DePodesta Blog on the Recent Draft

Amazingly and wonderfully, Paul DePodesta, now with the 'Dres, has a blog called "It Might be Dangerous... You Go First" and he recently blogged on the draft. Here are some interesting bits, plus some comments from me:

• "...a more detailed analysis can greatly inform a team's strategy for a particular draft. As I've mentioned before, due to the fact that baseball's draft does not immediately impact the Major League level, teams don't necessarily have to draft for need. This allows for more flexibility in the process on an annual basis." This makes a lot of sense. Still, when you draft, you still need to field a team in the lower levels, so at some point you do draft by need. But the main point is that when the probability of the prospect drafted is still visible without a microscope, you should draft best player available. So if Posey was gone and Matusz was available, Sabean probably would have pissed off a lot of fans here by drafting Matusz. I'm sure there will be a lot of Giants fans following what Smoak does in his career: a couple of years ago, Baseball America did an early (end of calendar year) rank for the draft next year, and the #10 prospect was Evan Longoria, and I've been following him ever since, even though we had no chance to draft him, and, I would say, did pretty well for ourselves by selecting Lincecum.
• "This year our scouting staff determined that the 2008 draft class was particularly deep in high quality college hitters, thin in college pitching, and a bit disappointing in terms of depth at the high school level. In fact, there seemed to be an unusually high number of high school players who weren't terribly interested in signing (demanding top half of the first round money in order to forego their college experience)." That would explain the Giants drafting so many college hitters, then going for college pitching (where they hopefully have some sort of competitive advantage in identifying talent), and didn't select high schoolers until later, which was quite a contrast to last year's 5 of 6 high school draftees up front.
• "Approximately five days before the draft we started having conversations about the reality of the draft class and began formulating our strategy. What if we didn't take a pitcher the entire first day? Given the class, we decided, we were prepared to do just that. We weren't going to take a pitcher just to seemingly balance our draft. There were pitchers were liked, of course, and we would be prepared to take them. However, there was an unusual number of attractive hitters, so we wanted to take advantage of the opportunity and land as many of them as possible. We certainly didn't want to pass on some of these hitters, who in other years may be at the top of the class, in order to take a pitcher who was very similar to pitchers available in the following few rounds (or in any other year). As I've stated many times in this blog, we know we're not going to be right every time, but if we could secure a full group of top hitters, not just one or two, we felt that at least some of them will meet or exceed our expectations." That's a good thing for fans to remember, the quality and depth of each year's draft is not the same. Sometimes one year's 3rd round pick is better in quality than last year's 2nd round pick. My draft study showed a lot of variability with each draft, but I didn't make a point of it, as I don't have that stats ability to prove that, but it is nice to see an outsider confirm what I noticed in my study. And you can take what he said at the end, change it to pitchers instead of hitters, and that has been basically the Giants draft philosophy for many years now.
• "One of the biggest complaints about most drafts is that certain players were "overdrafted". I'll be honest - I don't really believe in that concept. First of all, our knowledge in terms of where players will be selected is imperfect to put it mildly. Remember, it only takes one team out of thirty to step up and take a player, and then he's gone. There are no do-overs. We may really like a guy, think we can get him in the 4th round or so, and then he's gone in the second. It happens all the time. Therefore, I believe that if you like the player and want him in your system, just take him. My litmus test is how I'm going to react when I hear another team call the player's name: a) a grimace with a head bob, b) an audible "Gah!" with a twist of the neck, or c) nauseous. If (c), then take the player if he's available." That is something I had wondered about, just like any other fan, for years, but I have trusted Sabean's judgement on that since I feel that he has shown good judgement in the trades he had made previously and the drafts he made when he ran the Yankee's draft. Free agency, sometimes you have to take what is on the market, because you need it, not because you necessarily think he's really that good, sometimes you just have to roll the dice when the philosophy is to win now with Bonds, much like any other team scratching and clawing to win it all, now. You have to cross your fingers. Back to the point, this is something I've learned from participating in a few fantasy drafts, if you really want a player, you have to select him, i.e. "over-draft him", if you want him on your team. Thus I tried to overdraft no more than a round or two before I thought others would want him (Lincecum :^) because as he wrote, all you need is one team thinking they need him, and he is gone.
• "The first round is always the most difficult, especially when you're picking in the bottom third. It's very difficult to predict the pool of available players, and the 5-tool, can't miss types are generally gone in the top 5 or 10 picks. This year was no different. " This is exactly what my point has been when I write here and at other Giants and baseball blogs about the draft, the best players are gone after 5 to 10 picks, depending on the year. After that, it becomes a much harder exercise of identifying players who might be useful, let alone who might be good players. As he noted above, sometimes you just have to draft a whole bunch of similar players in hopes that one of them will make it.

This is a great firsthand description of what is happening for a MLB team when preparing for and executing the draft. I must give my kudos to DePodesta for being so open and reaching out to the public in this way.

Giants Thoughts

As I wrote long ago, the baseball draft is more of a crapshoot than thought. It is relatively old now, but nobody since has been able to see the point my study tried to make. Most have focused on the fact that there is value and that there is more value early on, but still value later.

I never said that there wasn't value, but that the chance of collecting on that value is very low. In fact, I've kind of assumed that there must be some value, else these baseball execs must be the biggest chumps in the world, running a draft for over 40 years and not getting any value out of it.

What the authors of these other studies and the people who point out these studies to me lack is a clear understanding of how statistics work when you have situations of low probability. Which is something most people lack because they are simply not used to such phenomenon. But anyone who has taken an introductory class on business stats would understand.

It is all about the difference between expected value, which is what all other studies that have duplicated my study has found, and probability. Expected value has to do with what you can expect to receive if you were able to run the same event over and over again, and calculate the average value received.

But that's not the point when you are dealing with reality and you just have that one shot. That's why I used craps and the lottery as examples to illustrate my point. You have that one roll, you have that one ticket, you either win or you lose, do or die, what is the result?

In baseball, in the back of the first round, you have around a 10% chance of finding a good ballplayer (forget about the fact that it would be another 4-6 years before he pays off). So 9 times out of 10, when you skip that back of the first round pick, you just threw away a player who most likely will never ever be a good player, 90% chance of that happening, based on the draft results over a 13 year period.

While the result of past draft picks have no bearing on the future of any particular draft pick you want to examine, it is my assumption that the population feeding into the draft is producing players at approximately the same rate over time. Just like over time there is a certain percentage of, say, left-handed people, and other population characteristics. It may go up, it may go down, other sports probably does siphon off talent, but today, where there are few athletes capable of playing multiple sports at the same level of excellence, they generally are smart enough to figure out which sport will make them the most money relative to the damage their body takes and the joy they get out of playing the sport, and so the level has been relatively about the same for many years now, now that football and basketball have risen to baseball's level - and perhaps surpassed them - in the national stage.

One day I will figure out how to explain it right, but not tonight. Hopefully one day.