Team LOB usually incorrect

[Redleg]

Just look at almost any game, count up the runners left on base and see if that number matches what the box score total says. Odds are the numbers will not be the same.

[BoomerSoonerAMH]

This isn’t something I’ve noticed and I’m a box score junkie, so here are a couple of things to check:

1. Are you looking at the LOB total for each batter as opposed to the team LOB? If so you aren’t going to match. For example, Abel leads off an inning with a walk. Baker and Charles strike followed by Daniel hitting a HR. Baker and Charles are each credited with a runner LOB, but the team LOB is 0. Conversely, if Daniel had also struck out, then he, Baker and Charles all would have been given 1 LOB but the team total would only be 1 LOB.

2. Are your box scores proving? The dormula for proving a box score is: AB+BB+HBP+SH+SF+CI=PO(by the opposing team, usually 27 but could be less if fewer than 9 full innings or more if extra innings)+team LOB+R. If you getting matching numbers on both sides of this equation, then there isn’t an issue.

[Redleg]

The problem is with Team LOB, not the sum of runners left on base by individual batters (there could be a problem there, too, but I haven't checked it).

It's very simple to confirm the issue, if you would like to do so: go through some game logs and add up the runners left on base in each inning and then see if that number matches the Team LOB number in the box score. Sometimes it does, sometimes it doesn't. Quite often, at least one of the teams' totals is in error in each game. And, no, the box scores will most certainly not prove in those instances.

[BoomerSoonerAMH]

So I took a look at 3 consecutive games, and while I know that is a small sample size, there was at least 1 instance of an incorrect LOB total in each game (1 game had incorrect totals for both teams). I'm going to keep digging to look for potential causes, but the one thing that was consistent was that the box score total came up short of what the LOB total should have been. Here is what I had:

Game 1
Visitor: Box Score 6 vs. actual 7
Home: Box Score 8 vs. actual 10

Game 2
Visitor: Box Score 7 vs. actual 7
Home: Box Score 8 vs. actual 9

Game 3
Visitor: Box Score 9 vs. actual 10
Home: Box Score 8 vs. actual 8

[Redleg]

Sample size would be relevant were we discussing averages or rates, but this is a different animal. Even if your sample size was one game, if the total was incorrectly calculated in that one case then that would confirm a problem, simply because this is an issue of arithmetic. You found problems in each of the first three games you looked at. Clearly, something is wrong here with the values going into the calculation.

The hunch I've had is that instead of the game adding up the total of runners left by the visiting team in each of its innings, one of the home team's innings is sneaking in there as part of the calculation. Or vice versa.

Anyway, thanks very much for taking some of your time to duplicate the problem and for sharing your results.

[BoomerSoonerAMH]

Concerning sample size, I agree with you that even 1 game shows that an error is occurring.

My point, albeit not well made, was that only 3 games may not be enough to identify any causality and certain doesn't verify that it will always calculate the box score LOB less than the actual LOB. Had I sampled 300 games and found they always calculated the box score less than the actual, the whole thing could be undone by 1 game with a box score LOB greater than the actual.

[RavensFan]

I played a 9 inning game and one team had 34 men left on base, a statistical impossibility.