Controlling the Running Game Is Overrated

This post isn’t going to be overly long, because the data mostly speaks for itself.

One of the primary traits that scouts look for in a catcher is a strong throwing arm, and catchers who can throw out opposing base stealers are often considered to be good defenders regardless of what else they do behind the plate. And, there’s no question that creating outs and intimidating runners into staying put is a useful skill, and a catcher who can shut down the running game can add value to his team.

However, this year’s Pittsburgh Pirates are proving just how small a part of overall run prevention that throwing out runners actually is. Here are the leaderboards for National League teams in opposing SB/CS:


Tm SB CS CS%
ARI 21 22 51%
STL 20 13 39%
LAD 36 21 37%
PHI 39 22 36%
SDP 59 25 30%
ATL 40 17 30%
CIN 33 14 30%
SFG 41 17 29%
COL 47 18 28%
LgAvg 41 16 28%
MIA 36 14 28%
MIL 52 17 25%
WSN 37 11 23%
NYM 43 12 22%
HOU 53 14 21%
CHC 53 11 17%
PIT 45 4 8%

The Pirates have thrown out four guys trying to steal all season long, and their CS% is less than half of the next worst team in the NL. Rod Barajas and Michael McKenry have been completely ineffective at throwing out runners, having allowed 41 net steals in the season’s first couple of months.

Despite their complete inability to gun down would-be-thieves, however, the Pirates are allowing just 3.6 runs per game, the third lowest total in the National League. So, despite being about as bad as a Major League team could possibly at this one aspect of defense, the Pirates overall run prevention is actually the strength of their franchise, and the reason the team is still contending for a playoff spot despite an historically bad offense. The story is similar over in the AL, where the Tampa Bay Rays are last in the league in throwing out base stealers (7 CS in 43 attempts), but allow the third fewest runs per game of any AL team.

This isn’t to say that allowing base stealers is good, or that you don’t want a catcher with a strong arm who can keep opponents from trying to take second or third – it’s just not really necessary for keeping opposing teams from scoring, and is more of an enhancement skill than anything. Throwing out base stealers is good, but it’s a very small part of a team’s overall run prevention. The emphasis placed on controlling the running game does not match the impact it actually has on a team’s ability to keep other teams from scoring.




Print This Post



Dave is a co-founder of USSMariner.com and contributes to the Wall Street Journal.

92 Responses to “Controlling the Running Game Is Overrated”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Cliff says:

    Let’s not actually try to quantify how valuable it is…

    +7 Vote -1 Vote +1

    • Dave Cameron says:

      We already know how valuable it is. A successful SB is worth about 0.25 runs, and a caught stealing is worth about -0.50 runs. Given 49 stolen base attempts, an average NL team would have allowed ~35 steals and thrown out ~14 runners. So, we’re looking at 10 extra SBs and 10 fewer CSs, so just move the decimal over and add the values together to get 7.5 runs lost due to their below average ability to throw out runners.

      In other words, not a big deal.

      Vote -1 Vote +1

      • L.UZR says:

        So 7.5 runs in 60 games = 0.125 runs per game, lowering the run prevention number to 3.475.

        Vote -1 Vote +1

      • Mark Himmelstein says:

        7.5 over less than half of a season seems like more than you’re implying here….if you entirely fault the catcher for those runs (not entirely fair), that would make a league average bat (for the position) roughly replacement level….no?

        Vote -1 Vote +1

      • Aaron (UK) says:

        The +0.25/-0.50 is fine, but a bit lazy really.

        Given we know the base/out situation for each steal attempt we can credit/debit players accordingly (though even then we’ll be missing important info like count). RE24 or even WPA are much more appropriate to use in analysing steals (as opposed to batted balls) because the baserunners have much more control over their intentions.

        This is not a specific dig at fangraphs but it does feel that sabermetrics generally has decided it doesn’t like steals much so it can’t be bothered to analyse them as thoroughly as it could.

        +13 Vote -1 Vote +1

      • drewggy says:

        That’s almost a full WAR over only 60 games. Not sure how that is “not a big deal” when they are now currently 1 game back in the standings.

        Vote -1 Vote +1

      • Dave Cameron says:

        It’s a two win pace over a full season for the worst a team could possibly be at this. No one’s saying this isn’t a thing that has any value, but if the range of possible outcomes at a team level is two wins over a full year, then the emphasis placed on throwing arm is clearly overstated.

        As for the leverage aspect of steal attempts, that cuts both ways. Yes, players are more incentivized to take a base in a close game, but they’re also incentivized to not make an out in that same situation. The value of a steal is higher and the cost of a caught stealing is higher. If you want to show that stolen bases are significantly more valuable than 0.25 runs apiece due to the average leverage of situations that they come in, feel free to do the legwork and prove your case.

        Vote -1 Vote +1

      • Mark Himmelstein says:

        But isn’t +/- 2 wins the maximum we’d expect from any single defensive position anyway? Obviously throwing out baserunners isn’t a defensive position, and some it has to be credited to the pitchers holding runners on, but if you credit half of that to the catcher, and say he can create +/- 1 win with his arm alone, isn’t that kind of a big deal? It’s not a big deal for any single pitcher, but for a catcher to add that much relative significance with merely one element of his defensive game does seem like a pretty big deal. No other defensive position can contribute a range nearly that big with just one element of their responsibility (i.e. a shortstop arm value, second baseman’s double play value, center fielder’s range value, etc).

        +19 Vote -1 Vote +1

      • L.UZR says:

        4-win pace though between Pitt and catcher for division rival StL.

        Vote -1 Vote +1

      • brendan says:

        dave, I know you are trying to make the point that catcher throwing is over-emphasized, but I’m not sure you’ve done so. As the other comments point out, +/- 1 win for catcher throwing is pretty significant, like the difference between a -10 UZR and +10 for a defensive player at another position. I think that it gets an appropriate amount of attention/emphasis right now.

        as for the fact that the pirates are still preventing runs effectively, I think it just speaks to their pitching/defense in other areas. not really so relevant.

        +6 Vote -1 Vote +1

      • RC says:

        “It’s a two win pace over a full season for the worst a team could possibly be at this. ”

        Its not that at all. We saw this a couple of years ago with the Red Sox, where they looked terrible… and then it got worse as teams started running more often.

        Also, 2 wins is worth what, $10M now? Thats probably more important in itself than who your manager is, etc. Its a big friggen deal.

        +13 Vote -1 Vote +1

      • Eric R says:

        “but if the range of possible outcomes at a team level is two wins over a full year, then the emphasis placed on throwing arm is clearly overstated.”

        Well that was just about worst possible vs average, which is only about half of the range of possible outcomes.

        2 wins between actual and average– isn’t that kind of like saying, “We have a replacement level shortstop already, so upgrading to an average SS is a waste of our time…”?

        Vote -1 Vote +1

      • Baltar says:

        Dave Cameron has taken absurd positions in each of last 2 articles and based each one on only 1 example (this one and the Ethier article).
        The theme of each is that small differences don’t matter much, which is eithere a tautology (one small difference doesn’t matter much) or false (small differences add up and do matter).
        I wonder whether he really believes what he has written in these articles, or whether he is deliberately taking crazy positions to stir controversy. Either way, I wish he would stop.

        Vote -1 Vote +1

  2. L.UZR says:

    Sorry but I don’t follow. Isn’t the real analysis to compare the impact (if any) of Pitts 8% CS rate on its 3.6 runs per game? For example, if I plug in the MLB average CS for Pitt, what would the RPG be?

    Vote -1 Vote +1

    • Dave Cameron says:

      They’ve been about eight runs below average in terms of throwing out runners, so their R/G with average CS% would be 3.47, 0.13 R/G less than it is now. That’s 21 runs over the course of a 162 game season. It’s not nothing, but given that this is as bad as a team could possibly perform in this area short of having Venus De Milo behind the plate, it’s clearly not that big of a deal.

      Vote -1 Vote +1

      • Justin says:

        Dave,

        We read articles on this site all the time about how a win or two can significantly impact the postseason race. Now you’re saying that a franchise like the Pirates, who clearly do not have the financial resources that other teams have, do not have to invest resources in this category because it’s overrated. Two wins is two wins. While the Pirates most likely won’t sustain this rate or their early success, clearly the running game is an important aspect of the game.

        +9 Vote -1 Vote +1

      • ImKeithHernandez says:

        Justin,

        I don’t think you’re working on the same assumption that Dave probably is. Specifically that one win (at most) can be made up by having a catcher who can’t throw at all, but can hit. At worst, a catcher that can hit but not throw should be equal to a good thrower that can’t hit, because not a lot of runs are actually prevented by caught stealings.

        Vote -1 Vote +1

      • Mark Himmelstein says:

        Right, but that’s true in pretty much every defensive situation. Catchers also have to block pitches, frame pitches, call games, and as you point out, hit. All of these things go into their value. Just like for a shortstop range, arm, turning double plays, and hitting all matter as well. But if its really a 20-40 run spread for the value of throwing out baserunners, of all singular defensive components, that’s easily one of the most potent. That’s not to say you play a catcher who can throw but can’t hit or block pitches, just that of various defensive components, catchers can create value with their throwing. I don’t particularly think this is overrated. There are plenty of bat-first catchers in the big leagues, and few catchers who play every day just because they can throw. They have to be able to do other things at least reasonably well.

        Vote -1 Vote +1

      • PXF says:

        That’s another example of how you could take context into account — how likely is a team to make the post-season if it picked up two additional wins. I realize the Pirates are 1 game out in the NL Central on June 13, but do you think they’ll be in contention at season’s end? More to the point — would you advocate that improving in this one area is the most cost-effective way to pick up two additional wins? (Maybe you would — maybe that’s a cheaper way to pick up WAR, I don’t know.)

        Vote -1 Vote +1

      • AF says:

        Of course, that would be equally true no matter how many runs the Pirates or Rays gave up per game overall. If you wanted to point out that stolen bases don’t account for that many runs, that would have been fine, even though it’s pretty well known. But you didn’t say that. You argued that stolen bases are unimportant because the teams that don’t throw out many runners happen to have good pitching staffs. Which is obviously fallacious.

        +11 Vote -1 Vote +1

  3. Joe says:

    doesn’t it make more sense to actually calculate the expected runs from each situation (before and after the stolen base attempt) and look at the difference to see what it’s actually costing the pirates? instead of pointing to their e.r.a. over a very small piece of the season and drawing conclusions from it, which seems shaky at best?

    Vote -1 Vote +1

    • Buddha6883 says:

      This is what I was going to say. We are drawing conclusions based on ERA and a two month sample size, when does that ever not get specifically denounced?

      Vote -1 Vote +1

      • Paul says:

        Agree with this comment, and just want to add that I think the phrase “specifically denounced” is criminally under-utilized.

        Yesterday a moron on talk radio projected out Billy Butler’s 1/3 season and said, “George Brett never hit that many home runs in a full season.”

        I would vote for Obama if he came out and said that if re-elected every time somebody says something like this, they go straight into the fire pit.

        Vote -1 Vote +1

      • Paul says:

        And just to add that SB/CS is one of those stats that has really super huge small sample biases in part because most of the stolen bases in the league are accounted for by a small number of players, and 1/3 of the way into the season teams have played one another one series only, for the most part.

        But I also think being so concerned with CS is just fundamentally wrong. Not only do catchers not bear full responsibility for this, one of the bigger factors that gets overlooked is the reputation of the catcher, and the willingness of teams to have average runners run on them. So overall, just looking at attempts and SBs would be more relevant.

        McKenry and Barajas actually have good reputations for throwing out runners, so I suspect they don’t get run on a lot by guys up and down the lineup. On the other hand, the Royals have a catcher who throws almost submarine style. Amazingly, he actually throws people out. But if Prince Fielder yells out to your 2B that’s he’s coming down on the first pitch when he sees your submarine throwing catcher back there, is that really comparable to even good base-stealers making sure they have the right jump and count before they run on Barajas? Don’t think so.

        Vote -1 Vote +1

  4. jj says:

    I’d be interested in seeing how many of those runners that stole a base came around to score vs what would have been expected to score if they did not steal. The fact is te Pirates have been outscored this year and any extra run prevention will be needed going forward if they continue to have such a horrible offense.

    Vote -1 Vote +1

  5. Aaron (UK) says:

    Wouldn’t net WPA for teams on SB/CS plays be a reasonable way of assessing the importance of this aspect of the game? Can anyone run that data?

    Vote -1 Vote +1

    • Baltar says:

      No, it would not. WPA for two identical events varies greatly by such factors as when the event happened and does not show the relative importance. For example, if a player hits 2 solo home runs in a game where his team won, in the 1st and the 9th, the second one has a huge WPA and the first one a small WPA. Yet the importance of each was exactly equal (one run).
      WPA is actually worthless for any purpose.

      Vote -1 Vote +1

  6. Krog says:

    Any reason why the Pirates have such low steal attempts against? If my counting is correct they rate in the bottom third in steals against. Perhaps their pitchers are doing a good job at limiting steals with pickoffs or quick deliveries.

    Vote -1 Vote +1

    • bada bing says:

      If they are only allowing 3.6 runs per game, I’m guessing they don’t allow too many baserunners, hence fewer opportunities for runners to attempt to swipe a bag.

      Vote -1 Vote +1

    • RC says:

      Pickoffs are counted as CS whether or not the guy was trying to steal.

      Vote -1 Vote +1

      • Mark Himmelstein says:

        That’s not true, pretty sure they’re only counted as CS if the runner attempts to go to second base. If the runner is picked off trying to get back to first they’re counted as pickoffs (PO).

        Vote -1 Vote +1

      • jpg says:

        RC is right. Pickoffs go down in the books as a CS no matter what.

        Vote -1 Vote +1

      • Steve says:

        RC is right but I think the point Krog was trying to make is that pickoffs keep a baserunner honest and limits his ability to even attempt a stolen base.

        Vote -1 Vote +1

  7. Jim Lahey says:

    It effects Win-Loss record more than scoring I imagine. 10 runs scored from baserunning over the course of the season means squat, but it could mean 0-10 wins if the situation was right. Players usually steal when they’re in tight games.

    Vote -1 Vote +1

  8. Aaron says:

    This is a ludicrously superficial argument. You could just as easily look at the Astros, Brewers, Mets and Cubs, who are all below average and rank 2nd, 3rd, 4th and 6th in Runs Allowed in the NL and make the opposite case using the same logic.

    +14 Vote -1 Vote +1

    • Edwin says:

      I thought the point was that even though the pirates are absolutely terrible at controlling the running game (be it by pitcher or catcher or combination), they are still able to prevent runs at a very successful rate. This seems to indicate that controlling the running game is a much smaller part of run prevention than defense or pitching. I think the data fits that.

      Vote -1 Vote +1

      • Buddha6883 says:

        But it’s focusing on one specific case. I could also look at one team over a two months period like, oh, the Pirates (hey that’s clever!) and say that run differential is overrated. We all know that isn’t true. It’s not a very good way to draw conclusions and go writing an article

        +6 Vote -1 Vote +1

    • Yeah, outliers are interesting, but you’d like to see a correlation for teams overall if you’re going to make the claims that Dave is making.

      Would it be absurd to think that a team could have a poor defensive CF, but still be good overall at preventing runs from scoring? Of course not. So then why make those claims based on Pit’s poor CS numbers?

      Vote -1 Vote +1

  9. Ian R. says:

    Huh. I look forward to a future post on how hitting home runs is overrated because the Dodgers are fourth in the NL in runs scored despite hitting the third-fewest home runs, and that hitting doubles is overrated because the Blue Jays and White Sox are third and fourth in the AL in runs scored despite being tied for last in the league in doubles.

    Saying that controlling the running game is overrated because the teams with unusual weaknesses in that area are still great at (other aspects of) run prevention is a flawed argument. The difference between the Pirates and a league-average team in this department (prorated to assume the same number of attempts, 49, as the Pirates’ opponents have actually attempted) is about 10 CS – probably about half a win, and that’s less than halfway into the season. That’s small, sure, but it’s not insignificant. It just looks tiny compared to the other things the Pirates are doing well.

    +26 Vote -1 Vote +1

    • drewggy says:

      Swinging at balls out of the strike zone is no big deal, because, hey, Vlad Guerrero!

      +8 Vote -1 Vote +1

    • Baltar says:

      Hitting home runs is overrated. The number of runs a team scores correlates more closely to the number of walks than to the number of home runs. (Compare the current rankings of teams in runs, HR’s and BB’s if you don’t believe me.)
      That does not mean that homeruns are irrelevant, nor that one HR is not more important than one BB.

      Vote -1 Vote +1

  10. Justin says:

    Pitchers affect the running game a lot more than catchers. Obviously, having a strong arm behind the plate is important, but if a pitcher’s delivery is very long a decent baserunner will be able to steal quite often.

    Vote -1 Vote +1

    • Richie says:

      I believe Bill James’ very earliest research showed that catchers were more important in throwing out stealers than pitchers were. You’ll have a handful of extreme pitchers (very easy or hard to run on) and otherwise it’s definitely more on the catcher than pitcher.

      Vote -1 Vote +1

      • fergie348 says:

        I usually really like Dave’s analysis, but this one baffles me. The strength and accuracy of a catcher’s throwing arm is only one part of the equation – The unload time of the pitcher plays at least as important a part, and the defense behind the pitcher also plays a part. I don’t think every situation is equivalent (for instance, stealing third with less than two outs is way more valuable than stealing second with two outs) and it seems that the actual expected run values should change accordingly.

        Sorry, this one clanked for me. Much more analysis needed to make any sort of coherent argument that ‘controlling the running game’ either is or is not important in preventing runs.

        Vote -1 Vote +1

  11. Stefan says:

    I wonder why there have been just 49 steal attempts against Pirates. Teams are trying to steal more often against the Dodgers, which are 3rd on the list. Only one team (Cardinals with Molina) have significant less steal attempts compared to a Pirates team throwing out just 8%. Why?

    Vote -1 Vote +1

    • SMITHERS!!!! says:

      Pirates are either barely winning or getting blown out. No one wants to run on a team they’re beating by 6 runs, it goes against the unwritten rules.

      Vote -1 Vote +1

    • Briks42 says:

      Im guessing the fact that their run prevention has been so good despite this terrible CS % probably means that they have allowed fewer baserunners than most teams, so there have been fewer steal opps for opponents?

      Vote -1 Vote +1

  12. Mike Grabowski says:

    I feel like this could be “proved” for any position on the diamond. For example, “It’s not really important to have a right fielder who racks up put outs.” Isn’t the overall strength of a team just the sum of its “enhancement” parts?

    Vote -1 Vote +1

  13. philosofool says:

    I think people are missing the point of this article. The point is this: that a catcher who has a CS% anywhere from 20 to 40 is acceptable, and the difference between 20 and 40 is not nearly as great as traditional ideas suggest.

    It would be much more valuable to have a 100wRC+, 20% CS catcher than a 70 wRC+ and 40% CS catcher.

    Vote -1 Vote +1

    • Mark Himmelstein says:

      Right, but isn’t that true of any offensive/defensive player comparison? Isn’t it better to have a 100 wRC+ SS with a -5 RngR than a 70 wRC SS with a +2 RngR? It seems like the value of throwing out baserunners compared to other defensive components is actually rather large, it just doesn’t break the idea that the spread of values of offensive components is larger than that of defensive components. In a defensive context, though, it seems throwing out baserunners can have a large impact compared to other defensive elements.

      +5 Vote -1 Vote +1

    • Buddha6883 says:

      The Cardinals have a 149 wRC+ catcher and are behind the Pirates by two games, catchers don’t matter! Seriously the point of the article would be fine if its remotely correct and there was a team that needed to listen (unless he’s trying to apply this to simply evaluating catchers) but we have no idea if it is correct and there is noone benching AJ perzinsky for Jose Molina (I know they aren’t on the same team, I was trying to exaggerate like you). We are drawing a conclusion based on only two months of data and using run prevention which can include pitching, other positions fielding, park, and luck, ignoring those factors and how other teams did in the same two months to draw a definitive conclusion. It is wrong on so many levels

      Vote -1 Vote +1

  14. Garrett says:

    Ah yes, the good ol Dave Cameron “I have a definitive answer to a question because of one nonconclusive outlier.” Always love those on a statistically based website.

    +5 Vote -1 Vote +1

    • Sam Samson says:

      In fairness to Dave, he also cites the Rays. So that’s two nonconclusive (is that a word?) outliers.

      Vote -1 Vote +1

  15. Richie says:

    1) Everyone says ‘such-and-such’ is really important if you want to not give up runs.

    2) Team A is as bad at ‘such-and-such’ as you can be.

    3) Yet Team A still isn’t giving up many runs. At all.

    There’s nothing wrong with pointing this out. Not every article’s gotta be chock full o’ numbers.

    Vote -1 Vote +1

    • Disagree says:

      There is something wrong with using this paper-thin reasoning to argue that controlling the running game is somehow overvalued. The analysis in this article is well below the depth/quality fangraphs usually delivers but it is presented as some kind of definitive argument when there are several flaws pointed out by people more knowledgeable than I.

      Vote -1 Vote +1

    • chuckb says:

      I agree to a point. However, a better argument is that the Pirates are good at run prevention despite being awful at preventing stolen bases not that preventing stolen bases just doesn’t matter that much.

      Vote -1 Vote +1

  16. Basil Ganglia says:

    I would guess that skills such as framing pitches and blocking errant pitches (***cough***Olivo***cough***) are significantly more important.

    Vote -1 Vote +1

  17. Andrew says:

    Does anyone else but me think that the Pirates LOB%, that is tied for the highest in the NL, is having more of an impact on their Runs Per Game? Regardless of a quality defender with a plus arm at Catcher, the less opposition batters that score the lower the Runs Per Game.

    Or am I simplifying it too much to think that it’s great to say even if you don’t throw runners out during a steal, if you have a 76.4% Strand Rate odds are they don’t score as much anyway?

    If the LOB% regresses to the league average, then the RPG increase and the whole purpose of this article is voided.

    Vote -1 Vote +1

  18. CircleChange11 says:

    Pretty worthless article that displays typical pitfalls of an amateur statistician

    Vote -1 Vote +1

    • Gabriel says:

      Actually, I found this exercise worthwhile. You are right, the original author did a weak job. However, the followup conversation was quite useful and is the sort of thing that could lead to interesting research.

      For example, there’s the possibility that the difference between how the Pirates stop the running game versus how the average NL team does is currently on pace to mean the equivalent of 2 games over the course of the year. The difference between the Pirates and the Cardinals would be 4 games — assuming the pace continues. That’s huge and actually much greater than I would have thought.

      Some lines of research that may exist out there based on this:
      1) In past years, what’s been the greatest differential between teams over the course of the year?

      2) What’s been the greatest differential in a tight pennant race?

      3) Is there any relationship between preventing stolen bases and how well opposing teams do in other baserunning metrics (1st to 3rd, etc)?

      Again, these may have already been written up on other sites. I guess my point is that though I wasn’t impressed by the thesis of the article’s author, the actual data and the followup critiques helped me learn more about this and think of ways they could be looked at.

      +5 Vote -1 Vote +1

    • 300ZXNA says:

      Considering you dislike his articles so much, why do you follow him around like a personal jeering section? If he offends you so much, why not stop reading his articles? Or are you jealous that he has attained so much in sabermetrics while you haven’t risen beyond the level of common heckler?

      Vote -1 Vote +1

  19. walt526 says:

    As pointed out earlier, the damage has only been modest because there have been so few steal attempts made against the Pirates. It was suggested that the Pirates are involved in more blowouts than the average team. Perhaps (this is an empirically testable proposition). It seems to me that a bigger story is that opposing teams have been negligent in exploiting this particular weakness of the Pirates.

    Yet if teams started running more on the Pirates–an 8% CS rate suggests that every opponent who is an above-average basestealer should be running nearly all the time whenever they have an opportunity–then that type of incompetence would really cost them. If the numbers were 184 SB against 16 CS, then the Pirates would be looking at about -38.5 runs (using Dave’s numbers for a quick back-of-the-envelope calculation), which would be a huge deal.

    Of course, given the small sample size, we can’t conclude that 8% CS rate is representative of Barajas/McKenry. Given prior performance, I think that we are reasonably certain that they are below league average (28%), but I’m not convinced that they are as historically bad as the raw 2012 numbers imply. As such, I agree with the overwhelming consensus of commenters that Dave’s overreaching with his use of a single outlier to make a rather grandiose conclusion that baseball conventional wisdom is wrong.

    Vote -1 Vote +1

    • Bookbook says:

      Totally what I was thinking. Is it true that teams can’t exploit this weakness very much, because most of the line up can’t steal a base, even against Venus de Milo? Or because the base stealers can’t get on often enough in win-relevant enough situations,that stealing would be advisable? Or maybe teams just aren’t eager to change their routines very much for one series, because it might throw them off of their sound judgement going forward?
      It’s very interesting, especially in the context of a new era where stolen bases are probably somewhat more valuable that the so-called steroids era that preceded.

      Vote -1 Vote +1

    • AF says:

      Exactly right. In short, preventing stolen bases is more important than throwing out runners. Accordingly, the relevant statistic is net SB against, not % of runners thrown out. A team that was truly “as bad as a Major League team could possibly be” would give up a lot more SB than the Pirates do.

      Vote -1 Vote +1

      • B N says:

        This. And it’s not like you have to dig real deep to find an example. Just last year the Red Sox gave up156 SB in 206 attempts. That translated to 14 runs lost, by the standard formulation. Clearly, that would have been the difference in a postseason berth.

        Moreover, they weren’t even the worst a team could possibly be. They allowed somewhere a bit shy of 2000 baserunners. So people still only tried to steal about 10% of the time. If the Sox had been much worse, people could have tried to steal… 15% of the time (21 runs)? 20% (28 runs)? It adds up real fast.

        The worst a team could be is letting over 3k attempts (since there’s usually two stealable bases). So you could have an attempt rate of 1.75 your OBP rate. With even allowing only a 75% success rate, you could give up something like 250 theoretical runs in a season.

        Needless to say, that would be bad. You caught what Dave conveniently ignored: it’s the increase in ATTEMPTS that can kill you with SB. The Pirates are actually not that bad and are probably just unlucky. If they truly were this bad (8%), the # of attempts would have skyrocketed and we’d be seeing a very different picture.

        Vote -1 Vote +1

  20. AF says:

    Similarly, having a productive offensive player at first base is overrated. The Red Sox first baseman, Adrian Gonzales, is 23rd in the majors among 1B in wRC+, but the Red Sox are second in the majors in runs scored.

    You know what else is overrated? Having your number one starter perform well. Tim Lincecum, the Giants’ number one starter, has an ERA of 6.00. But the Giants’ team FIP is 3.29, third best in the majors.

    Also overrated? Scoring runs. The Nationals and the Giants are near the bottom in runs scored, but they both have winning records and positive run differentials because they don’t give up very many runs.

    Vote -1 Vote +1

  21. Chummy Z says:

    Honestly, this data doesn’t speak for itself. Picking two teams with low CS a bit over 1/3 of the way into a season is not really…good. This would actually mean something if you went back and looked at longer stretches of data for teams with low CS totals relative to teams with average and high CS totals. Also, taking pitching and defense by the other 8 players into account would be nice as well. Heck, there are a ton of other variables and statistics that you could take into account. I love Fangraphs, but this is lazy work on what could be a very interesting topic.

    Vote -1 Vote +1

  22. This analysis just reeks of “correlation does not prove causation.” Many other variables need to be controlled before you can get any use out of such a study. You need to control for pitching, for one. The Pirates are above average in FIP. Fielding? Admittedly they are not very good there. What about the ballpark? Baseball reference says PNC is a very pitcher friendly ballpark.

    What we have here is very interesting, but the best conclusion that you can draw is “needs further study”. Actually doing a regression of CS% and RA shows a negative correlation. As expected, teams with higher CS% allow less runs. The difference is slight and and the r^2 value negligible, but it is there. Looking at only the Pirates while you have 29 other teams to examine isn’t a very thorough analysis. I’d like to see what happens when you control for pitching, ballparks, competition, and defense, but it really doesn’t matter. The amount of outs created by a catcher throwing out a runner is really negligible compared to the role of the defense and pitching.

    Vote -1 Vote +1

  23. MikeS says:

    “Proving” once again that hitting is the most important thing, pitching is second and defense and base-running are distant thirds.

    And that has nothing to do with the fact that hitting is the easiest thing to measure, pitching is second and defense and base-running are distant thirds.

    Vote -1 Vote +1

  24. Bob says:

    Lol @ “catchers who can throw out opposing base stealers are often considered to be good defenders regardless of what else they do behind the plate”.

    I don’t know what world Dave Cameron is living in.

    Vote -1 Vote +1

  25. mister_rob says:

    If you think controlling the running game is overrated, then you must think running games in general are overrated.
    Which I find fascinating, because you are the same guy who thinks the Crawford and Reyes deals were great bargains. 2 guys with middling power and inconsistant on base abilities, but who are known for their running games

    So if running games arent all that important, and range is not a factor for a LFer in fenway, what does Crawford supposedly do (did) that is worth 20m per year?

    Vote -1 Vote +1

  26. Oscar says:

    I honestly can’t believe Dave Cameron actually wrote this article. It’s, like, self-parody.

    Vote -1 Vote +1

  27. DJ Pauly Goldschmidt says:

    So maybe the Pirates are just getting lucky stranding RISP after they steal bases.

    Vote -1 Vote +1

    • cobradc23 says:

      This is what I was thinking. How much luck is involved? Are the Pirates letting baserunners steal at a 92% rate then stranding those particular runners 90% of the time (which would obviously be unsustainable)? How many of the stolen bases are stealing 2nd with 2 outs and the batter makes the third out? IMO, 1/3 of a season is not enough data to confirm what Dave is attempting to lay out.

      Vote -1 Vote +1

  28. Snowblind says:

    Perhaps this article was mis-posted, and it was really intended for NotGraphs.

    Apart from earlier points about how important a couple wins are to someone clawing their way to a division title or 2nd wild card, focusing on this one snippet is like saying we can poo-poo basestealing by measuring the average lead the baserunner takes. Of course the stat in isolation is near meaningless… but it contributes to a whole bunch of more fundamental things about how the catching and pitching is doing on a ballclub.

    The quality of the pitching in the first place, how the pitcher pitches from the stretch, handedness of the pitcher, pickoff moves, the pitcher’s comfort level with their catcher affecting their strategy (they might not throw certain pitches if they don’t think their catcher can keep a guy on, or don’t think their catcher can handle it at all), quality of the SS or 2b, and on and on… these things all effect and are affected by how the catcher controls the running game.

    And ability to control the running game is a partial proxy for the total catching ability and trustworthiness of the catcher, which affects the pitcher’s game plan. You can’t look at # of guys thrown out in isolation and map it to specific sabermetric stats. You look at ability to control the running game as representative of the intangibles the catcher brings to the pitching staff overall, not as some vital contributor to runs scored / runs allowed.

    Vote -1 Vote +1

  29. Keith Waters says:

    A player is not charged with a CS on a pickoff unless he tries to advance.

    Vote -1 Vote +1

  30. But says:

    Yadier Molina polished his ring while laughing at this.

    Vote -1 Vote +1

  31. Marver says:

    100% of me believes this article was published in order to make Jesus Montero look better.

    Vote -1 Vote +1

  32. Marver says:

    “This post isn’t going to be overly long, because the data mostly speaks for itself.”

    Clearly.

    Vote -1 Vote +1

  33. Choo says:

    It’s official: Miguel Olivo is the most worthless baseball player ever.

    Vote -1 Vote +1

  34. Joe says:

    “This post isn’t going to be overly long, because the data mostly speaks for itself.”

    Ironically, the author wasn’t listening…

    —-

    I nearly spit out my coffee when the connection of the Pirates run prevention was evidence that SB/CS are no big deal. If they were worst in the league in run prevention (and the absolute impact was the same) would it matter more? How about league average run prevention?

    So if there was a -20 LF on the same defense would it also not matter and not be worth looking at upgrading? Eh…. they give up few runs, why upgrade the defense to prevent even more runs? That would be silly, right?

    I realize this is more than a catcher issue and improving this could impact other areas (more slide steps could impact pitching effectiveness, different pitch calling with men on base, etc), but to say the impact is negligible is just flat out hilarious from a SABR inclined author.

    Vote -1 Vote +1

  35. Bip says:

    The quality-control that the fangraphs audience does on these articles is unrivaled, and I’m not being sarcastic.

    Vote -1 Vote +1

  36. Steve says:

    You cannot ever criticize a member of the media ever again for making perhaps incorrect cause and effect relationships. How is this different from the media praising Mike Scioscia’s use of small-ball because he wins a lot of games?

    Vote -1 Vote +1

  37. Dmitri says:

    What is the value of framing pitches? Any way that can be evaluated?

    Vote -1 Vote +1

    • B N says:

      Yes. But if information gets out about how people have been framing pitches, a lot of pitches serving sequential life sentences might be released back onto the streets.

      Vote -1 Vote +1

  38. rotofan says:

    I guess shortstop defence is overrated too: The difference last year between the best and worst fielding shortstops was 23.7 runs per 162 games while the difference in throwing out base runners alone this season projects to 20.25.

    You see, you can prevent runs even if you have the worst fielding shortstop in the league, so it’s more of an enhancement skill rather than a necessity.

    Or maybe there’s a better explanation. Maybe the pitcher plays a large role in preventing runs and maybe the fraction of run-preventions that’s left over is split 9 ways, or if you break it down to individual defensive traits such as range, errors, arm, etc, 30 or 40 ways, so that no one trait is essential.

    Vote -1 Vote +1

  39. Cidron says:

    not a factor… Gentlemen, Let me introduce you to Pudge Rodriguez and the howitzer on his shoulder. Dont tell me that his ability to gun down pretty much anybody didnt factor into the game. Sure it did. Moreso if the other team was a running team (ala the early 1980′s Cardinals) vs a slugging team. Today it isnt as big a factor due to the fact that there really are very few “running teams” out there.

    Vote -1 Vote +1

  40. spike says:

    Haven’t read all the comments yet, but it seems Dave is saying catcher throwing ability is over emphasized because the difference between the best and worst isn’t much in terms of run prevention.

    However, Dave, don’t you think that this is probably true BECAUSE there is so much emphasis placed on this ability? In other words, the reason why we don’t see a big gap between the best and worst is because it’s important enough that there just aren’t any catchers who are allowed to be so bad at it? Does that make sense?

    Vote -1 Vote +1

  41. I get the feeling that Dave wishes he could take this article back. He must see how silly and simplistic it is.

    Vote -1 Vote +1

  42. Excellent points made in this thread. I love the running game, I’m biased, and probably always will. I’ve seen way too many games decided by 1 run where great baserunning, or a late steal, or a bad throw by the catcher, or a pick off attempt that sailed over the 1st baseman’s head, changed everything.

    I realize this is a stat based site, and I love it, but this is why not all GM’s are dedicated “solely” to sabermetrics. The question in my mind would be “what would a GM pull from this data”?

    For instance, I’m sure the Pirates and Cubs are going to try to find guys who can defensively “close the gap” on base runners after this season. Smart guys will ask, at the expense of what? Always a good question.

    All readers need to do, in my opinion, is watch young Tony Campana and how much he affects the other team as one example. I think with most great base stealers, it’s not so much that what they do has little value, I think it has more to do with the team and hitters behind them.

    How many runs would Campana score, say, in the New York Yankees lineup?

    Vote -1 Vote +1

  43. Tree Climber says:

    The #6 org are poor at throwing out runners. Therefore, to fangraphs, it’s not an important stat.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *