Chaining draft picks

Let’s say you are in a draft and with your first-round pick you select second baseman Chase Utley. Not a bad choice, I’ve seen it done plenty of times before. That the remaining second basemen become much less valuable to your team is a concept most people understand.

The extent that the remaining second basemen drop in value depends on your league settings—whether there is only one second base roster spot or multiple positions you could stick a second baseman (e.g. a middle infield spot) is the determinate. And obviously the more positions you can stick a second baseman, the less each available second baseman drops in value to you. Only to you.

In the short term, most people are aware of this drop in value in drafts. I know this because rarely do you see someone take two second basemen early in a draft. Even when there are more than two spots to play second basemen on your roster, most people will hold off on a second one until at least the middle rounds, and when there are only two spots for second basemen (2B + Util spot) most people will not even take a second one.

Whereas people understand this in the short term, when it comes to putting together a full draft people forget that who you draft in the first round affects even who is most valuable to your team in the last round. I hate throwing the term value around like a curse word in a painfully unfunny Bob Saget comedy stand-up, so let me give you something more tangible to grasp.

Let’s say you are about to start a draft. At this point you know next to nothing about how it will end up looking—you don’t even know what pick you are going to have yet. All you have are your positional rankings and a list of sleepers to target at the end. Your top three sleepers are a shortstop, an outfielder and a pitcher.

Although you should not completely base your first few rounds on who you think you might will grab in the later rounds, it does make sense for your first three picks to not be a shortstop, outfielder and pitcher. You might draft a player from one or even two of these positions in the early rounds—if there is a great outfielder out there in the second round, go get him—but understanding how that affects the rest of your draft is important.

So you go into the drafting looking to target a first, second or third baseman early. The draft begins and you get your first and third baseman early, but a good second baseman eludes you as the draft heads into the dreaded middle rounds. With no second baseman on your sleeper list you’d be comfortable with in a starting gig, now is the perfect time to “reach” on a second baseman in the middle rounds, say Jose Lopez in the eighth round. Sure it might not be the best pick and sure his ADP is almost 30 picks later, but with the special need you have the pick is more than defensible.

Now, you do not want all of your middle-round picks to be this sort of defensive type, but if you are going to reach at some point on a player, reaching in this situation can be called ideal.

I understand that the concept discussed in this article is not something most people don’t know, but I do believe it is something people should be more consciously aware of in drafts. Understand that your late round targets affect your first round targets, and who you actually get in the first rounds affects the value of certain players later in the draft. Any position that gets lost in the shuffle can excusably be targeted in the middle rounds and when you chain the parts of a draft together in this way, you will put yourself in the best position to get the most out of a draft.

Ultimately, though, it is the individual players themselves who determine how good a draft was.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Who is the most valuable human asset in baseball?
Next: Beyond DIPS »


  1. Andrew P said...

    I disagree. 

    Unless you’re operating in a league without trades, or an extremely short bench, the last few rounds of the draft are almost always going to be bench players, whether they’re sleepers or not.  Unless you’re in an uncommonly tailored league, drafting your late round sleepers to be starter(s) on your team usually isn’t a very viable strategy.  It also means that you’d be taking backups at other positions before taking your starters at the sleeper positions.  Now, if you feel extremely strongly that your mid-round backups will perform above their draft position, there’s an argument to be made for taking them.  Nearly everyone, however, takes a more agnostic approach- there are guys we will reach for, but we want to draft people around their ADPs.

    If you’re reaching on Jose Lopez in the 8th round, you’re counting on your late-round sleepers, or late-round outfielder- to make up for the surfeit of value in your decision to draft Lopez.  You’re buying high on Jose Lopez, because you feel you can make up for it with one of your last picks.  Even if your sleeper performs well, you’re really only treading water, since you gave value on the Lopez pick.  You need your sleeper to REALLY perform well for that strategy to be worth it.  Even then, it’d probably have been better to simply draft two shortstops, or two outfielders, and take whatever 2B falls to you.  At least then you’ll have a viable trade option should your sleeper perform, rather than creating a hole that can only be filled by your sleeper.

    Furthermore, if you’re waiting until the last rounds to take your sleeper starter, you’re operating without a safety net.  What happens if someone takes your sleeper a round before you?  A pick before you?  You’re stuck with the dregs of the position.  So now you’re left with Jose Lopez (who you reached on), a waiver-wire shortstop (or OF… whatever), and a roster spot wasted on that waiver-wire player that could have otherwise gone to a sleeper.

    To me, this is a recipe for failure. 

    The more viable strategy is to examine the draft position of different tiers of players, and tailor the positions you’re looking for to certain rounds.  Obviously, you don’t know how things will play out in the draft, so you adjust when a particular position is going earlier or later than you’d expected- dynamism in positional scarcity during the draft.  If you’re less agnostic about certain players, and feel strongly about getting them, you’re still free to reach.  You do so at the potential cost of a lower-tier player at another position.  The cost of the reach, though, is offset by your projected gains in drafting that player.  Typically, you’ll fill out most or all of your starters before drafting your backup sleepers.  Occasionally, you might punt a position for awhile (during the draft or even the preseason) if there are no worthwhile players to be taken.  This would be analogous to not drafting a kicker in fantasy football.  If you punt a position, it’s because one player doesn’t have much value over a FA replacement.  It is much more flexible than reaching in the middle rounds, hoping both that your sleeper performs well, and that he will actually be there to be drafted.

  2. Werthless said...

    Andrew, I think Paul Singman is just emphasizing that your high ceiling sleepers should ideally be at positions where you don’t have a high value guy, and vice versa. You were a little hung up on “reaching” for a draft pick at a position, but maybe you can look at it as “passing” at another position.

    For example, let’s say you’ve labelled Scott Sizemore as a guy that you think will give you produce top 10 2B numbers (and thus be startable in your format). Well, you may pass up Utley for Longoria at the end of the first round, because you don’t project any 3B with Sizemore’s ADP to be a “startable” player. So you’ll “reach” for Longoria, based on the production you expect to have available to you in the latter rounds.

    Or, if one has targeted players/sleepers that will produce a highly variable outcome (ie. they could be all-star quality or stuck in the minors… David Price last year, Jason Heyward this year), perhaps one should fill out their roster leaving a spot open for that sleeper to have a shot. So, don’t draft 4 solid but unspectacular OFs in the top 8 rounds.

    In sum, when deciding between a handful of comparable values in the early rounds, consider where you may find value in later rounds. Everyone knows this. You could demonstrate with numbers and ranges of outcomes, but my post is already too long.

  3. Jimbo said...

    Andrew, I’m not sure if bench spots are part of most leagues.

    It will be interesting to see which direction the “middle round drafting” series will go. This is the sort of thought process I hope gets discussed.

    This is the benefit of doing mock drafts. I have my share of mid/late round sleepers, and can tell in a mock draft which ones I regret most when I miss them. Where I don’t think Andrew’s logic applies to the article example is that I most certainly WOULD expect to ‘make up’ any value-reach in round 7!

    Top 10 rounds, you do have expectations. After that, who is to say an “11th rounder” is categorically better than a “13th rounder”? It starts to become more about perception, team needs, positional runs, and so on. So I have no problem taking my sleeper in round 16 even if his ADP is round 20. I think people get too hung up on ADP sometimes.

    As for “reaching” for Jose Lopez, can’t he provide 7th round value from 2B?? I don’t think that’s a stretch. Even if he provides 9th round value, it’s still a good pick imo. To me, there’s a difference between an ADP reach and a “value” reach. If I project McCutchen to be a top 50 5×5 player, that’s 4th round value with 8th round ADP. Drafting him in the 5th round might be a huge ADP reach, but not much of a value reach. Of course I want to take him as late as possible, but any round after #4/before #8 makes him both a value and a reach eh? 

    Thanks to Mock Draft Central, I also understand now how deceiving ADP can be from a targeting perspective. Matt Wieters and Brad Hawpe might have the same ADP, but Wieters will have a far larger range of picks contributing to his average. So saying Jose Lopez went 30 picks too early (if he’s the last 2B you like and could easily be gone by your next turn) doesn’t convince me it is a bad pick. Depends on a host of variables, as the beginning of the article mentions.

  4. Larry said...

    This is an intriguing article.  I commend Paul on bringing these concepts to the conscious surface, even if they may be generally known.  That said, the idea of influencing your early round picks based upon perceived later-round sleepers at certain positions is definitely risky, as Andrew P notes.  BUT, as any past fantasy champion knows, there is always a fair amount of risk necessary to win (at least when competing owners are reasonably competent).

    The Jose Lopez example does confuse me a little in how it is termed.  Is this really a “reach”?  This strikes me as more of a standard draft day decision requiring the drafter to weigh value (above replacement) against the likelihood of other teams drafting your targets in coming picks/rounds.  Given that in the example it states there is no other viable starting 2B in sight, Lopez’s relative value at this point seems quite high and is, thus, not really a “reach” in my view.

    That said, I would enjoy seeing more articles discussing mid-round dilemmas that arise.  It seems at every draft for at least one pick, in the 3rd-10th rounds, the decision between players at two (and maybe three) positions can seem like splitting hairs.  An article discussing these types of scenarios would be highly interesting.

    Also, is there data on which positions, year-in-year-out, have historically seen the greatest number of off-the-radar players emerge to be worthy fantasy starters?  I have always thought it more likely to be OFs and Pitchers (given that there are just MORE of them), but that is just theory.  This would give a sense of where one could place less emphasis in mid-rounds.

  5. David B said...

    I think the problem with the article is that it assumes that the sleeper will pan out. Even if you think Scott Sizemore is going to be a top 10 2B, I still think it would be better to take Utley over other players if you think Utley is the most valuable player on the board. That way if Sizemore pans out you have a valuable trading chip in Utley and if he doesn’t you still have an above average starter at 2B. If one passes on Utley to start Sizemore and he tanks (a la popular sleepers like Chris Davis and Stephen Drew last year) then you have to use a replacement level 2B or trade something of value to get a good 2B.

  6. Paul Singman said...

    Great discussion guys. Andrew, I think Werthless’ post was a sufficient response to yours. Don’t get too hung up on the reaching part, Jose Lopez was just an example and perhaps “reaching” was not the best word choice.

    I’m not sure where some of you are getting the idea this strategy is “risky” or “depends on the sleepers” panning out. I’m not sure you could even classify it as risky or not. I started off making the assumption that one or two hitters that will start for your team will be drafted in the late rounds of your draft. Maybe not the last round, but late, and in the article I called them “sleepers.” And from there I said in an ideal draft, the players you draft early will play different positions than the players you can draft later yet still feel comfortable giving a starting job to. And from there, for any position that you fail to get early yet don’t have a great sleeper for, drafting a guy from that position in the middle rounds (and possibly reaching on him) would make sense.

    Admittedly the term sleeper was also somewhat misleading, but by it I mean any player you draft late in the draft that you plan on giving a starting job to. In any draft there will be a few of these players so I was, in the article, simply pointing out the best way to structure your draft around these players. It’s not a question of whether this is risky or not, but whether you are aware of it during a draft.

  7. Paul Singman said...

    David, a player like Lopez that you draft in the middle rounds doesn’t have to “gain value” to make this work.  Just think about any normal draft. You’ll probably draft some hitters early, some hitters in the middle rounds, and some hitters late. All I’m saying is since you have the most control over the hitters you can draft late, assuming you will get those hitters, ideally the hitters you draft early will have different position eligibility than your sleepers. You are going to draft players with these picks regardless, I’m detailing the ideal way it will occur.

    If there is a risky part to this strategy, it’s that you might not get the hitters late you planned on getting. That’s why, though, you do not completely base your first few rounds off of this concept, and also why you have more than one sleeper.

  8. David B said...

    I think the reason this could be termed “risky” is because your assumption that someone like Jose Lopes gains value during the draft, but it seems to me that a player can only lose value not gain value. A player’s value predraft is his value over the replacement level player at any position for which he is eligible to start on your team. When a team drafts Utley Jose Lopes loses value to you because you have a better option than Lopez to start at 2B and now he can only start at Util (assuming the league does not have a MI slot). Now when the team drafts a Util hitter Lopez’ value decreases to virtually 0 as his numbers mean nothing to your team.
    For teams who have not drafted a 2B his value is still constant, as his projected numbers have not changed and are still the same percentage above the 2B one can draft in the last round or pick off of waivers. Therefore, shouldn’t a team always draft the player (at a position where he would start) that is most above the replacement player at that position?

  9. Jimbo said...

    Any team that dominates wire-to-wire, from their drafted roster, will likely have consistent players up and down the team. Probably need a superstar or three, probably need a couple late round picks to pan out. Other than that, the bulk of any team will be Jose Lopez-caliber players (give or take).

    Had one guy just crush the league offensively in 2007. He didn’t have a bunch of 30/30 guys…just a consisent level of production across the board. Pull any three guys at random and their stats wouldn’t scream domination.

    So quibbling about one or two rounds, in hindsight, isn’t what makes a draft imo. It is more about getting the best combination of reliability and potential—in that order—over the course of the entire draft.

    If Jose Lopez is the third best 2B available, you can wait. Next round he’s the last guy you see in tier 2, now his relative value spikes…even though his player-specific value is unchanged.

  10. Andrew P said...

    i personally don’t care if my higher-ceiling sleepers are at a position i’ve already filled or not.  i want any player i draft to have the best combination of potential and likelihood to fulfill that potential as possible. if i have an extra startable player at a given position, i’ll just trade from a surplus to fill my positional need.  it not only presents less risk (you don’t need to gamble that your sleeper will fill a hole), but you don’t immediately give up 30 picks in value on ADP. 

    i realize that trade markets aren’t perfectly efficient, but it still strikes me as much more efficient than reaching on a player 30 picks early that has little upside over his ADP, and gambling that your hole-plugging sleeper will be available when you are ready to draft him.  i also realize that a player’s value to your team is both dynamic and relative to the construction of your team, but if i’m giving up 3 rounds of value, it’s already part-way through the season, and i know i’m not getting a better deal.  again, in my opinion, giving up value in the draft is a losing proposition unless you have compelling reasons to think that you’ll still net a positive ROI on that mid-round reach (and the “chained” picks thereafter).

    Werthless said the following:
    “In sum, when deciding between a handful of comparable values in the early rounds, consider where you may find value in later rounds. Everyone knows this. You could demonstrate with numbers and ranges of outcomes, but my post is already too long.”

    my issue is that taking jose lopez 30 picks early isn’t deciding between “a handful of comparable values.”

    mockdraftcentral has jose lopez going at pick 118.  if you take lopez 30 picks early (pick 88), here are the 2 players being drafted ahead of lopez, and the 5 players you’d be drafting behind him:
    86)adam jones 87)josh beckett 89)billy butler 90) carlos beltran 91)andrew mccutchen 92) hunter pence 93)nate mcclouth

    that’s actually a pretty risky group, as far as these things go.  there’s no doubt in my mind, however, that’d i’d rather draft the appropriate value (in my mind) than draft jose lopez 30 picks early, because i’m afraid of the drop to the next tier.

    Jimbo said the following:
    -“Andrew, I’m not sure if bench spots are part of most leagues.”
    maybe my experience isn’t the norm, but every single fantasy baseball (actually any fantasy sport) league i’ve played in has had bench spots.
    -“This is the benefit of doing mock drafts. I have my share of mid/late round sleepers, and can tell in a mock draft which ones I regret most when I miss them. Where I don’t think Andrew’s logic applies to the article example is that I most certainly WOULD expect to ‘make up’ any value-reach in round 7!”
    then you have greater foresight than i do.  i certainly have players that i’d reach on and pass on relative to their ADP, but if i’m hitting on my sleepers to the point where the average sleeper i pick is worth 30+ picks in mid-round ADP, i’m ecstatic.
    it’s also a false choice to say that you’re willing to give up 30 picks in mid-round ADP if it nets you a sleeper.  first, why not take the guy that fits the ADP in the mid-rounds AND take your sleeper?  second, you don’t know that you’ll even be able to draft that sleeper that will net you such a positive ROI.
    -“As for ‘reaching’ for Jose Lopez, can’t he provide 7th round value from 2B?? I don’t think that’s a stretch. Even if he provides 9th round value, it’s still a good pick imo. To me, there’s a difference between an ADP reach and a “value” reach. If I project McCutchen to be a top 50 5×5 player, that’s 4th round value with 8th round ADP. Drafting him in the 5th round might be a huge ADP reach, but not much of a value reach. Of course I want to take him as late as possible, but any round after #4/before #8 makes him both a value and a reach eh? “
    judging by MDC’s ADPs, singman’s example would be in a 10-team draft, not a 12-team, so i think you might change your argument if that was a misconception.
    what you said about ADP and value is 100% true, but jose lopez is probably a poor example.  how likely do you think it is that lopez will net a positive return on pick 88 in the draft?  what is his ceiling and relative likelihood of reaching each of those positive returns?  what is lopez’s likelihood of a negative return on his draft position?  what probabilities do you ascribe to each of the negative outcomes?  in my opinion, lopez is much more likely to produce a negative return on that draft position. 

    i realize i may be getting too hung up on lopez as an example, but i’m not sure what else to base my response on.  ADP already accounts for position scarcity, and drafting a player above that ADP based upon position scarcity during the draft has to be justified on a case-by-case basis in which their are inefficiencies not accounted for in ADP. 

    the lopez example, in my opinion, is only relevant if you think you can find a “sleeper” that can make up that 30-pick value disparity.  it further assumes that a player’s trade value has less utility than the combined net result accrued by taking both lopez and the sleeper pick.

  11. digglahhh said...

    Good discussion. Sorry I’m commenting a bit late.

    Andrew P.,

    I , as well, think you may be a bit hung up on the specifc Lopez example.
    First, let me just say that the strategy I generally preach in regard to late round picks is to swing for the fences and hope you connect a few times and then replace those who you missed on with replacement level FAs. This means that in the mid rounds I want solid base line production and to maximize my value. Therefore, I do not want to be reaching to fill a positional need as early as round 7-9. I’m usually still in pick the best player on the board mode.

    Even with that in mind, there are two points I think worth mentioning.

    First, and this was mentioned before, a player’s relative value does increase when he represents the last player at his position at a certain tier (…to the extent that the tiers prove to be correct evalutions of relative value). Really this is just a refraction of classic supply/demand. We often break positional depth down to determine scarcity, but really we could break each positons down further, so we wind up with more thinly sliced groups 2B(1), 2B(2), etc. As the players in each group come off the board, each remaining player within the tier sees his value increase because there are fewer players available to replicate his production. Supply your own analogy regarding trying to hook up at a party.

    Of course, the one thing to keep in mind is that many teams aren’t going to draft extra middle infielders, so once they fill up their starters at each position, the leftovers are likely to sit well beyond where their value dictates they should, because the are just surplus and their raw production doesn’t match that of the leftovers at other positions. So, you have to monitor the demand part of the equation as well as the supply.

    Second, and more to the Lopez point, once we get to the 90 – 120 rage of ADP, we are starting to flirt with the point where it just doesn’t mean all that much, except as a gauge of what OTHERS MIGHT do (as opposed to what YOU SHOULD do). I’m sure you can find a player in the 105-120 range that you would be fine taking 25 picks before his ADP, as well as a player in the 90-105 range who you wouldn’t want any part of 25 picks after his.
    Also, take the Matt Wieters/Brad Hawpe example from above. They may have a very similar ADP, but those picks are made for totally different reasons and reflect very different levels of certainty. With Hawpe that number reflects the general consensus of his worth. With Weiters, it reflects the average of a wide range of expectations.

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>