The 3 toughest outs??

[TigersSlappy]

It makes me want to scream even more about our closers over the years.....

Interesting article from JoeP:

"The other day, I heard an announcer call the ninth inning, “The toughest three outs to get in baseball.” I will not name the announcer for a variety of reasons, the main one being that I could probably turn on the television tonight, scan the DirecTV baseball package and within a few minutes hear another announcer say the same thing, almost word for word. This “toughest three outs to get are the last three outs” is pretty well engrained into the grand baseball conversation."

Joe Blogs: The Three Toughest Outs

[Who is the Drizzle?]

I'm not sure I'd say that they're the 3 toughest outs for a pitcher, if the pitcher executes, but I think it's the toughest inning (along with the 1st for a SP) to execute his pitches.

And as someone commented there, I'm not sure about the inning-by-inning run stats used; there are probably many instances where 2 or 3 guys are left on base with nobody out because the game has ended. And of course with the straight hitting stats, teams will use their best reliever and I'm sure there are more bunts and things that do not get executed.

But yeah, if you got a pitcher that can handle the pressure, it's not any more difficult.

[tiger337]

I think there are some pitchers that are intimidated by the ninth inning, so the last three outs would be tough for them. However, I don't think there are that many of them in the majors and don't think teams usually allow those pitchers to pitch the ninth inning. For everyone else, the ninth inning shouldn't be more difficult than other innings.

[Mr. Bigglesworth]

I don't think there are that many walk-off wins, which is what would be required for the runners to be left on base with less than 3 outs to be unaccounted for in ERA.

Probably less than 5% of wins are walk-off, and not all of those involve multiple runners left on base when the winning run scores. So we are talking about a fraction of a small fraction of innings that would be impacted.

As such, I think the impact of not accounting for the potential runs on base when the winning run scores on ERA is small enough that it would not alter the conclusions one arrives at.

[Mr. Bigglesworth]

Also, if the relatively poor batting line in the 9th is due to the effectiveness of closers as a group, then the implication is the benefit of using a closer is greater to much greater than the perceived added difficulty in getting the last 3 outs. Which goes to Joe's point, I think.

[Mr. Bigglesworth]

Last thought on this, I think.

I often wonder if sayings like this persist in part because there may be some truth to them at lower levels, like high school baseball. Far more have played high school ball than professional, let alone MLB. So I wonder if there aren't fans who heard this from their coach in HS, and their coach was accurate enough when referring to HS ball, yet transferring the thought to cover MLB becomes inaccurate. Plus, it is exactly the sort of thing a HS school coach would tell his/her players in an attempt to have them keep their focus in the last frame.

Random thought: I also wonder if more fielding errors / fewer DP might lead to more 9th inning rallies at the lower levels as well.

[Who is the Drizzle?]

I don't think there are that many walk-off wins, which is what would be required for the runners to be left on base with less than 3 outs to be unaccounted for in ERA.

Probably less than 5% of wins are walk-off, and not all of those involve multiple runners left on base when the winning run scores. So we are talking about a fraction of a small fraction of innings that would be impacted.

As such, I think the impact of not accounting for the potential runs on base when the winning run scores on ERA is small enough that it would not alter the conclusions one arrives at.

If you figure 1 run on average is missed for the 5% of games, that reduces the difference from .08 runs to .03 runs. I found an article that through August or something of last year, 10% of games ended with walk-offs, so take out extra inning games and I don't think .05 runs worth of difference is far-fetched.

[Mr. Bigglesworth]

If you figure 1 run on average is missed for the 5% of games, that reduces the difference from .08 runs to .03 runs. I found an article that through August or something of last year, 10% of games ended with walk-offs, so take out extra inning games and I don't think .05 runs worth of difference is far-fetched.

a) 10% seems high to me at first glance. Are the Tigers involved with a combination of 16 walk-off wins and losses in a typical year? I'd guess the number is closer to 10 most years.

b) Even if it is 0.05 runs lost due to this effect, does it fundamentally change the conclusions in Joe's article? All Joe is arguing is there doesn't appear to be data to support the observation that the 9th inning are the 3 hardest outs to get. Increasing the ERA of the inning by 0.05 doesn't change that conclusion, does it?

[tiger337]

I often wonder if sayings like this persist in part because there may be some truth to them at lower levels, like high school baseball. Far more have played high school ball than professional, let alone MLB. So I wonder if there aren't fans who heard this from their coach in HS, and their coach was accurate enough when referring to HS ball, yet transferring the thought to cover MLB becomes inaccurate. Plus, it is exactly the sort of thing a HS school coach would tell his/her players in an attempt to have them keep their focus in the last frame.

.

I think this happens a lot. Another example is that there are more errors and mental mistakes made at lower levels. So, if a team plays the game the "right way" taking advantage of other team's mistakes, running the bases aggressively to force errors, etc they can gain more on lower levels than they could in the majors.

[Mr. Bigglesworth]

You just described what I am trying to do with my son's little league team. Stuff like backing-up throws, having the right guy covering the base, hitting the cut-off man, taking the easy out, not throwing to a base if you can't nail the runner, running out all plays, etc. can win games fairly regularly at that level.

And, well, it is the right way to play. I know none of these kids are going to MLB, and some of them are pretty unathletic, but I feel just about anyone can learn to do these sorts of things to help a team win and be involved instead of always being stuck in right field picking at blades of grass.

[DaYooperASBDT]

Ballgames can be won or lost in any inning, but clearly the 9th inning can place more "mental pressure"
on the pitcher and the fielders - but the hitters would feel that pressure also. So perhaps it would then
balance out.

Mental aspects aside, it seems to me that closers have a huge advantage over SP's - they generally throw
under 30 pitches (except for the big boy) so tiring is not an issue and they can get away with throwing more
"arm stress" pitches like the slider, split, or screwball.

So maybe it's really the toughest inning for the batter, moreso than the pitcher.

[Who is the Drizzle?]

a) 10% seems high to me at first glance. Are the Tigers involved with a combination of 16 walk-off wins and losses in a typical year? I'd guess the number is closer to 10 most years.

b) Even if it is 0.05 runs lost due to this effect, does it fundamentally change the conclusions in Joe's article? All Joe is arguing is there doesn't appear to be data to support the observation that the 9th inning are the 3 hardest outs to get. Increasing the ERA of the inning by 0.05 doesn't change that conclusion, does it?

a) It was 220 in 2010 and in 2011 it was projected to be a bit higher as of August. With 2430 games, that's about 9%.

b) When someone is using statistics to make an argument, it better be done correctly, especially if they are criticizing those that use conventional wisdom in it's place. Based solely on this article, I would call his conclusion inconclusive. If you adjust for tight games and quality of pitcher, I would not be completely surprised if the outs are slightly tougher to come by. I don't *think* they are, but it's possible.

[Mr. Bigglesworth]

b) When someone is using statistics to make an argument, it better be done correctly, especially if they are criticizing those that use conventional wisdom in it's place. Based solely on this article, I would call his conclusion inconclusive. If you adjust for tight games and quality of pitcher, I would not be completely surprised if the outs are slightly tougher to come by. I don't *think* they are, but it's possible.

I suspected this argument was coming, but held off addressing it until it was made.

Joe's piece was a blog entry, not an academic piece. I believe the data provided was more to illustrate the point that batters as a group do not appear to be harder to get out as opposed to demonstrating it conclusively. I also think he qualified the data as having its' flaws upfront, as opposed to presenting the idea that it was undeniably conclusive.

I suspect if he were writing a piece for SABR or something, it probably would have been more thorough and better supported, but then it probably wouldn't have appealed to the casual reader as much, nor would it have been as concise. I think for the format and what I presume to be the readership targetted, the data provided, as well as how it was preented, was sufficient.

[Who is the Drizzle?]

I suspected this argument was coming, but held off addressing it until it was made.

Joe's piece was a blog entry, not an academic piece. I believe the data provided was more to illustrate the point that batters as a group do not appear to be harder to get out as opposed to demonstrating it conclusively. I also think he qualified the data as having its' flaws upfront, as opposed to presenting the idea that it was undeniably conclusive.

I suspect if he were writing a piece for SABR or something, it probably would have been more thorough and better supported, but then it probably wouldn't have appealed to the casual reader as much, nor would it have been as concise. I think for the format and what I presume to be the readership targetted, the data provided, as well as how it was preented, was sufficient.

I don't like blog posts that are dumbed down to the point where an obvious point that was not addressed sufficiently. I mean, it's not like I read it as an academic piece, I just skimmed the thing. And again, if you are going to imply that conventional wisdom recited during a ballgame is wrong, don't be equally lazy with your statistics. I'm sure if someone dug deeper with announcers you would get a more nuanced answer that TOUGHEST OUTS AAAHHHH. But no TV analyst can go any deeper during a game than the internet's equivalent of a blog entry, so to do likewise does nothing.

But I guess I was not his target audience.

[Mr. Bigglesworth]

a) It was 220 in 2010 and in 2011 it was projected to be a bit higher as of August. With 2430 games, that's about 9%.

I stand corrected.

Thank you for looking that up. I really ought to have done that myself.

[Mr. Bigglesworth]

I don't like blog posts that are dumbed down to the point where an obvious point that was not addressed sufficiently. I mean, it's not like I read it as an academic piece, I just skimmed the thing. And again, if you are going to imply that conventional wisdom recited during a ballgame is wrong, don't be equally lazy with your statistics. I'm sure if someone dug deeper with announcers you would get a more nuanced answer that TOUGHEST OUTS AAAHHHH. But no TV analyst can go any deeper during a game than the internet's equivalent of a blog entry, so to do likewise does nothing.

But I guess I was not his target audience.

I don't think his statistics were as lazy as conventional wisdom.

I think he did a reasonably good job illustrating that it does not appear the last 3 outs are the hardest to get, and I'd say his analysis, as simple as it was, was a step further than most baseball commentary on TV or radio, yet it could easily be worked into those broadcasts.

Had he considered runners left on base in a walk-off wins, would it impact his conclusion?

[tiger337]

b) When someone is using statistics to make an argument, it better be done correctly, especially if they are criticizing those that use conventional wisdom in it's place. Based solely on this article, I would call his conclusion inconclusive. If you adjust for tight games and quality of pitcher, I would not be completely surprised if the outs are slightly

This has been a common criticism of yours but, while I agree to an extent, sometimes you need to take it one step at a time. Unless Joe wants to lose 95% of his audience, he needs to start out with something simple. The next thing that might happen is the hardcore saber community takes the question to their more analytical sites and tackles it more aggressively. Joe's contribution was to get the discussion started. Bill James is very good at that too. He often gets ripped for oversimplifying things in the beginning, but by doing that he gets the ball rolling.

[Mr. Bigglesworth]

I always thought James was a writer first and a statistician second. I view Joe the same way.

I look at it this way - Joe is trying to communicate an idea as concisely as possible. The stats provided demonstrate reasonably well that there isn't much, if any, difference between the 9th inning and other innings offensively. Certainly not enough to support the idea that the 9th inning is the toughest inning to get outs at first glance.

Is it perfect? No. He'd probably be the first person to admit that.

Is it likely that if he included the extra analysis that the conclusion would change? I think no. So to me, the oversight is pretty insignificant.

[Who is the Drizzle?]

As I said in my initial post, I don't think there is any difference from inning 3-9 except for the pressure that comes from closing out a game, and typically the guys that would wilt under that are not pitching the 9th. And in fact there is also conventional wisdom that says certain guys pitch better in the 9th because of that pressure. I think if TV Guy X felt the need to go deeper for his audience, he would probably say it is the pressure that makes it tougher, and he would probably also say that a typical closer would pitch better under pressure and feed on his adrenaline. So in a sense, conventional wisdom would also say that for certain guys it's easier to pitch in the 9th. (Note that this is also a problem I have with trying to over-analyze CW, the very nature of analysis tends to focus only on one piece of data, and it tends to interpret it literally, where CW by definition is not often subject exclusive and not often taken as literal.)

[tiger337]

where CW by definition is not often subject exclusive and not often taken as literal.)

I think conventional wisdom is often taken literally and accepted widely as absolute truth.

[jackson_cannery]

Ugh, he used "your" incorrectly in the second-to-last paragraph. I think I just lost all respect for Joe.

Edit: I'm kidding, sort of. It was a good article until then.

[Gehringer_2]

Actually, the logic of Joe's critique in his very 1st example that he uses to set up the piece is seriously flawed. It is immaterial if Cano is the 1st or 50th hardest guy in the ML to strike out looking, the fact is that at a <2% rate, you could easily watch dozens of Yankee games and never see it happen, which is exactly what McCarver said, and about which he was absolutely correct (strange as that may seem for McCarver!). That there are many other guys who you might never see take a called third has nothing to do with whether you are likely to see Cano do it. So Joe starts right out confusing the concept of 'rare event' in absolute terms with that of "rarest event" in comparative terms.

[tiger337]

Actually, the logic of Joe's critique in his very 1st example that he uses to set up the piece is seriously flawed. It is immaterial if Cano is the 1st or 50th hardest guy in the ML to strike out looking, the fact is that at a <2% rate, you could easily watch dozens of Yankee games and never see it happen, which is exactly what McCarver said, and about which he was absolutely correct (strange as that may seem for McCarver!). That there are many other guys who you might never see take a called third has nothing to do with whether you are likely to see Cano do it. So Joe starts right out confusing the concept of 'rare event' in absolute terms with that of "rarest event" in comparative terms.

That's true, but if something is generally a rare event, isn't McCarver misleading the audience by singling out Robinson Cano? He is saying that Cano striking out looking is a rare event as if it is unique to him

[Who is the Drizzle?]

I think conventional wisdom is often taken literally and accepted widely as absolute truth.

Not if people are effective at seeing the forest from the trees.

[Gehringer_2]

That's true, but if something is generally a rare event, isn't McCarver misleading the audience by singling out Robinson Cano? He is saying that Cano striking out looking is a rare event as if it is unique to him

You could take it that way, but I think it is also perfectly fair just to take McCarver at the actual meaning of his words and lay the responsibility for any additional inferences taken on the hearer. If he had said "It's rare to see Porcello throw a curve ball", - also a rare event which might have been seen in a game at least once this season, would it imply RP was the only pitcher in the league that didn't throw the curve?

[tiger337]

Not if people are effective at seeing the forest from the trees.

But people often are not (or don't try to be).

[Mr. Bigglesworth]

Not if people are effective at seeing the forest from the trees.

Apologies in advance, but I find this post curious given that your argument against the stats Joe used seemed focused on the fact that ERA does not capture the runs that baserunners potentially would have scored had the game not been ended in walk-off fashion.

I mean, yeah, those runners could influence ERA, but it seems like a fairly small effect to seemingly be hung up on.

[Who is the Drizzle?]

But people often are not (or don't try to be).

This is because they are not taking conventional wisdom literally.

[Mr. Bigglesworth]

You could take it that way, but I think it is also perfectly fair just to take McCarver at the actual meaning of his words and lay the responsibility for any additional inferences taken on the hearer. If he had said "It's rare to see Porcello throw a curve ball", - also a rare event which might have been seen in a game at least once this season, would it imply RP was the only pitcher in the league that didn't throw the curve?

The way the article read, it sounded as though McCarver was emphasizing the fact it was Cano who took a called 3rd strike that made the event rare, as opposed to a generic called 3rd strike being rare.

I didn't hear the commentary to offer an opinion on which was emphasized.

[jaymo]

Couldn't you just look at OBP for all 54 (51) outs?

[tiger337]

This is because they are not taking conventional wisdom literally.

You may have more faith in the masses than I do. I think most people are smart enough to not take conventional wisdom as literal truth. However, I don't believe most bother to put the time into doing so unless they are challenged. It's easier just to accept it. Thus, I do like to see conventional wisdom critiqued.

[tiger337]

The way the article read, it sounded as though McCarver was emphasizing the fact it was Cano who took a called 3rd strike that made the event rare, as opposed to a generic called 3rd strike being rare.

I didn't hear the commentary to offer an opinion on which was emphasized.

How much variation there is across batters also comes into consideration. If the average batter is called out on strikes three percent of the time, I wouldn't call two percent especially rare. In the case of pitchers throwing curve balls, it's clear to me that there is a wide variation with some pitchers throwing zero curves.

[Gehringer_2]

the statistical argument is confounded by the fact that once the better relievers as a class have been chosen as closers, it's no longer valid to compare 9th inning offensive stats to those earlier in the game - the two population universes are no longer randomly distributed wrt pitchers and hitters. He would have done better to go far enough back into history to a point before closing was a specialty, but then there would be other changed circumstances I'm sure, mainly the higher incidence of good pitchers throwing complete games......

[Gehringer_2]

The way the article read, it sounded as though McCarver was emphasizing the fact it was Cano who took a called 3rd strike that made the event rare, as opposed to a generic called 3rd strike being rare.

I didn't hear the commentary to offer an opinion on which was emphasized.

The point holds either way. Cano *is* in the upper echelon wrt taking called thirds, but Joe seems to be demanding that McCarver only make the statement if Cano is the very toughest. That seems to be an unfair burden of his own device - not McCarver's.

[Mr. Bigglesworth]

Again, if batters are hitting worse as a group because closers are being used, then the implication is that the effect of using closers is greater than perceived increase in difficulty in getting the last 3 outs.

[Mr. Bigglesworth]

The point holds either way. Cano *is* in the upper echelon wrt taking called thirds, but Joe seems to be demanding that McCarver only make the statement if Cano is the very toughest. That seems to be an unfair burden of his own device - not McCarver's.

Obviously rare is a subjective term, but if Cano is the 20th least likely regular to watch a 3rd strike, I don't know if I would term that rare. I mean, many teams, if not most would have a guy was did it less frequently. OTOH, if he is consistently top 5, I personally think rare would be a fair term to use.

In any event, I think the larger critique being made is that Joe doesn't believe Tim had actual data to back his claim, and was making it solely based on intuition / gut feel.

[Mr. Bigglesworth]

Couldn't you just look at OBP for all 54 (51) outs?

I think so, though I don't know as that fans, or baseball men, as a group view guys with high OBP as being tough outs, necessarily, even if that is what it literally means.

[Gehringer_2]

Again, if batters are hitting worse as a group because closers are being used, then the implication is that the effect of using closers is greater than perceived increase in difficulty in getting the last 3 outs.

Clearly possibly. But how do you separate the relative magnitude of the two effects? If closers are really all really great pitchers who would have ERAs half of what other pitchers would have, then the suppression of offense would still inhere despite the offensive "potential" being somehow greater.

Or is that what you already meant and I'm misreading your sytnax?

[Mr. Bigglesworth]

If batters, as a group, are batting worse because of closers, and closers are used predominantly in the major leagues, then there really isn't a need to state the last outs are the toughest to get, because as a practical matter the effect doesn't have a significant impact in close games.

Also, closers aren't always used in the 9th.

[Who is the Drizzle?]

Obviously rare is a subjective term, but if Cano is the 20th least likely regular to watch a 3rd strike, I don't know if I would term that rare. I mean, many teams, if not most would have a guy was did it less frequently. OTOH, if he is consistently top 5, I personally think rare would be a fair term to use.

In any event, I think the larger critique being made is that Joe doesn't believe Tim had actual data to back his claim, and was making it solely based on intuition / gut feel.

When I read that, I inferred it more as Cano doesn't take a lot of called 3rd strikes for a guy that takes a lot of pitches.


That's not at all what he said literally, but taken with what I know of Cano that's how I took it. Of course Miguel Tejada and Vlad Guerrero are not going to take a called 3rd strike, cw tells me that they swing at everything.

Now if you show me that Cano doesn't actually take a lot of pitches or has a higher called 3rd strike K rate compared with guys that do take a lot of pitches I would call McCarver wrong. But it's a TV telecast, he's going to lose viewers if he goes too deep or qualifies every observation just like Joe would lose readers if he goes too SABR in his blog.