Draft Analysis of First 50 Years (a series): Definitions and Initial Analysis

As I've been saying I would do, I'm finally done collecting and doing initial analysis of the June MLB draft.  I collected the data from that great baseball resource, www.baseball-reference.com, after the 2018 season (so the data is up to date up to then, plus I bumped up a few players who look likely to move up to the next category).  But given that it takes a number of years for prospects to reach the majors, I thought I would just focus on the first 50 years of the draft, covering 1965-2014.

I'm not done with the more serious analytics, but there are a lot of random bits of info that I wanted to share now, as well as providing my methodology and definitions.  I've been releasing some of this in posts this year, covering the early parts of the draft, in relation to the Giants picks, as I had the early parts of the draft done before, but just recently finally went over and inputted the last picks I had captured, covering the first 200 picks overall in the draft.  As usual, my semi-long post has grown large.

ogc thoughts

First: An Observation:  Teams Rarely Regret Not Signing a Draftee

I'll start first with an observation, which I could back up with data, but that would take me learning more R to do that.  Still, I thought it was interesting enough to bring up without data:  roughly 1-10 out of the 50 players drafted in the first 200 picks overall did not sign with the team (there were a few where everyone signed, and there were a few over 10, but vast majority in that range), and as I went through each pick overall and tabulated the numbers (if I knew R, it would have been much faster; next project), I could see that the vast majority of players who did not sign eventually did not even reach the majors, let alone be good.

Sure, there were a few who didn't sign who went on to a great career (Tom Seaver came to mind; ugh, imagine how good the Dodgers would have been in the 70's had he signed with them?), and many who were useful, and even good, but for the vast majority, they either never made the majors nor did very much in the majors.

The takeaway there for teams is that while it's sad to lose a draft pick (even one punted by signing a free agent before the deadline), the draft, even in the first round, is a crapshoot, the vast majority of players never reach the majors, and the vast majority of players who make the majors are mediocre, at best.  The draft is a volume business of sifting through a lot of pyrite before you find that gold nugget of a prospect who is a good player.

Second:  Some Methodologies and Definitions

And that reminds me, I should share some definitions on what "good" is, at least for my study ,and what my methodology is. 

First and most importantly, why I'm doing my analysis the way I did, which is what I did with my first draft study.  The vast majority of draft studies focus on averages of production, like the average WAR per each draft pick.  While there has been some interesting findings with that methodology, I looked at categorizing each pick instead, because averages work best if the data is normalized, but as we'll see, the distribution of talent for each pick is heavily skewed towards the category "No Majors", meaning that they never played in the majors, and Mediocre, which is my category for those of value but not average value.  For example, starting with the 5th pick overall, roughly 30% of those picks never made the majors, and by the back of the first round, picks 20-30, it is up to 35-40%.  As a Stanford professor illustrates this problem:  if you put one hand in 0 degree water, and the other in boiling water, your hands are in average, comfortably warm water temperature, does the average convey the situation properly for your hands?

I approached it from the viewpoint of any team (or any fan):  what do we want from the draft.  Sure, we can get interesting players like Tim Alderson, whom the Giants selected after Bumgarner in that draft, and used him to get a productive veteran, like Freddy Sanchez, or a useful player like Joe Panik, but that's not why any of us are interested in the draft.

We are all looking for a good player, that next generation star who will lead our club in its next competitive iteration, or keep the good times going.  In any case, whether you have a trading piece or not is often determined by circumstances, as some who don't look that interesting (Tyler Herb) gets you a nice prospect (Mike Yastrzemski), while many other better prospects never net you anything (Giants have a long history with those AAAA types).  However, whether a drafted player ends up good or not is clear cut, once you have definitions and categorizations on the value of a player.

Thus, I'm still focused on the likelihood of a team finding a good player through the draft, much like I was in my first draft study, only today, I'm using modern metrics.

WAR Is the Modern Way

As my prior draft study used what was available (batting average and ERA), I wanted to use WAR with this new study, but what are the proper categorizations?  The major underlying concept about WAR is that an average player is still worth something, and therefore is valuable, and 2.0 WAR is defined as the value of an average player (Wins Above Replacement) relative to the 0.0 WAR that is defined to be a replacement player.

A replacement player is that player on the edge of making the majors.  It's the player that a team theoretically can procure easily from the minors, who would neither add wins to the team nor subtract wins.  In other words, add 0.0 WAR to the team by being on the 25-man (soon to be 26-man, that'll take some getting used to) roster.  Whereas, if you have an average player, he would add 2.0 WAR (or wins) to the roster if he took the place of the replacement player (for example, a team has a replacement player on their 25-man roster, but trades for an average player, kind of like when the Giants traded for Kevin Pillar and he replaced Reed on the roster; assuming Reed is a 0.0 WAR player, he was pretty bad in his short stint, so he could be less), which is valuable, but not what I would define as good.

WAR is not perfect yet, and probably will never be perfect, not like physics or calculus, but can be useful, like financial analysis, where certain metrics are good indicators of the good situations you are looking for.  In baseball, it's the combination of what a player can produce in offense and in defense, and, as the saying goes, the best ability is availability, so someone who can be healthy over a full season.  In other words, as a hitter on offense, or as a fielder and/or pitcher on defense, and someone who can stay on the field and produce, as WAR is a counting stat.

Offensive measurement has been ahead of defense for a long time now, and probably will always be ahead. Yet offensive measurement still had leaps and bounds in recent decades, as teams collectively realized that walks added a lot more value than thought.  And recent concepts like launch angle and exit velocity only came into vogue once technology could provide such metrics, although Ted Williams had been talking about these concepts in his great tome, The Science of Hitting, for over 50 years now (great book, I don't get anything for recommending anything in my blog, FYI).

Defense is an area that is still being explored, and thus the validity of the defensive values of WAR being used now is subject to potential change.  Pitching measurement has been much better than fielding, but the discovery of DIPS showed that there was still much to understand about pitching.  And even DIPS missed that fact that there ARE pitchers, like Matt Cain and Barry Zito, who could control BABIP, which saberists like Mike Fast, after he joined the Astros, stated in an interview that BABIP control is a skill.  Yet tools like FIP are still used, although it misses a significant skill advantage, that of depressing BABIP much better than other pitchers. So there is still areas where we can advance in understanding.

Fielding has only recently gotten out of the dark ages of fielding chances and error percentages.  New methodologies like zone range (leading to Ultimate Zone Rating or UZR) and new technology has provided analysts with new ways of defining fielding excellence, allowing analytics to get down to nitty gritty stuff like catcher framing, arm strength, positioning, and shifting.  A great source that is innovating frequently is the work being put out via The Fielding Bible, where there is a data table available for perusing who did what.  This is where WAR falls down in accurately measuring defense, but getting better all the time.  Statcast has started to collect fielding, as well as pitching and hitting metrics, and this source of data will definitely be the leading edge of analytics for the future.

So, overall, WAR is like most metrics:  flawed but it's well understood, the best we have available right now, and it provides enough information to allow analysts to influence decision making positively, i.e. make good decisions.  It is pretty good for figuring out who are good players and who are not, even if it is not definitive (and never will, given we have no advanced fielding stats over much of the history of baseball).

What is a Good Player?

So what would a good player be defined as?  Well, if average is roughly 1.5-2.5 WAR, and average is valuable but not what most fans and teams think of when they dream on a drafted prospect, I thought I would try 3.0 WAR as the definition of a good season.  And since it takes a player six full seasons to become a free agent, six times 3.0 WAR equals 18.0 WAR.  And that also works out to nine seasons at 2.0 WAR, as well, to recognize those who are pretty okay (average is still a good thing to be in the majors) but for a long time (again, the best ability is availability).  So I started with the idea of 18.0 WAR, but wasn't tied to it.

However, once I went digging into the numbers, what I found at 18.0 WAR, it did seem to be a dividing line between good and not good.  I'll admit a bit of bias here, because, while Dave Kingman was my first prospect love and I liked him as a player, I did not consider him to be good, overall, he just went homer happy after his first season or two, and he was under 18.0 WAR, but right on the border, he was 17.3 bWAR.  I could not define him as a good player.  Useful, yes, but good, no.

Still, there was something about 18.  I then started looking at the best elite players, those above 18.0 bWAR.  Again, 18 seemed to work, with 18.0  for good vs. 36.0 for great players vs. 54.0 for HoF players, this all seemed to be a good divide as well.  And 9.0 bWAR seemed to be a good dividing line between mediocre MLB players and useful MLB players.

I also liked that 9 is a key number in baseball:  9 innings, 9 players (18 innings for the two teams, 18 players as well), 162 games is divisible by 18, so you can break a season up into a 9 inning game with road and home halves.  But mostly, looking at the names in the list, 18.0 WAR worked in my subjective view of who is good and who was okay but not good.

Player Categories

So I came up with these Player Categories:

• Unsigned:  Did not sign when drafted that year
• Bust (Minors):  Never made the majors, only played in the minors
• -WAR:  Negative WAR
• Mediocre:  0.0 bWAR to 8.9 bWAR
• Useful:  9.0 bWAR to 17.9 bWAR (basically where average type players end up)
• Good:  18.0 bWAR and higher
• Great:  36.0 bWAR and higher
• Hall of Famer:  54.0 bWAR and higher

I have found one other researcher who analyzed the draft as I did (Royals Review).  Average also started at 9.0 WAR for him, but Good started at 15.0 WAR, Very Good at 21.0, and Great at 27.0.  He considered anyone under 9.0 WAR to be a bust, but I think Mediocre, for me, occupies a space between bust and average/useful, because they have some value to the team.  From my viewpoint, the draft bonuses are so low relative to the price of WAR, that positive WAR is still of value.

Positive WAR Value to Teams 

Here's how I see that. Matt Swartz has projected $12.4M per WAR, but other analysts, using the lull of the past two off-seasons when the top teams stopped spending to get under the CBT threshold, projects a lower value. I don't see why I would adjust lower based on purposefully reduced spending, unless we know that the top teams have stopped spending.

Looking at Bumgarner's contract, one might think there is some price deflation.  But I think he's just an outlier, he wanted to go to Arizona, it seems abundantly clear now, and it seems like he took less to get where he wanted to go, while Wheeler's 5 year, $118M contract ($23.6M AAV) and Ryu's 4 year, $80M contract ($20M AAV) look more in line with the market.   And, of course, there's Gerrit Cole's 9 year, $324M ($36M AAV) and Anthony Rendon's 7 year, $245M ($35M AAV), plus the Nats redoing of Stephen Strasburg's contract to 7 years, $245M ($35M AAV).

So I see the past couple of seasons as blips due to the CBT, not the actual full buying power, and that eventually teams will be spending freely, which appears to have started up again this off-season.  Even Keuchel, who couldn't get a satisfactory contract last season, got a 3 year, $55.5M ($18.5M AAV) contract.

And at $12.4M, even 0.1 WAR is worth $1.24M, which covers the bonuses of most draft picks nowadays. So I decided that Mediocre includes down to 0.0 bWAR, as it's so close that I feel that random luck probably separates those at 0.0 from those at 0.1, may as well include them into the Mediocre category. Especially since defense isn't that well defined yet, a swing of 0.1 WAR in either direction would not surprise me, as sabermetrics get better at defining the value of fielding.

Other Thoughts on Categories Definitions

Now those with negative WAR, I could have included them in Mediocre.  Most players don't ever get the chance to generate more than negative 1 or 2 WAR before they stop getting chances, especially today (unlike Johnny LeMaster, who was severely in the red but still had a long career in the majors).  I just felt like there should be some sort of dividing line, so I chose being negative WAR, just to see how that would affect the percentages in each category. It seems like teams don't let it ride with prospects who can't produce wins, they will get a certain amount of chances, and then it's on to the next prospect.  And, again, I felt that there should be a dividing line between those who produced poorly compared with those who produced some value.

Thinking about this now, after collecting the data, I probably should have broken up the Mediocre category into at least another 2 or 3 categories.  Anything up to 3.0, maybe I could have included them with up to -3.0 WAR, for a, say, Replacement category.  Maybe with more opportunity, these players could have gravitated towards 0.0 WAR.

Then maybe 3.0 to 5.9 could be Mediocre, and 6.0 to 8.9 would be below average, covering players who were average-ish over 3-5 seasons.

But if a player can be average (1.5 to 2.5 WAR) over 6 seasons, then they usually (teams are trending to not do this anymore, though) can move into free agency and get a nice contract, and those players are useful, which is why I defined 9.0 to 17.9 bWAR to be Useful.  I could have used the term average, but I wanted to convey the value of the categorized player via the label.

Third:  Initial Analysis

Looking at the percentages for each pick overall, I initially grouped the picks in the following buckets.  I plan on eventually testing each grouping and pairing each pick with the group it statistically tests with the Null Hypothesis.  But for now, I think this should do, via eyeballing (I expect the groupings to change later, but not significantly, especially for the first round picks, up to #30):

1. Pick 1:             48.9%
2. Pick 2-3:          26.8%
3. Picks 4-6:        19.6%
4. Picks 7-14:      15.6%
5. Picks 15-23:    10.8%
6. Picks 24-76:      5.20%
7. Picks 77-140:    2.14%
8. Picks 141-200:  1.29%

With 200 picks, that covers roughly the first six rounds, and into the seventh round, for today's draft with 30 teams.  My original study just covered 100 picks overall.  Aside from the work involved with capturing all that data from an online source, I was only interested in the first 2-3 rounds back then.  Knowing how bad the draft was from that study, I pushed myself to get to 200 picks which would get us to the full fifth round (150 would have been short), where some good players are still found (like Brandon Belt).  It helped immensely, too, that baseball-reference.com provides some tools to help make the data capture easier. 

In any case, with the split appearing to these eyes to be at 140-141, had I went with 150, I would have been very tempted to push it to 200, just to see if it was an aberration for 141-150 (0.71%) or if it was a new grouping.  As it was, it was an aberration, just not as big as it seemed with that range, reverting to mean from 151-170 (1.67%), bringing to roughly 1.35%, then 171-200 was roughly 1.24%, working out to 1.29% over that full range. 

Draft Odds Implications

First off, isn't it amazing that even for the first pick overall, it is still less than a coin flip odds that the team has found a good player?  Not that there aren't always some available, that's for sure, but even the team with the ability to pick anybody they want, fails more than 50% of the time to find a good player.  And it's all severe downhill after that, falling exponentially every two rounds, or less.

By the time we get to the top half of the majors, theoretically, the winning teams, the odds have already dropped to 10.8%.  To put that into perspective, that means that on average, a team that stays in that draft range within the majors would find one good player every 9 years, on average.  And for the division winners and contenders, the odds have dropped to 5.2%, which means finding one good player roughly every 20 years of the draft, if the team can somehow stay that good, while only finding one good player every 20 years.  This illustrates the lack of depth in good players in any particular draft, as well as the difficulty in identifying and developing that player into a good player.

This is why I say that it's extremely hard (if not impossible) to stay competitive indefinitely.  The rate of one good player every 9 to 20 years is not enough to keep a team competitive.  Money is (see the Yankees), but ultimately, you want win it all on top of simply being competitive.  The Yankees, in spite of their competitiveness and vast stores of financial might, has only won one World Series championship in the past 19 seasons.

Similarly, it is just as hard to find good players via International Free Agents as well.  Don't know what the current top list shows, but over a decade ago, when the Giants signed AnVil and RafRod to $2M+ deals, it was telling that the Top Ten list of top contracts showed only one good player, Miguel Cabrera, and 9 duds of a prospect.  There are no easy sources of good players via the draft or IFA, the two ways of finding a young amateur who might become a good ballplayer for a team.

And look at the odds for supplemental picks for the first round:  5.2%.  Again, that is roughly finding one good player every 20 seasons.  So if you lose 20 good players via free agency, out of those 20 picks that you get, you on average would have found 1 good player.  This is why I've never been so excited about getting draft pick compensation, sure, it's way better than being poked in the eye with a stick, but it's not going to materially affect your team success either.  It's even worse if you are getting a pick in the supplemental second round, which is what the Giants got for Bumgarner and Smith.

Odds A Team Ends Up With No Good Player Selected Over Five Drafts

Furthermore, these results illustrate why most competitive teams have a hard time staying competitive.  Based on the above odds, over, say, a 5 year period (like Zaidi's five year contract, or the five years he has been GM), following are the odds that a team will end up with no good player selected in that 5 year period of selecting in that range:

1. Pick 1:                 3.5%
2. Pick 2-3:            21.0%
3. Picks 4-6:          33.6%
4. Picks 7-14:        42.8%
5. Picks 15-23:      56.5%
6. Picks 24-76:      76.6%
7. Picks 77-140:    89.7%
8. Picks 141-200:  93.7%

Hence why teams are tanking.  With a Top 6 pick, over a 5 year period, you can expect to select a good player (or multiple) anywhere from 66.4% to 96.5% of the time (like the Nats tanking to get Strasburg and Harper; a team has to push hard to lose 100 or more games, as they did in those two seasons).

Meanwhile, consistent playoff contenders can expect to be shut out of finding a good player 76.6% of the time, and the remaining winning teams can expect to end up with nothing anywhere from 42.8% to 56.5% of the time, with their first round picks.  And once you get into the third round or so (pick 77 and later), you can expect (over a 5 year period) to be shut out of finding a good player 90% or more of the time, from those five picks in that round (and for each round).

Furthermore, taking that across multiple rounds, over a five year period for rounds 3 to 5, covering 15 picks, and using 1.86% as the odds (roughly two picks in 2.14% range, one pick in 1.29% range), a team would end up with nothing 75.5% of the time.

Using an example that has more meaning, Zaidi had a 10th overall pick in 2019 and will have the 13th pick overall in 2020, later this year.  Based on the odds for those picks, 26% of the time, he'll find one good player, but 71% of the time, he'll end up with nothing, just by random chance based prior history (a little over 2% of the time, he'll have two good players, very rare; Sabean found 3 in a row with Lincecum, Bumgarrner, Posey).

Baseball Competitive Cycle

Hence, why I believe that there are competitive cycles that all teams go through, a winning period, followed by a losing period, and eventually cycling back to winning if management is good enough.

But with no guarantee of when that will happen.  The Yankees over the last 40 years have won the World Series five times, which is pretty good.  However, four of those championships were won with the young players they picked up while losing from 1989-1995, led one their top pick in that period, Derek Jeter (Sabean was head of Yankee scouting then, FYI).   So if you remove 1989-2000 from that period, they have won only one championship in those 28 seasons, or roughly random.

Giants 2020 Draft Odds

Here are the Giants draft picks in 2020 per MLB Pipeline, with the percentage for that pick from the table above:

• 13 (first round):  odds of finding a good player:  15.6% (in 7-14 range)
• 50 (second round):  5.2%
• 70 (Bumgarner QO compensation):  5.2%
• 71 (Smith QO compensation):  5.2%
• 87 (third round):  2.1%
• 117 (fourth round):  2.1%
• 147 (fifth round):  1.3%
• 176 (sixth round):  1.3%

Altogether, these 8 picks add up to 38.0%, which means that, on average, in such a situation in prior drafts, there is an expected value of 0.38 of a good player.  Using these probabilities, that works out to a 67% chance of not finding a good player (assuming that the history of Good player distribution at that draft pick is accurate), leaving only 33% chance of finding a good player (or more than one, since these are not mutually exclusive; one day I'll figure out how to use Bayes).

That is roughly a third, which means that if the Giants do this again another two more seasons, with similar number and placement of picks (unlikely given the two QO picks, but if), then the Giants on average would find one good player in those three drafts, covering picks from round 1 to 6.  And since prospects take up to 4-6 years to develop, it is probably about 8-10 years before you have a good idea of whether a good player has been found (and even then, as we saw with Panik, a strong start does not mean a good career) in that 3 year draft period.

Sabean/Evans Drafting While Winning

Hence why I've been defending Evans and Sabean's legacy covering the past decade of draft picks.  Even with a relatively good pick in the first round in 2020, plus two QO picks, Zaidi most likely will not find a good player, and it could be up to 8-10 years before we know for sure that he succeeded or failed with his picks.

Meanwhile, Sabean and Evans had a bad run per common Giants social media complaints starting from 2010 (first pick after winning season) with Gary Brown (#24, 5.2%):

• 2010:  Gary Brown, #24, 5.2%
• 2011:  Joe Panik, #29, 5.2%; Kyle Crick, #49, 5.2%
• 2012:  Chris Stratton, #20, 10.8%
• 2013:  Christian Arroyo, #25, 5.2%
• 2014:  Tyler Beede, #13, 15.6%
• 2015:  Phil Bickford, #18, 10.8%; Chris Shaw, #31, 5.2%
• 2016:  None (lost from Shark signing)
• 2017:  Heliot Ramos, #19, 10.8%

That adds up to 0.74 expected good player among those picks.  And if you multiply the opposite odds (of not finding a good player, which is one minus the odds of finding a good player), there's a 45.9% chance that they don't find a good player among this bunch.  In other words, pretty good odds of nothing happening, and it's random, not poorer than average selecting.

The dichotomy is that there are some - slight - odds of finding multiple good players among the sample, and that raises the overall expected value to compensate for the 46% odds that no good player is found.  This illustrates what I was talking about above, that there are competitive cycles because the odds of finding a good player becomes prohibitive once you are playoff competitive or even mediocre, a .500-ish team.

In spite of 4 non-playoff level picks, the Giants are very likely (46% of the time) to end up with no good player, if they are as good as selecting as any other team over history.  No competitive team can stay competitive if after 8 seasons of first round picks, they are as likely to select no good player as they are to select one or more, they need more support than that, especially given how long it takes good players to matriculate to the majors.

Any Good Player Out of this Bunch?

Meanwhile, three prospects among these are still possibilities for becoming good:  Christian Arroyo, Tyler Beede, and Heliot Ramos.

Arroyo is still only 25 YO for next season, and hence why I still include him.  Rays didn't play him much in 2018-19, but taking his Rays performance and extrapolating out to a full-ish season, he produced at more than a 2.0 bWAR/600 PA rate.  He should be mediocre at worse, useful possibly (9 WAR is a high hurdle), and perhaps good if he can peak to 3-4 bWAR in his late 20's, while playing average for the next 6-8 seasons.  Very outside chance, have to see how he does in 2020 to get better feel, but he needs to take a big step this season if he's to reach good level.

Beede is negative at the moment, and only had 0.1 bWAR in half a season in 2019.  However, he improved greatly by the end of the season.  He appeared in 24 games, starting 22 games.   He struggled early, and in his first 9 games, 7 starts, he had a 6.45 ERA (5.85 FIP), 37.2 IP, 42 hits, 28 BB, 39 K's, 7 HR, .857 OPS, .327 BABIP, 60% strikes, 17% strikes looking, 11% strikes swinging, and average Game Score of 43.  In his last 15 starts, he had a 4.42 ERA (4.62 FIP), 79.1 IP, 85 hits, 18 BB, 74 K's, 15 HR, .774 OPS, .306 BABIP, 63% strikes, 16% strikes looking, 12% strikes swinging, and average Game Score of 50.  

Key stat for Beede's improvement is his drop in BB/9 from a horrendous 6.7 to an elite 2.0 BB/9.  Still giving up too many homers (horrendous 1.7 HR/9 in those 15 starts), which is why his OPS is so bad.  His batting line was .269/.315/.459/.774, which is a low OBP, and if he can reduce to 10 HR (keep in ballpark to doubles) and just do the same, the SLG would only be .427 and OPS only .742.  And if he can just repeat his performance with less homers, that's 2-3 WAR (two pitchers close to his 4.42/4.62 were Merrill Kelly and Tanner Roark, 1.4 bWAR and 2.0 bWAR, respectively), if not better, and since he's 27 YO next season, he would need to do that for 6-9 seasons to reach good status.  

Given how good this was, he looks likely to reach useful status, though if any injuries, he's mediocre. There's an outside chance of being good, as he has a pitcher's body, no prior injury history that I'm aware of, so if he can just continue what he was doing for half a season of 2019, he might be able to achieve that category.  For example, dropping to 10 HR vs. 15 HR drops his FIP to 3.80, and his ERA would be in that range as well:  bWAR for similar pitcher ERA/FIP area ranged from 2.5 to 6.0 bWAR.  This reminds me a bit of Lincecum's first season, he pitched really well for the most part, but one bad month made his overall ERA average.  Hopefully Beede can take the leap in 2020 and just continue doing what he did well in that second half of the season, but over a full season.

Heliot Appears to be The One

Heliot Ramos is the shining light of the bunch.  He hit .306/.385/.500/.885 in Advanced A ball as a 19 YO.  Examples of recent past 19-20 YO in California League:  Jo Adell as 19 YO in 2018 hit .290/.345/.546/.891, Gavin Lux as 20 YO in 2018 hit ..324/.396/.520/.916,  Luis Urias as 19 YO in 2016 hit .330/.397/.440/.836, Cody Bellinger as 19 YO in 2015 hit .264/.336/.538/.873.  Doesn't mean that he's a sure thing, but good recent precedence for 19 YO hitting in the California League then rising to majors to do well.

Right now, I would say the odds are likely that Ramos can be that good player in the bunch.   His level of performance relative to age in the league context is similar to other good hitters who were similarly young, and performed similarly.  Yes, strikeouts are a concern, and could be the deal breaker for his MLB prospects, but many hitters can make up for their strikeouts with their power, which he has a lot of.

Even if he doesn't make it, 46% of teams with similar selections would have ended up with nobody good as well, and hence why I can understand stating that the Giants declined because of this lack of success, but don't understand blaming Sabean/Evans/Barr, when the odds were against them in the first place.  It's like getting mad at someone because they didn't make the Olympics or win a Nobel Prize, the odds are just very high that they are going to fail, so blame is the wrong sentiment.

It is just a matter of the beast of winning, most teams in history have been beaten down by this fact of the draft (currently Cards and Dodgers are defying the odds, but then again, so was Sabean from Cain to Belt), and poor results is not a result necessarily of bad selections, but of simple reversion to the mean, the average result of all draft picks:  that you are more likely (much more likely after the first overall pick) to not find a good player than you are to find a good player.  The odds are stacked against the player and the team.

Could There Be Two Good Players?

And I like the odds of there being two good players.  With Beede having such a great ending to the 2019 season, that gives me a lot of hope, although not certainty, not close at all yet.  If he can simply repeat what he did (admittedly tough, see D-Rod's 2019; then again, Beede has stuff that D-Rod doesn't) in that 15 starts at the end, he'll probably be about a 2.5-3.5 bWAR player, over a full season, at age 27 in 2020.  With 5 more years of control, he would need to produce another 15.0 bWAR roughly, to age 32 season.

Which is not likely to happen - though you can say that about any player until they reach 18.0 bWAR, because of the attrition rate, and you can also say this about players who in the low 20's WAR range as they can fall back, as Lincecum did - but at least there is that hint of possibility, given how well Beede did late in 2019 (really well, he had a great 8.4 K/9, 2.0 BB/9, and 4.2 K/BB ratio, all sterling rates, he just needed to kill his HR rate - big deal, though, he needs to halve it to be okay - to become a top of rotation level pitcher).  He looks like he could become the ace of the rotation over the next few seasons, presumably once Shark and then Cueto are dealt for prospects sometime in 2020, once he figures out everything, perhaps paired up with Webb atop the rotation for the near term Giants.  Obviously, he needs to prove it in 2020, or the odds practically drop to zero.

Sidenote:  Shark and Cueto Probably Gone in 2020

I think both Shark and Cueto will be dealt before the 2020 trade deadline.

Shark has been a sturdy and mostly reliable starter.  For all the complaints about him, he's been above average the seasons he's been healthy, generating 7.6 bWAR, which is roughly just above what he's been paid in salary.  I think a contending team will be interested in trading for him in spring training, probably due to injury on their part, maybe because they missed out on the better starters in free agency, and he only is signed for 2020.   Given his odd 2018 season injury, a team would need to see him pitch well and look good at the start of spring training before committing to trading for him, I would think.

Cueto has to prove he's healthy, recovered fully and productive, from his TJS (not everyone recovers to be good again) so a team probably isn't interested until near the trade deadline, with half a season of proof (plus half a season less salary), and only a season and a half left of contract.  He's good enough that I think a team would accept his contract once they got enough proof.  Teams are always looking for good sturdy pitchers, and both of them have proven to be that for the most part, with little contract risk left for either.

Another reason for thinking this is the signing of Tyler Anderson.  He'll be out for the first half of the 2020 season, coming back around the trade deadline.  At that point, one or both of Shark and Cueto could be traded, plus at some point, they will need to cut back on starts for Webb (or any other young starters who happen to be pitching in the majors) or perhaps a young pitcher starts losing it, and need replacing.

Future Research

I'm open to ideas, so post them if you got one.  When I reached the end, I realized that I've done more than I had intended to start with, with this one post.  For further work, I'll need to learn to use R, and do analysis that way. 

However, doing the examples with the Giants drafts gave me the idea of covering the span of Sabean's leadership reign over the Giants, from 1997 to 2018.  Once he handed the reins over to Evans, it appears clear that he did not do much for the Giants other than scouting, his first love; it appears that he was burned out as a GM, tired of dealing with the modern younger GM's who did things differently.  Still, the people he hired to help run the team was still around, doing what they did before, particularly John Barr, head of scouting, who found us BCraw and Belt in the later rounds, as well as Duffy, and more recently, across all rounds, Beede, Webb, Ramos, Bart, and Corry. 

I'm not sure of the format yet, but thinking of covering one draft at a time, tallying up for that draft, and running a cumulative overall summary, to see where he stood after each draft.  But if you can think of a better format, I'm all ears.