Rookie Scores

October 24, 2017
 2017-55

Part One

 

              Let us suppose that we divide all rookies into ten classes, with the highest class of rookies being "Tens" and the lowest, most common class of rookies being "Ones".     As to why we are sorting rookies into ten classes. . .well, we’ll talk about that later.   As to how we are sorting rookies into ten classes. . .well, we’ll talk about that later.   First let’s just talk about Class One.

              I sorted all rookies in history. . .well, not quite.   I did not include

              1)  Pitchers, and

              2)  19th century players.

              I sorted all position-player rookies since 1900, which is a total of 15,907 rookies.   It’s not 15,907 different players, because many players have two rookie seasons.   Carson Kelly of the White Sox was a rookie in 2016 and a rookie again in 2017, and if he plays again in 2018 he will be a rookie again in 2018.   Geovany Soto won the National League Rookie of the Year Award (1988) in his fourth rookie season, and Tommie Agee won it in 1966 in his fifth rookie season.     

              So there are 15,907 rookie seasons since 1900, of which 7,490 are Level One rookie seasons.   It’s almost half; not quite half.  

              A Level One rookie season is a season is which the player

              1)  Is not especially young,

              2)  Is not a fast runner, and

              3)  Does not have a big year with the bat. 

              A player who has a completely undistinguished combination of these things.  

              A Level One rookie has about a 1 in 1,000 chance to be a Hall of Famer.    Nine Level-One Rookies have won a Rookie of the Year award, not in the Level-1 season but in a subsequent rookie season. Level One rookies are generally players who have limited playing time.   Only 3½ percent of them play in 100 games; only 26% of them have 100 plate appearances.   Of those who DO have a little playing time, they are almost always older players who don’t run well.    Less than 20% of them will have 1,000 plate appearances in their major league careers.

              When a Level One rookie DOES beat the odds and go on to a Hall of Fame career, it is almost always a catcher.   The only Level One rookies who went on to be Hall of Fame players were Mike Piazza, Rick Ferrell, Roy Campanella and Monte Irvin.    One other Level One Rookie, Edgar Martinez, is a good Hall of Fame candidate.   Other good players who reached the majors as Level One rookies include Jose Bautista, Bob Boone, Jay Buhner, Vinny Castilla, Ron Cey, Jeff Conine, Raul Ibanez, Ken Caminiti, Cecil Fielder, Aubrey Huff, Mike Lowell, Tino Martinez, Victor Martinez, Frank McCormick, Paul O’Neill, Jorge Posada, Hank Sauer, J.T. Snow, Tim Wallach, Cy Williams, Todd Zeile and Richie Zisk.  

              These are players who at first appeared to be nothing special, at least in the statistics; in some cases these players had impressive minor league records that we could have keyed on, but that’s not part of the data that I was working with.  These good players who emerged from the Level One Rookie group are less than 1% of the whole group, and almost all of them were classified Level One after a limited-at-bat season which was more a pre-rookie season than a true rookie season.   In 1971 and again in 1972, for example, both Ron Cey and Richie Zisk got a few at bats each season and were classified Level One, but remained rookies.   They had "true" rookie seasons in 1973, and for the 1973 season Cey was a Level Two rookie and Zisk was Level Three.  

              There were 94 Level One rookies in 2017.   By team, those were Juan Graterol and Nolan Fontana (Angels), Max Stassi and A. J. Reed (Astros), Bruce Maxwell, Jaycob Brugman, Joey Wendle, Jaff Decker and Ryan LaMarre (A’s), Raffy Lopez, Darrell Ceciliani, Mike Ohlman and Ian Parmley (Blue Jays), David Freitas (Braves), Manny Pina, Jesus Aguilar and Yadiel Rivera (Brewers), Luke Voit, Carson Kelly, Alex Mejia, Chad Huffman, Alberto Rosario and Breyvic Valera (Cardinals), Victor Caratini, Mike Freeman and Taylor Davis (Cubs), Christian Walker, Ildemaro Vargas and Jack Reinheimer (Diamondbacks), Rob Segedin, Kyle Farmer and O’Koyea Dickson (Dodgers), Chris Marrero, Carlos Moncrief, Jae-Gyun Hwang and Miguel Gomez (Giants), Guillermo Heredia, Dan Vogelbach, Jacob Hannemann, and Mike Marjama (Mariners), J. T. Riddle and Tomas Telis (Marlins), T. J. Rivera, Matt Reynods, Gavin Cecchini, Travis Taijeron, Philip Evans and Tomas Nido (Mets), Adrian Sanchez, Pedro Severino, Rafael Bautista and Raudy Read (Nationals), Francisco Pena and David Washington (Orioles), Carlos Asuaje, Jabari Blash, Dusty Coleman, Christian Villanueva and Rocky Gale (Padres), Andrew Knapp, Ty Kelly, Jorge Alfaro, Cam Perkins and Brock Stassi (Phillies), Elias Diaz, Jacob Stallings, Danny Ortiz and John Borman (Pirates), Brett Nicholas, Drew Robinson and A. J. Jimenez (Rangers), Cesar Puello (Rays), Steve Selsky (Red Sox), Patrick Kivlehan, Stuart Turner, Zach Vincej and Chad Wallach (Reds), Tom Murphy, Stephen Cardullo and Mike Tauchman (Rockies), Cam Gallagher (Royals), Dixon Machado and John Hicks (Tigers), Mitch Garver and Niko Goodrum (Tigers), Matt Davidson, Omar Narvaez, Kevan Smith, Rymer Liriano and Jacob May (White Sox), and Tyler Austin, Ji-Man Choi, Garrett Cooper and Kyle Higashioka (Yankees).   

              Of those 94 players, history suggests there might be one or two in there who have pretty good careers, but probably not more than that.    In that group of players, the three best seasons in 2017 were by Jesus Aguilar and Manny Pina (Brewers), and Matt Davidson (White Sox).   Aguilar hit .265 with 16 homers, 52 RBI in 279 At Bats.   Pina hit .279 with 9 homers, 43 RBI, and Davidson hit 26 homers.   But those are all older players who don’t run well, so I don’t see any of them as being likely to have a big future.  

              But if you know anything that contradicts that, please share it with us.   I mean. . .I know that Garrett Cooper is a big guy, of almost Judgian size, who hit .366 with Colorado Springs last year and .326 with the Yankees.   If you have reason to believe that any of these guys is better than I have represented him as being, please tell us about it.  

 

             

2.   First explanation of the method

              The system is set up so that there are 53% as many Level Two rookie as Level One, 53% as many Level Three rookies as Level Two, 53% as many Level Four rookies as Level Three. 

              The reason it is set up that way is that I made an a priori decision that I wanted to have 25 Level Ten Rookies in baseball history.   (Actually I initially made an a prior decision that I wanted to have just five (5) Level Ten Rookies, but after running the data I realized that was array didn’t really work, so I backed off and made it 25.)

              I wanted to have 25 rookie seasons at the highest level, and I wanted to have steadily diminishing numbers moving toward the highest level.    When I found that I had 15,907 rookie seasons to deal with, it then became a math puzzle:  what starting number and what diminishing ratio will give us 25 Level Ten seasons?    It turned out that if you start with 7,490 Level One rookies, and diminish that number by 47% with each step up, then you wind up with 15,907 total rookies and 25 Level Ten rookies—the combination that I was looking for.  

 

3.  The Top Rookies of 1900 to 1919

              In the Dead Ball Era (1900 to 1919) the highest-scoring rookies are Joe Jackson (1911), Sherry Magee (1904), Fred Snodgrass (1910), Josh Devore (1910) and Max Carey (1909).  

              By year, the top-scoring rookies are Sam Crawford (1900), Bill Coughlin (1901), Joe Tinker (1902), Jimmy Sebring (1903), Sherry Magee (1904), Ty Cobb (1905), Eddie Collins (1906), Tris Speaker (1907), Eddie Collins again (1908), Home Run Baker (1909), Fred Snodgrass (1910), Joe Jackson (1911), Ray Chapman (1912), Les Mann (1913), Benny Kauff (1914), Braggo Roth (1915), Sammy Bohne (1916), Ross Youngs (1917), Charlie Hollocher (1918) and Frankie Frisch (1919).    Nine of those 20 are in the Hall of Fame; would be ten if Joe Jackson were eligible. 

 

4.   Level Two Rookies of 2017

              There are 3,969 Level Two Rookies in history, of whom 19 are in the Hall of Fame; however, this is not 19 separate Hall of Famers.   Harmon Killebrew has three seasons as a Level Two Rookie (1954, 1955 and 1956).     The 19 seasons are by 13 different players: Dave Bancroft (1915), Joe Sewell (1920), Gabby Hartnett (1922-23), Bill Terry (1923-24), Hack Wilson (1923), Ernie Lombardi (1931), Harmon Killebrew (1954-55-56), Willie Stargell (1962-63), Tony Perez (1964-65), Carlton Fisk (1971), Jim Rice (1974), Cal Ripken (1981) and Wade Boggs (1982).  

              About 12% of Level Two rookies played 100 games or more, meaning that 88% of them did not.   Again, they’re mostly late-season callups, but they’re a little younger and faster than the Level One rookies, and they hit better in their late season callups. 

              Whereas less than 20% of Level One rookies will have 1,000 plate appearances in their careers, at least 44% of Level Two rookies will do so.   (It is impossible to know the exact percentage, because there are some players in the group who WILL have 1,000 PA but have not yet reached that level.)    Substantially less than 50% of Level Two rookies will have 1,000 career Plate Appearances.  Other good players who reached the majors as Level Two rookies include Rafael Palmeiro, Darrell Evans, Jim Thome, Graig Nettles, Fred McGriff, David Ortiz, Gary Gaetti, Manny Ramirez, Julio Franco, Ted Simmons, Paul Konerko, Steve Garvey, Todd Helton, John Olerud, Jason Giambi and Carlos Delgado.   If a level Two rookie does go on to become a star, it is almost certain that he will be a power hitter.  

              These are still long-odds guys.   Their shot at the Hall of Fame is less than 1%.    There were 67 Level Two rookies in 2017, who were Derek Fisher, Yulieski Gurriel, Tony Kemp and Colin Moran of the Astros, Matt Chapman, Chad Pinder, Boog Powell and Renato Nunez of the A’s, Teoscar Hernandez, Anthony Alford and Dwight Smith Jr. of the Blue Jays, Johan Camargo, Rio Ruiz, Lane Adams and Micah Johnson of the Braves, Lewis Brinson of the Brewers, Jose Martinez and Harrison Bader of the Cardinals, Mark Zaqunis of the Cubs, Rey Fuentes of the Diamondbacks, Austin Barnes, Alex Verdugo, Andrew Toles and Tim Locastro of the Dodgers, Ryder Jones, Christian Arroyo, Orlando Calixte and Austin Slater of the Giants, Erik Gonzalez, Yandy Diaz and Francisco Mejia of the Indians, Taylor Motter and Tyler Smith of the Mariners, Brian Anderson of the Marlins, Brandon Nimmo and Dominic Smith of the Mets, Andrew Stevenson of the Nationals, Austin Hays, Anthony Santander and Chance Sisco of the Orioles, Luis Orrens of the Padres, J. P. Crawford of the Phillies, Max Moroff, Jose Osuna, Gift Ngoepe, Jordan Luplow and Chris Bostick of the Pirates, Jared Hoying and Willie Calhoun of the Rangers, Marco Hernandez, Deven Marrero and Sam Travis of the Red Sox, Philip Ervin and Jesse Winkler of the Reds, Pat Valaika and Ryan McMahan of the Rockies, Jorge Bonifacio and Ramon Torres of the Royals, Jeimer Candelario and JaCoby Jones of the Tigers, Zack Granite of the Twins, Willy Garcia and Nick Delmonico of the White Sox, and Miguel Andujar, Dustin Fowler, Tyler Wade and Mason Williams of the Yankees.    These guys names will be more familiar to you than the previous list of 94, but they’re still not high-profile guys. 

              Four Level Two Rookies have been selected Rookie of the Year, three of those in seasons when there just really wasn’t anyone else.   Those four are Tommy Helms (1966), Lou Piniella (1969), Sandy Alomar Jr. (1990) and Eric Karros (1992).  

 

5.   Second Explanation of the Method

              There is a formula, and the formula has four parts.   The four elements of the formula are

              1)  The Season Score,

              2)  A hitting number based on the player’s OPS and Plate Appearances, designed to track OWAR but simpler,

              3)  A number based on the player’s age, and

              4)  A number based on the player’s speed score. 

             

 

6.  The Top Rookies of 1920 to 1939

 

              The highest-scoring rookies of 1920 to 1939 were Bob Meusel (1920), Sammy Bohne (1921), Pie Traynor (1922), George Grantham (1923), Kiki Cuyler (1924), Harry Rice (1925), Paul Waner (1926), Lloyd Waner (1927), Red Barnes (1928), Earl Averill (1929), Ben Chapman (1930), Joe Vosmik (1931), Arky Vaughan (1932), Joe Medwick (1933), Hal Trosky (1934), Phil Cavaretta (1935), Joe DiMaggio (1936), Hersh Martin (1937), Jeff Heath (1938) and Ted Williams (1939).    Nine of the 20 are Hall of Famers. 

              The top five high-scoring rookies of the 1920 to 1939 period are 1. Ted Williams, 2. Joe DiMaggio, 3. Paul Waner, 4. Jeff Heath, and 5. Hal Trosky. 

 

7.  Why are We Doing This?

              I have always searched for systematic ways of knowing who we should bet on, who we should invest in for the future.   The sensational 2017 season of Aaron Judge pushes this issue to the front.  

              Let us take. . .well, Walt Dropo and Ron Kittle.    Walt Dropo in 1950 had a superficially sensational season, hitting .322 with 34 homers, 144 RBI.   That’s a lot of RBI for a rookie.  Those are really fine triple crown stats, and Dropo was overwhelmingly selected as the American League’s rookie of the year.   But the thing is, he wasn’t REALLY that good, was he?   Dropo in 1950 was 27 years old, a first baseman, and slow.   Assessing his future, one can see some important negatives.

              Ron Kittle with the White Sox in 1983 hit .254, but with 35 homers, 100 RBI; the 35 homers, at the time, tied an American League record.   He won the Rookie of the Year Award, but. . .was he really that good?   Like Dropo, he was older, he was slow, and he struck out a lot.  

              You don’t IGNORE the fact that Dropo hit .322 with 144 RBI; you don’t ignore the fact that Kittle hit 35 homers.   If Dropo could simply have repeated his 1950 season a few times, he would be in the Hall of Fame.  

              Aaron Judge is a little bit like Kittle and Dropo, isn’t he?   He’s a big right-handed power hitter who strikes out a lot, runs pretty decent but he isn’t Byron Buxton.   How do we balance these things?   How much should we buy into Aaron Judge’s future?

              But it isn’t just Aaron Judge.   I was looking at the long, long list of rookies from 2017.  Which of these guys is REALLY good?   If I had a fantasy league team—which I don’t—but if I did, who should I value?    Who is hiding on that list, who was really good but not so obviously really good?  

              If we look back at 1962, in 1962 the top three finishers in the American League Rookie of the Year voting were Tom Tresh, Bob Rodgers and Bernie Allen.   In the National League only two players drew votes, those being Ken Hubbs and Donn Clendenon.    But if we look back at the 1962 season, Lou Brock was a rookie that year, and Willie Stargell and Boog Powell and Jim Fregosi.    I am trying to look at those players based on the information that was available in 1962—but using just that information, to focus on Brock and Stargell and Powell, rather than Tom Tresh and Bernie Allen.    

              Yes, of course we have scouting reports; we have our personal observations of these players.   But that’s disorganized information.   I am trying to create an organized, systematic approach to the problem. 

 

8.  The Level Three Rookies

              On the third level of rookies there are 2,103 rookies in history, and 19 rookies in 2017.   These are the players you would have heard of, most of them (not all of them), but these are the guys who did some good work and are regarded as possibly having a future.    Josh Bell and Paul DeJong and Albert Almora.  

              Of the 2,103 Level Three rookies, 32 are in the Hall of Fame, and some more will be.    The 32 rookie seasons by Hall of Famers are by 24 different players.   There are additional seasons in the group by Chipper Jones, David Ortiz, Miguel Cabrera and Jim Thome, as well as other Hall of Fame candidates like Fred McGriff, Omar Vizquel, Al Oliver, Steve Garvey, Miguel Tejada, Ted Simmons, Larry Walker, Dale Murphy and Keith Hernandez.   A Level Three Rookie has about a 2% chance to be a Hall of Famer, perhaps a little better than 2%. 

              A player CAN rank as a Level Three rookie (or higher) just based on youth and speed.   He doesn’t HAVE to be a regular player; he doesn’t have to play tremendously well as a rookie.   18% of Level Three rookies play 100 games or more, as opposed to 12% of Level Two rookies and 3½ % of Level One rookies.    About 63% of these players will have 1000 or more plate appearances in their careers.    About 50% of them will have 3000 or more plate appearances. 

              45 MVP Awards have been won by players who were at one point Level Three rookies.

              A lot of players are Level One Rookies for several years as they bounce up and down between the majors and minors.   That does not tend to happen, to the same extent, with Level Three Rookies.   These are better players, in general; they come up, and they stick.    But we are still not talking here about the serious Rookie of the Year candidates, in most seasons, or about players who would often go on to be All Stars. 

              The 19 Level Three Rookies of 2017 are Matt Olson of the A’s, Richard Urena of the Blue Jays, Dansby Swanson of the Braves, Brett Phillips of the Brewers, Paul DeJong and Magneuris Sierra of the Cardinals, Albert Almora Jr. of the Cubs, Greg Allen of the Indians, Mitch Haniger of the Mariners, Brian Goodwin and Victor Robles of the Nationals, Trey Mancini of the Orioles, Hunter Renfroe of the Padres, Rhys Hoskins and Nick Williams of the Phillies, Josh Bell of the Pirates, Tzu-Wei Lin of the Red Sox, Terrance Gore of the Royals, and Clint Frazier of the Yankees. 

              Fourteen Level Three Rookies have won the Rookie of the Year Award. 

 

9.   Let’s talk about some of the Flaws in the System

              Well, no; it’s too early for that.   I’ll do that later.  But Greg Allen and Terrance Gore shouldn’t be on the list above; those are flaws in the system, that I will acknowledge now and discuss later. 

 

10.  Third Explanation of the Method

 

              2)  A hitting number based on the player’s OPS and Plate Appearances, designed to track OWAR but simpler. 

              The formula actually is OPS minus .580, squared, times Plate Appearances, times 3, and never less than zero. The Season Score formula has a similar element. . . .one of the elements of the Season Score formula is based on OPS above a certain margin as well.   But this element is different because (a) the Season Score element is based on a hitter’s straight-line distance above the margin, whereas this is a hyperbolic measurement, and (b) the Season Score leavens those elements with playing time elements (runs scored, RBI, etc.), whereas this does not.  

              This is designed to give an advantage to. . .well, Willie McCovey in 1959, or Rhys Hoskins last year.   A player who hits really well in limited at bats, but not so limited as to be easily dismissed.    In Season Scores Rhys Hoskins and his teammate Nick Williams are almost even.   Hoskins had 37 runs scored, 44 RBI, 44 hits, 105 Total Bases; Williams had twice as many hits, more runs scored, RBI, and total bases (44, 55 and 148).  In season score they are almost even (141-139, Hoskins), but Hoskins leads in this calculation 120 to 55.    That’s why it is there; to put players like Hoskins ahead of players like Williams.  

 

11.  Technical Note

              The definition of who was a rookie in what season has changed several times over the years.   A friend of mine wrote wrecently that Rip Sewell pitched so little in 1938 that he was still considered a rookie in 1939.    That’s not EXACTLY true, of course; there was no such thing as being "considered" a rookie in 1939.   The concept of a rookie barely existed in 1939; one sportswriter might have considered Sewell a rookie, and another might not have, and another might have said "A rookie?  What does that mean?"   The term was not widely used, and, when used, was often spelled "rooky".   There was no rookie of the year award, no rules relevant to the subject. 

              Once there was an award and there were rules for who was eligible, the rules were perhaps not well thought out, and perhaps subject to vagaries of the schedule.   The rules intended to allow a player to be considered a rookie if he played a few games after a late-season callup, but the schedule expanded from 154 games to 162 and the season pushed longer into September, so the first set of rules didn’t hold up over time, and (I think) the second set of rules didn’t quite work, either.   Several players who were not considered rookies THEN might be considered rookies NOW. 

              Al Rosen had 58 at bats prior to the 1950 season, but was not considered a rookie in 1950.   Vada Pinson, the top rookie of 1959 and one of the greatest rookies of all time in 1959, was not considered a rookie at the time, 96 prior at bats.    Reggie Jackson had 118 at bats prior to his 1968 rookie season, so was not considered a rookie at the time.   Some of these exclusions may have been based on days on the major league roster rather than at bats; I’m not sure, and I don’t care.

              From the standpoint of this research it is not in any way helpful to worry about what the rookie eligibility rules were at the time.     That has nothing to do with what we’re talking about.   For purposes of this study I’ll consider Rosen, 1950, Pinson, 1959, Reggie, 1968 and all others like them to have been rookies.  

              Also. . .even smaller technical note, but I should explain.   For the Federal League players (1914-1915) I divided their scores in half, because it isn’t REALLY a major league; it’s a good minor league which is trying to become a major league.   The numbers are not to be trusted.

 

12.  The Rookies of 1940 to 1959

              The Rookie of the Year Award started in 1947.    From 1940 to 1946, which are mostly war-time seasons, the number one rookies by this method are Dom DiMaggio (1940), Phil Rizzuto (1941), Stan Musial (1942), Lou Klein (1943), Buddy Kerr (1944), Red Schoendienst (1945), and Del Ennis (1946).  

              Beginning in 1947 we have Rookie of the Year Awards to compare our results to.   In 1947 there was only one Rookie of the Year Award for the two leagues, which was won by Jackie Robinson, and Jackie is also our formula’s selection as the #1 rookie.    He was 28 years old and his score (915, Level 8) is lower than it would have been if he was younger, but it was still the best of 1947.

              In 1948 the Rookie of the Year was Alvin Dark, but in our system Dark ranks 7th among the rookies, behind (1) Richie Ashburn, (2) Larry Doby, (3) Dick Kokos, (4) Duke Snider, (5) Herb Adams, and (6) Hal Jeffcoat. 

              This is more a victory for our system than a failure; Hall of Famers Ashburn and Doby have moved to the front of the list, and another Hall of Famer (Duke Snider) has also moved ahead of Dark.   It is not an unmixed record.   Dark has also been passed by Dick Kokos, Herb Adams and Hal Jeffcoat, none of whom had a career as good as Dark’s, or even really comparable.  

              In 1949 the Rookies of the Year were Roy Sievers and a pitcher; Sievers has the highest score, so nothing noteworthy there.   In 1950 the Rookies of the Year were Walt Dropo and Sam Jethroe, a 32-year-old player coming over from the Negro Leagues.    By our mathematical system the highest-rated rookie was Gus Bell of Pittsburgh—a point for the system, since Bell went on to a better career than either Jethroe or Dropo.   (Five players split the vote in the National League.   Bell was not mentioned.) 

              In 1951 Willie Mays and Mickey Mantle both reached the major leagues, and also Minnie Minoso had an impressive rookie campaign, hitting .326.   The American League Rookie of the Year Award went to Gil McDougald, the National League to Mays.   These are the top rookies in the majors as seen by the formulas:

              1.  Willie Mays

              2.  Mickey Mantle

              3.  Minnie Minoso

              4.  Gil McDougald

             

              Mays and Mantle not only had the highest scores in the majors in 1951, but actually the two of them had the two highest scores that any rookie had had since Stan Musial had arrived on the scene nine years earlier, and also higher scores than anyone else would have until 1956.   So we have to count that as a complete success.   The system in 1951 completes its assignment of correctly recognizing the future superstars as soon as they arrived in the majors.  

              1952, not so much.   In 1952 the Rookies of the Year in both leagues were pitchers.   The best player in the rookie cadre, long-term, was Eddie Mathews, and our system does rank Mathews first among rookies with 100 or more plate appearances—but ranks Mathews behind Gene Stephens, a 19-year-old for the Boston Red Sox who played only 21 games, batting just 53 times.  

              Our formula ranks Stephens as the number one rookie of 1952, based on his youth and his speed; he was very fast.    He picks up so many points for being young and fast that he ranks as the #1 rookie of the 1952 group.   This is a poor selection. 

              It would be easy to say that the system should not name a player as a top rookie based simply on youth and speed, without any real input from performance numbers; it would be easy to say that, but I am not sure that it is true.   The nomination of Stephens as a valuable rookie turns out to be wrong, but all systems make mistakes; math is inherently stupid.   In 1953 the system makes Al Kaline, who had only 28 at bats and 7 hits, as one of the top rookies of 1953; it does that for the same reasons that it assigns value to Stephens.   In that case, it is not wrong to assign value based on age and speed, it is correct.   The question is, is it better to ignore Stephens AND Kaline, or better to assign value to both of them, and be right one time and wrong one time?

              I kind of think it is better to assign them value.   If you think about it from the standpoint of a fantasy league draft, do you want to get a notice that "this kid might be really good" about a kid who DOESN’T turn out to be really good, or do you want to overlook the young Al Kaline?

              I think you want to know.   I think you want him to pop to the top of the list.    If you overlook him, somebody else drafts him before you know he is there.   If his name is there, you have a fair chance to evaluate him by other information. 

              Or if you are talking about a REAL team, the same problem:  do you want to hear about him early and maybe unnecessarily, or do you want to let him go?    I think you have to want to hear about him.   If you focus on him, you can have your scouts look at him and decide what value you place on him.   If you’re negotiating a deal with the 1953 Detroit Tigers, you want to know that Kaline is there, that he is a name you can bring up.  

              Anyway. . . . in 1953 the Rookies of the Year are Jim Gilliam and Harvey Kuenn.    Our system says that Jim Gilliam is the top rookie candidate, so there is no conflict there (although the system does not like Harvey Kuenn very much.)    In 1954 the Rookies of the Year were Wally Moon and a pitcher, but our system says that the number one rookie is Henry Aaron.   Moon hit .304 with 193 hits, but was 24 years old.   Aaron hit .280 with 131 hits, but was four years younger.   Big win for the system. 

              In 1955 the Rookies of the Year were a pitcher and Bill Virdon, but our system says that the best rookie was Roberto Clemente, a younger player who didn’t hit anything much.    Again, big win for the system. 

              In 1956 Frank Robinson, only 20 years old, hit 38 home runs; Robinson is the highest-scoring rookie since Ted Williams in 1939.   The American League Rookie of the Year was Luis Aparicio.   The system says that the three top rookies are 1. Robinson, 2. Curt Flood, and 3. Aparicio—although Flood had only 1 at bat.    He’s another youth-and-speed selection, but the system is more right than wrong, since Flood did go on to have an outstanding career.

              In 1957 the Rookies of the Year were Tony Kubek and a pitcher.   Our system says that the 1-2 rookies are Flood (who still had only 3 at bats) and Kubek, so. . .no real information there.    In 1958 the Rookies of the Year were Orlando Cepeda and Albie Pearson.   Our system loves Cepeda but dismisses Pearson.   In 1959 Vada Pinson, not recognized as a rookie at the time, had one of the greatest rookie seasons of all time. 

 

 

13.  The Level Four Rookies

         There are 1,115 Level Four Rookie seasons in history.    Whereas a Level 3 Rookie has about a 2% chance to be a Hall of Famer, a Level 4 Rookie has a four to five percent chance.     Of the 1,115 seasons 40 are by Hall of Famers, and those are by 34 different players.   In addition to those, the group includes Hall of Fame candidates Gary Sheffield, Manny Ramirez, Sammy Sosa, Dave Parker, Vladimir Guerrero, Dwight Evans, Bobby Abreu, Al Oliver, Lou Whitaker, Joe Carter, Buddy Bell, Robinson Cano, Matt Holliday, Dave Concepcion, Mark Teixeira and still others.   One has to assume that some of those players are going to make it to Cooperstown. 

              There were nine Level Four Rookies in 2017:   Franklin Barreto of the Athletics, Ben Gamel of the Mariners, Amed Rosario of the Mets, Franchy Cordero of the Padres, Daniel Robertson of the Rays, Rafael Devers of the Red Sox, Raimel Tapia of the Rockies, and Adam Engel and Alen Hanson of the White Sox.    These are mostly players in their early 20s, and for the most part they are either very fast or have some proven ability as hitters.   About half of these players will emerge as long-term regulars.

 

14.   Fourth Explanation of the Method

              There are two elements to the mathematical system that I haven’t explained yet, which are the age and speed numbers.   The player gets "youth points" which are

              1)  32 minus the player’s age,

              2)  Squared,

              3)  Times 2.5,

              4)  But not to exceed 400, and

              5)  Obviously if a player is older than 32, this number is zero. 

 

              So if a rookie is 31 years old, he gets 2.5 points under this rule.   If he is 30 he gets 10 points; age 29, he gets 22.5 points, and at 28, he gets 40 points.   A 25-year-old rookie gets 122.5; a 22-year-old rookie gets 250.  

              The "speed points" work this way.   I score the player’s speed on a 1 to 10 scale, 10 being the fastest runners, and 1 being the slowest runners.   The speed points are the player’s speed score squared, times 5, so a "10" runner gets 500 points here, and a "5" runner—an average runner—gets 125.  

              I work with a spreadsheet that I have built up over the years called NPE 2017 (Non Pitchers Encyclopedia, 2017); it has batting stats for each player each season, and also it has a "Speed Rating" for each player in each season, on a 1 to 10 scale.   These speed ratings which are in the system are based on formulas but also on judgment.   At some point in history—it may have been 10 years ago, I don’t know—I "scored" each players speed each season AND IN HIS CAREER based on his stolen bases, caught stealing, triples, defensive position, and runs scored as a percentage of times on base, so that I have a speed score for each player each season.   But then I "made sense" of them by adjusting out-of-line and out-of-whack numbers.   A runner can’t run 4, 5, 4, 3, 10, 3, 3 in a series of seasons; that just doesn’t make sense.  It is obviously some sort of statistical fluke.  I set up a system to spot the out-of-line numbers, and then just manually corrected them so that they make sense.  

              In this study this creates a theoretical problem or a theoretical bad-science problem, because I could be, in theory, using data which was actually unknown at the time to evaluate the player as a rookie, and also because we have a subjective element in what is otherwise an objective system.  In other words, if a rookie came up and stole no bases in 25 games and grounded into 6 double plays, we might evaluate him as being slow, but if he then stole 40 bases two years later we would know that he WASN’T slow, so I may have corrected the rookie speed rating based on information that became known later.    This creates a bad-science problem, because what we are doing in this effort is trying to learn to evaluate rookies based on that information which is known at the time, but I am stealing information from the future here to accomplish my goal.  

              But I don’t think this is an ACTUAL problem, as opposed to a theoretical problem, for this reason.   In the modern world, we have very good information about how fast a player runs as soon as he gets to the major leagues.   The real issue is whether our method is going to work in the future.  In the past, we had to do what we had to do to evaluate player’s speed.   We no longer have to do those things.    It’s a data-contamination issue.  We might NOT have spotted Henry Aaron in 1954, because (since he stole only 2 bases as a rookie) we might have misjudged his speed.  The ACTUAL problem is that we could do a better job of retro-evaluating rookies if we had more accurate speed scores.   But going forward, it’s not really an issue. 

 

 

15.   Let’s talk about some of the Flaws in the System

              But there is a problem related to that issue, which is this.    Our system shows Greg Allen and Terrance Gore as Level Three rookies in 2017, which isn’t right; they’re Level One or Level Two guys, and we have had similar problems in the past with guys like Gene Stephens.   That problem has to do with combining youth and speed.   If a player gets very high scores in both of those areas, then he winds up with a high score, even if you or I could see that he’s just a "legs" guy who is in the majors as a pinch runner, and he probably doesn’t have a brilliant future.  

              That’s the "A" problem—the system sometimes sees great value in young, fast players who don’t really have great value—and there is a "Z" problem as well, which is that the system sometimes does NOT see the value that a young player has.   The obvious cases here are Miguel Cabrera and Rafael Devers.   In 2003 Miguel Cabrera was 20 years old.   He played 87 games for what were then the Florida Marlins, hitting .262 with 12 homers, 62 RBI.   The system sees him as being a Level Three rookie.   It’s obviously not right.   In 2003 I quite certainly knew that Miguel Cabrera was something special, and I am guessing that many of you did, as well.    The system says, "Well, he’s young, but he’s not a tremendous hitter and he is slow, so I’m not going to get excited about him."  The system does not know what you and I do know about Cabrera, or about Rafael Devers in 2017.   These guys shouldn’t be "threes" or "fours"; they should be sixes or sevens or eights.     Another one of those is Johnny Bench in 1968.   Our system sees him as being a "4", when in reality he should be an 8 or a 9.   We have difficulty recognizing exceptional ability in young players who are not fast.  

              I am trying to build an objective system which gets the right answers 100% of the time.   This system doesn’t get the right answer 100% of the time, or even 90% of the time.   It gets the answer that we want maybe 60, 65% of the time.   It does better than the Rookie of the Year voting does.  The system could never be improved so that it was 100% accurate, but it could certainly be refined and developed so that it was significantly more accurate than it is now.   It could be refined so that it recognizes that Rafael Devers is a much higher level rookie than Terrance Gore.     

              If I were, let us say, working for a University on a government grant, or if I were a 23-year-old intern working for the Milwaukee Brewers and I was assigned by my superiors to solve THIS problem, then I would work on this problem for a year, and I would emerge from that work with a much stronger version of this system.    On the other hand, it would have taken me a year to do that.

              That would be certainly be a worthwhile use of a year out of someone’s life, I think, but I’m just not like that.   I’ve never been that way.   I work on a problem for a week, two weeks, a couple of days, and then I write up what I have and go on to some other problem.   In all likelihood I will circle back to the same problem again in a couple of years, and probably I can make it work a little better then than I can now.  

              We might say that there are pioneers, and there are practitioners.  Practitioners are people who are patient, careful, meticulous; they address all of these nagging little issues BEFORE they publish anything about a subject.   Your college professors teach you to be good practitioners.  

              Many people get frustrated with me because I am not a good practitioner.   I have taken a tremendous amount of criticism in my career for not being a good practitioner of the analytical arts, and many people—many of my readers—get very ANGRY with me because I not a great practitioner.   People want to tell me how I SHOULD be doing these studies, goddammit.   Some of you on this site bitch at me for my failings along this line, and others of you pick, pick, pick at me for not doing this and not doing that and not doing some other thing.   I would tell you that you know who you are, but you don’t; you don’t have a clue.

              But you have to understand; I’m not essentially a practitioner.   I’m essentially a pioneer.   I’m an explorer.   I explore an idea for a week, and then I move on to some other idea.   Nobody taught me how to do this stuff; I just figured it out, made up my own rules as I went along.    If I didn’t do it the way that you would have done it, that is because I did it and you didn’t.  

 

16.   The Rookies of the 1960s

              The top rookies of the 1960s, by year, are Willie Davis, 1960 (Level 7), Willie Davis, 1961 (Level 8), Lou Brock, 1962 (Level 7), Joe Morgan, 1963 (Level 7), Tony Oliva, 1964 (Level 10), Joe Morgan, 1965 (Level 9), Tommie Agee, 1966 (Level 9), Reggie Smith, 1967 (Level 8), Reggie Jackson, 1969 (Level 9), and Larry Hisle, 1969 (Level 9).   There were two Level-10 Rookies during the 1960s, both from 1964--the two 1964 Rookies of the Year.

              In this article I am using Oliva’s age as it was listed at that time; at that time he was believed to be 23 years old in 1964.   It is now known that that was not Oliva’s actual age; he had come to the United States on his younger brother’s passport or something, and he was actually a couple of years older than that.   I guess I should correct my spreadsheet on that account. .  ..I’ll go do that now.

              Anyway, comparing our 1960s list to the 1960s Rookies of the Year.   In 1960 we are not clearly either right or wrong.   The 1960 Rookies of the Year were Ron Hansen and Frank Howard, whereas our system likes Willie Davis and Zoilo Versalles more.    It’s kind of a wash; Davis had about as good a career as Howard, and Versalles about as good a career as Hansen.     In ’61 our system likes Willie Davis again, whereas the Rookie of the Year voters preferred Billy Williams, and the voters were right about that one. 

              1962 I wrote about earlier; it’s a big win for our system, that it recognizes Lou Brock as the most valuable rookie property, rather than Ken Hubbs or Tom Tresh.    In 1963 the Rookies of the Year were Pete Rose and a pitcher.   Our system likes Joe Morgan better; Morgan played only 8 games, and obviously a player who plays only 8 games is not a Rookie of the Year candidate, but he had youth and speed.    In 1964 our system picks the same rookies as the Rookie of the Year vote. 

              In 1965 the top candidates for the National League Rookie of the Year award were two second basemen, Jim Lefebvre and Joe Morgan.   The voters chose Jim Lefebvre.   This was just stupid.   This was one of the worst award votes of all time.   Morgan simply PLAYED much better than Jim Lefebvre did; he was a vastly better all-around player.    Lefebvre won the vote because (1) he played in Los Angeles and (2) the Dodgers won the pennant.   It’s an indefensible vote.   I was 15 years old at the time, and I knew it was a stupid vote then. 

              In 1966 the Rookies of the Year were Tommie Agee and Tommy Helms; our system sees Agee as a Level 9 rookie, the best in the major leagues, and Helms as a Level 2 rookie, one of the worst Rookies of the Year ever.    In 1967 the top two rookie position players were both in the American League, Reggie Smith and Rod Carew.   Our system liked Reggie better; the voters like Carew better, and since he is in the Hall of Fame, we have to give that one to them, although you can’t go wrong with either player.

              1968 is more a failure for our system than a success.   The 1968 Rookies of the Year were Johnny Bench and a pitcher.   Our system doesn’t much like Bench—a failure for us—and picks Reggie Jackson as the best rookie of the season, which is OK because Reggie was a great player although he was not considered a rookie at that time.  

              In 1969 the Rookies of the Year were Ted Sizemore and Lou Piniella.  Our system prefers Larry Hisle and Bobby Murcer (who, again, was not considered a rookie at the time.) 

 

17.  The Level Five Rookies

              There are 591 Level 5 rookies in history.   Let’s organize the data for you.   In history we have 7,490 Level One rookies or common rookies, 3,970 Level Two rookies, 2,103 Level Three rookies, 1,115 on Level 4, and 591 on Level 5.  

              A Level One Rookie has about a one in one thousand chance of developing into a Hall of Famer.  A Level Two Rookie has a little less than 1% chance, a Level Three Rookie about a 2% chance, and a Level Four Rookie has a 4 to 5% chance of being a Hall of Fame Player.   A Level Five Rookie has about an 7% chance to be a Hall of Famer.   30 players who were Level 5 rookies are in the Hall of Fame, and Derek Jeter and Alex Rodriguez were both Level 5 rookies as late-season callups, in the years before they were in the lineup.  

              About 3 ½ percent of Level One rookies are players who played in 100 games or more, as opposed to 12% of Level Two Rookies, 18% of Level Three Rookies, 28% of Level 4, and 33% of Level 5.  46% of Level 6 rookies played in 100 games or more, 55% of Level 7, 83% of Level 8, 88% of Level 9, and 100% of Level 10 rookies are players who played in 100 or more games.

              About 18% of Level One Rookies will eventually get to 1,000 plate appearances in the majors, as opposed to 44% of Level Two Rookies, 63% of Level 3, 78% of Level 4, and 83% of Level 5.    90% of Level 6 rookies will get at least 1,000 plate appearances in the major leagues, 97% of Level 7 rookies, 98% of Level 8 Rookies, and 100% of Level 9 and Level 10 rookies.

              About 6 or 7% of Level One Rookies will get 3,000 plate appearances in the majors, as opposed to about 15% of Level Two Rookies, 34% of Level Three Rookies, 48% of Level Four Rookies, 55% of Level Five Rookies, 65% of Level 6, 74% of Level 7, 85% of Level 8, 91% of Level 9, and probably 100% of Level 10.   (There is a Level 10 rookie from 2017, and he is the only one of the group who has not yet made it to 3,000 career plate appearances.)

              Talking about a Level 5 rookie, then, we’re talking about a player who has about a 55% chance to get 3,000 career plate appearances, an 83% chance to get 1,000 plate appearances, and a 7% chance to be a Hall of Famer.  

              There were three Level Five rookies in the majors in 2017, who were Andrew Benintendi of the Red Sox, Wilmer Difo of the Nationals, and Yoan Moncada of the White Sox.  

 

18.  The Rookies of the 1970s

         The outstanding rookies of the 1970s, as selected by our system, were Cesar Cedeno (1970), Ralph Garr (1971), Garry Maddox (1972), Rich Coggins (1973), Robin Yount (1974), Fred Lynn (1975), Garry Templeton (1976), Mitchell Page (1977), Paul Molitor (1978) and Rickey Henderson (1979).   Interestingly enough, in the 1970s we have an almost complete difference of opinion from the voters as to who the best rookies were.   The only one of these players who won the Rookie of the Year Award was Fred Lynn, who was also the MVP.  

              Our system is completely wrong in 1977, and less dramatically wrong in 1973.   In 1977 our system picked Mitchell Page as the outstanding rookie in the majors over Eddie Murray (American League Rookie of the Year) and Andre Dawson (National League Rookie of the Year).   Obviously the voters were right, and my mathematical system is wrong. 

              1973 is sort of the same, not as much; our system picks Rich Coggins over Al Bumbry and Gary Mathews, who both turned out to be better players than Coggins.   Other than those two seasons, however, the system wins big through the 1970s, doing far better at picking winners than the award voters.   Understanding, of course, that the award voters were not TRYING to speculate about what a player would do in the future; they’re just acknowledging what has happened.

 

19.  The Level Six Rookies

              There are 313 Level Six Rookies in Baseball History, including three from 2017.   The three from 2017 are Ian Happ of the Cubs, Allen Cordoba of the Padres, and Bradley Zimmer of the Indians.   Not sure that I would agree with Bradley Zimmer ranking higher than Benintendi, but the system likes him because he can fly.    A Level 6 Rookie has about a 9% chance to develop into a Hall of Fame player.   22 Players who were Level 6 Rookies are in the Hall of Fame, and others who are candidates or comparable players include Pete Rose, Giancarlo Stanton, Mookie Betts, Carlos Beltran and Larry Walker.  

              On Benintendi and Zimmer. . .the system gives Benintendi a 160-point advantage because he is two years younger than Zimmer, but gives Zimmer a 255-point advantage because we scored Zimmer’s speed at "10" and Benintendi at "7".    But you know; it’s a little bit arbitrary.    If we scored Benintendi’s speed at "8" and Zimmer at "9", it might be more accurate, and it might give us a different ranking.   I don’t know that it is possible to avoid those kind of wrinkles.   My apologies to the Indians fans in the group.  

 

20.   The Rookies of the 1980s

              The top rookies of the 1980s, as selected by the system, are Leon Durham (1980), Tim Raines (1981), Ryne Sandberg (1982), Darryl Strawberry (1983), Juan Samuel (1984), Oddibe McDowell (1985), Barry Bonds (1986), Dave Martinez (1987), Roberto Alomar (1988) and Ken Griffey Jr. (1989).  

              Again, only one player was picked both by this mathematical system, and by the Rookie of the Year voters—Darryl Strawberry in 1983.   But whereas the mathematical system picks five Hall of Famers in 10 years (Raines, Sandberg, Bonds, Alomar and Griffey), the Rookie of the Year voters picked 20 players, and only one future Hall of Famer (Cal Ripken).     Our system does completely whiff on Ripken.   It lists his as the #7 rookie of the year, behind Sandberg, Sax (NL Rookie of the Year), Willie McGee, Von Hayes, David Green and Chili Davis. 

 

21.  The Level Seven Rookies

              There are 166 Level Seven Rookies in baseball history.   A Level 7 Rookie, historically, has about a 15% chance to be a Hall of Famer.

              There were two Level 7 rookies in the major leagues in 2017:  Manuel Margot of the Padres, and Ozzie Albies of the Braves.  

              A Level 7 Rookie is most often a strong Rookie of the Year contender if he gets enough playing time, and generally, most of the time, is headed toward a pretty good career.   Recent Level 7 rookies who completely bombed include Cameron Maybin (2007), Chris Coghlin (2009), Everth Cabrera (2009), Jose Tabata (2011) and Brett Lawrie (2011).    Coghlin in 2009 was chosen the National League Rookie of the Year over Andrew McCutchen (although not in our system.   In our system, Coghlin was a Level 7 rookie, McCutchen a Level 9.)  But other Level 7 rookies of the last twelve years include Ryan Zimmerman and Matt Kemp (2006), Jacoby Ellsbury (2008), Starlin Castro (2010), Kris Bryant, Carlos Correa and Eddie Rosario (2015) and Trevor Story and Byron Buxton (2016).   

 

22.  The Rookies of the 1990s

              In the early 1990s, for whatever reason, our system goes through a five-year slump in which it does a very poor job of spotting the best young players coming to the majors.   The best young players who came to the majors in that era were all power hitters—Frank Thomas, Chipper Jones, Mike Piazza, Jim Thome, Manny Ramirez, Jeff Bagwell, Juan Gonzalez; those, and Ivan Rodriguez, who is not most notable as a power hitter.  Our system struggles to identify the best young power hitters.  

              From 1990 to 1994 the players seen by our system as the best players in each group are Delino DeShields (1990), Chuck Knoblauch (1991), Reggie Sanders (1992), Tim Salmon (1993), and Raul Mondesi (1994).  There are better players that the system fails to find, and then there is A-Rod, who was in the majors as an 18- and 19-year-old, but (a) did not impress with the bat, and (b) was not exceptionally fast, consequently scores as a good but not great prospect.  In the late 1990s it does a little better, picking as the rookie to watch Johnny Damon (1995), Derek Jeter (1996), Nomar Garciaparra (1997), Mike Caruso (1998) and Carlos Beltran (1999).  

              The Award voters don’t do a lot better; they picked a lot of guys like Pat Listach, Bob Hamelin and Marty Cordova.  Knoblauch, Salmon and Mondesi were also Rookies of the Year. 

 

23.  The Level Eight Rookies

              There are 88 Level Eight Rookie Seasons in baseball history.   There were none in 2017.    Historically, about 23% of Level Eight Rookies will eventually wind up in the Hall of Fame. 

              I told you earlier that 98% of Level Eight Rookies will have at least 1,000 career plate appearances in the major leagues.   The only two Level 8 rookies who failed to make it to 1,000 career plate appearances were Al Gionfriddo (1945) and Red Barnes (1928).    Barnes was kind of a Lonnie Smith type guy.   I don’t know if you know this, but the reason that Smith had problems in the outfield was that he had very small hands and feet, for an athlete.   He fell down in the outfield—a lot—because, if you have small feet, you have a much smaller balance area; you have less tolerance for being off balance and maintaining your balance, and the ball came out of his hand when he threw sometimes because he had small hands, and it was harder for him to grip the ball.   Barnes was the same way; he was a good player and he made a lot of what were called circus catches in the outfield, but he had small hands and small feet, which caused him problems.  

 

24.   The Best Rookies of the 2000s

              The highest-scoring rookies of 2000 to 2009 were Rafael Furcal (2000), Albert Pujols (2001), Carl Crawford (2002), Rocco Baldelli (2003), Melvin Upton Jr. (2004), Rickie Weeks (2005), Hanley Ramirez (2006), Ryan Braun (2007), Jacoby Ellsbury (2008) and Andrew McCutchen (2009). 

              Rickie Weeks was the highest-scoring rookie of 2005, but with a score of 806, whereas there were three rookies who scored above 1,000 in 2001 (Pujols, Suzuki and Jimmie Rollins) and two who scored above 1,000 in 2007 (Braun and Tulowitski).  

              On a head-to-head basis, the system is outperforming the Rookie of the Year vote.  Furcal and Pujols are picked by both methods, as were Ryan Braun and Hanley Ramirez.  In 2002 the Rookies of the Year were a pitcher and Eric Hinske; our system picks Carl Crawford, who was not a great player, but better than Hinske.   Picking McCutchen rather than Chris Coghlan is a significant advantage.   In 2003 our system picks Rocco Baldelli; the voters picked Angel Berroa and Dontrelle Willis, which is kind of a wash, in that all three player’s careers sank like stones.

25.   The Level Nine Rookies

              There are 47 Level Nine Rookie Seasons in baseball history.    About 30% of these players are in the Hall of Fame, or are nearly certain to get there. 

              There was one Level 9 rookie in 2017, that being Cody Bellinger. 

              Four Level Nine Rookies did not make it to 3,000 career plate appearances:  Billy Grabarkewitz, Mitchell Page, Rocco Baldelli and Mike Caruso.  

 

26.  Recent Seasons

              Since 2010 the highest scoring rookies have been Jason Heyward (2010), Brett Lawrie (2011), Mike Trout (2012), Yasiel Puig (2013), Danny Santana (2014), Francisco Lindor (2015) and Trea Turner (2016).  

 

27.  The Level 10 Seasons

         OK, initially I was going to have five seasons at the highest level.   To score at this level requires virtually a perfect combination of elements.   You have to be young and fast and a good hitter; at the least you have to have an impressive combination of these attributes.   The five highest-scoring seasons ever, who thus would have been the five Level 10 Seasons, are Joe Jackson, 1911, Joe DiMaggio, 1936, Ted Williams, 1939, Vada Pinson, 1959, and Mike Trout, 2012.  

              But I didn’t go with 5, but 25.   Aaron Judge is the first name to be added to this list since 2012, when there were two super-rookies. 

              Chronologically, the 25 Level 10 Rookies are Joe Jackson (1911), Kiki Cuyler (1924), Paul Waner (1926), Ben Chapman (1930), Hal Trosky (1934), Joe DiMaggio (1936), Buddy Lewis (1936), Jeff Heath (1938), Ted Williams (1939), Barney McCosky (1939), Willie Mays (1951), Frank Robinson (1956), Vada Pinson (1959), Dick Allen (1964), Tony Oliva (1964), Fred Lynn (1975), Nomar Garciaparra (1997), Carlos Beltran (1999), Albert Pujols (2001), Hanley Ramirez (2006), Ryan Braun (2007), Jason Heyward (2010), Mike Trout (2012), Bryce Harper (2012) and Aaron Judge.  

              Of the first 24 Super Rookies, only six are now in the Hall of Fame (Cuyler, Waner, DiMaggio, Williams, Mays and Frank Robinson).   Two others are almost certain to get to the Hall of Fame, Pujols and Trout, so that’s 8.    Joe Jackson was a Hall of Fame caliber player who is banned from being selected; that’s 9.  Five others had careers which were good enough that they COULD be in the Hall of Fame, and may be selected sometime; those are Pinson, Dick Allen, Oliva, Lynn and Beltran.    That brings our count to 14, so 14 of the other 24 players had what you could plausibly describe as Hall of Fame type careers.  Trosky had a health issue that kept him from reaching his full potential; that’s 15.

              Two players, Buddy Lewis and Barney McCosky, lost prime seasons and had their careers strongly negatively impacted by World War II; that gives us 17.   Ben Chapman, Jeff Heath and Nomar Garciaparra were great players and had fantastic seasons, but just didn’t quite get there for one reason or another.   That’s 20.

              The other four players are still active.   Jason Heyward may have the most disappointing career of any super-rookie ever.  Harper, Braun and Ramirez have neither had Hall of Fame careers nor played themselves out of the conversation; they could still be in or out. 

              Thanks for reading.   I’ll open this up for comments tomorrow. 

 

 
 

COMMENTS (30 Comments, most recent shown first)

mikeclaw
Is Jason Heyward the new Claudell Washington?
5:27 PM Oct 29th
 
bjjp2
Marisfan: oops my bad. Duh.
4:57 PM Oct 27th
 
bhalbleib
Just so I can get a concrete idea of how this works with a player I am very familiar with, George Brett's age 20 season (when he basically didn't hit at all in 41 plate appearances) would be a 2? Or a 3? He is very young, obviously, he probably had above average speed at that point (my Royals fan memory doesn't go back to 1973, I was born the same year as the Royals) but he didn't hit a lick (again in a very small sample size). And since I presume his Age 21 season is also a Rookie Season under your system, what did he rate for that year, when he still was very young, presumably still above average speed and he did hit a little, although his OPS+ was still only 91. A 4, a 5?

thanks.
3:34 PM Oct 27th
 
MarisFan61
(typo -- ah screw it) :-)
3:23 PM Oct 27th
 
MarisFan61
bjjp: He mention up-top that this doesn't include pitchers.
3:23 PM Oct 27th
 
Arrojo
I am wondering if the Bill James projections in the Stats handbook would be useful - combined with age somehow to project rookies. I use OTS (onbase times slugging) from the Handbook all the time for my purposes. OTS is much closer to runs created than OPS of course.

For example, the BJ OTS projection for Andrew Benintendi was .16 coming into 2017 (I round to two decimals since 3 decimals implies a degree of accuracy that isn't really there) as a 22 year old. Max Kepler also had an OTS projection of .16 but he's 24, so clearly Benintendi is the better prospect.

Other 2017 OTS rookie projections: Bregman .18, Dahl .16, Mazara .15, D. Swanson .15, Story .18, W. Contreras .19

I haven't hit on a formula to combine age with OTS but I also haven't tried very hard.
3:18 PM Oct 27th
 
bjjp2
Surprised to see no mention of Seaver, one of the greatest pitchers ever and ROY in 67.
3:08 PM Oct 27th
 
MarisFan61
Well, I think I know the answer. I also think it's a thing that jogs the mind a bit, and I'd be interested if it jogs anyone else's.

What I'm saying is, the Fowler example points up that a player can get some above-the-minimum rating even without doing anything at all in the majors; he can be a Level 2 or Level 3 (or higher) without any credit at all from anything actually done in the majors. This seems tantamount to saying that a similar player -- a very young, fast player (even, as it happens, one who just blew out his knee, but forget that) -- who never even got called up could also rank the same, because he has the same basic positives, except for not having been brought up.

I guess the answer may be, "Yes, so what?" Plus, "Of course it matters that a guy got called up, as opposed to a guy who didn't."
The reason it jogs my mind is, generally when we consider such things, it does depend on the guy having done something in the majors; it's not just about his age and tools. (I do realize that Bill isn't bound by "generally.")
1:13 PM Oct 27th
 
MarisFan61
For convenience, so nobody has to go and look back:

"The fact that the system puts him anywhere above Level 1, or even if it put him merely at Level 1 rather than 0, comes very close to implying that any fast very-young minor leaguer who didn't even come up to the majors is at least a "Level 2 rookie," despite not having even come up, doesn't it?.....
The Fowler example 'feels' like a strong suggestion that it might be better to give more weight to actual major league performance in that year."
1:50 AM Oct 27th
 
MarisFan61
......At risk of lowering my apparent IQ further: I feel like insisting that Fowler's getting ANY 'group' rating here absolutely DOES have to do with his having played an inning (actually less, but that's irrelevant) .....does have to do with his having played an inning, as opposed to not having come up at all which was my point -- which maybe I didn't make clear enough.

If you want to say that arguably he should even be a 3, I'd say it magnifies my point further. I'll just leave it at that for now, rather than taking up yet more space with it -- but I do think you misunderstood the point.
1:07 AM Oct 27th
 
MarisFan61
Actually I stay up at night trying TO understand things! If I fail, it's not from lack of good intent. I'm satisfied that overall I do a fairly decent job. :-)
I do appreciate the answer. I'll take another/closer look and see if I get it better.
12:36 AM Oct 27th
 
bjames
Blefary was a 7, Bowens a 3. So yes, Blefary was far ahead.
11:23 PM Oct 26th
 
bjames
The "Dustin Fowler problem" is very interesting. The fact that the system puts him anywhere above Level 1, or even if it put him merely at Level 1 rather than 0, comes very close to implying that any fast very-young minor leaguer who didn't even come up to the majors is at least a "Level 2 rookie," despite not having even come up, doesn't it?


Maris, sometimes I wonder if you stay up at night figuring out how not to understand things. If anything, Level 2 is too low for Dustin Fowler. His playing an inning doesn't have shit to do with it, frankly. If I had a more robust, broad-based system, he probably should be a 3.
11:21 PM Oct 26th
 
Fireball Wenz
If the system messed up about Gene Stephens, it was at least in line with public perception. My father would frequently remark that everyone thought Stephens was going to be a star.
4:00 PM Oct 26th
 
MarisFan61
The "Dustin Fowler problem" is very interesting. The fact that the system puts him anywhere above Level 1, or even if it put him merely at Level 1 rather than 0, comes very close to implying that any fast very-young minor leaguer who didn't even come up to the majors is at least a "Level 2 rookie," despite not having even come up, doesn't it?
(BTW he actually played less than 1 inning. It was a terribly heartbreaking moment. Anyone watching it had to be feeling for him. That "electrical box" along the right field wall is insane.)

I realize it could be argued that a player's having been brought up by the major league team marks him significantly, and that therefore it makes sense for him to be 'eligible' for such groupings even if he hardly played. But I wonder -- as the article does too, a little. The Fowler example 'feels' like a strong suggestion that it might be better to give more weight to actual major league performance in that year.

An additional interesting thing in the instance of Fowler is that it was a leg injury that put him out, and a quite severe one, and so it might be considered questionable how much 'credit' for speed he should get in any future projection because it can't be assumed that his speed won't be affected. But I do realize that this is a very specific example, kind of a one-off, and that we shouldn't look for a method to take account of such a thing.

I ought to add, I do know who I am :-) :-) .....I greatly appreciate the chance to spew off things like this under these articles and in "Hey Bill," and I never intend my comments as anything more than 'maybe maybe this may be of interest.'I should add also, I love the article and the method. To me, Bill's work on rookies has been some of the most interesting and memorable of all his writing, and this is right in there. I still remember very well and often refer to the big article on rookies in one of the early Abstracts, marveling at the Twins' great crop of rookies in '82, looking at the great "class of '64" (or maybe was that a separate article?? I think it was), showing the overriding importance of age, looking at how the future development of the players varied according to position (with second base being surprisingly unfavorable), and even looking at ethnicity and coming out with some interesting findings on that. When I express wonderment about this-or-that, usually it means I simply love the rest.
2:07 AM Oct 26th
 
brewer09
I wrote about rookie sensations this past week. Not the same thing, of course, but both sorted rookies into different levels of success.

https://raleighco.com/sports/judging-aaron-history-rookie-sensations/
1:49 AM Oct 26th
 
BobGill
Speaking of 1964, what about Sam Bowens, as compared to Curt Blefary in 1965? This used to be a semi-famous comparison: same team, almost identical Triple Crown stats, but Blefary won the Rookie of the Year award while Bowens was lost amid the great Class of 1964. Looking at it now, though, I notice that Blefary walked twice as much and was also two or three years younger -- and of course he had a significantly longer career. It seems that he must've reached a higher level under this system, too. Is that right?

7:36 PM Oct 25th
 
337
Seems to me Jose Reyes should have scored pretty well--19 years old, played more than half the season (June 10th on), fast as hell, pretty fair hitter. No?
7:07 PM Oct 25th
 
bjames
1964-- Conigliaro is a Level 8. Jose Cardenal (just 20 games, 15 at bats) is a Level 7. Bert Campaneris (Cardenal's first cousin) is a 6, as are a couple of other guys. Jim Ray Hart was a 5, and Rico Carty a 5.
6:17 PM Oct 25th
 
bjames
Dustin Fowler has a Speed Score of 9. We have a lot of information about his speed. He hit 15 triples in the minors in 2016, and 8 in 70 games this year. He grounded into only 2 double plays in 297 at bats in the minors this year--one-third the normal DP rate. He plays the outfield. It is clear that he runs well.
6:13 PM Oct 25th
 
JohnPontoon
I'd just like to chime in that your, um, grumpiness near the end of part 15 made me grin from ear to ear. It was hilarious and, of course, well written. In addition to being a pioneer, you're also a genuine capital-A Artist in my book, Mr. James. Not that I've written a book.
4:24 PM Oct 25th
 
evanecurb
I'm curious about the levels of all of the top rookies of the amazing 1964 rookie class: Tony Conigliaro, Rico Carty, Jim Ray Hart, Wally Bunker. Did I miss anyone?
4:12 PM Oct 25th
 
voxpoptart
I really like this! I'm not sure anything can be done that catches Greg Allen/ Terrance Gore "false positives" without also failing to recognize Kaline/ Brock. But for the "Z problem", I can think of possibilities.

1) Cabrera or Devers, say: the hitting wasn't impressive, but it was impressive for their age, beyond what the age adjustment caught. Would it help any to make the .580 "replacement level" in part two adjustable, so that really young players were being compared to, I don't know, .530 or .480 or something?

2) The system seems to underrate catchers -- and I have a hunch it might underestimate the Dale Murphy/ Carlos Delgado/ Mike Sweeney types who were mistakenly begun at catcher too. How about increase a player's score by the percentage of his starts he records at catcher? 100% at catcher equals double score, 50% at catcher equals 50% increase, etc. Seems it might help.
3:02 PM Oct 25th
 
337
Which is unfortunate, in that age is the one factor that can most easily be fudged. I mean, you're never gonna hear "He batted .335 in AAA, but it was later revealed that it was actually .287 in AA ball..." or "He was listed at 6'3" but it turned out that he used his brother's body to be measured, and he was in fact himself only 5'8"..."
2:08 PM Oct 25th
 
hotstatrat
As I recall, over thirty years ago, Bill did write that age was BY FAR the most important determining factor about whether a rookie was going to have a Hall of Fame career or not.
1:33 PM Oct 25th
 
W.T.Mons10
I'm bemused by Dustin Fowler made it to group 2 by playing one inning, with no batting stats and no speed score. Apparently just being 22 is good enough.
12:57 PM Oct 25th
 
danfeinstein
Unless I've missed it, I don't see how the elements are put together to come up with a final rookie score. How much would it affect Tony Oliva to use the correct age? Another player listed as the best in a particular year also may have had the same issue - Rafael Furcal's age was corrected a couple years after his initial call up making him 22 instead of 19 in his rookie season.
9:59 AM Oct 25th
 
hotstatrat
I am wondering why speed is considered such a huge factor in your system - about equal to that of age. It seems the vast majority of the occasions your system misses is due to overrating speedy rookies and underrating slow ones.
8:02 AM Oct 25th
 
jemanji
Along another line, you encouraged us to say something if --- > we suspected that the system was letting players slip through the cracks.

From Seattle let me nominate Guillermo Heredia as a player who's likely to see 3,000 AB's. He's got a special skill, defense; Jerry Dipoto says his break on the ball is super-elite.

He played through a serious right shoulder injury and his eye for the zone is good.

So, may be an example of a Level One player to be aware of. His glove will keep him around for absolute sure, his batting eye is liable to justify it, and there seems to be upside in his bat.

My two cents, anyway.
2:10 AM Oct 25th
 
jemanji
++ But you have to understand; I’m not essentially a practitioner. I’m essentially a pioneer. I’m an explorer. I explore an idea for a week, and then I move on to some other idea. Nobody taught me how to do this stuff; I just figured it out, made up my own rules as I went along. If I didn’t do it the way that you would have done it, that is because I did it and you didn’t. ++

Brav-O.

You've got the right to say this about once per article, though this is the first time we've ever noticed you moved enough to object about it. You've got a guy in his late 60's giving the freshest, most fundamental insights on the 'net ... and most guys' reactions seem to be to compete with it.

But you said it better.

There are modern chess players who could beat masters of the 19th century because they're sharp with a calculator and patterns. Arpad Elo (who invented the rating system) says this is akin to a high school chemistry student regarding himself superior to Isaac Newton because he's got a 2017 textbook.

The value is in the flash of insight, the creative genius. Not the fourth order of precision.

Thanks for saying.
Jeff


2:07 AM Oct 25th
 
 
©2017 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy