Why a save is not a save

Earlier this winter, The Hardball Times offered prospective fantasy baseball writers the opportunity to compete in a Hardball Times fantasy league. Entrants wrote fantasy baseball articles, the best of which would be chosen as our winner. While we could only choose one winner to play in the league (congratulations, Dave Chenok), we had so many great articles that we have decided to publish some of the best. This is one of those submissions.

There are a few things I look forward to each year as we move from winter into spring. Longer days and warmer weather. Crocuses popping up through the snow. Grover Cleveland’s birthday (which happens to be the same day as my niece’s). And, especially, the inevitable article that appears on every Fantasy website imploring would-be league champions: “don’t pay for saves.”

The typical article explains why it is folly to waste money (or high draft picks) on closers. “Saves are unpredictable,” they tell us. “Closers can lose their job at any moment.” The article may tell us how Joe Borowski got more saves one year than Trevor Hoffman, Mariano Rivera and K-Rod combined, and you could have gotten Borowski 137 rounds later. The article asserts that saves are always out there on the waiver wire. “And after all,” they smugly conclude, “a save is a save, no matter who gets it.”

It all reminds me a little of Robin Williams whispering “carpe diem” to his minions in “Dead Poet’s Society”; the boys listen with bated breath and nod gravely.

I look forward to this annual article because, like lemmings, people follow the advice. And that clears the playing field, allowing me to do exactly what the experts advise against: take strong closers in earlier rounds.* The experts are missing something pretty fundamental: a save is not a save.

Why not? The key concept here is so simple that it amazes me it gets consistently ignored: Closers contribute to scoring categories besides saves. ERA. WHIP. Ks. Joe Borowski may well get as many saves as Mo one year, but Joe is probably going to hurt you, relative to Mo, in all the other scoring categories. “Oh,” I hear the experts saying, “but that is silly. Closers don’t pitch enough innings to impact those categories meaningfully. Solid starters will more than make up for any ERA or WHIP effect you get from having Joe versus Mo.”

The problem is: It isn’t true. It’s like saying that eating a chocolate bar each day won’t affect your weight, because you eat a lot of other food, and it’s only one little chocolate bar. Right.

Look, in a given week, three relief pitchers are probably the equivalent of one starter in terms of innings. Over the course of a full season, the difference in the non-save scoring categories between having, the equivalent of six innings a week of Josh Johnson (which three good relievers will give you) versus having the equivalent of six innings a week of Joe Blanton (which three weaker ones will give you) is nothing to sneeze at. You’ll do well enough in saves, and help your position in the other pitching scoring categories.

Best of all, you don’t have to sacrifice quality starters to assemble an elite relief corps; in most mixed leagues, starting pitching is so deep that—assuming you know what you are doing—you can find a starter in Round 17 who is statistically equal to one you could add in Round 12.

*The rest of this article is written from the perspective of a snake draft league, but the principles and analyses apply equally well to auction leagues

Let’s illustrate with an oversimplified example. Assume a league with none active pitchers, six of whom are starters and three of whom are relievers. I’ll use the full-season stats of six starting pitchers I actually had in one league to model that part of the equation:

Core starters	        INN     K      ERA      WHIP
Wainwright, Adam SP STL	230.1	213	2.423	1.051
Myers, Brett SP HOU	223.2	180	3.139	1.243
Wilson, CJ RP TEX	204.0	170	3.353	1.245
Scherzer, Max SP DET	195.2	184	3.496	1.247
Santana, Ervin SP ANA	222.2	169	3.921	1.320
Gallardo, Yovani SP MIL	185.0	200	3.843	1.368
SubTotal	       1261.1	1116	3.339	1.241

Now let’s look at the impact of adding three “early round” relievers per my strategy…
Scen 1: “Top-drawer closers”

      	                  INN   K	  ERA    WHIP
Wainwright, Adam SP STL	230.1	213	2.423	1.051
Myers, Brett SP HOU	223.2	180	3.139	1.243
Wilson, CJ RP TEX	204.0	170	3.353	1.245
Scherzer, Max SP DET	195.2	184	3.496	1.247
Santana, Ervin SP ANA	222.2	169	3.921	1.320
Gallardo, Yovani SP MIL	185.0	200	3.843	1.368
Bell, Heath RP SD	70.0	86	1.929	1.200
Wilson, Brian RP SF	74.2	93	1.808	1.179
Soria, Joakim RP KC	65.2	71	1.782	1.051
SubTotal	      1471.2  1366	3.125	1.227

…versus the impact of waiting and taking less attractive closers per “conventional wisdom.”
Scen 2: “Don’t pay for saves”

                         INN	  K      ERA    WHIP
Wainwright, Adam SP STL	230.1	213	2.423	1.051
Myers, Brett SP HOU	223.2	180	3.139	1.243
Wilson, CJ RP TEX	204.0	170	3.353	1.245
Scherzer, Max SP DET	195.2	184	3.496	1.247
Santana, Ervin SP ANA	222.2	169	3.921	1.320
Gallardo, Yovani SP MIL	185.0	200	3.843	1.368
Jenks, Bobby RP CHW	52.2	61	4.443	1.367
Gregg, Kevin RP TOR	59.0	58	3.509	1.390
Capps, Matt RP MIN	73.0	59	2.466	1.260
SubTotal	      1446.0  1294	3.342	1.252

Whoa. The difference is fairly significant in all three of the non-save categories modeled. Think you won’t score more points with a 3.125 ERA than a 3.342 ERA? Yeah, you will. In my main league last year, a 0.217 ERA differential was worth up to six points.

“But wait,” I can hear the experts protesting. “You picked three guys you knew had great stats for Scenario 1, and three guys with lousy stats for Scenario 2. You cherry picked.” Well, not really. I saw these exact combinations (or their equivalents) in several leagues I participated in last year. Maybe the difference wouldn’t be as dramatic if I’d used Jon Papelbon instead of Joakim Soria, but it’d be even greater if I’d used Matt Lindstrom instead of Matt Capps. Frankly, guys who follow the “don’t pay for saves” mantra don’t wind up with relievers as good as Bobby Jenks, Kevin Gregg and Capps.

“But wait,” I can hear the experts chortling. “What if you’d picked Joe Nathan or Jonathan Broxton—you’d have been hosed with this strategy.” Well, that’s true, but anyone can get injured, as the folks who used an early pick on Chase Utley well know. Besides, you have to be a little bit smart in executing any strategy—Nathan has had arm trouble in the past. And Broxton was outstanding in 2009, but it was his first full year as a closer—you don’t want any early-round strategy focused on guys without a multiyear record (see also Pablo Sandoval).

“But wait,” I can hear the experts spluttering. “If you wasted early round picks on Heath Bell, Brfian Wilson and Soria, you would never have had the six starting pitchers you did. Your Scenario 1 starters would not be as strong as your Scenario 2 starters to start with, and that would wash out the impact of the closers.” Again, not true—I did take Adam Wainwright early in this league, but I picked up four of the other six guys after Round 15 or off waivers, and my core starting pitching was statistically superior to most teams in my leagues. You need to do your homework, but you can assemble a statistically equivalent set of starters waiting several rounds to take your last three or four. Your straters may not be as good as someone else’s, but if you play your cards right there’s an equal chance they’ll better.

“But wait,” I can hear the experts croaking. “If you used early round picks for closers, you can’t possibly have had enough hitting—you must have sacrificed points there.” Well it’s hard to model what I didn’t do, but… assuming you use the first four rounds of your draft on strong hitters at weaker positions, there are generally enough corner infielders and outfielders left in rounds 8-12 to build a very solid overall hitting lineup. You can have your cake and eat it too.

Look, there are no guarantees in any of this. I am not saying that prioritizing closers guarantees you’ll win your league. What I am saying is that the non-save scoring statistics of closers have more impact on your team’s overall pitching performance than conventional wisdom would have you believe. So when this year’s draft rolls around, think twice before you congratulate yourself for your fantasy acumen in picking up Fernando Rodney in Round 17. You may think you didn’t “pay for saves,” but actually you just did pay—the opportunity cost of the money you could have won, which is now flying into the stands along with the last home run Rodney gave up. But—oh yeah—he still got the save. Trust me, a save is not a save.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Twisting Oliver: joining the elite
Next: Read and react, part 1 »

Comments

  1. Ben Pritchett said...

    @Dave Chenok- I see your point. When I draft I always pay a little not alot. I usually shoot in the middle. “Not paying for saves” really never works. I usually pick two or three guys I like that fit this profile, usually undervalued. Last year it was Billy Wagner and Andrew Bailey. Then I like to get a reach for my third closer.

    Great read, Mr. Chenok. We, “experts”,  are stupid sometimes. Winners don’t overthink the room.

  2. Chris said...

    I dunno, I won my league last season while completely punting saves all together. It was a 20 team keeper and by the end of it I was using Clay Hensley as my sole RP and several SP’s with RP eligibility. In fact, if I remember right, the only other true RP that I had was Rafael Betancourt, though Medlen was a reliever early on before going to the rotation.

    I ended up with one of the better pitching staffs in the league and the best overall offense (lead in every cat other than SB). I’m not sure how well this strategy will work next year when I have fewer RP eligible pitchers (JDLR wont qualify, not sure if Masterson will), but elite relief at the cost of positional depth can really hurt.

  3. Mark said...

    Markets are always evolving, even in fantasy baseball.  When I first started playing a few years ago, I remember the popular trend (and a major topic of discussion in snake draft analyses) was the concept of a “closer run”.  Closers were overvalued, and it manifested in often in an early-middle round that would get dedicated by as much as 60-70% closer selections (it would often happen another time or two later in the draft as well), because as the top guys would go off the board, owners later in the round would panic and feel they needed a secure closer.  I’ve always felt this was where the “don’t take closers” meme was born.  Job security was viewed as a more important issue than impact in other categories, which was obviously a flawed valuation of the market.

    But now, many “experts” have uncovered the flip side of the coin.  The fact that so many closer jobs are unsecured means that cheap saves will exist in the waiver wire throughout the season.  The result is the “closer run” phenomenon is dying, as every league now has at least a handful of owners who have enough knowledge and nerve not to cave.  This is a process though, and at least in recent years, I’ve found the overall result is that—though there will still be good and bad value picks at the position—the closer market is much closer to where it belongs than leaning in either specific direction (under vs over valuation).  The pendulum is still swinging though and its likely that closers will continue to see their perceived value drop in the coming years, creating the type of market ineffeciency you’re attempting to expose. 

    The problem with this year is that the position looks incredibly shallow on draft day, much moreso than in other recent seasons.  But, if anything, the overvaluation of closers in the early and mid aughts taught us that a lack of depth at the closer position is a superficial issue at best, and so in a situation like this patience is probably once again a better strategy than anything.  By mid season, there is very little chance the position will appear nearly as shallow as it does now.  So while I agree with your analysis in concept, this is not the year I will necessarily start targeting elite closers early in my draft.  There are too many teams with an abundance of good relief pitchers who don’t have the position settled yet (though this could obviously change during March).  If one of the few truly elite options falls a bit for some reason in my draft I won’t pass him up, but once again I won’t be gunning for saves in the single digit or even early double digit rounds this season.

  4. Will said...

    While I’m a bit envious that I missed a chance to step up to a high level of competition, I’m glad such a contrarian piece won the contest. You really give food for thought…

    I tend to play this in another way, taking the cheapest closers who are pretty sure to keep their role, but also use my reserve spots to build a composite-ace in the way you describe by paying just a few bucks for three elite setup guys. That way, I’ll get saves, as well as respectable ratios while also having cash left over to pad my other categories.

    Oh, and just a random point: I think it’s wise to realize that often the closers on the best teams get fewer saves than others.

    Congrats!

  5. Edwin said...

    Dave, thanks for a good read. This point has been made before, however. I recall writing an article about it back in the days of Seamless Baseball on MVN.

    Obviously the Eric Karabell’s of the world are not math whizzes. Joe Borowski kills you every time. The value of the “never pay for saves” adage lies not in blindly waiting for garbage Fernando Rodneys at the end of the draft, but in the awareness of the volatility of saves in general and the ability to acquire quality relievers who acquire the role midseason (Axford).

    In short, pay for some saves at the draft, don’t draft Rodney, and be a hawk during the season and score the Axfords when they come.

  6. Sal said...

    All well and good, but you’re severely impacting hitting and two other counting categories (Ks, somewhat predictable, and Wins, are they as unpredictable as saves?).

    Although I partially disagree with your strategy, I do think that the pendulum has swung so far that savvy owners should be on the lookout for that top tier reliever that does fall far enough on draft day that you can snap up at a great value.

    Regardless, great article. It’s tough to write about RPs and saves. Nice work.

  7. Dave Chenok said...

    @Chris: it is certainly possible to punt a category—saves or otherwise—and still win your league, but I don’t recommend it as a strategy.  Too many things must go right in all the other categories.  I also note your league format, and that is an important consideration in any strategy one puts together—the first question I always ask about a league is, “how does the scoring work, and how many teams.”

    @Will: I like your strategy.  Again, its applicability depends a little on league format, including how many Ps you can roster at any one time.  The more Ps you can start—and the more flexibility about how you deploy starters vs. relievers, innings limits, etc.—the better your approach works.

    @Mark: thanks for your thoughtful commentary.  One thing I’d say though is that your approach risks exactly what I’m cautioning AGAINST.  Will there be a bunch of saves available mid-year,  especially with the closer mess going into this year?  Absolutely!  But will they be more like Mo, or more like Joe?  Often the mid-year guys end up being Guitierrez types, guys who get tried out as an experiment, do well for a few games, get anointed as closer, and then see their peripgherals deteriorate as the league catches up with them.  It is specifically BECAUSE of the volatility in the closer ranks this year that I will be drafting elite closers early in many of my drafts.

  8. Bryce said...

    It is a delicate balance for sure, but like all snake drafts you are at the mercy of that one guy that takes the first SP and then the one guy that takes the first RP. As soon as Halladay goes, Felix and company are usually right behind him regardless of what hitting studs are on the board. Likewise when Mo goes, Marmol and the crew start flying off the board. You’re that guy that I cannot stand! 

    Seriously though, ADP has Mo going #61 (earliest #54) and Mamol #71 (earliest #57) right now. Where would you usually take your first RPs at?

  9. Mark said...

    Just to expand on my point above about how shallow RP is at the top, here are a few specific examples:

    -Mariano Rivera (61 ADP):  Despite another apparently dominant season, the 41 year old actually showed a few signs of chinks in the armor last year.  Career or near-career lows in xFIP (3.65) and K/9 (6.75) along with a BABIP (.222) nearly .040 below his career average and a HR/FB (3.6%) nearly 3% below his career average show he may not be quite the same uber-closer he’s been through his career, and yet he’s still the first guy off the board in most drafts.

    -Carlos Marmol (70):  He’ll sure help you a ton in Strikeouts, but the offset WHIP killing he’ll give you make him a pretty unbalanced choice for second closer off the board.  He’s a three category player whose really only excellent in one category, and its offset but how he can hurt you in another.  You can do much better value-wise in the sixth round by choosing an SP or a position player without paying for it later. 

    -Neftali Feliz (109):  Excellent numbers and little reason to doubt he can repeat, but decent chance he’s tried in the rotation at some point. 

    -Johnathan Papelbon (113):  A weakish season last year and the presence of Daniel Bard means he carries more risk than in previous seasons. Yes, he’s being drafted as such, but there are similar situations in terms of risk and upside much later on in the draft (Johnathan Broxton comes to mind).

    -Andrew Bailey (132):  Precipitous drop in K/9, flyballing ways, 91.1% strand rate, and a balky elbow make him a very iffy choice for the 8th closer off the board, especially with all the bet hedging Billy Beane appears to be doing behind him.

    -Jose Valverde (138):  Downside of WHIP killer a la Marmol but without having anything particularly special about him otherwise (aside from relative security, albiet with some injury risk as well).

    -Francisco Rodriguez (151):  Good year last year, but if he finishes 55 games he gets paid like a 3.5 win player in 2012, and Sandy Alderson may pressure Terry Collins to avoid this situation.  He still should get plenty of saves, but there’s enough risk here that he should be treated as a guy with as much downside as upside even at this draft position.

    All these guys are getting drafted in 100% of leagues, meanwhile you have some very undervalued upside in the later rounds and even getting left off some drafts:  John Axford, Joe Nathan, Brad Lidge, Jonathan Broxton, Craig Kimbrel, Matt Thorton, Frank Francisco, Kevin Gregg, Joel Hanrahan.  Are there some job security risks here?  Sure, but in most cases there’s excellent upside, little performance downside compared to the upper level guys, and a very passable understudy in place who isn’t even getting drafted in the majority of standard format drafts (Hong-Chih Kuo, Matt Capps, Johnny Venters, Chris Sale, Evan Meek, Octavio Dotel, Koji Uehara). 

    This may be the last year where its really true, but it still seems like a prime year to go the patience route with closers and reap serious profit with that strategy.  The only upper level guys I’ll consider this year are Marmol, Brian Wilson, Heath Bell, or Feliz and probably none unless they fall past the 8th round.

  10. Mark said...

    Dave,

    Thank you for the response (my last post was in the works before I saw it).  Here’s why I disagree:  There are too many situations where there are teams deep in quality RP without a settled closer, or with a high risk but huge upside guy, too much upside in the back half of the draft, and too little security at the top.  The back end of the closer ranks actually seem pretty deep, much deeper than the top end.  Most of my reasoning is in the post above, but just wanted to respond directly as well.  How different is Jonathan Broxton than Jonathan Papelbon?  Broxton’s situation seems so binary and decryptable (we should know what he is pretty quickly in March), and he has such a potent understudy in Kuo, that I would much rather draft him today and see what he does in the Spring than draft Papelbon in the 9th or 10th round.  The risk, of course, is that someone else drafts Kuo and Broxton’s velocity dip and results continue to suffer as they did in the second half last year.  But there are so many other fluid situations with multiple closing candidates with legit MULTI-CATEGORY upside like this right now that the risk seems mitigated.  Would Dotel really be that different from Valverde if he had the job?  How about Sale stacked up against Bailey or Feliz?

    Again, I agree with your thesis, but I think it only works when you can mitigate your risk with true lack of downside.  I just don’t see that guy this year.  The only three candidates Soria, Bell and Wilson and only Soria qualifies as nearly as elite as Mo, Broxton, K-Rod, Papelbon, Wagner, etc. have looked in previous years when they were at the top of their games.  And even though I think team quality is overrated in closer selection, the Royals could be this year’s version of the Pirates and go months without a decent run of save opportunities the way the Pirates did in the second half of 2010

  11. Samuel Lingle said...

    The flip side of this coin is that you can still get closers who post good stats in the late rounds.

    Maybe you won’t get guys who are at a 1 ERA, but its quite possible Jake McGee has an ERA in the low 2’s with 70 strikeouts and 35 saves, or something like that.

    Relievers are very volatile so even when you spend to get that closer with good job security and great ratios its still hard to really tell what will come of it. Marmol, for example, is the 2nd closer on many lists but its always possible he loses a little control and blows up.

    In the past you had Rivera, Nathan, Soria, etc who you could all count on to provide solid numbers assuming no injuries, but to me it seems this year there are more questions than ever with the closer position.

  12. Red Sox Talk said...

    Dave, kudos for standing against “established wisdom” and giving some evidence with your argument. I have been saying for years that elite relievers are worth high draft picks. I have drafted the likes of Papelbon and Rivera in the 3-5 rounds for years, and enjoyed the strikeouts, ERA and WHIP benefits of doing so.

    As some of the other commenters note, fantasy has rapidly evolved and people are out there trying different strategies more and more. So it’s hard to say what the value of elite closers is for your particular league. You could punt saves in a league where the competition is very fierce for closers and do well. You might not do so well in a league where other owners employ the same strategy, or are content with one closer rather than three.

    My strategy is basically to maximize overall fantasy value early (5 categories), fill in deficiencies in the middle rounds, and target sleepers late.

  13. Paul said...

    Nice post Dave & it mostly squares with the team you drafted in the mock (good MI early and a few good closers).

    The point is that closers used to be overvalued in drafts and the smart play was to wait, now it’s common to do this, so the ‘zig’ in an expert league is to get a soria type early – because you have 11+ other wire hawks to compete with for the Axfords of the year + you are less likely able to absorb the rate stats of a JoeBor in a good league.

    In a less competitive league, you can probably wait on closers as you are more likely to strike ww gold.

    The other point is that if we are now in another SP zone like last year, then there are better SP around, so everyone gets a better era/whip – meaning that if you are guy stuck with the JoeBor types then you are further behind on the rate stats.

  14. kevin said...

    It seems like you’re banking on a lot going right elsewhere with this strategy. You say use the first four rounds on strong hitters at weaker positions and you can have your cake and eat it too, but what if one of the said hitters is the same chase utley you mentioned?

  15. Dave Chenok said...

    @Kevin: if you drafted Utley as your first pick, you probably lost regardless of what you did with closers in later rounds.

  16. jeffreygross said...

    (oh, and it should be apparent that I often play in deeper pitching leagues, so the numbers may not translate as well to shallow leagues)

  17. Dave Chenok said...

    @Bryce: It’s not that I’d take them much earlier, but I would probably take 2 top drawer closers in successive rounds at about 6 or 7, and then pick up another at round 9 or 10 (depending on what mischief my first few picks cause).  When our mock draft commentary comes out over the next few weeks you will see that there is also a relationship between my closers strategy and draft position.

    @Mark-I’m certainly not saying you are wrong.  There are a lot of ways to win a fantasy league.  But at least you’re THINKING about this.  A lot of guys look like they came out of the hypnotist’s office vacantly repeating “don’t pay for saves.”  I mean, I watched a guy last year (well, actually he was texting, but you get the idea) laughing about how he’d “stolen” Rodney in Round 17, and how he was going to get as many saves as I was going to get with Mo in Round 7.  maybe, but…

  18. Dave Chenok said...

    @Jeff: we’re kind of like Obama and the Tea Party here I guess…in principle I agree with the point that a 4 category player is worth more than a 1 category player.  I probably value the 210 innings of peripherals put up by 3 good RPs more than you do; while RPs only DIRECTLY influence one category, they INDIRECTLY influence 4 others in a 5×5 scoring format, and the instinct of small effect size is wrong—the difference can be worth a LOT of points.

    There is admittedly a certain arrogance to my strategy.  It relies on my ability to “outpick” competitors around deep (SP, OF) positions in the teen rounds of a snake (or after much of the league’s auction budget has been spent).  My experience has been that I can put together a starting rotation and/or OF that are more or less as strong as any competitor’s without using Round 8-12 picks for them, hence cleaning up on RPs in the earlier rounds is a winning strategy.  Some of this is described more in my mock commentary, so let me stop for now to keep everyone interested.

  19. Jeffrey Gross said...

    I think its all quite interesting. I focus on the fact that bad RP dont hurt you, but agree a good one can certainly help

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>