Why discretion is the better part of valor

I’d like to follow up on some issues surrounding Jonathan Halket’s most recent column about backward induction. One of the questions raised by the column is that of holding off on high-round options at positions you’ve confidently targeted sleepers for. How do you balance the appeal of getting high value late in the draft against the risk of not acquiring your targeted sleeper and being left high and dry?

The nightmare scenario would go something like this. An owner pegs Julio Borbon as a great value late in the draft. In the third round, at pick 26, this owner notices Carl Crawford is still on the board. The owner considers Crawford here, as he thinks this is a good value. However, he thinks Borbon could be an even better value a hundred picks later (great players are at a disadvantage here because there’s only so far a highly ranked player can outproduce his cost) and doesn’t want to clog that outfield-speed slot. He thinks rostering both Crawford and Borbon would be overkill and leave him power-deficient. So, the owner drafts Kevin Youkilis instead. Then, several rounds later, somebody else drafts Borbon a round before this owner was planning to. Now, this owner is behind in steals and doesn’t have another player to fit the archetype he needs. He ends up drafting Coco Crisp and hoping he surprises and fills some of the need he was hoping Borbon would (and he knew Crawford would.) To make matters worse, Crawford outproduces Youkilis and other quality corner bats were available the round following the Youkilis pick.

(I realize I could have just used Jonathan’s Ryan Doumit/Victor Martinez/Rod Barajas example here, but then I’d be tangentially entertaining the “should I invest highly in a catcher argument” and if you read my columns you know that I’m anti-high pick catcher. Unless I’m in an AL-only league, Joe Mauer could be sitting on the board in the third round and I’ll most likely pass without giving a second thought.)

Now, I can’t tell anybody how to balance the risk of not getting Borbon versus the perceived value gap between Crawford in his draft position and Borbon in his. This is a rather abstract concept to try to quantify. The answer depends on a number of dynamics, as discussed in Jonathan’s article and the comments section. What other options will be there if I get sniped on my sleeper and how do I feel about those options? How likely is it that others are on the same wavelength as I am regarding my sleeper targets?

There’s another important question though—the most obvious one—that somehow often gets lost in the mix of the more abstract components of this discussion: How do the elite option and my sleeper compare in terms of absolute value and what are these respective players’ realistic downside potential?

To prime this discussion, it seems that a refresher on the actual workings of the market in relation to elite players is in order. Jonathan is actually an economist; I, on the other hand, am just a pedantic know-it-all, so I hope I’m not speaking out of school here.

Everybody must be aware of two factors that go into the price of an elite player. One is dependability. You or I may interpret Borbon as a great value, but it’s also possible that he’s demoted to the minors before the All-Star break, or that he can’t hit lefties at the major league level (he barely even played against them last year) and therefore winds up only getting 400 ABs for the season. I’m not trying to throw water on Borbon optimism, just expressing legitimate possibilities that people tend to totally dismiss once they’ve bought in to a player. Our evaluations don’t matter—only his manager’s and organization’s.

Those risks don’t exist for Carl Crawford. There are no risks that apply to Crawford that don’t apply to Borbon.

So, Borbon has significantly more downside risk than Crawford. This is an important determinant of the price points of the respective players. But, wait … isn’t some of that downside potential difference mitigated by the price I pay for the commodity? Isn’t Crawford’s smaller risk of not producing elite (by his standard) numbers magnified by the amount I invest in him, while Borbon’s greater downside risk cushioned by the fact that I invested less in him?

In an economic sense, yes. That’s an astute observation. (See how I pretended to compliment you, but I really complimented myself—I told you I’m kind of a douche.) But not so fast there, imaginary devil’s advocate.

The goal isn’t really to just accrue the best bang for you buck; the objective is to win your league. Maximizing your value per dollar (or pick) invested is just a tactic by which you do that, not an objective in and of itself. Commenter Jonathan Sher makes a rather poignant point about this in relation to his strategy for one of his leagues, which we will get to a bit later.

If you are relying for Borbon for big-time production, you are still out that production if Borbon flops, or even just performs acceptably but fails to exceed his draft position. What you invested in him is less important than what you depended on him for.

Here’s a real world example: David Wright is a borderline MVP-caliber player, but he only happened to make $7.75M last season. His hugely disappointing 2009 campaign was probably “worth” his salary, but that doesn’t mean the Mets were satisfied with his performance. And more directly, the Mets certainly didn’t do a lot of winning last year! So, in practice, the value consolation is largely irrelevant to a struggling team.

Plus, there’s the third-party production that enters the final equation, which is the opportunity cost of taking Youkilis over Crawford (plus the added value the owner behind you got in Crawford over what would have been his pick. Drafts are particularly conducive settings to see “the butterfly effect” in action.

If Crawford performs exactly as expected, Borbon performs 20% better than expected (and still substantially below Crawford), but you had “budgeted” for 40% better, and Youkilis performs 10% better—are you ahead of where you’d be if you just took Crawford (and some other undefined player in Borbon’s round who performed as expected)? … Hmm … In overall value, you probably are. But do you have the right mix of categories or is some of your value tied up in categorical surplus irrelevant to the standings?

This gets very complicated, so I’m just going to end this part of the discussion and let it stand as a rhetorical. I can only carry on a debate with one imaginary devil’s advocate at a time without devolving into full blown schizophrenia. So, let’s move on to the second market force.

The other market force relevant here is the non-linear value of luxury goods (there’s probably a 10 dollar economics term for this, but I don’t know it). Superstars are luxury goods. If you are buying a car, for example, there’s something of a “you get what you pay for model” for most options along the price point chain—Ford, Honda, Lexus. However, as you approach the extremes this somewhat linear relationship falls apart. It’s virtually impossible for a Bentley to be something like six times the car a Lexus is. In fantasy baseball, there reaches a point of production at which there are so few players who can replace that player’s value, his production value versus his cost ceases to be linear.

This happens with players in the real game too. A-Rod is nowhere near the best value in terms of, say, cost per win share. But, how many players can produce at his level? The way the market works in real life is that as a player’s production rises, the cost per unit of production rises a bit too. Hypothetically, win share 1–10 costs one price per win share; each win share between 10–20 an escalated price per unit, 20–30 even more, and then the price explodes at the superstar level. If you need one player to produce 35+ win shares, there aren’t many options. So, you pay a premium for A-Rod’s ability to produce virtually unequivocal value (and to protect against other teams usurping that value from you—this is a facet many never bother to consider).

So, what does all this mean in relation to our original question about studs and sleepers? Why did I go through all of that? Do I just like to run on and take advantage of the attention spans of the five people who would actually listen to me? Well, yes, but I also have a point, though it’s not one I haven’t made here before.

You should not let your late-round sleeper targets affect the highest rounds of your draft. At least through the first 50 picks, you should be selecting who you determine to be the best overall player available to you at your pick. Personally, I extent this rule notably further than the top 50, but I think the top 50 is the shortest acceptable period to apply this principle.

Here are a few reasons why I preach this:

First, you need elite producers to win. Everybody is going to have some luxury goods, so pick the ones that are best compared their alternatives. If you are in the market for a $350,000 luxury sedan (which you are, by definition, in the top 25) you should be weighing your options in that price range. You should be comparison shopping the Rolls Royce Phantom against the GT Bentley. The fact that you think a Lexus LS430 represents a far better value than either isn’t relevant. … Your debut rap album isn’t going platinum with a Lexus on the cover; this isn’t 1994 (unfortunately!).

Just to nitpick the metaphor, if you’re considering Ellsbury or Crawford you are really more in the market for a luxury sports car, right? OK, I’m officially getting to close to the Bill Simmons zone now, aren’t I?

So, here’s a big piece of advice I try to give to fantasy players. Don’t outsmart yourself!. Often times, us more advanced players get too wrapped up in our ability to spot value that we forget that leagues are won by absolute production. Don’t romanticize your sleepers to the point that you think they can actually replace studs. Occasionally it happens, but if there was a good chance a player could provide legitimately elite value, he likely wouldn’t be pre-ranked outside the top 100.

By the way, the outsmarting yourself phenomenon is even more dangerous and frequent in fantasy football. I don’t know how many times I had to talk friends off ledges this past year, as they were considering benching their first-round running back or second-round wide receiver for some injury replacement or secondary target blessed with an “attractive match-up.”

Here’s Jonathan Sher in the comments section to Halket’s column, as he talks about this issue in relation to an auction league (emphasis mine):

One’s willingness to [pass on a stud in favor of a sleeper] should also be related to one’s risk tolerance, and not in some generic sense, but the circumstances of one’s auction. This year, for example, I have what I believe is the best keeper list in my league and certainly one of the top 3—high value guys at very low salaries. So rather than risk waiting for a sleeper, I’m much more inclined to pay inflation-adjusted prices for top talent like Longoria than count on Beltre being under-valued. Beltre probably has a greater upside compared to what I expect others will value of him. But I don’t need more high value guys to win. Rather, I need high production guys (since I have most of my budget) and can afford inflation-adjusted prices or even more and still win or finish near the top. My greater risk is running out of available talent before I run out of money.

Second, what is precluding you from taking a stud and your sleeper, even if they are similar in make-up? Drafting is about both building a team and stockpiling value. If your sleeper pans out as you predicted, you have a valuable trade chip either in him or the higher-priced similar player. You can cash either of those chips in for elite production where you need it. In this specific case, you can jump out to a huge lead in SBs, dump one of the players and not even have it affect your standings in that category too much, as the low volume of the steals category makes large leads difficult to overcome.

Third, you should have sleeper targets for all categories. So, if picking the best available player for six or seven picks starts to develop clear categorical needs for your squad, you should have a plan to address them, whatever they are. I much prefer a large list of less profound sleepers than a few pet sleepers to whom I am extremely committed. You will never be able to draft your whole list, so the way I work is to build the best team I can with my top picks and then start to go after the sleepers whose skill sets make the most sense for my team. Obviously, there are exceptions. I drafted Josh Johnson regardless of my pitching staff’s strength in literally every league I played in last year.

Finally, here’s another question for you guys to ponder. I avoided temptation to get into this in this column because my columns are already long enough. But … what exactly even constitutes a “sleeper”? Borbon was discussed a fair amount both in Jonathan’s article (including the comments section) and mine. He’s pre-ranked 113th on Yahoo’s draft; is he even a sleeper? Can you be a sleeper while projected to go in the 10th round of a 12-team draft? I always thought sleepers were much more obscure. Scott Sizemore is a sleeper; Julio Borbon is not. That’s how I see it at least.

While others use the term to refer generically to anybody who they feel will considerably outperform draft pre-draft rank, I tend to reserve the term for either players few know about, or those most have written off as being unable to fulfill their potential or clearly past their tenure of usefulness.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Petco Park revisited
Next: This annotated week in baseball history: Feb. 14 – Feb. 20, 1874 »

Comments

  1. John K said...

    Loved this column, thanks a lot.  Addresses something I struggle with every year and has already helped me work out my draft strategy a bit more. 

    Separately I’m answering your rhetorical question and throwing in my two cents: a sleeper is someone who has a good chance of significantly outperforming their draft slot.  I think the term “deep sleepers” is typically used for players that fit your definition.

    Finally, I just want to register this: my day is very slightly brightened every time I read “Derek Ambrosino aspires to one day, like Dan Quisenberry, find a delivery in his flaw”

  2. Jonathan said...

    First of all, thanks for following up on my article, Derek.  Some of your examples are much better than the ones I came up with. 

    As for relative vs. absolute value or production, I will say this: Apart from some perhaps important quirks, he who wins the relative value race will win the league (and the absolute value race). 

    That is, he who earns the most ‘profit’ will win the league (assuming that the profit is well distributed in all stats – but of course this applies to absolute production as well).  Needless to say, the ‘most profit’ should be measured as the sum of all profits and not the sum of something like %profit (obviously a 100% profit on a $1 player is not the same a 100% on a $20 player).  Also, players must be able to earn negative profit and be able to have negative realized values – no lower bound of $0. 

    Finally, all owners must have the same budget (this is relatively obvious in auction settings) Then it is just an accounting identity:
    relative profit for a player = end of year value – concensus start of year value (or price paid in auction)
    so:
    Sum of relative profits = Sum (end of yr value- start of year value) = Sum (end of yr value) – Sum(start of year values) = total Absolute value – start of year budget.

    So the guy with the highest sum of relative profits is also the guy with highest tot abs values or productions, since the start of year budgets are the same for all teams…

  3. Jonathan Sher said...

    Derek,

    Thanks for expanding so well on the point I made in response to Jonathan’s piece. A few comments:

    (1) I don’t play in a draft league but your advice about the top-50 (or more) picks makes sense. Bob Gibson made a similar remark about his pitching philosophy—don’t outsmart yourself by not throwing your best pitch.

    (2) The only thing that precludes you from getting a stud and sleeper in a category is the limit on roster spots for each position. If you’ve filled two of three corner infield spots, your favorite sleeper is still available and one of the last of the established corners is nominated at an auction, what do you do? Do you allow the availability of that sleeper to affect how high you bid on that corner? Do you discount what you would otherwise bid, and if so, by how much?

    (3) I am so glad you made a point about what does and doesn’t constitute a sleeper because that issue is too not examined critically by those who dole out fantasy advice.  Here are some thoughts:

      (a) There is much written about sleepers before the auction/draft and that attention sometimes obliterates the chance some of those sleepers will get undervalued.

      (b) I try to factor in (a) in coming up with a list of players I think will be under-valued by other owners in my league.

      (c) The approach in (b) often leads me to target:
    - good but not great veterans coming off a year worse than their career norms (Adrian Beltre, for example)
    - players who may benefit from a change in ballpark or batting order position (Beltre again)
    -  young players whose prospect bloom has faded because they performed below expectation, didn’t get an opportunity or were injured (Phil Hughes)
    - prospects who aren’t top echelon but will get a chance to play (Sizemore)

      (d)The answer to what is a sleeper is first and foremost league specific. In many leagues, Scott Sizemore would be a sleeper because his skill set has never been of the sort that got him labeled a top prospect, his production in the minors suggest he could be decent in the majors and he will get amply chance to show that this year. But in my auction league, in which half the owners are Tigers’ fans, Sizemore is not so much a sleeper.

      (4) The question also affects trading strategy. I have Julio Borbon on my roster. The more he is hyped as a sleeper and the more that hype is believed by some owners in my league, the greater his trade value and the more likely that I will be able to peddle him before the auction for more than what I expect his actual value to be. So please keep hyping Borbon.

  4. Jonathan Sher said...

    Jonathan,

    It is true whoever end up with the most relative profits, as you define it, wins. And it is equally true that whoever ends up with the most absolute profits wins.

    But there is one assumption that isn’t always true in an auction: that the start -of-the-year budgets for all teams is the same.

    Let’s say, for discussion purposes only, that an owner needs to accumulate $300 worth of performance to win a league with a conventional $260 salary cap. Now let’s examine two owners who have very different keeper lists heading to auction:

    Team A owner has followed the sage advice of Hardballtimes and has a superior keeper list whose salary is only $60 but whose absolute value at season’s end will be $200.

    Team B owner sleeps with a framed picture of Joe Morgan under his pillow and has a keeper list whose salary is $160 whose absolute value at season’s end will be $100.

    Assuming, for a moment, that Team B owner has an epiphany before the auction, trashes the Morgan portrait and replaces in its frame a picture of Dave Cameron, and now enters the auction with all the acumen of Team owner A. What do each have to achieve in order to win?

    Team A owner needs to spend $200 to buy $100 of value. Sleeper picks are of little value to him and in excess are harmful because he may run out of roster spots before he’s accumulated $100 of value. So he instead picks high value guys at inflated prices—double their absolute value—getting $100 of production for $200 cost.

    Team B owner needs to spend $100 to buy $200 of value. He comes armed with a sleeper list and successfully squeezes every dollar well, spending $100 and getting $190 of value.

    Team B owner gets more bang for his buck at auction, nearly four times as much as Team A owner. But Team A owner wins the league and spends the subsequent off-season sending anonymous emails to Dave Cameron that simply read: You may be smarter than Joe Morgan but that alone is not enough.

  5. Rotodog said...

    All the points you guys make are certainly relevant. Very good, technical points.  Here is my one issue with this line of thinking. Most of the article discusses Profit and profit potential. Yes, profit is needed to win.  But by focusing solely on profits, i think we OUTSMART ourselves.

    Give me a guy that takes the most valuable player everytime at a position he needs and I will show you a guy that rarely wins.

    Jonathon eluded to it in this statement tucked into the whole response.

    “That is, he who earns the most ‘profit’ will win the league (assuming that the profit is well distributed in all stats – but of course this applies to absolute production as well).

    Now, in a rational world and a rational market, that statement is pretty much true. BUT, the biggest thing standing in the way is the comment in parenthesis . “assuming that the profit is well distributed in all stats”. 

    This is rarely achieved in a profit only seeking drafter. It takes a lot of work to construct a roster of equally distributed profits. Many people do not realize that roster construction should be the #1 focus, while seeking profits within that original focus. But you must assemble the best and most rounded set of stats you can for the money. 

    One simple example.And please, this is an exercise, I am not considering trades and other things…

    You own Carl Crawford, Jacob Ellsbury and David Wright.  Jose reyes is nominated and his price stalls at 15 bucks. More than half of what he is worth. What a Bargain and what a profit on a STUD!!  But is he right for you? Is that the best way to allocate your dollars? Even at half price?

    I know the example was very simple, but I want to reiterate the fact that the best bargains dont always equate to success.. The best bargains within the structure of a properly drafted team will usually win out.

    I see many players consider profits as a top priority with little regard to what they are buying and when..  just Something to consider…

  6. Jimbo said...

    Love the line about Crawford’s risk vs Borbon’s. Very true how infatuation can take over. There are a couple things I’m not ready to concede yet, and would love your feedback. Seems your primary objections are with the risk of sleepers being available and the potential loss of production from passing on an elite player.
    1. What constitutes a “safe” sleeper. I make my list of emergency picks, some undervalued bounceback candidates, and “blue-chip” breakout candidates. There are two times later picks influence my first 5-8 rounds: if there are ample targets available who present low ‘availability’ risk—eg I wouldn’t count on landing Beltre in a specific round, but if I like him and Cantu and other similar-tiered options then I might not fill up the position early; the other is when I identify one player who I simply want on my team next year—making sure I take him before anyone else would think about it. I have no problem looking like a fool to draft a player I want…just don’t want to do that very often.
    2. Not sure we agree on the application of this backward logic. I wouldn’t suggest taking a clearly inferior player because Borbon might be on my team later. Insead of Crawford v. Youkilis, how about Crawford v. Kinsler v. Votto? Their stats aren’t very comparable on the surface, with different distribution of strength, so how do I evaluate the best pick in round 2 if I deem them all to have similar total value? There is position depth to consider, floors and ceilings to think about, category strengths, my specific draft strategy, league inefficiencies, and yes…alternatives for that specific spot on my roster. If I don’t like many 1B after Votto; I do like some 2B after Kinsler; and I like a lot more OF talent than I have room for on my team? Votto may be the best choice.

    If I “rule out” any early SS because of who I like later in the draft, it isn’t solely the fault of those sleepers/value picks. It is a summary judgement that early SS won’t have more value than their ADP peers (eg corner power or OF studs) given my approach to the draft and review of positional alternatives across the board.

    Without some semblance of a bottoms-up view of the draft, I don’t know how people make certain decisions. Choosing between Lincecum, Mauer, Wright, Crawford, Kinsler, Tulowitzki, Reynolds with the 12th and 13th picks can create too much anxiety and ‘chaos’ for me. All those players are just about in order of ADP and cover all roster positions. Being able to rule out a need for a certain position is a start. Now my decision is a bit easier.

    Your post seemed to infer that there is always an “obvious” value being passed on to execute this strategy, but for me it’s about simplifying the genuinely close calls and maximizing total team profit.

    Sleeper = anyone valued significantly (pretend that’s in bold) more than market. Granderson in the 3rd could qualify imo.

  7. Jonathan Sher said...

    Derek,

    I had Melvin Mora at 3B as a keeper no less so I cry no tears for you and Garrett Atkins,

    Draft strategy and auction strategy are quite different. While I don’t play in a draft league, I do mock drafts for fun and generally wait on corners because there are many who don’t go to middle rounds who I especially like—two of my favorites are Billy Butler because of what he’s done as a hitter at such a young age and Gordon Beckham because I love that he will quickly qualify for 2B too.

    But in an auction, it’s not a question of when to take someone since you only control 8% of the player order assuming a 12-owner league. It’s a question of how much to bid. If you have Ryan Howard in hand, you might discount your ceiling for bidding for Teixeira but you still bid and you still need to quantify how much that discount will be.

  8. Derek Ambrosino said...

    Jimbo,

    General positional depth is certainly a dynamic overshadowing all early round decisions, but theoretically that demand is largely reflected in each player’s price to begin with. In cases that are much more clear cut, I will consider my expectations of the later rounds of a draft heavily when determining my earlier actions. For example, I’m not interested in touching Mauer in a shallow league because my positional floor is basically set. People aren’t going to be drafting multiple catchers, so the worst case scenario (in the absolute sense) is that I end up with the 10th or 12th best catcher. At the point at which I draft that player (very last rounds) he’s actually likely to be a value.

    In general, it’s helpful to at least vaguely organize positions into tiers, but that high in the draft, there are still plenty of quality options at every position – even the thin ones – so I suggest drafting the best overall player. Value over replacement is part of that equation, but the more adequate replacements that remain, the less important it is.

    Rotodog,

    If there is one thing I want readers to take away from this piece it is that which you highlighted – beware of outsmarting yourself.

    We all must admit that it makes us feel good to pull out an expected draft pick, and have others look at you like, “what’s this guy doing,” only to see that player perform, while you sit smug, cementing the perception that you have more finely tuned sensibilities than your friends. But, only make that pick because you think that is the best pick to make, and not for e-props. I’ve warned before that I suspect several experts get caught up in this as well, because these great calls are what they build their rep on. It’s just like the flashy “stock analysts” on TV; they rarely tell you to invest in the steady growing blue chips. They want to be able to add to their resume that they correctly identified the next big thing, and you don’t do that by encouraging people to invest in the boring, reliable blue chips. But, when you look at the portfolios of those swimming in wealth, there’s a ton of blue chippers in there.

    Jonathan S,

    Re q#2: Obviously, these are subjective questions. I’m going to say something here that may make me sound a little hypocritical: I don’t like to lock up most of my corners or OFs too early in the draft. I know I say to take the best player regardless, but I like to retain versatility as well. So, I’m highly unlikely to do something like taking Ryan Howard and Mark Teixeira back-to-back in round one and two. The reason is basically that the value discrepancy and reliability among groups of top players (say ranks 12 – 18) aren’t huge. So, while my second 1B may be “the best on the board,” he may only be a 3% better than the outfielder next to him. Meanwhile, I see that the lower second tier/higher third tier corners often stick around to the point at which they become very good values. This because many teams load up on those slugging corners. So, if I take OFs and MIs in rounds 2 – 4 and come back to those corners in round 5, the corner available there might represent a greater absolute value over other available players than Howard did at pick 15. The difference here, as opposed to waiting on a sleeper, is that we’re still operating in the realm of established, reliable players. So, there is less risk of totally busting.

    Of course, with that said, I wound up with about three teams last year that had Garrett Atkins as its starting 3B…

  9. Penguin said...

    On a similar topic, when selecting keepers my quandary is this. Do i keep a player with an ADP in the 7th round where i only lose an 18th round selection, or do i keep the stud who is going in the 2nd round where i give up my 4th pick?

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>