It’s how often a player reaches base that’s important, not batting average, not RBI.
Baseball doesn’t have a clock in the sense that football or basketball does, but it has outs, 27 of them, and each one an offense spends brings the game closer to extinction. The players who reach base most often are the ones most likely to put off the inevitable death of the offensive effort. The more your players get on base, the more your players get a chance to hit, meaning you score more runs.

I have been reading some of the comments from people over the weekend. I see a few derogatory remarks about OPS. I agree OPS is not the perfect stat, but its a hell of a lot better than batting average. Batting average is a useless stat, its nearly as useless as RBI when judging the usefullness of an individual hitter.

RBI are opportunistic; RBI are a team stat and are not indicative of a player’s ability.
In 1985 Don Mattingly had a great year. The Yankees often batted Rickey Henderson first and Mattingly second. Henderson was having an even better year than Mattingly, reaching base 42% of the time and putting himself in scoring position constantly thanks to his 28 doubles, five triples, and 80 stolen bases–the last of which cost the Yankees only 10 caught stealing. At his peak, Henderson was the rare player where the rewards of stealing handily outweighed the risks. Hitting .324/.371/.567 behind this on-base dynamo, Mattingly drove in 145 runs and won the MVP award.

The next year, Mattingly was even better, improving his numbers to .352/.394/.573. Oddly, he drove in 32 fewer runs. The problem was Henderson, who saw his OBP drop to .358 in 1986, meaning he was on base less often. Better Mattingly + Worse Henderson = fewer RBI opportunities for Mattingly. If RBI were an expression of a player’s ability, we should hold the shortfall against Mattingly despite his being better than the year before. That doesn’t make much sense.

I know most people on the Redlegnation understand production, and for the most part I am preaching to the choir, but I would like one of the Redzoners “people who think batting average and RBI are useful stats” here to ENLIGHTEN me. Tell me why Battinag average is better than OPS, teach me the reasons RBI are a stat that can tell me how good a hitter is.

UPDATE: Brian makes a good point in his comments, I didn’t mean to disparage a whole class of people (Redzoners)

15 Responses

  1. Brian Erts

    Umm… Matt, I’d say that your “Redzoner” tag is misaimed.

    That poster doesn’t represent Redzone on any level when it comes to baseball acuman.

    Nor does he represent anything more than an gutless troll who slammed me and one of my posts without putting in a real name… hence the Redzoner moniker.

    BTW that argument is useless here IMO, We all know the deal… I also believe that we are all able to point someone who actually cares about learning more in the right direction.

  2. Joel

    I agree with Brian. Personally, I don’t have a problem if someone doesn’t want to take the time to learn about OPS or some of the more complicated stats. Everyone has the right to enjoy the game however they want to enjoy it. I do get tired of people who make weak arguments and disparage others because they have actually try to learn more than what is on the back of a baseball card.

    If you don’t want to be a “stathead,” fine. But if you going to go into an argument and ignore the mountain of evidence that is out there, well, you are the foolish one.

  3. Brian Erts

    The whole RBI thing as a measurement for a players skill is whack anyway, Deron Johnson 1965 Reds is my best example of what happens when a guy plays in a stacked lineup and has a career BA and SLG.

    130 RBI’s… never got to 100 again or before.

  4. Jeremy

    Here’s a down and dirty response to your request for someone to tell you why batting average and RBI are important. I don’t think that those two stats are as useful as OPS overall, but there are times when they should be examined.

    We can all understand why getting hits is important. If your whole team had OBPs of .400, but only ever walked, it would be tough to score runs, despite the high OBP. Your odds are less than 50/50 to get on-base, and you’d need to get on base 4 times before you got out three times, and that just don’t add up. Scoring runs is the goal of the offense, lest any of us have forgotten.

    Hits allow you to drive runs in from home, first, second, or third – not just third with the bases loaded. Hits = good. Of course OPS deals with this in a way, by looking at slugging percentage, which involves hitting. But slugging gives you as much credit for one HR as it does for four singles. Imagine a scenario where a guy comes up 8 times with a man in scoring position, and gets one HR, 3 walks, and 4 outs. That’s an OPS of 1.500, 2 rbi, and maybe 2 runs scored if he gets in on one of his walks.

    A guy who gets 3 singles, a double, and 4 outs in those same situations has an OPS of 1.125, but with 4 rbi, one time where he put himself if scoring position, and the same OBP (so the same potential for scoring runs being driven in by teammates). Certainly that’s more productive, so slugging can be deceptive in terms of production in some situations.

    Someone needs to knock the runs in, plain and simple. It is possible to have a high OPS based on walks and HRs alone, and these are not the most productive ways to knock runs in. Maybe those guys should be high power leadoff men, but they are not reliable rbi producers, which every team needs. This is just common sense, so I’ll provide some statistical backing for it, in case the stat mongers don’t trust their senses anymore.

    This season the top 20 rbi men at this time average 52.3 rbi, with a .299/.379/.538/.916 line. This is very good, as is to be expected from the top rbi producers in a given year. But what’s interesting to note is that the standard deviation from the batting average is less than that from the OPS (.033 to .101). Which means that the top rbi guys have a more consistently high batting average than OPS. Sure they are all good, but to knock in runs (which is how runs score) you need to get hits.

    Of course many of these numbers are situational, and a much more in depth analysis would be necessary to tease out the relationship between RBI and OPS and BA, outside of opportunities, park factors, pitching, and defenses. But the relationship between a high batting average and rbi’s has face validity (it makes sense), and there is at least burgeoning statistical evidence to support the relationship. In the face of that, to blindly examine OPS alone seems like folly.

    I know that Bill James has been doing more work on situational/clutch hitting as an important indicator, and though I’m not familiar with it, it seems like a move in this direction.

    And don’t get me wrong, if I had the option of a bunch of high OPS guys over a bunch of high BA guys, I’d take the OPS guys, because they usually very productive. The top 20 OPS guys – average of .974- have an average BA of .315, which I think is where so much of this confusion comes from. In most cases the high OPS guys are the high BA guys too, so it’s tough to s see that BA might have a role in production outside of what OPS covers. In comparison the top 20 BA guys – average of .327- have an average OPS of .927, so it’s a better bet to go with the OPS guys. With the high OPS guys in this sample you loose only .012 point of BA and gain .047 in OPS.

    But in some situations, OPS can be misleading.

  5. Chris

    While I’m a stat-appreciator, I’m not a stat-figure-outer. Does your calculation for standard deviation control for the fact that OPS has a much bigger range than BA, generally? (in other words, batting average, for all players, ranges from .185 to .350 – OPS ranges from 400 to 1400). And I still think that validating ANY stat to RBI is useless. PERHAPS looking at “RBI per RBI opportunity” would have some merit.

  6. Jeremy

    good point. The OPS and BA STDV are about the same proportion of the range, while slugging varies more and OBP varies less. hmmm… i’ll have to check it out with risp. It’s too hard for me to watch a guy like adam dunn and not think that BA is related to RBI, and it’s just common sense. But we’ll see.

  7. Brian Erts

    Here’s some wood for the fire

    In 70 PA’s with RISP Dunn is 1st in BB Casey is first in Batting Average

    casey has 26 RBI’s Dunn 18

    Caseys line .344/.416/.406/.822

    Dunns .204/.466/.469/.935

    Casey RC/27 in RISP 5.16
    Dunn RC/27 in RISP 8.30

    Casey 1 RBI every 2.4 ab’s Dunn every 2.77

    Casey 22 hits in that situation 4 were doubles

    Dunn 10 hits in that situation 4 were HR’s 1 was a double

    Dunn 21 BB in that situation Casey 10

    Dunn scored 28 times after that situation Casey 21

  8. Pinski

    If a team could actually get 8 straight at-bats with runners in scoring position that would be awesome, but its not likely.

    So instead lets say that there is a runner on first in each of those at-bats. Which one is better? The guy with the HR and the 3 walks? (2RBIS) Or the guy with the 3 singles and the double? (Maybe 1 RBI, as long as its not Casey or Griffey or LaRue on 1st)

    Thats the whole point. If you want to look at specific comparable situations, runners on first, etc, Then fine use batting average, RBIS etc. Thats okay. But when you don’t want to go that deep into the stats, and try to compare them in the exact situations, you need something better. OPS is that.

  9. Jeremy

    The example was not 8 straight PAs with risp for a team,just a sample of 8 for a particular player. Dunn’s already had 70 as posted above, so i don’t think it’s unreasonable to consider a sample of 8.

    Also, i don’t really think looking at production with RISP, is really that deep in the stats when you look at how far people go with RC and winshares and VORP. Chris Welsh can handle it.

    I just think it’s important to note that BA and RBI can be important stats, and that we shouldn’t throw them out the window just because OPS is en vogue. That leads to people claiming Adam Dunn is way ahead of where he is right now IMO, which is a ton of potential.

    I’d rather have David Ortiz or Miguel Cabrera than Adam Dunn, and he leads both of them in OPS.

  10. Pinski

    Like I said. Yes Dunn is leading overall in OPS. But not with RISP. Thats what I was trying to point out.

    But Dunn, has more Runs than those two other guys. And this gets to the chicken and the egg issue. Yes you need a guy to hit you in to score runs, but in the end if you can score 100 runs batting 250, who cares about BA?

  11. Blue


    vs. lefties: OBP-.278 SLG- .397 BA- .162
    vs. righties: OBP- .434 SLG- .639 BA- .284

    Late innings of close games: OBP- .323 SLG- .500 BA- .192

    Just some information for you to chew on.

    I have a theory that players with high OPS and low batting average achieve this by knocking the crap out of really bad pitchers, whereas hitters with high average are more consistent against good pitchers. This could explain why the Reds, with high OPS and low BA destroyed the D-Rays but couldn’t touch the Red Sox, with Clement, Arroyo, and Wells. Thoughts?

  12. Greg

    First, with any of these stats, you have to quantify # of ABs or PAs. An OPS or any other stat doesn’t mean much if you don’t surround it with a sample size parameter.

    I have theorized in the past that the recent A’s teams have had problems advancing in the playoffs because of the disparity between BA and OBP, without really making comment on OPS or SLG.

    It would seem that the better pitcher you are facing (which is often the case in the playoffs), and the better control that pitcher has, the less likely a player is to draw a walk….the less likely a player is to draw a walk, the more likely his OBP will approach his BA.

    Unfortunately, we are talking about small samples here and a non-uniform way of classifying control or “good” pitchers.

  13. Chris

    Here’s something I think we’re missing: These criticisms of Dunn would be valid if the goal of the offense was to score exactly one (no more, no less) run in every situation where there’s a RISP. If that was the goal, a walk is much less valuable, and the heavy-BA guy is much more valuable than the OBP-Slugger. BUT THE GOAL ISN’T TO SCORE ONE RUN – IT’S TO SCORE LOTS OF RUNS. Dunn getting on via walk for example, is valuable both in what it is (another baserunner), and what it isn’t (an out).

    Some expected runs information from 2004:

    Runner on 2nd only:
    0 Outs: 1.1596 were scored, on average
    1 Out : 0.7104 runs
    2 Outs: 0.3359 runs

    But when you add the second runner:
    O Outs: 1.4669
    1 Out : 0.9577
    2 Outs: 0.4605

    The probability of scoring at least one run (all >0 outcomes treated equally):

    Runner on 2nd:
    O Outs: 61.8% of the time
    1 Out : 41 % of the time
    2 Outs: 22.5% of the time

    Runners on 1st & 2nd:
    O Outs: 63.4% of the time
    1 Out : 42.3% of the time
    2 Outs: 22.4% of the time

    This is what I thought it would say – adding a second man on first behind a runner on 2nd doesn’t do much to help you score A run, but it does a lot to help you score MULTIPLE runs.

  14. Joel

    I have a theory that players with high OPS and low batting average achieve this by knocking the crap out of really bad pitchers, whereas hitters with high average are more consistent against good pitchers. This could explain why the Reds, with high OPS and low BA destroyed the D-Rays but couldn’t touch the Red Sox, with Clement, Arroyo, and Wells. Thoughts?

    I just did a brief glimpse at some of the top pitchers that Dunn has faced this year. My list included 37 pitchers and I will admit that I just eyeballed names to determine who was good and not (quick & dirty) but names included players like Smoltz, Clement, Prior, Wood, Zambrano, Milwood, Sabathia, Oliver Perez, Oswalt, Clemens, Pedro, Kip & David Wells, Peavy, Mulder and others. I tried to take guys that were at the worst identifiably average or a “tough lefthander”. While I did pick and choose with no statistical criteria, I did not look at the Dunn’s stats against each pitcher until after I had chosen my list.

    Here are Dunn’s (non-scientific) numbers against the top pitchers he has faced this year:
    121 PA, .268 avg, .413 OBP, .660 Slg, 9 HR, 19 RBI, 24 BB, 34 K.

    Again, this is just 2005 data, so the results aren’t definitive, but at the very least it is interesting to note that Dunn is hitting the better pitchers pretty well, especially considering that they are some of the best pitchers in the NL. I didn’t do this for the whole team, but I figured Dunn is usually the example for this type of argument and it was pretty easy for me to gather his numbers together.

  15. Jeremy

    That’s an interesting point, but it’s not like guys who have high batting averages don’t have reasonably high OBPs too. I doubt that the difference in runs would make up for the lack of rbi production.

    Using the expected runs from above, let’s look at FeLo and Dunn. With RISP Felo has an OBP about 170 points lower than Dunn. Ouch. over 150 PAs with RISP over the season, that would be 26 times Dunn got on that FeLo didn’t. If those came equally with 0, 1, and 2, outs, we could expect 6 more runs to score because of Dunn’s walks.

    At this rate, if both of them could play 150 games this year, Felo would have 80 RBI with RISP, and Dunn would have 50. And there it is.