Men Left On Base
You may recall that during the team's worst losing streak before this one, the team had gone into Pittsburgh and gotten swept. Paul Maholm threw the best game he'd ever thrown--and may ever throw--in the first tilt, but then the Astros went out and left 18 men and 13 men respectively in the last two fracases, practically ensuring their own defeat.
And when they left 13 on vs. Milwaukee three days later, the question occurred to me whether the Astros had stranded more runners than any other team in the National League.
Given the deep nature of the current funk, by now, I seriously doubt that's the case, but it appears that there's no place to find that information anyway. I tried generating it myself, but it looks like there's no quick way of doing that, either. It seems no website I'm aware of keeps current or historical team-by-team left on base data.
But while futilely searching for information on whether or not Houston may or may not have had more stranded baserunners than anyone else, I received the following thought-provoking reply from Charlie at Bucs Dugout when I had asked aloud if Houston indeed had left more runners than anyone else:
If you think it will help, you can tell [your readers] that fans of every bad offensive team think this. Pirates fans say this every year. Even their manager thinks that their offensive problems are the result of not enough key hits, as opposed to just offensive ineptitude.
And an old post I found at Baseball Musings fleshes Charlies' thoughts out a little further.
[These bloggers Pinto was referring to] approach the Left On Base stat as a bad thing. That's not really true. Leaving lots of men on base is often a sign of strong offense, one that puts lots of men on base!
So, then, there appears to be this belief floating about that, contrary to what we always thought, men left on base are actually a good thing, that they are simply the unavoidable residue of runs scored.
I'm not sure if I should term this a sabermetric idea, per se, but it's definitely an internet idea, if you can judge the droppings by the habitat. And certainly the theme is one that is palatable to sabermetricians, who perhaps more than anything else, have sought to abolish the myth of timely hitting.
So if it's not their idea, this stranded men are good men concept, it's certainly up their alley, one they could put on their mantels as they seek to convince us that all our old-fashioned ideas about sacrifice flies and clutchness and situational hitting are just so much superstition.
Me, I want more data before I take sides, although of course I didn't bring much with me. All I got are the 2007 Astros.
| In All Games | In Wins | In Losses | |
| Left On | 356 | 150 | 206 |
| Average/Game | 7.12 | 7.14 | 7.10 |
| Fewest Men Left - | 2, April 24, at Pittsburgh (L) |
| Most Men Left On - | 18, April 25, at Pittsburgh (L) |
| Record When Leaving 0 - 5: | 7 - | 10 |
| When Leaving 6 - 10: | 12 - | 16 |
| When Leaving 11+: | 2 - | 3 |
|
|
From this tiny smidgen of a sample, from one lousy offensive team, for two months, I will take the tentative conclusion that men left on base have no relationship at all--not positive, not negative--with winning ball games. The Astros are a .420 team right now a I write,and that same flavor of mediocrity is reflected in their winning percentage when they strand just a few (.411), when they strand a medium amount (.428) and when they strand a whole mess (.400).
It doesn't appear to matter how many the Astros strand. Their tendency to play .420 baseball seems to shine through that particular detail as if it were transparent.
Maybe it is.
Let's take a look at runs scored, again the microscopic sample of our own Astros, the only team I love enough that I would go to the trouble of counting team left on base boxscore-by-boxscore:

I'm open to suggested interpretations, but I don't see anything. When the Astros have left five, they've scored one--and they've scored ten. Leaving six might mean scoring one measly run, or it might mean scoring thirteen.
There appears to me to be no relationship at all.
Although I will say that Microsft calls that black line running downward from left to right a "trendline," and it does suggest that although the effect is weak, your uncle Louie has been right all these years: the more left on base, the fewer runs.
With all my fancy Excel sheets, I'm pretty good at coming up with reams of useless data. Where I'm less useful is figuring out what it means. And that is true with this particular project more than usual.
Lots of data, no analysis. But I will turn to you, gentle reader, for deductions I've missed, and for suggestions on where to find info on whether the conclusions suggested by our own anemically hitting baseball team are true for the rest of the baseball world.
You know, the one outside East Texas, the one that regularly hits for power and for average.
0 recs |
14
comments
Comments
Hmmm.
by Stros Bro on May 28, 2007 7:15 AM CDT 0 recs
I'm not surprised at the results
My initial reaction is that attempting to find a correllation between winning a game and LOB in the game is not the right approach. But I could be wrong.
You certainly can find or calculate total LOB (for all games played), since that can be determined from Runs Scored minus Base Runners. I would be very surprised if the best offenses aren't near the top in total LOB. But that doesn't tell us much, since offenses with high on base rates should have a lot more runners left...even after a big innning. I think the more relevant analysis might be some sort of LOB rate....maybe runs scored per runner left on base.
by clack on May 28, 2007 10:01 AM CDT 0 recs
The Problem
The first two are easy, but the third is not. Outs could be figured by (Games * 27) + (extra innings * 3) + outs walked off, and I could do that for the 'Stros, but how am I supposed to know how many walkoff victories the Red Sox have had?
Another way you might caluculate total outs is to add groundouts and flyouts and strikeouts. MLB's site promises these numbers but doesn't deliver.
And I found a number on Baseball Prospectus' site called total outs, but using it just didn't crosscheck with the LOB number I had for Houston. I don;t know why.
And absolutely. Comparing LOB/TPA and LOB/RS for all National League teams at least would be much more illuminating than what I had here.
It's something that I would be glad to follow up with, if an easier way of calculating LOB can be found.
by rastronomicals on
May 28, 2007 10:34 AM CDT
up
0 recs
Perhaps a more accurate
by bwhite2323 on
May 28, 2007 10:37 AM CDT
up
0 recs
Out Stretching & Picked Off
H + BB + HBP + CI - CS - GDP - R for Houston comes within 3 of the figure I know to be correct for Houston's LOB, and I guess those three were either out stretching or picked off.
I think pick offs are listed for players, and maybe you could add those from a team page,
But we're not likely to find out stretching listed anywhere. Or am I missing something?
by rastronomicals on
May 28, 2007 11:45 AM CDT
up
0 recs
Pickoffs Are At Prospectus
And now (H + BB + HBP + CI + ROE - CS - GDP - Picked - R) is high by 12. Out stretching is probably some of that, but the other thing perhaps not covered here is runners thrown out at the plate. When after a flyball caught, that's probably not covered under double plays, is my thinking, though I guess runners errantly waved home by Doug Mansolino are a subcategory of the stretching thing.
But I don't know. It might be possible to calculate this accurately, but it's clearly rather complicated. . . .
by rastronomicals on
May 28, 2007 12:20 PM CDT
up
0 recs
I think
by TimStros on
May 28, 2007 1:19 PM CDT
up
0 recs
If you can get that close
My high school coach (back in the days when aluminum was only for kitchen foil) used to mandate as a goal scoring half of the base runners we created. I don't know how well that really correlates with winning, but when we lost he was all over our butts about not scoring half of our base runners (and we ran many many laps).
We won lots of games.
by bwhite2323 on
May 28, 2007 1:37 PM CDT
up
0 recs
some good points
by clack on
May 28, 2007 4:20 PM CDT
up
0 recs
My gut feeling is...
I don't get the feeling that low or average LOB actually indicates anything.
by AstroAndy on May 28, 2007 10:06 AM CDT 0 recs
A couple of unrelated thoughts here...
And from a logical standpoint, if low LOBs are a good thing, then wouldn't that make hitting into double plays a good thing too, since that eliminates LOBs? Same for caught stealing, picked off, base running errors, etc.
by bwhite2323 on May 28, 2007 10:33 AM CDT 0 recs
Very interesting discussion
by littlevisigoth on May 29, 2007 9:50 AM CDT 0 recs
Runs/LOB
5 runs/ 1 LOB = 5 (very good)
1 run/ 18 LOB = .055 (terrible)
This could be a stat that shows more realistic trends of how a team is doing.
by jharmz on May 29, 2007 3:31 PM CDT 0 recs













