4374 vs. 3 x 9 x 162 is the real new school vs. old school, guys with computers vs. guys who watch games or whatever you want to call it. Using stats that evaluate events based on total runs scored in a season will value those events differently than watching runs scored in 3 outs and games won in 27 outs.
If each team was given 4374 outs to score as many runs as possible, then stolen bases, bunts and Sac flies are worthless. KOs are just the same as any out, productive outs are worthless and by far the most important thing is avoiding outs.
Outs canít be avoided, and a teamís total runs are not randomly distributed over a teamís 9 x 3 out segments. Managers should not make in game decisions (including lineup construction), based on scoring the most possible runs in a season. What makes baseball compelling is that teams have to try score enough runs to win each game, not just wait around until they avoid enough outs in a row to score a lot of runs.
Statistics are good. The ability of modern teams to compile and analyze stats has made modern teams better. When statistics show baseball professionals consistently making bad decisions, step back from the stats. Compare the results of the statistical analysis with common sense and what you see in the real world to determine if/why the analysis may be wrong.
Basing decisions on total runs scored is only correct, if the distribution of runs is random. Baseball is a game of skill. As an observer, we have to assume not random. Skill not luck. Results are distributed normally, so given large enough sample size, we can use stats to evaluate players. Normal distribution does not mean random. The best decision in a specific situation may not be the choice that statistical analysis says will lead to most runs scored in a season.