Productive outs

[brettac1]

The amount of times that the situation you mentioned actually occurs is so small that it is statictically irrelevant, though. More games are won with a 3-run 1st inning bomb by a power hitter than with a weak out to the 2nd baseman in the 8th inning.

 

Emotions are usually quite easy to identify.

 

Sure, but their minimal at best effects on games aren't.

[BrewCityUnit]

Brian, I did read the article. It is one guy's point of view. That is like watching a Michael Moore documentary and thinking it isn't biased.

 

Brian, sorry for calling him Ashlee Simpson but he does have the picture. I just hope you understand I enjoy arguing with you over these things. I respect your opinion and know you have great knowledge of the game. The beauty of this website is to have discussions like that. I know I am new here and don't have the same opinions as others on here but I can only bite my lip for so long.

 

I don't like the arrogance of brett or rulz and someone needs to challenge them every once in a while.

 

If the team was winning, I'm sure we would all be getting a long much better but then again Braynan wasn't productive last night (just kidding)

[Brian the Automator]

[ Brian, I did read the article. It is one guy's point of view. That is like watching a Michael Moore documentary and thinking it isn't biased. ]

 

So, what do you say about the facts he presents? What in particular do you disagree with?

[brettac1]

I don't like the arrogance of brett or rulz and someone needs to challenge them every once in a while.

 

I agree, like when you challenged rluz's percieved arrogance with some arrogance of your own:

 

One other thing that we differ on, I can remember all at bats.

[rluzinski]

One other thing that we differ on, I can remember all at bats.

 

If you can remember the situation and outcome of the 4000+ plate appearances the Brewers have had this year, you are either an idiot savant or a robot. Then you have to be able to compile those 4000+ plate appearances and try to coorelate certain events with the outcome of the game.

 

Not possible, except for maybe rainman.

[BrewCityUnit]

First of all he said that the A's went on to beat the Angels and then beat S.F. in the world series. That never happened, so he lost a lot of credibility right there.

 

Also, yeah it didn't work out for the Twins in that game but by swinging away and not bunting were they guaranteed to win? In my mind you can throw all the numbers away. They don't mean that much to me. If I was a Twins fan, I don't think I would have wanted them to bunt up one, down one yes, but not up one. But he did get the bunt down and they had two chances to score the runner so the decision wouldn't have bothered me too much.

 

I just don't think you can count every bat equally whether you are comparing OPS, or productive outs. Every at bat is a differnent situation where I would hope for best case senarios and worst case senarios. Obviously, players are going to disappoint and not do everything expected of them.

 

I think you need both OPS and productive outs. I would never say that a player that makes a lot of productive outs is great the same way I think Braynan sucks. I prefer players who have the ability to change their approach based on the situation and count. I don't know if that answered your question but when he started bashing Buster and started throwing out tons of numbers I started to lose interest.

[brettac1]

First of all he said that the A's went on to beat the Angels and then beat S.F. in the world series. That never happened, so he lost a lot of credibility right there.

 

His article was a spoof of the article Olney wrote on "Productive Outs", which the author links to in his article.

[Brian the Automator]

[ First of all he said that the A's went on to beat the Angels and then beat S.F. in the world series. That never happened, so he lost a lot of credibility right there. ]

 

Forget it. That's just proof that you only skimmed the article. I'm outta this argument.

[danzig6767]

The problem wih the productive outs therory is baseball is a 162 game season.It's not like football where 3-4 plays can decide whether a team is 10-6 and in the playoffs or 8-8 sitting home.

 

Sure in certain games a "productive out" can be very important,but in the grand scheme of a long baseball season it's far more important to just have alot of hitters that get on base/have power.Over the roughly 5000 at bats a team has,what they do in the 97% of the time where a "productive out" wouldnt even factor in,that will dictate far more how productive an offense is.

[GoodGod262]

Of course there are times throughout a 162 game schedule that there is such thing as a productive out and one needs to be able to do it. Does that mean I want to pile up a bunch of productive outs in a game, heck no

 

The problem I see with this team is when you need a sac fly or get a guy to third with less than two outs in a tight game they are incapable. Does it mean I think they should keep practicing these things in games, hell no they make too many outs to begin with.

[BrewCityUnit]

Sorry, I never read Moneyball and maybe his whole "A's beating the Giants" went right over my head. Did I miss something?

[endaround]

My biggest problem with the whole productive out thing is that people seem to treat them as something players should be always able to do unlike every other skill in baseball. Complaining that they didn't move the runner over is like complaining they didn't hit a home run. If the pitcher knows you are trying to move the runner over to third and you're left handed, chances are you'll see alot of pitches away that when you try to pull you hit to short. Or you'll see a bunch of inside pitches that you can't catch up to if you're right handed. If a man's on third the pitcher may go with high fastballs to keep you from getting somethign to lift. Players don't have this ability to flip a switch that always leads to a "successful" out.

[bjkrautk]

The problem with productive outs is essentially the same one that existed when we debated the Olney article: it is always preferrable to not make an out than it is to make a productive one.

 

I'd be more interested in tracking % of ABs yielding non-productive outs with runners on (strikeouts, pop-ups, foul outs, ground balls hit in front of the runner, etc.), and see how that correlates with runs scored.

 

 

I could repeat the arguments I've made in the past about how stats don't measure situational hitting (that is - if people could actually agree what the situation calls for). I'll save all the descriptions but one that I'm hoping 90% of people here will agree with:

 

If there's a runner at third, the hitter's primary job is to get that run home.

 

There are nuances to that scenario (if a team is down multiple runs, the number of outs, and the point in the game), but generally speaking, a run is preferrable to an out.

 

The reason that Branyan is such a frustrating hitter for so many of us to watch is that, in these situations, he strikes out half the time.

 

That's not an exaggeration, either.

 

Splits -- 2002-2005

Man on 3rd, < 2 outs: 22 K's in 40 AB

Man on 3rd*: 48 Ks in 96 AB

RISP: 115 Ks in 230 AB

 

*-Total of runners on 3rd, 1st & 3rd, 2nd & 3rd, and Bases Loaded.

 

I won't say that he scores "meaningless runs," or any of the observation-based criteria that make the more sabermetric among us cringe. I will say that his high number of non-productive outs in run scoring position negate most (if not all) of the production you'd expect from a player with his kind of OPS. Strikeouts are never productive; that's why some of us hate watching Russell play.

[BrewCityUnit]

Thank you bjkrautk. Your the man

[BrewCityUnit]

what those splits show is that when the pitcher needs an out, Braynan has no chance. He is not that good of a hitter. His OPS rises when no one is on base or it is a meaningless at bat.

[endaround]

Except it doesn't show that at all. Branyan 2002-2004 with man on 3rd<2:

 

.281 .362 .688 1.050

 

2005:

 

.250 .455 .625 1.080

 

But hey--keep pretending! What it does show is that pitchers fear Branyan in that situation. In 51 PAs, he has walked 11 times.

[Brian the Automator]

[ I will say that his high number of non-productive outs in run scoring position negate most (if not all) of the production you'd expect from a player with his kind of OPS. Strikeouts are never productive; that's why some of us hate watching Russell play. ]

 

If you're of the opinion that strikeouts are detrimental (I'm of that opinion, but my guess is "less so" than others), the quesiton is how many points should we "subtract" from Russell's OPS to account for these things?

 

I agree to the notion that if you took Branyan and his .900 OPS and swapped him a guy with an equal OPS and less strikeouts, obviously I would go with player B... The question is, how much of a detriment are Branyan's K's, and how do you quantify that?... because to me, a .900 OPS (and more importantly, a 1.001 OPS against righties) can't be nullified THAT much by k's.

[bjkrautk]

what those splits show is that when the pitcher needs an out, Braynan has no chance. He is not that good of a hitter. His OPS rises when no one is on base or it is a meaningless at bat.

 

Not....quite...

 

Actually in some of the situations I've mentioned, Branyan has a lofty OPS. With a runner on 3rd and fewer than two outs, his OPS is over 1.000. He's been very good with RISP this year in OPS terms, which is offset by being slightly below-his-average in the three preceeding years. (In case it's not clear by now, I'm aggregating the 1 & 3 year splits from ESPN.com -- I don't feel like recalculating for OPS.)

 

It's the inconsistency in Branyan's production that's highly frustrating, not necessarily the aggregate results. For people who consider situational hitting to be fundamental, watching a professional seem oblivious to this aspect of the game is aggravating. That's why he's ever the flashpoint for the 'observational' v. 'statistical' analysis types.

[wisportswriter]

I've neither the time or energy to read all the statistic posts in-between, but Automator, I couldn't agree more with your original post on this thread.

[Brian the Automator]

One important split:

vs. Lefties: 13 K in 20 AB (65% strikeout rate)

vs. Righties: 52 K in 142 AB (36% strikeout rate)

 

Given the concept that Branyan should be strictly a platoon player, I don't think a 36% strikeout rate is the end of the world.

[rluzinski]

I will say that his high number of non-productive outs in run scoring position negate most (if not all) of the production you'd expect from a player with his kind of OPS. Strikeouts are never productive; that's why some of us hate watching Russell play.

 

As has been beaten to death a million times before, while strikouts are certainly bad in some situations, overall they are barely worse at all. His do NOT nagate "most" or "all" of his production, as you would say.

 

"Runs Created" takes into account the negative effect of K's and Russell is STILL leading the team in RC/G ro anyone with meaningful amount of AB:

 

LINK

 

Are his K's self defeating? Sure, but not NEARLY to the effect that most people think. Everyone thinks that, with the bases loaded, all yo have to do is put the ball in play and good things will happen. That's just not the case in reality. When looking at actual play-by-play data and seeing how different baseball events affect runs scored, a K is about 20% worse than a normal out in an average situation. Again, this is using ACTUAL game information.

 

For people who consider situational hitting to be fundamental, watching a professional seem oblivious to this aspect of the game is aggravating. That's why he's ever the flashpoint for the 'observational' v. 'statistical' analysis types.

 

Very well said. Unfortunetely, it seems like the times Branyan laces a 2 RBI down the line is forgotten, over the times he K's w/RISP.

[bjkrautk]

If you're of the opinion that strikeouts are detrimental (I'm of that opinion, but my guess is "less so" than others), the quesiton is how many points should we "subtract" from Russell's OPS to account for these things?

 

I agree to the notion that if you took Branyan and his .900 OPS and swapped him a guy with an equal OPS and less strikeouts, obviously I would go with player B... The question is, how much of a detriment are Branyan's K's, and how do you quantify that?... because to me, a .900 OPS (and more importantly, a 1.001 OPS against righties) can't be nullified THAT much by k's.

 

Give me 1,000 years, a high-powered computer, play-by-play data of every game played in the 1990s, and a roomfull of monkeys with typewriters....and I may be able to give you an answer.

 

One of the problems with thinking of hitting as being dictated by situation is that it doesn't fit the pure statistics we commonly think about. I can say that I think it's possible to maximize runs scored by moving runners and taking advantage of situations, but that's about where my math runs out.

 

I will say that I don't think it's coincidental that, in the last four years, the two times that the A's have outproduced their expected runs created, it's been the two years in which strikeouts per AB were lowest. Actually, if you rank the prior three years...the fewer the strikeouts, the better they perform relative to OPS. (Projecting the current 2005 ratio over a full season's worth of ABs makes the relationship true for this year as well.)

 

To try to make any conclusive evidence of 'slappymetrics,' though, I'd have to crunch far more datasets to see if the same +/- relationship occurs. If it does, then I might be able to figure out a adequate discount rate for K's. (Even then, the discount would be based on team, and not individual, performance.)

[rluzinski]

But linear weights does exactly what you are talking about. It essentially looks at play-by-play data and figures out what every event's effecton runs scored is, based on different situations. It takes the weighted average of the effects of an event, based on the 24 base/out states (no on, 1 out, runner 1st, no out, etc...) and uses an average value of each event:

 

LINK

 

Comparing the relative negative value of a K and an out, you can see it varies depending on a sitation. For instance, with 2 out they are equal for obvious reasons. With a runner at third and less than 2 out, it worse to K than any other out. With a runner at 1st it's actually BETTER to K, since it's the ground out dp isn't possible then.

 

ANYWAY, taking a weighted average of all those events based on what situations occur most often (no one on and no outs is weighted the highest, for instance) what are the values of each? A K is worth -.31 outs, every other out is worth -.299. Hardly a difference at all.

 

In conclusion, while Branyan's K's especially hurt with runners on and less than 2 out, his K's are hardly worse in an average situation.

[dadofandrew]

Teams like the RedSox, Cards, Braves, Twins, Angels, would have found a way to get atleast one run if not two runs last night in the 8th inning without getting a hit

 

I'd just like to point out that the Twins have scored fewer runs than the Crew...even though they use the DH and not a P.

 

Makes everything else said moot to me.

[bjamesw22]

But linear weights does exactly what you are talking about.

 

 

Using linear weights the value of a single plus the value of a stolen base isn't equal to the value of a double. That makes no sense to me. There is no perfect stat that states exactly anything. They all have flaws one way or another. They aren't as cut and dried as 4 - 2 = 2.