The problem wasn't Thomas, it was the guys hitting behind him. After Aaron Hill it got scary ... usually a bunch of low OBP/SLG guys.

Yes, it takes a lot to get him around the bases so it's incumbent that John Gibbons puts guys behind him who can do that.

Best Regards

John

Which is worse - a slow guy on base after drawing a walk, or a fast guy on the bench after popping out?

If you can't get the right answer to that question, you have no business managing anything above a T-ball team.

Which is worse - a slow guy on base after drawing a walk, or a fast guy on the bench after popping out?

Obviously the slow guy on first caused the fast guy to pop up in trying to avoid a double play.

But the opposing pitcher didn't have to worry about him stealing, plus it didn't put pressure on the defensive players. You can't use numbers to figure out how this would have changed the game in the Blue Jays favor. </sarcasm>

But the opposing pitcher didn't have to worry about him stealing

Huh? I seem to remember Juan Guzman spending a lot of time keeping Cecil Fielder close to the bag at first.

Drove me nuts watching that. When the ground starts to shake, you throw to second--simple.

Best Regards

John

I've always thought clogging up the bases was a little silly. You might not be able to take an extra base on a hit (1st to 3rd on a single for example), but I don't think one runner is going to slow another one down behind him. It'd be pretty tough for a batter to hit a double and not have a runner on 1st make it to 3rd.

The bigger issue with slow runners is hitting into double plays.

Bobby Crosby batted fifth in 92 games and scored 40 runs which is worse than Thomas.

That can't be right... [looks it up].... According to B-ref, Crosby batted fifth in only 13 games.

Last season with the A's, Frank was very slow (only 11 doubles in 466 AB), but he still finished second on the team with 77 runs scored.

A player's value is the sum total of his skill sets. There are essentially 3 - hitting, fielding, and baserunning. Yes, OPS ignores the baserunning part of offensive and is not the end-all of a player's offensive value. (Not withstanding the fact that OPS is not a great indicator of a player's offensive context-neutral value in the first place - but we'll pretend that it is the end-all).

First of all, we can easily guess as to a playe's baserunning ability. Speed usually, but not always, equals baserunning value. (Of course, the SB/CS part of baserunning value, we can easily look up and then use a lwts formula like SB*.18-CS*.45 = net SB runs.) The question is not whether Thomas' lack of speed decreases his value or Pierre's speed increases his value (from baserunning of course - the rest of their speed value is already acconted for) - that is obvious - the question is by how much.

We can easily quantify how much their hitting differs in theoretical runs. This year, Pierre's lwts per 150 games was -14 and Thomas' was +18, a difference of 32 runs or around 3 wins. The next step is simply to see (measure) what the difference is in their baserunning value in runs. Then we can stop using words like, "Yeah, but he clogs the bases," or, "But what about Pierre's speed on the bases," etc. that only confuse and confound the issue. As someoen smart once said, "Once the facts are on the table, most arguments cease to be rational."

The important thing is to get an idea of the typical magnitude of baserunning value as compared to offensive and defensive value. That way the silly "words" arguments can cease as well, if we know what a typical fast or slow (good or bad actually) baserunner brings to the table. Is it 20 runs, 30, 10, 5? How much difference does speed really make that is not already measured by a player's offense (lwts) and SB/CS numbers? (Keep in mind that all the good research that I am aware of about the impact of stolen base threats on the bases indicates strongly that it is a wash between affecting the pitcher and the batter, with some interesting things that I won't get into (you can read about them in The Book if you want).

It is easy to figure out someone's baserunning value. Someone at BP (I think) and I forgot who it is, does that occasionally. Bill James and Dewan do it I think, and I have been doing it for a while now as part of my Superlwts player evaluation system. I did not follow what this guy wrote in the article above, but I assume he basically did the same thing. You simply look at how often a runner addbances the extra base or not (or gets thrown out) on a single or double to the various sections of the OF, as compared to the average baserunner. You can also look at how often he gets forced at second on a DP ball, how often he tags and advances on fly balls and how often he advances from second to third on various ground balls. Some people (the BP guy I think) also look at how often a a player advances on a WP or PB, although I don't like that since most players either advance or they don't and it usually has very little to do with speed or ability. Of course, by definition a WP or PB have to be a baserunner advance so I don't what data he is using for that (pitches in the dirt or ones that get away from the catcher?).

Anyway, if we add all that up, while it is not perfect, it gives us a darn good idea, like offensive lwts and UZR (or other PBP defensive metric) of the exact value of a baserunner, especially if we have multiple seasons worth of data to smooth out the fluctuations and reduce the noise.

Here are the leaders and trailers for 2007 in baserunning lwts (runs above/below average) in total lwts (not per any number of games). Keep in mind that these are one year totals and do NOT represent the true one year value of baserunnng. To get that, we need to look at the range of numbers ovre multiple years and then regress that (shrink it toward zero) a little bit. Also keep in mind that the range of offensive runs in true ability is like -20 runs per season for a replacement hitter (although there are hitters like Neifi Perez, Brad Ausum, and John McDonald, who are true -30's, and either sleep with someone's daughter, have incriminatng pictures of a GM or manager or are really good defensively) to +40 or 50 for a hitter like A-Rod (+70 for a once in a lifetime hitter like Bonds), and -20 to +20 on defense, again with occasional outliers beyond these limits. These are true talent values. You will see numbers outside these ranges in any small periods of time all the time.

So what is the range for defense?

2007 leaders and trailers in baserunning lwts:

Best

Sizemore +7
Figgins +6
Zimmerman +6
Suzuki +6
Rollins +6
Castillo +5
A-Rod +5
Holliday +5

Pretty much speedy runners with a few who are not known as speedsters. Either the non-speedsters got a little lucky in 07 or they are very smart runners, or both. In the long run you will see that almost all of the leaders are fast and almost all of the trailers are slow. BTW, Sizemore is a MONSTER overall player.

Here are the trailers:

Garko, Ryan -6
Griffey Jr., Ken -6
Delgado, Carlos -6
Kotchman, Casey -6
Feliz, Pedro -6
Lowell, Mike -6
Posada, Jorge -6
Thomas, Frank -6

As you can see, all dead slow players. Yes, Griffey is dead slow! BTW, with his defense and baserunning, he is one of the worst players in baseball - well below replacement level. One big reason why CIN is such a bad team - paying a below replacement level player superstar money! That should be a felony.

So Thomas was -6 runs total last year. Pierre by the way was +3, a 10 run difference.

Now, let's look at the leaders and trailes for 4 years combined (04-07). That gives us a better idea as to the true range of values, like the -30 to +40 for hitting and -20 to +20 for fielding.

There are only 2 players who have more than 2000 PA and are more than +4 per 150 in baserunning:

Figgins +5
Taveras +5

Both speedsters of course.

Marcus Giles and Damon are the only qualifiers (over 2000 PA) who are +4.

I think we can safely say that the upper limit in baseruning true talent is +5 runs per season. By the way, 1 run is about 5 points in OPS (I think of the top of my head). So you can add anywhere from 10 to 25 points in OPS to a good or great baserunner.

The worst, with over 1500 PA over the last 4 years are Ben Molina, Griffey, Piazza, and Giambi, all slower than dirt. They are at -5 runs per 150. Molina is -6. So again, figure that -5 is the lower range of baserunning. Again, subtract 25 points in OPS for the few worst runners and generally 10-20 for a typical bad one.

Thomas and Ortiz are BTW -4 per 150 over the last 4 years and Pierre is +3.

That is all I have to say on that matter.

The bigger issue with slow runners is hitting into double plays.

As is often the case, when people make shizit up, they tend to be wrong. It ain't (the bigger issue).

The biggest determinant of hitting into a DP is whether you are a lefty or righty and your GB/FB ratio.

Other than that, the spread among the best and worst batters hitting into the GDP, is very small:

The 4-year leaders:

Too numerous, all at +2 runs per 150.

The trailers:

Too numerous at -2 per 150.

Do those baserunning runs numbers include SB/CS or not?

If I had 9 Frank Thomases, I'd be a hell of a lot more worried about the defense then the fact that he was "clogging the Bases".

If I had 9 Frank Thomases, I'd be a hell of a lot more worried about the defense then the fact that he was "clogging the Bases".

I wonder what the scores of those games would be ...

If I had 9 Frank Thomases, I'd be a hell of a lot more worried about the defense then the fact that he was "clogging the Bases".

I wonder what the scores of those games would be ...

Especially if you were in the NL.

Great post MGL. Usually I don't understand half the things you say ( I know it's very complicated stuff), but this was very well written.

Yes, Griffey is dead slow!

And was 6 out of 7 in stolen bases last year -- the fact that Mr. Hamstring-Injury even attempted 7 SBs last year is pretty shocking.

And of course this does not change the fact that he is dead slow. But it does suggest that every manager should have a list the pitcher/catcher combinations Griffey stole on, and should give a green light to anyone who isn't in a full body cast if they get on base against that duo.

#14 thanks. #10, no, only "baserunning." As I said, SB/CS are easily available and to convert to runs, just use SB*.18 - CS * .46 or something close to that. The nice thing about slow guys is that they can't "cost" you anything in SB/CS runs although these days the average basestealer is positive I think (so 0 SB and CS is in fact a little less than average - not much). In the old days (a few years ago), the average stolen base attempt, including busted hit and runs, were net negative so that a 0,0 player actually was a plus!

I think that maybe Griffey can run fairly fast if he wants to. Ditto for Bonds. Or maybe occasionally when their legs aren't bothering them is a better phrase. Either way, I think that their poor baserunning numbers are of a somewhat different ilk than guys like Molina, Piazza, Posada, etc. IOW, I don't think that Benji Molina can steal off of those same catcher/pitchers. Plus since those guys used to steal bases they know how to. Didn't Dunn have a bunch of successes also?

I think that maybe Griffey can run fairly fast if he wants to. Ditto for Bonds. Or maybe occasionally when their legs aren't bothering them is a better phrase.

Ditto, I've noticed that when fast guys "lose" their speed to injury rather than to age/weight that they can every now and then rev the motor up pretty good, more so than the guys who have always been slow.

even someone perceived as a cripple, like Buckner in his 30s, could get going surprisingly fast when he gritted his teeth and just got going. (Buckner could seemingly only go full tilt running to first or stealing a base, he couldn't turn a corner running at full speed- his ankles probably would have snapped if he tried...)

mgl, have you been able to factor in how often Frank and his ilk pull up at first when many guys would be able to leg out a double?

Yes - Griffey is dead slow! BTW, with his defense and baserunning, he is one of the worst players in baseball - well below replacement level.

I find this hard to believe. Most defense measurements that I've read have him in the middle of the pack in '07 amongst major-league right fielders. Care to prove how, exactly, he is "one of the worst players in baseball"? That is quite an interesting claim.

MGL, if you don't mind, what are Beltran's and Reyes' baserunning numbers?

#19, no I don't care to "prove" it, unless you are planning to take me to court, in which case I might have put together some proof. You sound like a formidable opponent.

No I don't know hof often slow guys get singles that might be doubles from faster guys. I am sure it is a lot. But of course that is already included in everyone's offensive numbers, so it doesn't really interest me.

Reyes and Beltran are both very good baserunners, as you probably suspected. Beltran was +4 in 07 and +3 average for last 4 years. Reyes was +3 this year and +2 average in the majors.

Reyes and Beltran are both very good baserunners, as you probably suspected.

Yes I did suspect. Thanks.

No I don't know hof often slow guys get singles that might be doubles from faster guys. I am sure it is a lot. But of course that is already included in everyone's offensive numbers, so it doesn't really interest me.

Hmm. Clearly part of the whole "clogging the basepaths" issue...

Best

Sizemore +7
Figgins +6
Zimmerman +6
Suzuki +6
Rollins +6
Castillo +5
A-Rod +5
Holliday +5


One of the interesting things about that list is how many good, even outstanding, hitters are on that list, even good power. Castillo, Figgins and to a lesser extent Ichiro are the only "typical" speedy types.

MGL forgot to mention Tippett's study which was the first one I ever saw. I think it was probably Ichiro's MVP year (and if memory serves he compared him to Giambi). I don't remember if he went very far beyond those two players, though I think he went far enough to say Ichiro was the top that year. His findings were consistent with what MGL states. What I think I found most interesting was how few opportunities Ichiro had to take the extra base. Between guys not getting hits behind him to guys, yes, "clogging the bases" in front of him, he only had a handful of chances -- was it maybe as few as 15-20? Anyway, it was low.

So it's not so much that speed doesn't have great "value" in and of itself, it's that there are so few (non-base stealing) opportunities to use it.

It would be interesting -- not useful really but interesting -- to look at baserunning vs. hitting vs. fielding "value" on a "per opportunity" basis. Yes, the range of hitting "true talent" is far greater in raw run terms but you also get 600-700 opportunities to amass those raw run differentials. That is, the range in baserunning is about 10 runs (best to worst); the range in hitting is about 70-80 runs. But if there are only on average, oh, 50 opportunities per 150 to take the extra base vs. 600 PAs per 150, that actually makes baserunning more valuable on a "per opportunity" basis.

I suppose that wouldn't actually be the case given how MGL calculates it (as I understand it). Every single without a man on first provides an opportunity to stretch to a double, it's just that opportunity is weighted by the league average of anyone taking the extra base -- and for many singles, that will be very low. But it's still an opportunity.

Like I said, not really useful given that what you want to know is their overall value and you have little/no control over these types of baserunning opportunities.

Excellent post, Walt.

Tom Tippett's study on Ichiro and speed.

I do use opportunities, as you define them. I am not including singles that could be doubles, since as I said, they are already reflected in offensive numbers.

Here is another thing I forgot to mention. The "downside" to being a fast player: More IF singles. Why is that a downside you ask?

They are listed as merely singles in a player's offensive line of course, and thus they are given the same weight/value as an OF single in computing OPS, lwts, EQA, etc. In fact, they are worth (a little) less than an OF single, as any runners generally only advance one base. When I do (and report) my offensive lwts for each player, I compute the IF singles separately (with a lower value than an OF single), so I don't have to worry about that. (I also compute K, GO, and FO separately, with different values for LH and RH hitters; for example, a ground ball from a LH hitter has a higher value than for a RH hitter, as they move runners over from 2B more often and lead to fewer GDP).

In fact, they are worth (a little) less than an OF single, as any runners generally only advance one base.

Also, infield singles tend to occur with bases empty more often than does a typical single - in large part because a GB with runners on that might be a infield single with bases empty is often turned into a force play or an out advancing (e.g. a slow GB into the SS hole where a runner on 2B tries to advance and is put out 6-5).

-- MWE

One big reason why CIN is such a bad team - paying a below replacement level player superstar money!

Is that even legal?

Cincinnati's problem isn't so much that they didn't evaluate a player well - Griffey was every bit a superstar when he signed. Just that they gave him way too many years. His contract per season was probably below market at 12 million considering Jeter and A-Rod signed about the same time for 18-25 million. They just gambled and lost, and the last few years were pretty much just stuck with him.

I do use opportunities, as you define them. I am not including singles that could be doubles, since as I said, they are already reflected in offensive numbers.

There could be an upside to Frank Thomas hits as well - A double by Thomas is almost certain to clear the bases, unless its ground rule. I wonder if Thomas has driven in any runners from first on long singles?

I find this hard to believe. Most defense measurements that I've read have him in the middle of the pack in '07 amongst major-league right fielders. Care to prove how, exactly, he is "one of the worst players in baseball"? That is quite an interesting claim.

MGL is probably basing his assertion on UZR. UZR had Griffey as one of the worst RFs in the NL, at -16 runs.

Last 3 years Griffey has averaged +15 runs hitting. Throw in fielding and baserunning, and a -5 position adjustment, Griffey comes in about 1 win below average, though still above replacement level. He will be 38, so aging might knock him down to replacement.

Well, Griffey's hitting level doesn't seem appreciably worse now than it was six years ago; moving him to RF at least minimized how much he costs you defensively; and in any event, he's a first-ballot Hall of Famer, and those guys tend to age pretty darn well. I can believe that he's not all that great right now when you factor everything in, but I see no reason to think he's about to hit a substantial decline either.

but I see no reason to think he's about to hit a substantial decline either.

I can. Last year he played more games and had more PA than any year since his first with the Reds. I don't think he stays as healthy this year.

MGL is probably basing his assertion on UZR. UZR had Griffey as one of the worst RFs in the NL, at -16 runs.

Looking at RZR (Revised Zone Ranking) at Hardball Times, this year Griffey was eighth in the majors amongst right fielders. As I said yesterday, this seems to be middle-of-the-pack to me. Hitting-wise, he had the sixth highest OPS among right fielders.

Combine his fielding and his hitting with his baserunning, MGL says, and Griffey was "one of the worst players in baseball." His baserunning must really be atrocious, then.

Also, infield singles tend to occur with bases empty more often than does a typical single - in large part because a GB with runners on that might be a infield single with bases empty is often turned into a force play or an out advancing (e.g. a slow GB into the SS hole where a runner on 2B tries to advance and is put out 6-5).

Of course in these situations, they're as valuable as a "regular" single.

I've always been fascinated with Ichiro in this regard. I don't have infield singles, but we can look at some splits.

With a man on 2nd only, for his career, Ichiro is an outstanding 375/505/452 (about 40% of those walks are intentional). He has 112 hits in 299 AB. He has 11 doubles, 3 triples and 2 HRs, all of which presumably drove the runner home (plus himself twice) accounting for 18 RBI and 16 hits.

So he has 98 singles ... which have driven the runner home 45 times.

Obviously there could be a lot of context around that. The Ms have frequently had Cs at the bottom of their lineup during his tenure, so maybe slow runners. You'd want to control for the number of outs -- why risk scoring the guy if there are no outs? On the other hand, some of these may have been to the SS hole and the runner didn't even advance. And of course maybe that sort of percentage is more typical than I think.

So out of the blue, I checked out Jimmy Rollins -- 61 singles produced 36 RBI. Better but not huge. How about more of a slugger? Derrek Lee -- 88 singles, 39 RBI.

Since when has a man on 2nd not been in scoring position? What's the point of putting speedsters at the top of the lineup if Lee can't drive them in with a single? I don't think they're playing him shallow.

OK, 3 data points ... but, y'know, I'm just looking at the shape of the stat line. :-)

mgl: does the data you need for the baserunning values you've been citing in this thread extend back through the 1980's? Tim Raines is eligible for this year's Hall of Fame election, and it would be nice to be able to give an appropriate estimate for that portion of his value.

I don't know if you get into arguments like this, but where would you rank Raines as compared to Tony Gwynn?