Sunday, February 2, 2014

Thoughts on Drafting Part II

Another favorite pastime of prospect watchers and draftniks is looking to see what players got taken after your favorite team's pick that you were not so crazy about.  I admit I do it too!  It can be both fun and re-affirming that maybe you might be up to being a GM too, especially if the successful player taken later is one you had your eye on before the draft.  It is one way to look at the draft and its OK, as long as it is not the ONLY way you look at it.

The flaw in second guessing type analysis is that you are looking only at the poor picks your favorite team made and only the good picks another team made.  If you ask yourself why the Giants took Gary Brown and Jarrett Parker in the first two rounds of the 2010 draft when they could have had Joc Pederson who the Dodgers took in the 11'th round, and why the Giants took Gary Brown at #24 overall when they could have taken Taijuan Walker who was taken by Seattle at somewhere around #46, well, pretty soon you will have yourself believing that the Giants are the worst drafting team in all of baseball!

There are basically two ways of balancing the analysis out:  The back-at-ya, schoolyard, na-na na-na na na!, and the mathematical/statistical breakdown.  In back-at-ya, you counter with an awesome factoid that ogc turned me onto recently:  There are 4 MLB organizations not named the Giants who could have drafted all 3 of Tim Lincecum, Madison Bumgarner and Buster Posey!  That's right, 4!  Do you think maybe THOSE TEAMS fans are doing some second guessing and saying how much smarter Brian Sabean is than their GM?  OK, maybe not, but they SHOULD be!  What about the teams who could have had Brandon Belt about 50 picks higher than where the Giants took him in the 5'th round?  What about Brandon Crawford in the 4'th round?  What about Brian Wilson and Sergio Romo taken in rounds 30-something?  See, the game cuts both ways and the Giants have some pretty serious bragging rights of their own!  What about Matt Cain after pretty much every draft analyst at the time called him a reach?

From a mathematical viewpoint, I will again credit ogc for pointing out numerous times that once you get past the top 10-15 picks in the draft, the success rate drops off dramatically.  By the time you get to the supplemental first round, the odds are barely a 1 in 10 success rate.  Yes, the odds continue to drop off after that, but you still find scattered success stories all the way down into the 30 + rounds.  Since the odds of success in the supplemental and second round are so low, and the odds of success in the late rounds not that much lower, it is highly probable that you will find failures in the early rounds to compare to successes in the 38 later rounds.  When you multiply the total number of draft picks after round 2 by 30 teams, 38X30 you have 1140 chances to find a later pick that is better than your team's second round pic, or even late first round pick.  In other words, if you go looking with that mindset, you are virtually guaranteed to successfully uncover a later pick that makes your GM(and favorite whipping boy) look very bad.

Again, the MLB amateur draft lasts 40 rounds.  As I pointed out in  Part I, an organization that produces 1 major contributor per draft is doing very well.  That means that 39 out of 40 draftees in a given year fail to "make it" in the show.  With those kinds of odds and the rate of MLB job openings, an organization that can add just one more "success" every other year greatly increases their competitive advantage.  It's a whole lot of very hard work and a lot of money spent for an incremental improvement in your odds that has a huge payoff if it actually pays off.  I believe there is ample evidence that the Giants are one team that has bent the curve in their favor.  The Cardinals are another.  I can't really think of any other organizations that have done it as well.  I can think of at least 4 organizations who have done it a whole lot worse!


  1. The Red Sox and Rangers seem to do a great job of winning while maintaining a pipeline of minor league talent. The Rays have their own secret formula.

    You mean people called Matt Cain a reach? Ridiculous. Was there any one or two players in last years draft that you thought the Giants should have taken instead of Arroyo?

    Back to Joc Pederson, the numbers as a 21 year old in AA are impressive: .278/.381/.497, 22 HRs, 31 SBs. A lot of teams missed out on him. Who knows how he fits in the Dodgers org? Possibly trade bait.

    1. The Red Sox had a down cycle when they emphasized high-ceiling, high risk HS talent too much. The Rangers have grossly overspent in the international market for a long time. We'll see how they do now that there are limits and penalties in place. The Rays had a long run of top 5 draft picks and hit on Longoria and Price. Beyond that, I'm not sure what you see there. They are in the midst of a long dry spell in terms of draft picks.

      Yes, Matt Cain was widely considered a reach when he was drafted. I'm pretty happy with Arroyo so far, although it seems like the Giants are accumulating a huge surplus of second basemen disguised as shortstops. He wasn't on my radar until late in the predraft process, but once his name came up, it seemed like such a typical Giants pick, I went on record as saying I thought that there was some "fire behind the smoke." I will have to look at the draft list from last year and see who I thought they might go for before it became clear they were going after Arroyo.

    2. Giants took Arroyo at #25 overall. I would have been very happy with Philip Ervin, OF who was taken at #27 and with Justin Williams who was grabbed by Arizona at #52 overall. But I'm not complaining about Arroyo, at least so far.

    3. Yeah, Cain was considered a reach, kind of on par with how people thought Joe Panik was a reach: both were considered supplemental first round material, back 20-30 picks later, than where they were picked by the Giants.

      Not only that, but for years, it seemed to me, every year the Dodgers had a better pitching prospect than Cain, or so said all the prospect experts were pronouncing it. Yet they all fell to the wayside, year by year, until Cain was left standing with the Giants all these years, almost all the Dodger's so-called better pitching prospects (got Logan White a LOT of press about his genius for keeping their farm system stocked) ended up either elsewhere (Jackson) or injured (numerous) with only Billingsley the only one to make and stay with the Dodgers, and even he missed a season with, I believe, TJS. Cain stands alone compared to all those Dodger prospects.

      The Rays secret formula seems to be this: losing big time. As DrB noted, their picks haven't been as good lately, and I would note, basically once they started winning: which is one of my points that I made in the draft study that DrB was referring to that I had done previously. When you lose big time, it is much easier to look like a draft genius (though as the Pirates, Royals, Orioles, and even the Rays pre-Friedman showed, it is no slam dunk, you need some scouting skills too). Once you start reaching .500, your odds drop by half, and by the time you compete for the playoffs, drop half again, of finding that great franchise starter that most draftniks are looking for from the draft. So with that one in ten odds, if a team is playoff competitive over a 10 year period, on average they will find one good player with their first round pick in that ten year period. That's why most playoff teams fizzle out at some point and need to rebuild.

      Unfortunately, most draft studies have been doing it wrong, starting with BP's study that come out a while after mine went public (then most have copied them since). They look at the average WAR by pick, not realizing the importance of the distribution. I recall looking at the average WAR that came out of BP's study for the first pick and they didn't realize that basically on average, the first pick of the draft was somewhere between Michael Tucker and Marquis Grissom (I looked up the two of them because it was around then when they were Giants and their WAR fit those averages). I don't know about you, but for my first pick, I don't want to be picking up no Tucker or Grissom, if I get them, while it is a nice useful pick (and better than getting someone who never makes the majors), I would consider that a lost opportunity to find a good to great starting player. You want a Buster Posey, not a Jason Grilli, not a Tucker or Grissom.

      Just think of it this way, tell me who else the Dodgers drafted before they got Joc Pederson, and whether they would still pick them over Pederson today. Once you get out that far, the odds are just very against you finding good prospects. Basically the population of prospects selected out that far who become good players is very few, so there is a lot of luck involved with selecting any one of them. We can't tell who was luckier, the Dodgers finding Pederson that far back or the Giants finding Belt or Blackburn that far back.

    4. I like DrB's method of looking for one success every couple of years as a good way to judge success. I'm not sure that is the litmus test, but it feels about right for a starting point for comparison of teams (and to Shankbone's point that he has made in his posts on his blog about the need to place success in finding players via player development firmly into the mix for comparing farm systems) and in terms of replenishing a team.

      For the Giants model, basically there's three good starting pitchers, one closer, and three good hitting position players, which is seven players. That means roughly over a 14 year period, then you start over again. If you have good players, they should be lasting roughly about that long, if you hold onto them, though that seems a little short on talent, may be more like three every five years, making replenishment roughly an 11 year cycle.

      But that's a quibble, every two years for the Giants with the core that they have now, I think would be sufficient to keep us competitive into the 2020 decade.

  2. Dr. B: All salient analysis.
    I'm interested in the forum's thoughts to the following questions:
    1) Do successful teams seem to pick right (find the gem in the draft), or develop better?
    2) How do you view the acumen of Giants coaching as it applies to developing or improving the existing talent of pitchers vs. position players?

    1. 1. About 70% drafting the right player, 30% development.

      2. I think the Giants acumen for developing pitchers is overrated and for developing hitters underrated. I think they have been very good at drafting great pitchers and leaving them alone. I would point to Brandon Belt as an example of a hitter who they completely reworked his batting stance and grip. He has consistently improved after receiving and applying batting instruction.

    2. 1. I like DrB's ratio. Talent is so so important, that's why I think it's important to tank seasons as soon as you realize that you are not playoff likely. That way you can end up with at least a Top 10 pick and ideally a Top 5 pick, if not Top 3. That's why the Giants sucked in the 70's and 80's, they always fought their way up to mediocrity, which got them lousy picks in the middle of the first round. That is why we sucked for so long (plus, when we picked high schooler Barry Bonds with our first round pick, our GM didn't sign him).

      When you are picking in the Top 3-5, it's no guarantee, but you are fishing in a barrel for the fish you want. When you are picking in the middle of the first round, you are already fishing in a pond for the fish everyone wants, and by the end of the first round, you are already fishing in a lake. And essentially after the first round, you are finishing in the ocean, with no guidance of any sort, good luck.

      The odds are so bad that I can't say definitively which teams are right and/or develop better. From observation, I see that the Cards have been pretty good at adding new people and continuing the good times, Atlanta Braves too. Red Sox, I could go either way, but would note that if LA didn't take on all their bad contracts, they most likely would not have won last season. Then it would be almost a decade since their first one.

      The Yankees are still benefiting from the players Sabean helped them select when he was their scouting director, in the early 90's when Steinbrenner was forced out of baseball for some illegal political payments, and the baseball people could operate unhindered. Under Sabean, the Yankees acquired Jeter (much like Posey, he fell into their hands and they grabbed immediately), Pettitte, Rivera, Posada (he also got Snow for them too, but they didn't keep him), basically the core that won them all their 1990's titles, then as the poor draft position killed their farm, they had to keep on buying more and more players to try to keep things going. You can buy your title every once in a while (like Arizona did that one year, basically bankrupted them, but as Shankbone likes to remind, flags fly forever, or Yankees last one) but you really can't do that reliably, not enough talent in the free agent market to do that consistently, especially today when teams sign their best players long-term for the most part (except for Boras' clients; that should depress their market value in the draft eventually).

    3. 2. I generally agree with DrB that they are overrated, but (and I think he'll agree) I think they are good at developing pitchers, just not the miracle maker that some make them out to be. It is just that when you spend all your best draft bullets on picking pitchers, you are going to find a lot more good pitchers than good position players. I would also add that it helps that they appear to be sabermetrically ahead of the curve among MLB teams in defensive metrics (were happy and eager to see Field F/X data, AT&T got the pilot test), I read about how they position players based on the hitters and how their pitchers pitch to them (which will help Lincecum in 2014 now that he's on board), so that helps their pitchers look better. And until last season, they had usually been among the leaders in defensive runs saved, as another indicator of how their good defense helps them have good pitching. Plus a Fangraph study found that somehow Giants pitching staffs, year to year, are able to suppress the number of homers they give up, significantly below the 10% HR/FB ratio that saber studies have determined that "all" pitchers should regress to. All adds up to "good to great pitching".

      I would qualify the leaving pitchers alone comment. With Lincecum, they left him alone. They tried to modify Bumgarner, but soon just let him return to his regular mechanics and he dominated since. Wheeler they tried hard to modify, and frankly his stats with us wasn't that great, but once the Mets got him, they let him go back to his old ways and that's what appears to be the improvement from us to them for him. I recall Alfredo Simon, we got him from the Phillies, was good with them, but once we got him, they tried to change him and he was totally messed up. He left our system and eventually did well, I think, for Orioles. So the Giants do leave some pitchers alone, but others they work on changing, and others they try but then give up. I have no idea what the ratio might be, so I accept DrB's expert opinion that they generally leave pitcher alone.

      For hitters, I would submit a THT study that looked into how hitters improve or get worse when they switch teams and join a certain manager. What that study found is that hitters joining Bochy-managed teams tended to hit better once they join him, on average adding one win to a Bochy-led team. Belt is the prize student of their video training system that they talked about a few years ago, but I haven't heard of since. I know that Noonan got some training through that and said that he started hitting line drives better after that, but as you see with his MLB stats, it is not a miracle worker like it was with Belt, and even Belt took time to acclimate. Plus he resisted changing over a 2-3 season period, until finally adjusting his stance, moving back in the box to give him more time, and changing his grip. While Belt is only one example, given that we know bits and pieces of what happened, it sounds like the Giants know what works and what doesn't, but sometimes the student doesn't listen, and other times, well, you can't make silk purse out of a sow's ear, as the old saying goes.

      And as we all know, the Giants had spent most of their first round picks on pitching over the Sabean era, putting more pitching talent into their farm system, until recently when they started picking more position players. If it's around finding one good player in ten years for back of first round picks, it is probably around finding one good player in 20-30 years for second round picks. Then it gets really gets bad... :^)

      It really is a volume business, the odds aren't great, but you need to play the player development game if you hope to be good enough to win a championship.

    4. We also may have to consider the year-to-year element of drafting.

      For example - If Tidrow or Sabean or Bar see a set of great hitting/position prospects that they believe they can obtain in the 2 - 5 rounds (Belt, Susac, Crawford), then they may opt to grab their favorite pitching pick with the 1st rounder, in order to maintain balance in the system. It's clear to me that, while you always hear "draft for BPA, not for organizational need," the management understands that you want to balance your statistical chance of scoring both hitters and pitchers from the draft, and you maximize this by intentionally using your picks on both position and pitching prospects, and not just 1. Making trades based on organizational surplus decreases the value, because you're in a position of "need."

      However, this is just a thought. Perhaps I should be looking back into team's draft strategies over the years and see if anyone actually is taking ALL hitters or ALL pitchers with their 1-10 round picks. It probably happens, but I'd count it less likely. Also, of course, BPA will naturally flip flop from pitcher to position player on it's own, making the balance in drafting happen fairly naturally.

    5. Historically, even though baseball rosters are heavy on hitters vs. pitchers (13 vs. 12 for Giants in recent years, but at the start of Sabean's tenure, I think it was 14 vs. 11), the Giants draft has always (from what I recall) had more pitchers than hitters, I can't remember one year that it wasn't. That suggests that the Giants do value pitching over hitting in general, when drafting.

      And I do agree that there are times when organizational needs trump BPA, but I would add that once you get beyond the top 30-50 prospects of any draft, the BPA is probably so close to a lot of other players that a team could ideally pick BPA as well as for organizational need as well. That's when they start going for the extra pitchers, I'm guessing.

      And I agree that there is a year-to-year element of drafting. Each year has a different level of tiering of talent, different proportions of positional value, different proportions of HS vs. College talent, so the mix will always be different. That's why teams need to focus on BPA, but factor in organizational needs as well when the opportunity presents itself. If you got two prospects rated to be pretty much equal, then you may as well factor in org needs. But for PR purposes, it is a lot better to praise the prospect than say, well, we needed a catcher.

      I also agree that making trades based on organizational surplus decreases value, but not just because you are in a position of need (because ultimately, usually the other team is also in a position of need). As many recent studies have shown trades are generally made with players who are lacking in some way to the team currently owning them. Teams know what they got, and generally they keep the prospects and players that they think are the keepers and trade away the ones they don't think of as being as good. Same with free agents they keep vs. ones they let go of. But if you have a surplus, you are then forced to trade good value for lower value, just because that is how the process works, generally, so you lose value in that way.

      That is why I've been writing about the Giants genius of focusing on pitching for over ten years now. The flexibility that gives your team is enormous. If you have three great 1B, you basically have to trade off two of them and hopefully you made the right decision on who to keep. Texas did, keeping Teixiera while trading away Hofner and A-Gon. But the players they got back were not as good as the players they gave up, for example. However, if you have three great starting pitchers, you now have a great rotation, with two more slots you can fit in any new finds. Or some fail, like Isringhausen, or get injured, like Wood, and then they become great closers or relievers. And we got 7 spots to fill there before you have surplus. No need to trade off surplus, unless, that is, you don't think that surplus is worth keeping, for whatever reason.

  3. I think Sabean is better than at least half of all the GMs.

    More than fans of 4 teams should be jealous.

    I agree with most here that maybe fans of 20 or 25 teams, maybe even more, should be envious, even if he is not perfect, as no one is perfect, even Hall of Fame GMs.

    As for later round gems, it would be rare if every team doesn't have at least one, if not every year, at least often enough. I would say, that, it is more than just you get to roll the dice more often, but has also to do with player commitment, injury risk, etc.

    In the NFL, there are only 6 or 7 rounds, not 40. You can get All Pros talent/potetial (not just 1 round talent, not just making the team) in the 2 or 3 round or later for all kinds of reasons - discipline, injury recovery, etc (but not so much high school kid player commitment). Just look at Lattimore and Tank of the 49ers this past draft (hopefully they become All-Pros. with their All-Pro talent or potential).

    Not that football is like baseball, but late round gems come from 1. more chances and 2. other factors. For example, if, hypothetically, they rule out all injured players (like Wilson with TJ, for example) untill they have shown complete recovery, and high school kids with college offers (they will go much higher next year or when they don't have that commitment issue anymore), we will likely to have fewer late round gems, even if we keep 40 rounds of picks.