Uniqueness of MLB Draft
However, the MLB amateur draft is not like any other professional sports' draft. Baseball teams cannot trade their draft picks to anyone: they are forced to draft someone. And that player usually has a dollar figure that they want in order to sign with the team, a figure that is usually driven by what the guy who was drafted in his spot got last year and by what the guys drafted before and after him got. In other words, it has nothing to do with how good he is as a baseball player relative to past drafts, which is just like what happens in all the other sports' drafts too.
Which just doesn't make sense. For example, in basketball, if Walt Chamberlain, Kareem Abdul Jabbar, or Shaq is drafted number 1, there's no way in hell the #1 guy next year should get a contract close to what they got, they were once in a generation type of players. For Warrior fans, we saw that in action when we got Joe Smith for our #1 pick. Similarly for baseball, there is no reason to believe that the quality of draftees is the same each year at the same pick. But the MLB teams are forced to pay approximately that same amount for players whose quality fluctuate year by year or be forced to draft for "signability" because draft picks cannot be traded to another team before or after the draft.
That gets us to another difference in the MLB draft: virtually all MLB draftees do NOT join the team as a contributing player, they all start out in the minors and some first rounders never even make the majors. While the football and basketball draftees are expected to contribute to the team, significantly changing the fortunes of the team for the top 5 picks, it is so rare for players to go straight from the draft to playing in the majors that they occur about once in a generation. And yet they all are paid millions of dollars up front for their potential which may never be reached and may take years to blossom, if ever.
Method to My Madness
In my first article, "Sabean's Draft Record: Better Than Fans Think," I examined how the Giants did compared to other teams who were forced by their success to draft late in the first round and were known for their ability to draft well. These teams included the New York Yankees, the Atlanta Braves, and the Oakland A's. What I found was that the Giants had not been unsuccessful relative to these teams and arguably was better when drafting late in the first round. However, I only covered Sabean's term as GM of the Giants and from that meager data set, it seemed obvious that it takes at least four years for draft picks to "ripen on the vine" and start to contribute at the major league level.
Consequently, for this article, I greatly expanded my data set. I pulled from the Baseball Cube – great site for players' minor league data – the first 100 draft picks for each round from 1986 to 2003. That is the period from Sabean's first year as director of scouting for the New York Yankees to the latest draft information at the time that I collected the data in mid-season 2004. This gave me a very good set of data to work from.
Now I got a life, so I couldn't manually check each and ever major leaguer for comparison so I took a shortcut that those more sabermetrically minded might object to. What I did was use the Baseball Cube's provided stats for each draftee who made the major leagues – games played plus either batting average or ERA – and used that as a guide to classifying each player. I then tweaked from my own personal knowledge because batting average is not the best for measuring who is the better player.
Also, there is a problem that as players age, they might fall back down a level, especially as they reach their 30's, but I did not take that into account. I did, however, keep the same standard across most of the years, though I did tweak for recent draftees who wouldn't have the years necessary to qualify for the top categories, basically those drafts under 5 seasons ago. Lastly, I did not capture the college vs. high school distinction, which undoubtedly affects the rate of contribution on a per year basis, though apparently it evens out at around 5-6 years.
I classified each player into five categories though, upon reflection, I probably could have done OK with three. The five are the following: Star; Good; Useful; Marginal; and the Rest. For Stars, he had to hit over .295 or had an ERA under 3.30 and played at least 6 seasons, unless I recognized him as a star despite poorer stats then I would check his full career stats to see if his other stats look like a star. Yes, subjective but otherwise data collection would be a bear and, I believe, not worth the extra margin of detail. A good player had to hit from .275 to .295 or had an ERA under 4.25 plus over 6 seasons played. Also, players who had under 6 seasons played and had these stats plus who were among the top of the 100 draftees in games played, would count as a good player. To me, a good player not only has good stats but longevity and healthiness as well.
A Useful player is just that: useful. Useful enough for a team to keep him around for more than 3 seasons worth of games, meaning that a team either was willing to go through arbitration with him or to use him instead of a prospect. A Marginal player had played 0.5 to 3 seasons with stats that are not good (as above). And the Rest played less than half a season's worth of games in the MLB, including those who never made the majors.
Round by Round, Year by Year
First, I examined round by round. Obviously, teams did better in the first round than the second or third. In fact, so much better that I initially stopped there in my analysis. It did not seem like I had to go much further to see that lack of success in picking players started very early in the first round. But then I realized I should expand, first to 75, then to 100, to examine these picks more deeply. More on this later. For the 18 year period, 4% of Round 1 picks were Stars, 8% Good, 17% Useful. In contrast, 1% of Round 2 were Stars, 2% Good, 9% Useful.
By examining each draft year by year there is clearly three groupings of drafts. For the first four years after a draft, there are a handful of players making strong contributions. Then the number of players contributing something ramps upward in the fifth season and next the marginal return of draft picks appear to approach zero, meaning most prospects have a Star or Good performance within 4-5 seasons. This confirms the general impression I got in my previous article that used less data. Then from 6 to 10 years, the useful players start percolating upward, and sometime around that period, the number of Star, Good, and Useful players seems to have reached a steady point. However, as I noted, some may fall back but I did not capture those cases, these were all from a snapshot of time ending with the 2003 season.
Pick by Pick, Grouping Methodology
Because of the severe lack of performance from years 1 to 5, I removed those data points and next examined the success rate pick by pick for the years 1986 to 1998. First, I grouped the picks by 10 to ease data analysis, though I did examine picks 1-5 and 6-10, as there appeared to be a difference there when each pick was individually examined. Then I tried to group each 10 into groups of 10, to try to determine approximately where in the drafting the chances of success appears to have shifted downward.
Pick by Pick, Grouped by 10's
Because the data being used for categorizing the players, particularly the batters, is only a rough guide, the grouping of Stars, Good, and Useful players is the most accurate data set to be examined, but Star and Good is probably the most interesting to people. So, I will also discuss Star and Star/Good groupings as well as all three because they are interesting groupings.
Star: The data grouped into these sets: 1-10, 11-20, 21-60, 60-100. The percentage of picks in each set, respectively, are: 11.5%, 6.2%, 1.9%, and 0.4%.
Star/Good: The data grouped into these sets: 1-5, 6-20, 21-30, 31-90, 91-100. The percentage of picks in each set, respectively, are: 43.1%, 18.5%, 10.8%, 4.0%, and 1.5%.
Star/Good/Useful: The data grouped into these sets: 1-5, 6-20, 21-30, 31-50, 51-90, 91-100. The percentage of picks in each set, respectively, are: 73.8%, 42.1%, 31.5%, 21.9%, 12.9%, 6.9%
I broke out these in groups by examining each set of 10's standard error and determining whether, at 95% confidence level, whether that set of 10 is part of the population above or the start of a new grouping.
As one can see, the odds of picking a star start poorly and goes south very quickly. Even for teams picking in the first 10 picks, there is about a 1 in 9 chance of picking a Star. Given that these top picks get anywhere from $3-5 million in bonus and contract, a regularly lousy team could end up spending $30 million just to get that one star in 9 years of drafting. If you are a perennially good team, drafting in the 21-30 pick range, you are really screwed, as you pick a star about once every 25 years. That, coincidentally, works out to about $30-35 million spent on bonuses and contracts per star player.
It gets a lot better fishing for Stars and Good players in the draft. If you have a top 5 pick, you have about a 50% chance of picking at least a good player, if not a star. This grouping is better than just using Stars because of the unreliability of the cut off line between a Star and a Good player due to the use of batting average as the comparative stat. But if you are a good team, picking in the 21-30 slot, you are still looking at poor odds still, about 1 in 9, of picking a good or better player. At approximately $1.5 million bonus per pick you are looking at spending $13.5 million to find your Star/Good player. Again, coincidentally, that is approximately what a team picking in the top 5 would pay to get their Star/Good player, about $10 million (approximately $5 million per pick).
Almost like shooting fish out of the water now. A team has a 3 in 4 chance of getting at least a useful player in the first 5 picks, and still 1 in 3 chance if picking in 21-30. However, a useful player is not particularly valuable to a team. A nice addition, sure, but not one to push the team to greater or sustained heights. For this player, the Top 5 picker can expect to spend $7 million to get this player while the 21-30 picker ends up paying an expected $5 million.
Why Kiss a Lot of Frogs, Go Sign the Prince
Teams amazingly pay bonuses very closely related to the expected payoff of picking a Star, Good, or a Useful Player. Teams pay approximately $30-35 million in order to get a star, about $10-15 million in order to draft at least a good player, and $5-7 million in order to draft at least a useful player. Thus, in essence, teams ante up their bonus money, like a Christmas savings account, and over time pay about the same amount of money to get at least that type of player, though obviously your mileage will vary. However, unlike a Christmas account, a team may find after spending their $30M that all they have is a lump of coal in their stocking.
The impetus of these articles was the criticism of Sabean's practice of signing free agents before the deadline and giving up at least a first round draft pick. Many Giants fans have criticized this action, as well as some of the media. I was the only Giants fan I could find who believed in the logic of this, based on what I saw on the Giants community board on Yahoo. Thus I researched this to see if my logic was correct or not.
To me, this move makes sense because of Sabean's "peculiar" drafts. He typically drafts players who draw the following reactions from the experts and media commentators: "Who?"; "Drafted higher than thought."; "Could have waited to draft him"; "Why did they draft him?" Now if this is what regularly happens, then why not give up the first pick because then you would pick the same player for less money with the second (or later) round pick. You end up with the same players picked overall but then pay less money, money you can use on free agents. That was the basic logic that got me on board with this tactic of losing your first round draft pick.
But then, given the data I collected, his tactic of losing first round draft picks makes even more sense. Would you rather ante in over time $30 million in bonuses to get your Star (with no guarantee you would ever get one) or would you rather DIRECTLY pay your free agent star? Would you rather ante in $10-15 million in bonuses to get your Good player (with no guarantee you would ever get one) or just pay him?
Remember, the $30 million is not even paid to your star, that goes to your bonus babies, you still need to pay him his star salary. And, as noted, it is only expected, based on the odds, to pay $30 million, there is no guarantee. If you are a lousy picker or just plain unlucky, you could have spend that $30 million in bonuses and still end up with no Star player. Some lottery ticket, eh?
Michael Tucker Case Study
Taking Michael Tucker as an example, he is the classic useful player. Would you rather buy a $1.5 million lottery ticket (i.e. pay your first round pick in the 21-30 range) in hopes that you'll get a star (about 1 in 20 chance) plus have to wait up to 5 years for him to breakout? Or, would you rather spend it on a already proven useful player like Tucker? In total, it could be up to 5 seasons before you actually pick a Useful player, then another 5 to 10 years before he is actually performing like a useful player in the major leagues. Tucker contributed immediately.
And that's just for a useful player. A good or star player would take many more seasons before contributing to the team, on average. Drafting that far back in the first round, it could take an average of 9 drafts before you find a good player and an average of 25 drafts before you find a star player, then the up to 5 seasons to contribute at that level. Plus that's for Round 1 picks, obviously the odds are horribly lower as you move into the latter rounds. It is down to at most 1% by the 100th pick to get a Star/Good player.
Sabean's Astute Move
Sabean's tactic of actively choosing to give up draft picks makes sense on a number of levels. Most importantly of all, it makes sense on a money level. Why spend all that money in hopes of getting a player who may not contribute to the team until up to 10 years later for just a useful player? It also makes sense from a drafting perspective. Because Sabean normally drafts players that other teams overlook, he might lose the pick but he would still most likely pick the same player with a later pick. Thus the net effect could be that he Giants will end up signing the same players but spend many millions less than they would have by following this tactic of signing free agents that cost you a draft pick.
Another factor would be the overall abundance (or lack thereof) of draftable baseball talent. If Sabean sees a dry year for prospects, giving up draft picks would not only save money but it would save time from having to deal with players who most probably will never make the big club. Yet these top picks are considered "high" draft picks with the attention that gets (and are called failures if they don't advance). Lastly, it makes sense from a personnel level. Why pay players who may never spend one minute on the major league roster a million dollars when you can get a reasonably productive player - NOW - for the same amount?
For an example of surprise picks, the Nate Schierholtz pick in the 3rd round of 2003 caused a lot of head scratching on draft day but after he tore up every league he was in that season, people probably wondered why he wasn't drafted higher. He probably could have been still drafted much later than the 3rd round without fear of being picked since he was such a stealth pick, based on the impression I got from draft day commentary. And that would have saved even more money since later rounds get lower bonuses. But because baseball teams cannot trade a draft pick spot, they are forced to play the bonus baby draft game and pay their draftee more than they would have had they drafted him later, or worse, draft someone without the intention of signing the draftee.
Thus Sabean's tactic of losing his first round draft pick is brilliant. He diverts player development budget dollars into the 25 man payroll budget by signing a free agent. He does this by paying less bonus money in total to his draftees and probably drafted 95-99% of the same players they would have had they drafted in the first round. By doing so, he gets an immediate contribution to the major league team without significantly hurting the farm system's chances of developing young major leaguers versus waiting for years for a probably minor contribution.
In my last article, I will examine Brian Sabean's draft history, starting with his days as head of the Yankee's scouting department, and covering his years as the Giants GM. I will also discuss my thoughts on Sabean's apparent strategy for drafting and on the future of the Giants farm system.
Martin Lee writes 'A Biased Giant's Fanatic's View' for SFDugout.com when the mood and muse strikes him. He wants to teach and share his love of baseball and, in particular, his love for the San Francisco Giants. He will believe to his dying days that Bobby Bonds was robbed of being the first 40-40 player and should be in Cooperstown for bringing the combination of power and speed to the game.Please feel free to comment on his blog, http://biasedgiantsfanatic.blogspot.com/, if you have a question or comment on this article.
The views expressed in the columns do not necessarily reflect the opinions of the site's publisher, writers, or other staff members. The content on this site may not be redistributed without the expressed consent of SFDugout.com.