Joe Posnanski with more data showing that closers simply don't matter

[scott91575]

I agree wholeheartedly that a run in the 8th is equal to the same amount as a run in the 1st. However I think runs later in the game can be much more difficult to come by. What gets lost in the entire argument is pressure and the sense of finality that teams face. If you have what is percieved as a lock down closer it makes a manager force the issue in the 6th, 7th and 8th innings. You are forced to take more chances in a close game because you "think" the realities of scoring off an "elite" closer is difficult. Go look at game threads when its a tight ball game and see all the nerves pouring themselves over the posts. Humans playing the game feel those nerves too.



I agree with you and think that managers already are doing that. Lets just take a look at the way the Reds were constructed down the stretch last season. Chapman was getting the ball in the 9th. But the Reds had two top tier relievers who were used in a high leverage situation. Marshall in many situations was used against a teams best left handed hitter and Broxton was used in high leverage situations in the 7th or 8th. Every manager slots his relievers in a certain way, often times setting aside their best relievers for hte highest leverage situation.

In reality until a manager decides to buck the trend, we will not have any "proof" that closers don't matter. Until we have a manager who decides to go closer by committee we won't know if that way works better than what has become standard over the past decades.

Wouldn't your argument go against a closer? If runs are harder to come by late, then that is where you put your worst pitcher. It's a self fulfilling prophecy. If teams put their best pitcher on the mound in the 9th, then of course it makes it harder to score runs. If they do it in the 8th, then the 8th becomes the harder inning.

[scott91575]

Exactly.

The reason why later innings are higher leverage is because there is less time for a team to come back. All runs are not equal. Runs scored late in a game are worth more.

If you score a run in the first inning of a tied game, it does little to change the odds of who will win the game. Score a run in the ninth of a tie game, and you've greatly changed the odds of who will win the game. Again, not all runs are equal, and not all innings are equal. Closers matter, big time.

This is so wrong I have no idea how to go into all of the statistics. Runs are runs are runs.

Here is the deal, if you reduce the number of high leverage situations where you can lose the game (by doing better in earlier innings), then your win percentage goes up even if you get worse in the 9th inning.

Let's say you score 5 runs in the first inning and win the game 5-2. Now, let's say in the same exact game, you are down 2-0 in the 9th and score 5 runs. You are telling me those runs in the 9th were way more valuable? Hell no, you won the game either way. It doesn't matter when you got them.

[scott91575]

Because baseball is played in the real world and not on paper or in a computer program

You are so right. I never thought of it this way. I engineer in the real world. Time to throw away all those stats I use. I am sure it would have no affect on the braking systems I have designed. Don't worry, the next time you press on the brake pedal, I am sure all will work out since I will be engineering in the real world instead of using those hockey computer programs and stats. Idiots, using data to make conclusions. Pfft.

[Kc61]

This is so wrong I have no idea how to go into all of the statistics. Runs are runs are runs.

Here is the deal, if you reduce the number of high leverage situations where you can lose the game (by doing better in earlier innings), then your win percentage goes up even if you get worse in the 9th inning.

Let's say you score 5 runs in the first inning and win the game 5-2. Now, let's say in the same exact game, you are down 2-0 in the 9th and score 5 runs. You are telling me those runs in the 9th were way more valuable? Hell no, you won the game either way. It doesn't matter when you got them.

The value of runs is not the issue IMO.

Anyone on here who has managed people knows that there are certain tools that get you better performance. These tools sometimes are less than perfect.

In managing 7 relievers, a team must consider more than leverage. They must consider workload, repeatability of tasks, ability to work multiple innings, effectiveness, perhaps handling pressure, ability to pitch consecutive days. There are probably a dozen variables.

The best way to deal with this problem is depth. If you have depth of good relievers, then defined roles will be fine because whoever you throw out there in any inning will be good.

I believe that in the regular season managers believe that more clearly defined and repeatable roles overall gets you better results, and that managing by "leverage" leads to overuse of top guys and other problems which I identified in earlier posts.

In a perfect world the "leverage" argument would be correct, but it's not a perfect world, and managers instead usually allocate work in a somewhat different way.

[757690]

This is so wrong I have no idea how to go into all of the statistics. Runs are runs are runs.

Here is the deal, if you reduce the number of high leverage situations where you can lose the game (by doing better in earlier innings), then your win percentage goes up even if you get worse in the 9th inning.

Let's say you score 5 runs in the first inning and win the game 5-2. Now, let's say in the same exact game, you are down 2-0 in the 9th and score 5 runs. You are telling me those runs in the 9th were way more valuable? Hell no, you won the game either way. It doesn't matter when you got them.

I take it none of you guys arguing against me have ever taken a logic course, lol.

Actually, in the example you presented, the runs scored in the first inning were less important, at the time that the were scored, than the runs scored in the ninth inning, at the time that they were scored.

Let look at the reverse of this. Preventing runs in the later innings is more important than preventing runs in earlier innings. Both are important, but in later innings, it's more important, because doing so does more to increase your chances of winning.

Look at a non-baseball analogy.

You're at the blackjack table. You win $5 on your first hand. Later, when you are down to your last $5, you win that hand as well. That last win was more valuable than the first one, because it kept you playing. Granted, winning the first hand an important part of you being able to keep playing as well, but not as important as winning your last hand. If you made a mistake on your first hand, you had time to change your strategy (lower you bets, be more or less agressive with doubling down, etc) to make up for that lost hand. You don't have that chance on your last hand. You have to win it it keep playing. It is more important. Period.

Same with baseball. Same with everything in life.

Events don't happen in a vacuum, the situations and circumstances surrounding every event alters it and makes it different from a similar event in different circumstances. I really don't understand why this is so hard to grasp.

[AtomicDumpling]

I take it none of you guys arguing against me have ever taken a logic course, lol.

Actually, in the example you presented, the runs scored in the first inning were less important, at the time that the were scored, than the runs scored in the ninth inning, at the time that they were scored.

Let look at the reverse of this. Preventing runs in the later innings is more important than preventing runs in earlier innings. Both are important, but in later innings, it's more important, because doing so does more to increase your chances of winning.

Look at a non-baseball analogy.

You're at the blackjack table. You win $5 on your first hand. Later, when you are down to your last $5, you win that hand as well. That last win was more valuable than the first one, because it kept you playing. Granted, winning the first hand an important part of you being able to keep playing as well, but not as important as winning your last hand. If you made a mistake on your first hand, you had time to change your strategy (lower you bets, be more or less agressive with doubling down, etc) to make up for that lost hand. You don't have that chance on your last hand. You have to win it it keep playing. It is more important. Period.

Same with baseball. Same with everything in life.

Events don't happen in a vacuum, the situations and circumstances surrounding every event alters it and makes it different from a similar event in different circumstances. I really don't understand why this is so hard to grasp.

The bolded part just isn't accurate. As I explained earlier, you are only considering games that are close in the late innings. You are totally ignoring games that are not close in the late innings. In games that are not close, those late runs have far less value than the early runs that won the game. If you include all games and all innings there is no difference in value between early runs and late runs.

In a game that does end up being close in the late innings, those late runs are very valuable as you say, but they are no more valuable than the runs scored earlier because without those early runs the game wouldn't be close anyway.

Your blackjack analogy works the same way. If you had won more hands early on you wouldn't have run out of money in the first place and never would have been down to your last $5.

I think perhaps (and I could be wrong) your understanding of the Win% numbers from FanGraphs is not quite right. That data is very interesting, but I am guessing it doesn't mean what you think it means. I don't have much time to explain it all right now, but I will take a short stab at it. Here is the definition from FanGraphs: Win Expectancy is “The percent chance a particular team will win the game based on the score, inning, outs, runners on base, and the run environment. These percentages are calculated using historical results of thousands of real MLB games, meaning if a team is losing and has a 24% win expectancy, only 24% of teams in similar situations in the past have ever come back to win."

Notice that the definition does not say anything about the value of a run, because you don't know the actual value of the run until the game is over. You mentioned earlier that the Win%'s can change faster in the late innings and this is true if the game is close, but it is not true if the game is lopsided. An early run doesn't swing the percentages as much as a late run in a close game, but this is only because at the time you don't know how many runs will be scored later. Historically, any given situation was followed by a wide range of possible outcomes. Sometimes that run was the last run scored in the game. Sometimes there were many more runs scored in the game. Because there is such a wide range of possible outcomes you can't say with any certainty how much that run is worth until the game is over.

In a 1 run victory, every run is worth the exact same amount because you needed every single one of them in order to win the game. In a lopsided game technically all the runs are still worth the same as each other, but essentially the early runs were the most valuable because they were the only runs you needed to win the game.

If this doesn't make sense to you I guess we will just have to agree to disagree on this issue, which is fine.

Let me ask you the same question I asked bucksfan2:
Question: You are managing the Reds tonight and are given a choice between A) scoring 5 runs in the game, all coming in the 1st inning (when they are supposedly less valuable, giving the other team the rest of the game to come back); or B) scoring 4 runs in the game all coming in the bottom of the 9th inning (when they are supposedly the most valuable). Which option would you choose?

[bigredmechanism]

A run scored in the first inning of a 1-0 shutout is worth exactly the same as a run scored in the ninth inning of a 1-0 shutout.

[757690]

A run scored in the first inning of a 1-0 shutout is worth exactly the same as a run scored in the ninth inning of a 1-0 shutout.

After the game is over. But not at the time that the runs are scored. And when making game time decisions, that's what's important.

[MrRedLegger]

If outs mean more in later innings, why wouldn't runs?

[757690]

The bolded part just isn't accurate. As I explained earlier, you are only considering games that are close in the late innings. You are totally ignoring games that are not close in the late innings. In games that are not close, those late runs have far less value than the early runs that won the game. If you include all games and all innings there is no difference in value between early runs and late runs.

It's very accurate, as it only applies to that very specific example to which i was replying. In that example, I am only considering those two specific games, and no others. You are correct, every game is different, but that has no barring on that one example of two specific games.

In a game that does end up being close in the late innings, those late runs are very valuable as you say, but they are no more valuable than the runs scored earlier because without those early runs the game wouldn't be close anyway.

That is logical fallacy, for reasons I explained earlier. I'll try again.

Every event that happens in a game effects the the outcome of that game. A batter getting a walk in the third inning eventually causes the batter who got the game winning hit to bat in the ninth inning. Without that walk, the batter who got the game winning hit never bats and the game would be lost.

Every event matters. However, some events matter more than others. Some events have a stronger affect on the outcome than others. It is a necessary part of the dynamics of a game that some runs are more important than others, just like it is necessary part of the dynamics of the game that some hits are more important than others. It is logically impossible for every event to have the same importance, for every run to have the same importance in a game.

Your blackjack analogy works the same way. If you had won more hands early on you wouldn't have run out of money in the first place and never would have been down to your last $5.

As I stated earlier, the win on the first hand absolutely affected all future hands, including the last one. However, it did not affect the final outcome (running out of money) as much as the final hand did.

With the first hand, had I lost that hand, I would have adjusted my strategy (how much to bet, when to double down, when to hit or stand, etc) to increase my odds of not running out of money. Losing or winning that hand had very little affect on whether or not I ran out of money, since there were so many more options available to me to make sure that I didn't.

With the last hand, when it's my last $5, I have no options to get back into the game if I lose it. I have to win this hand in order to keep playing, so it is more important that I win this hand, than it was that I won the first hand.

More importantly, and more relevant to this discussion, I play this hand differently than I played the first hand. With the first hand, I play the odds. I hit when it statistically is to my advantage to hit, and stand when it is statistically to my advantage to stand. With the last hand, I have to win, so I take bigger chances, I throw the long term odds out the window and play to win just that hand, right now. Why? Because winning that hand is more important than winning previous hands.

I think perhaps (and I could be wrong) your understanding of the Win% numbers from FanGraphs is not quite right. That data is very interesting, but I am guessing it doesn't mean what you think it means. I don't have much time to explain it all right now, but I will take a short stab at it. Here is the definition from FanGraphs: Win Expectancy is “The percent chance a particular team will win the game based on the score, inning, outs, runners on base, and the run environment. These percentages are calculated using historical results of thousands of real MLB games, meaning if a team is losing and has a 24% win expectancy, only 24% of teams in similar situations in the past have ever come back to win."

Notice that the definition does not say anything about the value of a run, because you don't know the actual value of the run until the game is over. You mentioned earlier that the Win%'s can change faster in the late innings and this is true if the game is close, but it is not true if the game is lopsided. An early run doesn't swing the percentages as much as a late run in a close game, but this is only because at the time you don't know how many runs will be scored later. Historically, any given situation was followed by a wide range of possible outcomes. Sometimes that run was the last run scored in the game. Sometimes there were many more runs scored in the game. Because there is such a wide range of possible outcomes you can't say with any certainty how much that run is worth until the game is over.

In a 1 run victory, every run is worth the exact same amount because you needed every single one of them in order to win the game. In a lopsided game technically all the runs are still worth the same as each other, but essentially the early runs were the most valuable because they were the only runs you needed to win the game.

If this doesn't make sense to you I guess we will just have to agree to disagree on this issue, which is fine.

1. In a lopsided game, every run does matter, it just matters less than in a tight game. And the value of a run changes as the circumstances and situations change throughout the game. You have to look at the value of each and every run individually. I agree that in more lopsided games, runs are less valuable than in tight games. But that doesn't change the fact that runs at the end of a game are worth more than runs at the beginning of the game.

If a team scores five runs in the first inning, the first run is worth a certain amount, and the second run is worth more, since it increase the odds of the team winning the game more than the first run did. The third run is worth more than the second run, and so on. When comparing the value of runs scored early in a blowout to runs scored late in a tight game, you have to compare the extra value the fifth run in the first inning gives you over the value of the fourth run in the first inning, to the extra value that the first run of the ninth inning gives you over the value of zero runs scored in the ninth inning.

To put another way, you can't compare the five runs in the first inning to the final run in the ninth inning. You have to compare the fifth run of the first inning to the final run of the last inning.

Or to explain another way. If you are willing to admit that a run in blowout is worth less than a run in a tight game, then why can't you admit that a run at the end of the game is worth more than a run at the beginning of a game? It's the same logic.

Let me ask you the same question I asked bucksfan2:
Question: You are managing the Reds tonight and are given a choice between A) scoring 5 runs in the game, all coming in the 1st inning (when they are supposedly less valuable, giving the other team the rest of the game to come back); or B) scoring 4 runs in the game all coming in the bottom of the 9th inning (when they are supposedly the most valuable). Which option would you choose?

That's a good question. I don't know. I would have to look at the odds. Is it more likely that a team holds a five run lead after one inning, or is it more likely that a team is down by less than four runs in the bottom ninth? What I do know is that the odds are not the same, and therefore, the value of those runs is not the same.

[Griffey012]

I am saying it matters how many runs you score in the game, not when you score them.

Similarly, it only matters how many runs you allow in a game, not when you allow them.

You want to focus only on the rare game that is tied in the 8th inning. I want to take into account all types of games.

If you ask any player, coach, manager, etc. if they would rather have a lead in the 8th inning or be tied in the 8th inning without hesitation they would tell you the lead would be better.

If you use your best pitchers early you are more likely to be ahead in the 8th inning instead of tied.
If you use your stud pitcher for 5 innings instead of 1 inning you are more likely to be ahead in the 8th inning instead of tied.


Question: You are managing the Reds tonight and are given a choice between A) scoring 5 runs in the game, all coming in the 1st inning (giving the other team the rest of the game to come back); or B) scoring 4 runs in the game all coming in the bottom of the 9th inning. Which option would you choose?

Everyone is always going to take scoring 5 runs in the game instead of 4, but that is not the argument here. 5 is greater than 4, we all know that. What about going down 5-0 then scoring 5 to tie in the bottom of the 9th? Which team would you rather be a part of, the one with momentum or without? Or do you not believe in momentum?

I ask, how do you choose when and where to use your best pitcher? You cannot use him every day, and you cannot call on him out of thin air? What if your best pitcher can only throw 3 innings, are the 1st 3 innings the best situation to use him? No. Because he cannot throw the next three days now.

[Griffey012]

Every event in a baseball game leads to it's outcome. Take away any event, and the outcome will be different. Take away a HBP in the 2nd inning, and then the batter who drove in the winning run in the 9th, never bats in the 9th, and the game is lost. Take away the grounds crew raking the infield, and certain groundballs become errors, and lead to runs and change the score of the game.

Just because two events both lead to a game's outcome doesn't mean they are worth the same, or were equally important to that game's outcome. A run scored in the first inning is very important, as are all runs. However, it is not as important as one that is scored in later innings.

I agree starters are more valuable than most relievers, because of the numbers of innings pitcher. However, Cameron at Fangraph's did a piece last season on the value of closers once you factor in the leverage of their innings. He discovered that really good closers were as valuable as #1 starters.

This is beautiful. Thank you.

[gilpdawg]

\ Many times you'll see closer come in in a non-save situation to "just get some work" and stay on schedule and they can't pitch w/ 5 run lead. They need to be walking that razors edge in order to throw at their best.

You know, I hear this all the time, and I just don't buy it. I'm going to try and run some data and see what I can find out on this. I'm hypothesizing if you use a big enough sample that there's no difference.

[919191]

"Closers don't matter".

Tell that to the fans of teams who've blown alot of 9th inning leads.

David Weathers. Aroldis Chapman. David Weathers. Aroldis Chapman. David Weathers. Aroldis Chapman.

Hmmmmm.

This is a little off topic. Back when Weathers had a really awful stretch- i forget if he was a closer then or set up, but he was awful for a while. Before a game then, he was standing in the outfield with LaRue and someone else I forget next to the right field foul pole during batting practice. A ball made it to them and Weathers picked it up. He tossed it to my son who was pretty little then. He looked at me and said "That;'s the first time I've hit my target in 6 weeks". I always liked him after that.

[Raisor]

Score ten runs in the first and who is "closing" doesn't matter 99 percent of the time.