Runs Created and Lucky Wins

We’ve all seen baseball games where one team outhits the other by a healthy margin, yet manages to lose. It may be one or two key errors, opponents bunching their hits and walks or some other rare occurence that makes the difference. But very often the team that gets the best production — a combination of singles, extra-base hits and walks — comes out on top.

When I think of offensive production, I also think of run estimation formulas (like Base Runs, Linear Weights or Runs Created). These formulas convert the building blocks of run scoring — hits, walks and the like — into an estimate of the likely number of runs a team would expect to score.

Basic Runs Created — introduced about 25 years ago by Bill James — requires only hits, walks, total bases and at-bats as inputs. The formula has a simple design as well: H+W multiplied by Total Bases, divided by AB+W. Despite using just four statistical categories, basic RC does a very good job of predicting how many runs a given major league team will score in a season.

On a single-game basis, basic RC isn’t very good at predicting runs scored. On the other hand, it can serve as a rough approximation for value of basic offensive production. Using the nifty game log data provided on a daily basis to THT by Baseball Info Solutions, it’s possible to calculate basic RC for both teams in each game played in 2004.

I did this (substituting team plate appearances for the AB+W denominator) and it turned out the team that won the RC battle on a given day also won the game the majority of the time — no surprise there. How often did that happen?

Basic RC Differential by Game
RC diff. Games Road Home Total WPCT
0.00 to 1.00 86 29-21 26-10 55-31 .640
1.01 to 1.99 82 28-9 38-7 66-16 .805
2.00 plus 234 105-9 112-8 217-17 .927
All games 402 162-39 176-25 338-64 .841

The Road column is the win-loss record when the road team has the RC edge; the Home column is the win-loss record when the home team has the RC edge.

So far this season, teams winning the RC battle have won the game 84% of the time. When the RC margin was 2 or greater, the more “productive” team won nearly 93% of the time. That leaves a little over 7% when a team lost despite a 2+ edge in basic RC. That’s only 17 games (through May 5th) — here’s a table of the games in question.

Stolen or Squandered?
Game Date Away Runs RC Home Runs RC RC diff.
4/05/2004 San Diego 8 6.8 Los Angeles 2 9.1 2.3
4/05/2004 Cleveland 4 9.9 Minnesota 7 5.2 -4.7
4/11/2004 Minnesota 5 7.2 Detroit 6 5.1 -2.1
4/14/2004 Kansas City 9 10.0 Chi Sox 10 7.9 -2.1
4/17/2004 Kansas City 4 6.6 Minnesota 8 2.7 -3.9
4/18/2004 Cincinnati 11 10.6 Chi Cubs 10 13.3 2.7
4/22/2004 Kansas City 4 6.1 Cleveland 5 2.1 -4.0
4/24/2004 Houston 8 5.5 Colorado 5 7.9 2.4
4/27/2004 Oakland 8 8.8 NY Yankees 10 4.9 -3.9
4/28/2004 Chi Cubs 4 3.3 Arizona 3 6.3 3.1
4/28/2004 Texas 3 7.0 Kansas City 5 4.5 -2.5
4/30/2004 Chi Cubs 3 4.6 Saint Louis 4 2.6 -2.0
4/30/2004 Oakland 4 1.6 Tampa Bay 2 4.5 3.0
5/01/2004 Boston 3 4.3 Texas 4 1.6 -2.7
5/04/2004 Los Angeles 4 2.1 Florida 3 5.1 3.0
5/04/2004 Saint Louis 6 3.0 Philadelphia 5 6.3 3.3
5/05/2004 Chi Sox 6 6.3 Baltimore 5 8.3 2.0
The final column — RC diff. is negative if the away team has the RC edge and positive if the home team has the RC edge.

The largest RC advantage that did not result in a win occured on April 5th. The Cleveland Indians should have had a cakewalk on Opening Day; instead the Twins came away with a 7-4 extra-inning victory. In that game, the Indians held a 4-0 lead going into the 8th. The Twins tied it with a four-spot, winning it in the 11th with a three-run walk-off home run by Shannon Stewart. The Indians outhit the Twins 17-10 and outhomered them 3 to 2 (though Minnesota walked six times to Cleveland’s one).

Were these fluke victories? Was the losing team merely unfortunate on the day? Or is there a set of skills or defects involved in the results of these peculiar games? That’s a debate for another time.

Comments are closed.