THE CRANE POOL FORUM thecranepool.net (.com)


Forum Home

Master Index of Archived Threads


Stealin'

Edgy DC
Oct 04 2005 07:38 AM
Edited 1 time(s), most recently on Oct 04 2005 01:52 PM

TeamSBCSSB%
Philadephia1162781.1%
New York1534079.3%
Cincinnati722375.8%
Atlanta923274.2%
Houston1154472.3%
Arizona672672.0%
Florida963871.6%
Pittsburgh733070.9%
Milwaukee793469.9%
St. Louis833669.7%
San Diego994469.2%
Colorado653267.0%
San Francisco713567.0%
Chcago653962.5%
Los Angeles583562.4%
Washington454550.0%
The Mets not only lapped the field in stolen bases this season, but finished second in the league, behind Philadelphia, in stolen-base percentage. (Maybe Jimmy Rollins should be tasting some MVP support this year. Kenny Lofton also seems to have re-found his legs this year.)

Anyhow what's compelling about the Met numbers is that (1) unlike the Phils, they didn't have Mike Piazza to take advantage of, and (2) they did what they did with only one featured base stealer, plus small but significant contributions from almost everywhere else.

I can only imagine what Randolph would do with a lineup with a healthier Beltran, Milledge, maybe Cameron, maybe Raffy Freakin' Furcal or maybe just Anderson Hernandez.

sharpie
Oct 04 2005 07:49 AM

Nats at 45/45. Yikes.

Edgy DC
Oct 04 2005 07:52 AM

Yeah, clearly Randolph can run himself into a championship or run himself out of one.

How long before an MLB lawyer visits me to talk about that hologram thingie on the image I linked to?

Benjamin Grimm
Oct 04 2005 07:58 AM

Is that a hologram?

I thought Reyes was carrying a large magnifying glass in his back pocket. (You never know when one of those will come in handy.)

seawolf17
Oct 04 2005 01:44 PM

In case he wants to see his on-base percentage.

Edgy DC
Oct 04 2005 01:50 PM

Ooh, snap!

Vic Sage
Oct 04 2005 03:11 PM

thanks for another piece of evidence about the irrelevance of base stealing to (1) run scoring, and/or (2) winning.

Valadius
Oct 04 2005 03:13 PM

I don't know... every team with at least 80 steals finished over .500.

Vic Sage
Oct 04 2005 03:53 PM

[u:e14b5e7ba3]Team SB CS SB% W-L (League rank in Runs Scored)[/u:e14b5e7ba3]
Philadephia 116 27 81.1% 88-74 (2/16)
New York 153 40 79.3% 83-79 (7/16)
Cincinnati 72 23 75.8% 73-89 (1/16)
Atlanta 92 32 74.2% 90-72* (4/16)
Houston 115 44 72.3% 89-73* (11/16)
Arizona 67 26 72.0% 77-85 (10/16)
Florida 96 38 71.6% 83-79 (8/16)
Pittsburgh 73 30 70.9% 67-95 (14/16)
Milwaukee 79 34 69.9% 81-81 (6/16)
St. Louis 83 36 69.7% 100-62 (3/16)
San Diego 99 44 69.2% 82-80* (13/16)
Colorado 65 32 67.0% 67-95 (5/16)
San Francisco 71 35 67.0% 75-87 (15/16)
Chcago 65 39 62.5% 79-83 (9/16)
Los Angeles 58 35 62.4% 71-91 (12/16)
Washington 45 45 50.0% 81-81 (16/16)

Valadius
Oct 04 2005 03:58 PM

Hey, I never said anything about the percentage, just the number.

And you forgot the star next to St. Louis.

Vic Sage
Oct 04 2005 04:04 PM

cincinnati scored the most runs, but was 10/16 in SB.
Colorado was 5th in RS, but only 14/16 in SBs.
Houston had a ton of SBs, but was 11/16 in Runs Scored.
Mets creamed the league in SBs, but was 7/16 in RS
The Nats managed to be a .500 team while dead last in SBs and SB%.

St.Louis was the best team in the NL, but was 7/16 in SB (10/16 in SB%)

I don't know.
It looks like a pretty random correlation to me.

metsmarathon
Oct 04 2005 05:17 PM

sample size(?). if i just look at this season, i could make the argument that stolen bases HURT offensive production, seeing as philly and st. louis scored the same # of runs, but philly had a higher OPS (on the back of a higher OBP - hteir SLG was nearly identical). in fact, walks and stolen bases aside, their offensive stats all looked eerily similar.

yet we'v come to believe that OPS is the best correlator to runs. (and if not then OBP should be weighted heigher). so something MUST be dragging down philly's offense, no? gotta be the extra stolen bases! or maybe fewer GIDPs...?

of course, then, what must boost the mets' offense?

unless, of course, its random variation, and the overall correlation holds, but is masked by the random variation of a small sample size, and the contribution of other independent variables... (clutch/timely hitting? lineup construction? poor baserunning? conservative 3rd base coaching? who knows...)

all other things being equal, the team that steals more bases and steals more efficiently, should score more runs than a team that does everything else the same, but steals less, and less efficiently.

seawolf17
Oct 04 2005 08:00 PM

metsmarathon wrote:
of course, then, what must boost the mets' offense?

That would be the 350 plate appearances they gave to Miguel Cairo.

Frayed Knot
Oct 04 2005 09:14 PM

There have been correlation studies done with steals & runs scored
in the past with samples a lot larger than just one year.
Essentially they come out showing the relationship to be positive
although only somewhat, much closer to random than causal.

Edgy DC
Oct 05 2005 07:27 AM

See, the thing about stealing is that if you're not that good at it, you don't have to do it. It's an optional weapon.

If you're not good at hitting, you still have to hit. So it's a lot easier to show an independent correlation of doubles with runs, or homers with winning. Because bad hitters will most certainly unmake their team's chances far more than bad runners, who can mostly hide their lack of skill in conservative baserunning.

Nymr83
Oct 05 2005 09:04 PM

i doubt basestealing is meaningless, but in the long run it is not as beneficial as some would like to believe abd is only beneficial for as long as your percentage is high...i dont feel like checking the studies for the correct number but i'd imagine its in the 75% area.
this doesnt take into account situations where you get credited for a CS when you werent stealing (failed hit and run.) it also doesn't take into account situations where the necessary success % is lower (2 outs runner on 1st, stealing 2nd comes to mind) or higher (1st and 3rd 1 out, where getting caught at second prevents the guy on 3rd from scoring on an out)

in the end i'd build a team around pitching, defense, and obp. if i get players (like carlos beltran, david wright) who can do these things while also adding some steals then great. but if my guys do the other things well but cant steal (john olerud, keither hernandez both come to mind) i could care less about a lack of steals.
put another way- if given the choice, i'd rather see reyes lose a bunch of steals and get on base alot more often.

Vic Sage
Oct 06 2005 12:57 PM

building a team around "pitching and defense"... its an expression you hear all the time. And its just plain wrong.

Pitching IS defense... or at least a significant part of it. The other part of defense is FIELDING.

So, if you say you want to build a team around "pitching and fielding", at least its a correctly STATED position... even though its one i disagree with, since winning requires both defense AND offense.

Now, if you say you want to build around pitching, fielding and OB%... i could definitely get behind that, as it addresses both sides of the ball. The reality, however, is that you often have to choose between fielding and hitting. As for me, I'd take the hitting (especially OB%) over fielding, because good pitching can make up for the defensive deficiencies.

So for me, building a winner is about pitching and OB%... with power, fielding and speed (in that order) important accessories. I'd rather have a platoon of Hojo and Kevin Mitchell at SS than a season's worth of Rafael Santana, if Doc Gooden is on the mound.

MFS62
Oct 06 2005 01:16 PM

I remember reading some stats a while ago that said that a runner on second has a 20% higher chance of scoring than a runner on first.(Not sure if that meant with the same number of outs.) But if that ratio is still true, then it seems that a successful steal should have a positive effect on scoring.

Later

Benjamin Grimm
Oct 06 2005 01:23 PM

I don't think there's any doubt that a successful steal generally increases the chance of scoring a run. (An exception might be when it causes the next guy to be intentionally walked, thereby taking the bat out of the hands of your best hitter.)

But you have to figure the caught stealings, too. Every time a player is caught, it becomes impossible for him to score. So if a success increases the chance of scoring, and a caught eliminates it, where's the tipping point? I'm sure a guy with a 99% success rate is helping his team. A guy with a 1% rate is hurting it. The balance tilts somewhere in between, and it may not be at the 50% mark.

Nymr83
Oct 06 2005 01:27 PM

it is DEFINETALY far above 50%. an out is more valuable to the defense than a base is to the offense.

Frayed Knot
Oct 06 2005 02:08 PM

OK, here's a "Run Expectation" chart for 2004

Runners0 OUTS1 OUT2 OUTS
Empty.538.287.114
1st.926.550.246
2nd1.160.710.336
1st & 2nd1.467.958.461
3rd1.454.972.362
1st & 3rd1.8541.224.522
2nd & 3rd2.1341.472.618
Loaded2.2551.595.808




What this is saying is that the average number of runs that the average team scores after getting the lead runner on is .926 or, put another way, given that situation 100 times you'll score about 93 runs.
The bad news comes when you see that while the successful steal pumps that up to 116 times in every 100 tries (runner on 2nd/none out), the UNsuccessful steal drops you to all the way to 29 (none on / 1 out).


Here's some hypotheticals which show that even an 80% steal rate is only somewhat helpful:

Leadoff x 100 = 93 runs
Leadoff + steal x 100 = 116
Leadoff + (steal * .80) x 100 = 99 {stealing at 80% success}
Leadoff + (steal * .70) x 100 = 90
Leadoff + (steal * .60) x 100 = 81
Leadoff + (steal * .50) x 100 = 72

Edgy DC
Oct 06 2005 02:22 PM

Excellent ame your source, man.

OK, I'm probably doing this wrong, but I calculate that with none out and a runner on first, the general break-even point in stealing second is 73.2%.

With one out, 73.2% also.

With two out 55.9%.

That must be wrong. I don't have time to recaculate now so somebody else do it.

Anyhow I italicize "general" because clearly it's a better bet to go with Gerald Williams up than with David Wright up, or down by one rather than down by four. And base-stealers know that to a large extent.

metsmarathon
Oct 06 2005 02:25 PM

i've got different numbers....

you're forgetting (i think) the run expectancy that results from a caught stealing.

for instance:

with no outs:
no steal: 0.926
100% * 1.160 + 0% * 0.287 = 1.160
90% * 1.160 + 10% * 0.287 = 1.07
80% * 1.160 + 20% * 0.287 = 0.99
70% * 1.160 + 30% * 0.287 = 0.90
60% * 1.160 + 40% * 0.287 = 0.71

with one out:
no steal: 0.550
100% * 0.710 + 0% * 0.114 = 0.710
90% * 0.710 + 10% * 0.114 = 0.650
80% * 0.710 + 20% * 0.114 = 0.591
70% * 0.710 + 30% * 0.114 = 0.531
60% * 0.710 + 40% * 0.114 = 0.472

with two outs:
no steal: 0.246
100% * 0.336 + 0% * 0 = 0.336
90% * 0.336 + 10% * 0 = 0.302
80% * 0.336 + 20% * 0 = 0.269
70% * 0.336 + 30% * 0 = 0.235
60% * 0.336 + 40% * 0 = 0.202

so in each of these cases, it looks like you break even at about a 75% stolen base rate. below 75%, and you decrease your run expectancy (hurt your team/offense). above 75% and you improve your run expectancy (help your team/offense).

Nymr83
Oct 06 2005 04:04 PM

excellent work by frayed knot and mets marathon, i love when i dont have to look the numbers up myself.