THE EIGHT BASIC PRINCIPLES OF RUN CREATION AND WINNING BASEBALL GAMES ===================================================================== BASIC PRINCIPLE #1: ------------------- • "A batter's job is to create runs for his team." - Bill James - Principle 1a: Rule 1.02 of the Official Major League Baseball Rulebook states that "The offensive team’s objective is to have its batter become a runner, and its runners advance." - Principle 1b: Rule 1.04 states that "When a batter becomes a runner and touches all bases legally he shall score one run for his team." - Principle 1c: Rule 1.05 states that "The objective of each team is to win by scoring more runs than the opponent." - Principle 1d: Rule 1.06 states that "The winner of the game shall be that team which shall have scored, in accordance with these rules, the greater number of runs at the conclusion of a regulation game." -> NOTE 1: The rules of MLB are cited because the name of the game is to make runs to win. BASIC PRINCIPLE #2: ------------------- • Runs created are made up of three factors: On-Base, Advancement and Opportunities. Part of creating runs is to get on base, while the other main part is to advance runners already on base or get as far on the base paths as possible. Both need to be done in the fewest possible plate appearances. - Principle 2a: The rulebook once again states that two things need to happen: a batter has to get on base, and then go around the bases, before scoring a run – and, as Rule 5.08(a) says, they must do so "before three men are put out to end the inning." - Principle 2b: A plate appearance can end in several ways: a player can make an out (fly out, ground out, strike out); a player can reach base safely (base hit, base on balls, hit batsmen, error, or awarded first); a player can make an out but advance a baserunner (sacrifice bunts, flies, fielder's choice, dropped third strike); or a player can reach base safely but make an out (fielder's choice). - Principle 2c: The main statistics used to determine on-base, advancement, and opportunities are as follows: hits, bases on balls, hit batsmen, steals, caught stealing, groundouts into double plays, total bases, sacrifices (hits and flies), intentional walks, and plate appearances. -> NOTE 2: James uses the "new" version of the RC formula to adjust for the player's own stats versus the team/league-wide numbers. This total includes a small adjustment for strikeouts. The formula for this new version is provided below. ------------------------------------------------------- ***** RUNS CREATED and the OFFENSIVE W-L FORMULAS ***** ------------------------------------------------------- RC = [(OBF x ADVF) / PAF ] – (LRCAF x PA) Where: • OBF = (2.4 x PA) + (H + BB + HB – CS – GIDP) • ADVF = (3 x PA) + (TB + 0.24 x (BB + HB – IW) + 0.62 x SB + 0.5 x (SH + SF) – 0.03 x K) • PAF = (9 x PA) • LRCAF = {League RS – League [(OBF x ADVF) / PAF]} / League PA Outs = AB – H + SH + SF + CS + GIDP RC/G (also known as RC27) = (RC x 27) / Outs Offensive Context = (Team RS + Opp. RS) / (2 x Team GP) Offensive W-L Pct. = (RCG)2 / [(RCG)2 + (Off. Context)2] Games RF = (Outs/27) Off. Wins = Games RF x Offensive W-L Pct. Off. Wins over Replacement (OWoR) = Games RF x (Offensive W-L Pct. – .294) ------------------------------------------------------------------------------------------------ ABBREVIATIONS: RC - Runs Created; HB - Hit Batsmen; CS - Caught Stealing H - Hits; BB - Bases on Balls; TB - Total Bases; GIDP - Ground into Double Play; IW - Intentional Walks; AB - At-Bats; SH - Sacrifice Hits; SF - Sacrifice Flies; SB - Stolen Bases; RC/G - Runs Created per Game; Outs - Outs made by player; TeamRS - Team Runs Scored; Opp.RS - Opponent Runs Scored; TeamGP - Team Games Played; GamesRF - Games Responsible For; (W-L)Pct. - Offensive Win-Loss Percentage; Off.Wins - Offensive Wins. ------------------------------------------------------------------------------------------------ NOTE: There are two simpler versions of the RC formula that were in use for a long time prior to James revising the formula to remove the player's influence on league run scoring. They are: - Basic (Stolen Base version): (H + BB – CS) x (TB + .55 x SB) / (AB + BB) - Technical (TECH-1): [H + HB + BB – CS – GIDP] x [TB + .26 (HP + BB – IW) + .52 (SH + SF + SB)] / PA ------------------------------------------------------------------------------------------------ BASIC PRINCIPLE #3: ------------------- • A player can be assigned a number of games according to how many outs they make. Using this, we can determine how many runs a "team" of the particular player would create (on average) in a single game. - Principle 3a: A team uses 27 outs in a typical game (actually less in most seasons). - Principle 3b: Comparing runs to outs is actually a rate stat akin to ERA or points per game, and can be used as such. BASIC PRINCIPLE #4: ------------------- • By using the player's team(s) totals for runs scored and runs allowed, you can compare his runs per game total to the league(s) as a whole – without having to adjust for how he played at home as compared to other ballparks. - Principle 4a: A player doesn't see the same number of runs scored or allowed by his team as every other player in the major leagues (except for teammates, of course). - Principle 4b: The simplest way of comparing a player to the league without having to go through adjustments for park factors and the like is to use the "offensive context" that the batter finds himself playing within; that is, the total number of runs scored and allowed by the team (or teams) he has played for. -> NOTE 3: This method essentially compares the player to other players in the same situation as theirs; same pitchers, same opponents, same ballparks. BASIC PRINCIPLE #5: ------------------- • According to Bill James, the Pythagorean Formula suggests that there is a direct relationship between wins and runs in baseball, as described below [the ^ sign indicates raising the number to a power]: (Wins) / (Losses) = (Runs Scored)^2 / (Runs Allowed)^2 By using a player's RC/G totals and the team's run totals per game (scored and allowed, combined), the W/L Percentage for a particular player can be determined, along with how many wins they actually created offensively. - Principle 5a: The formula is so named as it is similar to that of the geometric equation a^2 + b^2 = c^2, which is the means of finding the distance of the longest side of a right angle. That formula was developed by Pythagoras in ancient Greece. - Principle 5b: To find a player's offensive won-loss percentage, square the player's runs created per game total. Divide that total from the sum of the square of runs created per game and the square of the player's team offensive context. - Principle 5c: The number of games a player is responsible for is equal to the number of total outs made (using the statistics in Principle 2c), divided by 27. This is essentially nine innings – the same measure used for pitchers for ERA. - Principle 5d: Offensive Wins are found by multiplying the offensive won-lost record with the games responsible for total. Offensive Losses are determined by multiplying one minus the offensive won-lost record with games responsible. This gives the player's total offensive wins and losses – akin to a pitcher's won-loss record. - Principle 5e: 67.97% of all major league teams (through 2024) have had their expected won-loss totals (based on runs scored and allowed) fall between plus or minus four wins from their actual total. This means that a team that is projected to win 80 games based on their XW% will win no more than 84 and no less than 76 at least two-thirds of the time. In 2024, 80% of the teams in MLB (24 of 30) had XW% that were less than four games from their actual totals; all 30 teams had an XW% within 7.5 games of their actual totals. As there were 741 different batters in MLB in 2024, this would mean that any player is extremely unlikely to exceed any win total based on the above formula by more than 0.01 wins. BASIC PRINCIPLE #6: ------------------- • A player's offensive value to a team is based on how much it would take to replace him on the roster – and not how much better he is over an average player. - Principle 6a: A player that is "average" in terms of offensive wins is still significantly better than a player in Triple-A or the other minor leagues. For example: In 2023, Andrew Knizner of the St. Louis Cardinals had a .402 OW% as the backup catcher to Willson Contreras, for a 2.68-3.99 OW-L record. Among non-pitchers in 2024, 617 of 651 players (94.8%) had an offensive winning percentage below .500. -> NOTE 4: This is one of three schools of thought on comparisons. The two others are league average (comparing the player to a theoretical league average player), and league "zero" (comparing a player to the worst player in the league). The concept of "replacement level" is essentially: "If a player had to be replaced by a team, either with a minor leaguer, by trade, or with a free agent, what amount of value would the team need to replace them?" BASIC PRINCIPLE #7: ------------------- • The point where a player's value to his team is below that of a player from the minor leagues is around a .294 winning percentage, or roughly 47.628 wins per 162 games. - Principle 7a: 97.85% of all non-pitchers had an OW% above .294 in 2024 (637 of 651). - Principle 7b: In the expansion era (since 1961), there have been only five teams that posted winning percentages below .294: the 1962 Mets, the 2003 Tigers, the 2018 Orioles, the 2019 Tigers, and (of course) the 2024 White Sox. A sixth team, the 1961 Phillies, won only 47 games in 155 contests – and it's highly debatable that they would have won any more had they played seven additional games. -> NOTE 5: The .294 number was chosen by Fan Graphs and Baseball Reference for Wins Above Replacement; the number was halfway between FG's and BR's previous totals (.265 and .320, respectively). This number goes hand-in-hand with the fact pointed out in 7b – even the worst teams in the majors will win 47-48 games in any given season, unless there are extenuating circumstances. BASIC PRINCIPLE #8: ------------------- • A player's Offensive Wins over Replacement level (OWoR) is his offensive won-loss percentage minus .294, then multiplied by games responsible for. - Principle 8a: This number provides an indication of what kind of offensive production would be needed to replace the particular player. - Principle 8b: Players with negative OWoR totals may provide more value to their teams by excellent defensive numbers, like a high range factor, high OF assist totals, or high Caught Stealing totals. (Either that, or they had injury issues, or played only in a short time frame with their club.) - Principle 8c: OWoR, along with other metrics, can assist in constructing a lineup and determining a player's role on a roster. Generally, 2.0+ WAR is for starters, while 1.0-1.9 WAR is for bench or platoon players, and the backups are 0 to 0.9 WAR. Baseball Reference.com suggests that a player whose WAR total is 2.1 to 4.9 is a solid starter, 5.0 to 7.9 is All-Star level, and 8.0 or above is MVP-level. - Principle 8d: The highest OWoR for a single team in MLB history is the 1927 Yankees (53.8). The 2019 Astros (52.3), the 1943 Cardinals (51.9), the 2017 Indians (51.7), the 1911 Giants and the 1906 Naps (51.0) are the only other teams with 51+ OWoR. (The 2023 Atlanta Braves finished with an OWoR of 46.33.) - Principle 8e: The 2024 White Sox posted the lowest team OwoR of the expansion era with 1.58 OwoR. No team since the 1899 Cleveland Spiders (-2.6) have posted a negative OWoR total. -> NOTE 6a: For reference, of the 95,570 players who have posted a plate appearance from 1871 through 2024, only 61 have had a single-season OWoR total of 8 or more (0.064%). -> NOTE 6b: Only two players have ever posted a single-season OWoR total over nine – and they both did so in 2022. Aaron Judge had a 9.34 OWoR in his 62-HR season for the Yankees, while Freddie Freeman had a 9.27 OWoR for the Dodgers. Prior to 2022, Henry Aaron had the best single-season OWoR with an 8.80 total in 1963 for the Milwaukee Braves. Shohei Ohtani posted an 8.71 OWoR in 2024. -> NOTE 6c: Baseball Reference uses a different method to determine Wins Above Replacement level, as does FanGraphs. These are usually differentiated by the abbreviations bWAR for Baseball Reference, and fWAR for FanGraphs. ================================== FIGURING PLAYER VALUE FOR PITCHERS ---------------------------------- Tom Tango developed a formula based on Voros McCracken's ideas about how Balls in Play affect pitching performance. Tango's formula, known as FIPS, took the core tenet of Defense Independent Pitching Runs, then used a constant that "re-centered" the league-average FIP to match the league-average ERA. The formula is: FIPR = { [(13 x HR) + 3 x (BB + HBP) – 2 x K] / IP } + FIP constant Where the FIP constant is League ERA minus the league totals in FIP. For example, in 2024 there were 5,453 home runs hit over 43,116.1 innings pitched. There were also 14,929 base on balls, 2,020 hit batsmen, and 41,197 strikeouts. This comes out to a total of 0.9125. The major league ERA was 4.0721, which makes for a major league FIP constant of 3.1596. What FIPR (or FiRA) can be used for should be obvious: it can be used as an approximation of what a pitcher's won-loss record would be if he had a league-average defense behind him. In fact, by substituting FIPR for Runs Created per Game in the reverse (i.e., instead of using the square of FIPR in the numerator, we use the square of the Offensive Context there), we can determine a pitcher's "Fielding Independent" Won-Loss record. And, as with OWoR, we can determine a pitcher's Fielding Independent Wins over Replacement level – or FIWoR. In general, pitchers are more likely to have higher WoR levels than batters, because of the larger quantity of outs per game they are responsible for. In 2024, Chris Sale led MLB with a 9.81 FIWoR, while Tarik Skubal paced the AL with a 9.29 FIWoR. They were the only pitchers with a FIWoR of 9 or more. Over the history of MLB, 2,635 of 53,183 pitchers that faced at least one batter (4.95%) have posted a 9+ FIWoR season. Since 1971, only 485 of 33,594 pitchers have posted a 9+ FIWoR (1.44% of all pitchers).