Giants Draft Analysis: The Sabean Era Drafts Overall (fourth and last in a series)

In this series, I broke up the Sabean era into three distinct periods - basically pre-Dynasty, Dynasty Building, and post-Dynasty - and analyzed each one based on the probabilities for finding Good and Great players, that I came up with in my recent study of the first 50 years of the draft.  It examined the odds of ending up with nothing in any particular draft and the cumulative probabilities of how many good and great players a baseball operations leader would have in a random draft (that is, if the leader has an average clue, as good as the average GM, as to which player is good, and thus is randomly choosing these prospects in the draft, based on this average expertise, randomly based on past history).

In this final blog post in this series, I look at the entirety of the Sabean era, see what the odds are of where the results are now, and where he could be if some of the current top prospects end up being good players.  As I've noted, it could take a decade or more for players to retire before we know the final results of any GM's draft results, if there are any players still playing and look like they could reach Good status.

And while this is the last of this series, I'm probably going to be updating this particular post after every season or two, see where he is, based on progress or decline of the remaining remnants of his drafts.  Obviously, there will be declines where we can say that a player won't be making it most probably, and there will be breakout seasons where he puts himself back into play.

ogc thoughts on The Sabean Era

Finding Good Players

Looking at the distribution spread for the 22 draft period based on the overall failure probability, I came up with a 0.2337% probability of failure, that works out to 3.8326% probability of finding a good player, across all 142 picks (assuming equal probabilities; a Monte Carlo simulation distribution would be more accurate, but I don't know how to do that yet, but this should be close enough to use).

Based on that success rate, using binomial probability, here's the probabilities based on 0.2337% chance of zero Good players over 142 picks:

1. 1.5% of finding one Good player
2. 4.5%
3. 9.0%
4. 14.7%
5. 16.5%
6. 16.4%
7. 13.9%
8. 10.2%
9. 6.6%
10. 3.9%
11. Odds of finding 11 Good players or more is 3.7%

At six Good players, Sabean and gang were in the middle of the pack.  There is 45.4% chance of finding less than six Good players, 38.2% chance of finding more, so he's a little above average.  Basically around average, which is the total sum of probabilities:  5.8 Good players for the 142 picks.

They look likely to come up with at least two more Good players.  Bart looks likely to reach, and I think another is likely among Ramos, Webb, Beede, Corry, and Reynolds.  Getting to 8 Good players would put Sabean's era in the top 24.3% of probabilities based on the overall failure rate.  Luck may be involved, but that's pretty good place to be, overall, looking more like skill than luck.

Is three doable?  Possible, but I don't know how likely.  If his reign gets there, that would put him in the top 14.1%.  At four, top 7.4%.  At these percentages, either he's extremely lucky, or extremely good at finding Good players via the draft, but more likely the latter than the former.

Finding Great Players

Looking at the distribution spread for the 22 draft period based on the overall failure probability, I came up with a 9.2648% probability of failure to find a great player over those drafts, and that works out to 1.66% probability of finding a great player with each pick, across all 142 picks (assuming equal probabilities; a Monte Carlo simulation distribution would be more accurate, but I don't know how to do that yet, but this should be close enough to see what's possible).

Based on that success rate, using binomial probability, here's the probabilities based on 9.2548% chance of zero Good players over 142 picks:

1. 22.2% of finding one Great player
2. 26.5%
3. 20.9%
4. 12.3%
5. 5.7%
6. Odds of finding 6 Good players or more is 3.2%

At two great players (Posey and Bumgarner are both over 36.0 bWAR), Sabean and gang were in the middle of the pack.  There is 31.5% chance of finding less than two Great players, 42.0% chance of finding more.  Basically around average, though a little under, which is the total sum of probabilities:  2.3 Good players for the 142 picks.  And 2 great players is the mode.

Too hard to predict great players, so I don't have any great feel for whether any will do it.  Both Bart and Ramos at least look on the right track, by performance and age, similar to Posey and Bumgarner.  Everyone else is pretty much out of the picture (it is 36.0 bWAR, after all, which is a lot, that's 18 average seasons, 12 pretty good 3.0 bWAR seasons, 10 seasons of All-Star 3-4 WAR seasons).

If one of them become Great, that would make 3, which would put Sabean slightly above average, with 48.7% chance of lower, 21.2% of higher.  If two are able to rise to Great, that would put him at 4, significantly above average, 69.6% chance of lower, 8.9% chance of higher.  It would require two great players to say that Sabean was good at finding great players, but even finding one would put him in good company, basically the 3rd quartile.  Who might be that next/last Great player?

Could Ramos Be the Next Great Player?

If I were pushed to chose one, it would be Heliot Ramos.  He'll be in his 20 YO season in 2020, and with a great start to the season, could force his way up to the majors by mid-season.  His performances in the minors so far have been above average and similar to other young prospects who made the majors at a young age too, performing as they did at a similar young age (Heliot was only 18 YO for the A-ball season, 19 YO for last season in Advanced-A and AA.  As I noted in my last post, Ramos, while not showing a great batting line in AA, he still produced above average for the league, although he was 5 years younger on average).

I've compared him against 18 YO's in A-ball and he looked good in comparison back then.  Let's compare what he did in Advanced A against others, first against league, then against his age, in prior years:

• Ramos: .306/.385/.500/.885 vs CAL .251/.329/.391/.719 (6th in OPS, 123% of CAL OPS)

Other 19 YO in CAL from 2014-2018 (minimum 250 PA):

• Jo Adell:  .290/.345/.546/.891 (18th in OPS, 121% of CAL OPS)
• Hudson Potts: .281/.350/.498/.847 (25th in OPS, 107%)
• Luis Urias:  .330/.397/.440/,836 (26th in OPS, 113%)
• Cody Bellinger: .264/.336/.538/.873 (30th in OPS, 119%)
• Franklin Barreto: .302/.333/.520/.833 (44th in OPS, 113%)
• Carlos Correa: .325/.416/.510/.926 (17th in OPS, 121%)

Other 19 YO in FSL from 2014-18:

• Ronald Acuna Jr (only 126 PA):  .287/.336/.478/.814 (12th in OPS, 120% of FSL OPS)
• Jake Bauers (249 PA):  .267/.357/.433/.791 (13th in OPS, 122%)

Other 19 YO in Carolina League from 2014-18:

• Victor Robles (20 YO):  .289/.377/.495/.872 (8th in OPS, 124% of Carolina OPS)
• Gleyber Torres:  .275/.359/.433/.791 (35th in OPS, 110%)

Here is Ramos stats ranking in the California League:  he was 4th in BA, 8th in OBP, 12th in SLG, and 6th in OPS in Advanced A-ball, while 3.5 years younger than the other hitters there.  So Ramos did pretty well, basically among Top 10 in batting line, and 123%, was not beat out by any 19 YO, and close to Jo Adell, Cody Bellinger, Carlos Correa, Ronald Acuna Jr, Jake Bauers, and Victor Robles (who was 20 YO, but thought I would include, as he rose quickly to the majors).

This is how long these players took to reach the majors, after their Advanced A performance:

• Jo Adell:  has not reached majors yet (was in Advanced A in 2018, so if reach in 2020, that would be two years)
• Cody Bellinger:  reached majors two years after Advanced A
• Carlos Correa:  reached majors one year after Advanced A
• Ronald Acuna Jr: reached majors one year after Advanced A
• Jake Bauers:  reached majors three years after Advanced A
• Victor Robles:  reached majors the same season he hit well in Advanced A (he was 20YO)

So Ramos, who hit roughly as well as the above in Advanced A, relative to the league OPS, if he duplicated their ascent, should reach the majors in 2020-2021, one or two seasons after, at age 20 or 21.

Here is Ramos' AA stats, didn't want to compare to other 19 YO as he only had 106 PA:

• AA:  Ramos hit .242/.321/.421/.742 vs EL .238/.311/.366/.677 (110% of EL OPS)
• He was 90th in BA, 75th in OBP, 32nd in SLG, and 42nd in OPS

Now, that was only 106 PA over 25 games, but if you split his AA by 12 games and 13 games, he had these stats (SSS but still, I think shows a great difference):

• 12 games:  .186/.286/.302/.588 (had hits in 5 out of 12 games)
• 13 games:  .288/.351/.519/.870 (had hits in 10 out of 13 games) (129% of EL OPS)

If you ranked this improved batting line, as he figured out how to hit in AA, against the leaders, Ramos was tied for 8th in BA, tied for 22nd in OBP, 5th in SLG, and 7th in OPS.  Small sampling size, of course, but still, some players need to adjust.

For comparison, Sandoval in 2008 at age 21 hit .337/.364/.549/.913 in AA, 123% of the EL average OPS, he was 3.5 years younger than the average position player there.  Ramos was 5.1 years younger than average (again, he was 19 YO for the season).  Sandoval that season, was promoted straight to the majors from AA, and, if not for his weight issues, probably wouldn't have seen the minors again.

In addition, Ramos' injury that he had cost him about a month or two of AA time.  He did well in the California League to start the season, hitting .270/.422/.587/1.009 in his first 19 games.  Then he had his injury, with a LCL sprain in his left knee.  He missed a little over a month of games, and when he came back, only hit .270/.313/.427/.739 in his first 23 games back.  Then he found his groove back, and in his final 35 games, hit .345/.409/.507/.916.  Had he not had that injury, he likely would have been promoted to AA sooner, by at least a month, and up to two.

Now, for whatever reasons, which I think relates mostly to his high strikeout rate though it could also relate to his likelihood of having to shift to RF at some point, Ramos is not ranked very high on any of the Top 100 lists, except for BP, and even there, 34th.  In fact, his striking out is high among the complaints about his hitting, which I would argue is pretty good since he's so much younger than the competition.  So my choice of Ramos might surprise some.

But based on the above, he has compared very well with other recent Advanced A 19 YO high performers, so I like his chances more, compared to Bart, based on age (he's 20-21 when he starts his MLB career; unlike Bart, who is already in his 23 YO season) and performance relative to young peers (Bart didn't even outhit his much younger teammate, Ramos, even looking at his before injury, post-injury return rustiness, and then post-injury hot bat).  In addition, Minor League Baseball Analyst, a prospect book I get every year, listed Ramos as a potentially elite player (Bart as well), in their 2019 edition (2020 not out until start of March), and that is what a great player is, elite.

So while I'm not even sure Ramos will be good, if I had to bet on one being great, I would bet on him, mainly because his performances so far compares well with young prospects who did similarly and rose fast to the majors.  In addition, I believe so because there is scouting that sees Bart as merely good overall, and only great at catching and power (I think he's going to surprise some there as well, I think he'll be good; just don't know how great he'll be), so I have Ramos ranked higher in potential than Bart (I understand why prospect rankers have Bart higher, he's older, more experienced, probably closer to the majors, got a lot of good tools; but based on comparisons with other 19 YO minor leaguers, Ramos' name is in great company).

Lesson in Randomness

All this is a good lesson in understanding probabilities.  Just because Sabean (or any other GM/leader) should on the high side in finding Good players, as this exercise shows, even at such very low odds, if you have an infinite number of monkey GM's randomly (relatively, not truly random, but at the average intelligence level across all the GM's) choosing at the same rate as in the past, there could be that lucky monkey GM who drafts a bunch of Good players.

There's good odds of finding 5 or 6 good players (32.9%) even with this low success rate, good chance of over (38.3%), so that one cannot say definitively that the GM is necessarily a genius if he's above average, even though he's beating the odds.  What one can say, is that he's been better at picking than other GM's with similar picks, but that there is a good chance of random luck too.  No way to distinguish for sure (or, at least, beyond my elemental knowledge of statistics).

But it's like hypothesis testing, the higher the improbability, the likelier that the GM is good at talent evaluation.  So right now, Sabean is right at average (though slightly above), which means his record of drafts is not much better than randomness of the average GM, he's been neither good nor bad over his time as the Giants business operations leader, though a bit above average in finding good players.  But if he reaches 7 good players, he's in the top 38%, which starts to look like separation.  At 8 good players, he is in the top 24%, which is pretty good.  And at 9 good players, he's top 14%, which is pretty rare.  Could be luck, but it's looking more and more like there's skill involved if more good players matriculate.  And I think the odds are good that at least two will prove to be good players.

Other Signs of Talent Evaluation Skills

That's why you can't just rely on the draft to say whether the GM is good or bad in talent evaluation.  He could be bad, he could be good, or he could just be very lucky, one way or other.   The further out his results separates out from the average, the better it looks that he's skilled, but there's also those odds that luck could have played a strong hand.

This brings me back full circle in my quest to understand Brian Sabean the GM.  Long ago, I was happy with the way he acquired talented players via trades, but his draft was the worse, and so I decided to research the draft to show how bad he was, compared to the best drafting teams back then.  What I learned was that the draft was a crapshoot, and that Sabean's record wasn't any different from other long term contenders like the Yankees, the Braves, or the A's, once they started winning consistently.  Losing (and badly in the Braves and A's cases) is what got them to their nice competitive status back then.

And that was because the draft odds were horribly bad, even in the first round, which everyone used to say was good (even today, people write about first rounders as if they expect all of them to become good), but my study back then, and this new study confirms using modern measurement, shows that the odds are actually really bad for finding a good player, no matter which pick you got, it's been less than a coin flip, but the odds of finding a good player is exponentially better with a Top 5-6 pick, and get exponentially worse, by two degrees, by the end of the first round, where the odds go from 49% for the first pick overall, 16% for teams on the bad side of the middle (7-14), to roughly 11% for playoff competitive teams (15-22), to roughly 5% for playoff teams (23 on).  Playoff teams have one-tenth the chance of the first pick overall, and still only a third of the chance for middling teams, and half that of playoff competitive teams, historically.

Distilling Talent Evaluation

The reason I've had faith, for the most part, in Brian Sabean has been his ability to find good to great players in trade with other teams, starting with his first unofficial trade to pick up Kirk Rueter, to his first blockbuster trade to trade Matt Williams to pick up a bunch of players which included Jeff Kent, to stealing Robb Nen, to trading for Jason Schmdt.

People get too tied up by free agency mistakes, blaming Sabean for them.  That's why I've written about the choices being made by a GM, and choices being made for the GM, and why I've been a huge advocate for using the draft for focusing on a strategy.  Often, when you need to fill the gaps in your roster (and no team has ever filled their team just with their draft picks to be competitive, so you fill through other means), you are stuck with the supply at the moment.  That's not his fault, though some like to turn fault on him for not finding a good player via the draft, but they don't understand the horrible odds of finding a good player, as I showed above, there is almost no way to ensure finding any particular position, other than to select a lot of players at that position (which, as I've noted often before, Sabean focused a lot of draft picks, and especially first round picks, on pitchers, until Barr joined the front office).

If you go the free agency route, you are at the will of the marketplace, and if you need a power hitting OF, and Aaron Rowand is the best available/willing, you have to sign him.   It's not ideal, but if you wait for ideal, you are sacrificing the opportunity to win that season. And while you know who might be available as a free agent, you never know if he signs with his team at the last moment, or he gets traded, and that new team signs him.

Same for the trading route, you again are at the will of the marketplace, and even worse, each team has a different philosophy of how they want to trade with you.  Some are dumb, look at, say, Baseball America's Top 30 for your team, and be happy with a combo-plate of prospects.  That's both good and bad, good if you can give him a bunch of players you don't care for but who are ranked, bad if they focus only on your top player, hoping to wait you out until you get desperate enough to give them your top prospects.  Others are willing to negotiate better, but maybe their player isn't exactly what you are looking for, he has flaws that you have to accept, as there is no other option.

Still, free agency is often made at the convenience of what you need which only costs money, but the supply is horribly limited.  It's not like a great hitter or pitcher is always available when you have the money to buy one, you either accept that you sign someone you need vs. someone you would rather have, had he been available.

Meaning you take what you can get, you won't get exactly what you think is ideal.  Trading opens up a lot more options, and, for me, is more indicative of how good a talent evaluator the GM is.  Picking up players via the waiver wire or as minor league free agents, as well.

Sabean Talent Evaluation

This was what enabled me to keep the faith during the late 2000's when almost everyone on the Giants board hated Sabean and booed his contract extensions:  his ability to find players via trades and other means.

And I can tell when the commenter is biased against Sabean because invariably they point out things like him picking up Jose Guillen or Ryan Garko, and other players who disappointed when acquired.  Every GM has mistakes, you can't know everything when you trade for a player, and thus focusing on his mistakes is just a convenient way for that person to feel like they know more than Sabean, and confirms their bias against him.

To show the difference, let's see what his top pick ups were and his worse trades.

Best pick ups were (eyeballed players for list):

1. Jeff Kent:  31.5 bWAR
2. Jason Schmidt:  22.0 bWAR
3. Randy Winn:  13.4 bWAR
4. Hunter Pence:  12.7 bWAR
5. JT Snow:  12.4 bWAR
6. Robb Nen:  11.8 bWAR
7. Ellis Burks:  9.7 bWAR
8. Santiago Casilla:  8.4 bWAR
9. Felix Rodriguez:  8.3 bWAR
10. Andres Torres:  8.1 bWAR
11. Gregor Blanco:  6.4 bWAR
12. Angel Pagan:  5.7 bWAR
13. Javier Lopez:  5.4 bWAR
14. Juan Uribe:  5.0 bWAR
15. Melky Cabrera:  4.6 bWAR
16. Hunter Strickland:  4.2 bWAR
17. Ryan Vogelsong:  4.0 bWAR
18. Scott Eyre:  3.8 bWAR
19. George Kontos:  3.6 bWAR
20. David Bell:  3.2 bWAR
21. Pat Burrell:  3.1 bWAR
22. Freddy Sanchez:  2.9 bWAR
23. Jean Machi:  2.1 bWAR
24. Dustin Hermanson: 2.0 bWAR
25. Kenny Lofton:  1.8 bWAR
26. Yusmeiro Petit:  1.8 bWAR
27. Brad Penny:  1.3 bWAR
28. Joaquin Arias: 1.3 bWAR

Worse trade/give aways (covering years they were with team traded to or pre-free agency):

1. Joe Nathan:  18.4 bWAR with Twins
2. Keith Foulke:  17.9 bWAR with ChiSox and A's
3. Livan Hernandez:  16.4 bWAR with Expos/Nats (salary dump)
4. AJ Pierzynski:  12.9 bWAR with ChiSox
5. Matt Williams:  12.2 bWAR with Tigers/D-backs (included both since he forced the trade)
6. Zack Wheeler:  10.3 bWAR with Mets
7. Chris Singleton:  10.0 bWAR with ChiSox and Orioles
8. Francisco Liriano:  9.3 bWAR with Twins
9. Steve Reed:  8.8 bWAR with Rockies
10. Rajai Davis:  7.6 bWAR with A's and Blue Jays
11. Scott Linebrink:  7.6 bWAR with Astros
12. Dan Otero:  5.4 bWAR with A's and Indians
13. Bob Howry:  5.3 bWAR with ChiSox
14. Russ Ortiz:  4.9 bWAR (salary dump) with Braves
15. Bill Mueller:  3.6 bWAR with Cubs
16. Doug Mirabelli:  3.5 bWAR with Rangers and RedSox
17. Carlos Villanueva:  2.5 bWAR with Brewers
18. David Aardsma:  2.3 bWAR with four clubs
19. Clay Hensley:  2.0 bWAR with Padres
20. Allen Watson:  1.8 bWAR with Angels
21. Erasmo Ramirez:  1.7 bWAR with Rangers
22. Yorvit Torrealba:  1.4 bWAR with Rockies (traded by Mariners)
23. Travis Blackley:  1.3 bWAR with A's
24. Alfredo Simon:  1.2 bWAR with three clubs
25. Joe Biagini:  0.6 bWAR with Blue Jays and Astros
26. Pat Misch:  0.6 bWAR with Mets
27. Brian Buscher:  0.5 bWAR with Twins

Going through all the trades over the years, Sabean was at +71.0 bWAR over the 18 years he was GM, roughly +4 bWAR per season.  He also added roughly 50 bWAR through waiver pickups and minor league free agent signings, like Burrell and Blanco (did not subtract the bad ones though).  

And through the draft (including players who moved on to other teams, just inclusive of all draft picks), Sabean-led drafts have totaled 281.8 bWAR over the 22 drafts led by Sabean hires (him, Tidrow, Barr), adding an average of 12.8 bWAR per draft, and counting, as others like Bart, Ramos, Corry, Webb, Hjelle, ascends to the majors.  And out of the first 17 drafts, which seems pretty set in results, 11 of the drafts produced at least 9.0 bWAR or more, 4 produced 18.0 bWAR or more, and 3 produced 36 bWAR or more.  

For now, out of those 22 drafts, only two drafts ended up being negative (2000 and 2013), though 2013 could become positive depending on what Arroyo and Rogers do in their careers, plus 2014 to 2018 are still much in play, as prospects rise to the majors, and either produce or not.  2014 look likely to be positive, as it includes Beede, Webb, and Coonrod.  2015 still has many prospects on the edge of the majors, plus Duggar and Cyr, who could produce enough to ensure it being positive.  2016 has Reynolds, and barring any sudden decline on his part, likely to be positive.  And 2017 has Ramos and Corry, while 2018 has Bart, Hjelle, Rivera, and others too soon to discern, but given Ramos, Corry, Bart, and Hjelle, likely to be positive, if not a lot more.  

I should also note here that Sabean led the scouting department for the Yankees, first as Director of Scouting and next as Director of Player Development, from 1987 to 1992.  During that time period, the Yankees drafted the following players, many of which contributed to their 4 in 5 dynasty, as well as other productive players:  Gerald Williams, Brad Ausmus, Deion Sanderes, JT Snow (yes, he was a Yankee first), Carl Everett, Andy Pettitte, Jorge Posada, Derek Jeter.  And Mariano Rivera was signed while Sabean led scouting, as well.

Overall, there is a preponderance of evidence that Sabean is a good talent evaluator, added evidence to his successes in draft, which suggests that wherever he ends up when his drafts are finalized, there is evidence beyond the draft that he was a good talent evaluator, if not a great one.