leagues for allowing players like Juan Pierre to ‘earn more points than his actual performance was worth in pure merit?’

  11. bob said...

    Here’s a quicker summary of my point:

    Numbers don’t lie, but they are frequently misinterpreted.

  12. bob said...

    We’re not on the same wavelength. I’ll close with main points:

    - It takes skill to increase, and ideally maximize the potential for success in probability based endeavors. This is the main skill in successful investing, poker and pitching. There is no such thing as a pitch with 100% probability for success.

    - “If you watch more baseball than I do, I really hope you get paid to do so (as sometimes I actually do as well, btw), otherwise, while I’m sure it’s a badge of honor in the myopic context of this specific conversation, it’s probably not all that reflective of too much I’d be envious of otherwise.”

    There is so much wrong with this line and your willingness to share it publicly, particularly on your own baseball site. Since you decided to get “personal” though, money is the worst reason for why I do anything. A) I’m fortunate to have been given a reasonable opportunity to adopt this view. B) Word of advice. Don’t talk about money with strangers.

    - I didn’t bridge the gap from your post to my response because I’m too lazy.

    Long story short. In my opinion, your conclusions indicate that you are both using an insufficient data set, and that you have a superficial understanding of the craft you are writing about. You would greatly enhance the degree of predictability and understanding of the game by expanding your data set.

    You wouldn’t calculate the probability of rain tomorrow by dividing the # of times it rained last year by the # of days in that year for obvious reasons. However, this is analogous to what most internet sabers do when issuing opinions about the game.

    Academic theory is worthless if it’s developed with a lack of appreciation for practical application.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>