I've forgotten the goal for teams playing MLB games

[jay]

My only real point is that I won't chastise him much for not doing it once in early April. I'm sure there will be truly awe-inspiring facepalm moments that better deserve my wrath in the months to come.

 

Fair enough. I think we're eye-to-eye here now.

 

Just please avoid resorting to making fun of my spreadsheets. Some folks get lost in them, but I haven't seen many here that are blind to reality.

[jay]

Couple points:

 

... just when I thought we were finally over this!

[Mr. Brooks]

Couple points: 1) Because Fangraphs has a formula that they believe can accurately determine "leverage", and publish it as an index, does not mean they have cracked the code on ANYthing. It's one person's opinion, and an opinion based on...what? Personally, I'd rather use my best reliever to pitch the bottom of the 9th in a game where I have a one run lead, and a scoreless inning guarantees a win, than use him in a supposedly "higher leverage" situation that may or may not result in a win. 2) Even supposing the "leverage index" is somehow useful, I go back to my earlier point...it's pretty much impossible to predict when these situations will occur. People tend to think relievers can just be summoned from the bullpen when needed. They can't. They need to warm up first, which means you have to decide to warm them up some time in advance of when a hypothetical situation might occur. Warming them up, and then not using them, has it's own set of problems. 3) This all comes back to this (yet again): every reliever on your staff is going to face situations where he needs to get hitters out in a close game. The answer isn't using Perkins or Burton every time the game is close after the 6th inning. The answer is to have a deeper bullpen. If Fien doesn't give up an 0-2 single to start the 8th, or then allow the following hitter to reach base, we're not having this conversation. 4) Which leads to this: Why do people assume that when the Twins bullpen can't protect a lead in the 8th, they'll somehow magically perform better in the 9th, when Perkins is no longer available because you've used him somewhere else already? Why do people assume that if Perkins had gone 2 innings on Friday, these same relievers that failed in the 8th Friday would succeed in the 9th Saturday, when Perkins wouldn't have been available at all?

 

A) Perhaps because if Perkins gets out of that jam, its the bottom of the O's lineup due in the 9th, instead of the top and heart of the order that was up in the 8th?

 

Now you are just using hindsight. The games on Friday and Saturday could just as likely been blowouts, either way. I doubt Gardy has a crystal ball that told him Friday and Saturday would be close games.

[jay]

 

1) Leverage is calculated from Win Expectancy. Hard to believe, but alll those hundreds of thousands of games that have been played can give you some pretty exact likelihoods of potential outcomes for any situation. You can disagree, but I will most certainly prefer my best pitcher on the mound in the situations that have the highest likelihood of impacting the chances of a win.

 

2) I addressed this point by highlighting how the situation could have played out. Simply, runners on base with a 1-run lead is far more dangerous than no runners on base with a lead.

 

3) Deeper bullpen, of course we all want that. Using Perk/Burt every game... again, no, and you're bastardizing the point to summarize it as such.

[Guest USAFChief]

A) Perhaps because if Perkins gets out of that jam, its the bottom of the O's lineup due in the 9th, instead of the top and heart of the order that was up in the 8th?

 

Now you are just using hindsight. The games on Friday and Saturday could just as likely been blowouts, either way. I doubt Gardy has a crystal ball that told him Friday and Saturday would be close games.

A) Due up in the bottom of the 8th for Baltimore that game was 9-1-2. Seems like a good enough time to me NOT to use your best reliever. Had Perkins started the 8th, and got the 9-1-2 guys out, now you have 3-4-5 up in the 9th and you're pretty much obligated to use Perkins for a 2 inning save, or you have some other reliever trying to get through the heart of the order for a save in the 9th. If you bring Perkins in at any point in the 8th after runners are on, you're in the same situation, although now you're using your best reliever with multiple runners on base, with no guarantee of getting through the inning without giving up run(s). I can see the argument for either, but IMO there are equally good, if not better, reasons to do it the way Gardy did. No hindsight required. Most MLB games are close. Assuming you're going to be in many close games is how MLB managers should go about their business, and how MLB GMs should go about assembling their rosters.

[Mr. Brooks]

A) Due up in the bottom of the 8th for Baltimore that game was 9-1-2. Seems like a good enough time to me NOT to use your best reliever. Had Perkins started the 8th, and got the 9-1-2 guys out, now you have 3-4-5 up in the 9th and you're pretty much obligated to use Perkins for a 2 inning save, or you have some other reliever trying to get through the heart of the order for a save in the 9th. If you bring Perkins in at any point in the 8th after runners are on, you're in the same situation, although now you're using your best reliever with multiple runners on base, with no guarantee of getting through the inning without giving up run(s). I can see the argument for either, but IMO there are equally good, if not better, reasons to do it the way Gardy did. No hindsight required. Most MLB games are close. Assuming you're going to be in many close games is how MLB managers should go about their business, and how MLB GMs should go about assembling their rosters.

 

I was talking about bringing Perkins in when Davis came up w/ bases loaded. I never said bring him in to start the inning.

If Perkins gets Davis and the next guy out, that leaves someone to start a clean 9th with the 6,7,8 hitters up I believe?

[jay]

If you bring Perkins in at any point in the 8th after runners are on, you're in the same situation, although now you're using your best reliever with multiple runners on base, with no guarantee of getting through the inning without giving up run(s).

 

That's exactly the point. So, use your best reliever to get you through that situation. You can't win in the 9th if you already gave up that lead. We can fundamentally disagree on which pitcher should be used there, but I want Perkins.

 

We agree Perk is our best BP pitcher. So, should Perk pitch in the 9th inning predominantely? Sure. This index table shows that the potential range of leverage scores goes higher in the later innings. The more important question: Should he only pitch in the 9th? Absolutely not... even if that mean someone else pitches the 9th and gets the glorified "save".

[Mr. Brooks]

A) Due up in the bottom of the 8th for Baltimore that game was 9-1-2. Seems like a good enough time to me NOT to use your best reliever. Had Perkins started the 8th, and got the 9-1-2 guys out, now you have 3-4-5 up in the 9th and you're pretty much obligated to use Perkins for a 2 inning save, or you have some other reliever trying to get through the heart of the order for a save in the 9th. If you bring Perkins in at any point in the 8th after runners are on, you're in the same situation, although now you're using your best reliever with multiple runners on base, with no guarantee of getting through the inning without giving up run(s). I can see the argument for either, but IMO there are equally good, if not better, reasons to do it the way Gardy did. No hindsight required. Most MLB games are close. Assuming you're going to be in many close games is how MLB managers should go about their business, and how MLB GMs should go about assembling their rosters.

 

That is debatable. In the 158,301 games from 1900 to 2010, the average margin of victory was 3.3 runs.

In 2010, 40% of the games were decided by 4 runs or more.

[Guest USAFChief]

1) Leverage is calculated from Win Expectancy. Hard to believe, but alll those hundreds of thousands of games that have been played can give you some pretty exact likelihoods of potential outcomes for any situation. You can disagree, but I will most certainly prefer my best pitcher on the mound in the situations that have the highest likelihood of impacting the chances of a win.

 

2) I addressed this point by highlighting how the situation could have played out. Simply, runners on base with a 1-run lead is far more dangerous than no runners on base with a lead.

 

3) Deeper bullpen, of course we all want that. Using Perk/Burt every game... again, no, and you're bastardizing the point to summarize it as such.

1) If you think all those hundreds of thousands of games have much relation to Baltimore vs Minnesota in early April 2013, knock yourself out. I pay no more attention to "win expectancy" than I do to "run expectancy." Each asks you to believe that an average amassed over a long period of time is relevant to a specific situation. I don't believe that 2013 Aaron Hicks, hitting with 2 on and 1 out, vs Justin Verlander, is equal to Albert Pujols hitting with 2 on and 1 out, against random crappy reliever. If you believe in a run expectancy chart, you by definition think those two situations are exactly alike. Same for "win expectancy," only on a smaller scale. 2) Talk about hindsight. The 8th inning started with nobody on base. It only became "high leverage" when runners reached base. 3) I'm not bastardizing the point in any way. It seems sort of...convenient...to insist "we're not saying use Perk/Burt in every game, but we're sure going to point it out every time Gardy doesn't."

[Guest USAFChief]

False. In the 158,301 games from 1900 to 2010, the average margin of victory was 3.3 runs.

In 2010, 40% of the games were decided by 4 runs or more.

I'm not sure what your point is. "Average" is not equal to "most often," and your 2010 data proves my point. 60% of games were "close," decided by 3 or fewer runs. Is not 6 of every 10 "most?"

[Mr. Brooks]

I'm not sure what your point is. "Average" is not equal to "most often," and your 2010 data proves my point. 60% of games were "close," decided by 3 or fewer runs. Is not 6 of every 10 "most?"

 

I wouldn't call a 3 run lead with only 3 outs to get close.

But, even if you do, that is still only 6 of 10 times. Hardly predictable that the next 2 are both going to be close, as you claim when you say, "no hindsight".

[jay]

I pay no more attention to "win expectancy" than I do to "run expectancy." Each asks you to believe that an average amassed over a long period of time is relevant to a specific situation. I don't believe that 2013 Aaron Hicks, hitting with 2 on and 1 out, vs Justin Verlander, is equal to Albert Pujols hitting with 2 on and 1 out, against random crappy reliever. If you believe in a run expectancy chart, you by definition think those two situations are exactly alike. Same for "win expectancy," only on a smaller scale.

 

I don't know anyone who would call those hypothetical situations the same. However, all those averages amassed can certainly tell us plenty and you're missing a lot to so readily dismiss them. Yes, context always has to be applied.

 

I can readily admit and identify a stat's shortcomings. Absolutely true, leverage (and WE and RE) can't capture the current game situation. The point remains that a manager should make his decision off both leverage and the current context (ie middle of the order, L/R matchups). Even all that said, the fact remains I'd rather see Perkins there, as soon as possible, with runners on base to preserve a lead in the 8th.

 

As has been pointed out, even with context in consideration, why would we want to save Perkins for the bottom half of the order in the 9th?

[jay]

It seems sort of...convenient...to insist "we're not saying use Perk/Burt in every game, but we're sure going to point it out every time Gardy doesn't."

 

I'd be shocked to hear you say Perk wasn't going to be pitching if we did still have that lead in the 9th. So, the real question becomes... when is the better time to use him? I'll take him in the 8th with runners on.

[Brock Beauchamp]

Fair enough. I think we're eye-to-eye here now.

 

Just please avoid resorting to making fun of my spreadsheets. Some folks get lost in them, but I haven't seen many here that are blind to reality.

 

I'm one of the more sabr-heavy posters on the forum, I think. 90% of the time, I will rely on sabr-stuff to make a point... I just think some real-world grounding needs to factor in occasionally... Particularly when you're talking about "grand scheme of things" items like bullpen management. A stat sheet won't tell you that the Twins play six consecutive games on the road and that your "bullpen ace" had pitched two nights before, warmed up the previous night, and that the weather is cold in early April... Stuff that *should* factor into any manager's situational assessment.

[Mr. Brooks]

I'm one of the more sabr-heavy posters on the forum, I think. 90% of the time, I will rely on sabr-stuff to make a point... I just think some real-world grounding needs to factor in occasionally... Particularly when you're talking about "grand scheme of things" items like bullpen management. A stat sheet won't tell you that the Twins play six consecutive games on the road and that your "bullpen ace" had pitched two nights before, warmed up the previous night, and that the weather is cold in early April... Stuff that *should* factor into any manager's situational assessment.

 

My point is that if the Twins have a 3 run lead in the 9th, with the bottom of the order due up, Gardy is STILL going to use Perkins regardless of those variables you mentioned.

Therefore I don't think those variables are very valid points in this specific discussion.

In a general bullpen management discussion, sure, but when discussing this specific occurance, I say no, because those were not the reason Gardy didn't use Perkins there.

[Guest USAFChief]

My point is that if the Twins have a 3 run lead in the 9th, with the bottom of the order due up, Gardy is STILL going to use Perkins regardless of those variables you mentioned.

Therefore I don't think those variables are very valid points in this specific discussion.

In a general bullpen management discussion, sure, but when discussing this specific occurance, I say no, because those were not the reason Gardy didn't use Perkins there.

If the Twins have a 3 run lead in the bottom of the ninth, couldn't one make the assumption Gardy's bullpen management has worked OK in that game prior to the 9th? At what point in this game should he have used Perkins?

[Mr. Brooks]

If the Twins have a 3 run lead in the bottom of the ninth, couldn't one make the assumption Gardy's bullpen management has worked OK in that game prior to the 9th? At what point in this game should he have used Perkins?

 

I think you missed my point.

My point was that Gardy didn't pass on using Perkins because of his availability or because of the air temperature.

If a save situation were to have occurred in that game, Perk would have gotten the ball, so I don't think its apt to this discussion to claim that those variables should have a bearing on this specific discussion.

[Guest USAFChief]

I'd be shocked to hear you say Perk wasn't going to be pitching if we did still have that lead in the 9th. So, the real question becomes... when is the better time to use him? I'll take him in the 8th with runners on.

I think the more realistic options are to play it like he did, or start the 8th with Perkins and get 2 innings. Either way is defensible. Bringing in Perkins after the 8th gets away from you is, in this case, a lot tougher to do than you think. In this specific case, it could have been done, but IMO bringing him in to face Davis was too late (already tied, only one out), and bringing him in earlier but after the start of the inning was pretty tough to do. I also don't know how well Perkins would respond to multi-inning appearances, multiple times. Some relievers are fine, some don't respond well. As it turns out, it for sure would have meant a different pen usage Saturday.

[jay]

I think the more realistic options are to play it like he did, or start the 8th with Perkins and get 2 innings. Either way is defensible. Bringing in Perkins after the 8th gets away from you is, in this case, a lot tougher to do than you think. In this specific case, it could have been done, but IMO bringing him in to face Davis was too late (already tied, only one out), and bringing him in earlier but after the start of the inning was pretty tough to do. I also don't know how well Perkins would respond to multi-inning appearances, multiple times. Some relievers are fine, some don't respond well. As it turns out, it for sure would have meant a different pen usage Saturday.

 

Your entire line of thinking about his utilization is based on believing that Perk has to get the last out of the 9th due to the "save" stat. I disagree.

 

Separately, as we went over earlier, I don't believe facing Davis would have been the best time to bring him in. I think he should have been warming up as soon as someone got on base in the 8th to be ready to go in if anyone else got on or else just have to stay warm a bit longer to pitch the 9th.

[Guest USAFChief]

Your entire line of thinking about his utilization is based on believing that Perk has to get the last out of the 9th due to the "save" stat. I disagree.

 

Separately, as we went over earlier, I don't believe facing Davis would have been the best time to bring him in. I think he should have been warming up as soon as someone got on base in the 8th to be ready to go in if anyone else got on or else just have to stay warm a bit longer to pitch the 9th.

1) I haven't mentioned the save stat anywhere in this thread. Perkins was likely to be used in that game, I'm trying to figure out what would be the best way to use him. 2) I don't know if Perkins can warm in the 8th after the inning starts, and then still be ready to pitch the ninth if it turns out no other runners get on base in the 8th. If he can, and it doesn't mean he's not available the next day, then I can agree that would be one way to play it. I also apologize for thinking you were the one arguing Perkins should have been brought in to face Davis instead of whats-his-name. I lost track of separate conversations.

[ashbury]

Your entire line of thinking about his utilization is based on believing that Perk has to get the last out of the 9th due to the "save" stat.

 

I wade back into this conversation with trepidation, but it seems to me that sending Perkins to the showers after pitching to one or two batters in the ninth saves very little wear on the arm, compared to having him warm up in the eighth, come into the game, sit down and wait perhaps 15 minutes through the top of the ninth, go out and warm up again, and pitch to those one or two batters. To me, that's what's wrong with the current stats-based analysis; and like Pocket Pig I consider myself on the "analytic" side of any discussion by nature.

[Mr. Brooks]

I wade back into this conversation with trepidation, but it seems to me that sending Perkins to the showers after pitching to one or two batters in the ninth saves very little wear on the arm, compared to having him warm up in the eighth, come into the game, sit down and wait perhaps 15 minutes through the top of the ninth, go out and warm up again, and pitch to those one or two batters. To me, that's what's wrong with the current stats-based analysis; and like Pocket Pig I consider myself on the "analytic" side of any discussion by nature.

 

Why does he have to pitch the 9th too?

Why can't he get you out of the 8th inning jam, then hit the showers?

[Willihammer]

Couple points: 1) Because Fangraphs has a formula that they believe can accurately determine "leverage", and publish it as an index, does not mean they have cracked the code on ANYthing. It's one person's opinion, and an opinion based on...what?

Crucial Situations from The Hardball Times

 

2) Even supposing the "leverage index" is somehow useful, I go back to my earlier point...it's pretty much impossible to predict when these situations will occur.

In any given inning, the 4th hitter will have the most opportunities to drive in a runner. Because if he reaches the plate then there is at least 1 runner on base, by definition. In a 1 run game, you play to the 4th hitter (Markakis in this case). And that's exactly what Gardy did. Reimold got his leadoff single, then Gardy got Robertson warming up immediately. He simply passed on Perkins in favor of Robertson.

 

4) Which leads to this: Why do people assume that when the Twins bullpen can't protect a lead in the 8th, they'll somehow magically perform better in the 9th with your boppers up there, when Perkins is no longer available because you've used him somewhere else already?

 

Because you haven't played the 9th inning yet. You don't know that you won't score again in the top of the 9th. You don't know that you'll give up 2 cheap groundball singles again in the home half. The only think you know is that the first runner is on in a 5-4 game in the home 8th, with a very good lefthanded hitter batting 4th.

 

The reason Gardy opted for Robertson instead of Perkins, I suspect, is because he expected the leadoff man man to sac bunt at that point. To play for the tie, basically, and try again in the bottom of the 9th. Because that's what he would have done. Instead, Buck let Machado swing away and by the time Markakis stepped up there were runners on 2nd and 3rd with 1st base open.

 

For the record, the O's had so many things go right that inning that its hard for me to take it out on Gardy. And even so, as others have said, if Robertson can't get one lefthander out, then what is he on the team for? More than anything, I would blame the loss on the GM for leaving Gardy 1 very goo... rock soli... pretty goo... one passable MLB lefthander short to start the year.

[jokin]

 

 

And even so, as others have said, if Robertson can't get one lefthander out, then what is he on the team for? More than anything, I would blame the loss on the GM for leaving Gardy 1 very goo... rock soli... pretty goo... one passable MLB lefthander short to start the year.

 

This. There were plenty of LH options in the offseason, FA options, trade options, waiver options, even Rule 5 options. Minor league signee Eric Bedard even got a 3 inning save against the Rangers last week ( I think he's still in the Astro rotation, though ,either way, I wish the Twins could have gotten him for either role- and for peanuts, oversight with low downside risk).

[ashbury]

Why does he have to pitch the 9th too?

Why can't he get you out of the 8th inning jam, then hit the showers?

 

We currently have motions on the floor specifying he enters the game at about 3 different points, and other motions on the floor to have him depart in 3 different points. On an open forum anyone can chime in, but this much cross talk was one of the reasons for my trepidation.

[Mr. Brooks]

We currently have motions on the floor specifying he enters the game at about 3 different points, and other motions on the floor to have him depart in 3 different points. On an open forum anyone can chime in, but this much cross talk was one of the reasons for my trepidation.

 

Yeah I hear you.

I just hate to see the argument dismissed by saying that you don't want to wear out his arm by making him throw multiple innings.

To me what happens in the 9th is a completely different discussion, and shouldnt have much bearing on this one.

[Guest USAFChief]

Yeah I hear you.

I just hate to see the argument dismissed by saying that you don't want to wear out his arm by making him throw multiple innings.

To me what happens in the 9th is a completely different discussion, and shouldnt have much bearing on this one.

It seems to me you're more than willing to wear out his arm, if you're making the argument Perkins should be used in every game's "highest leverage situation." 1) There are an awful lot of high leverage situations in a season, and 2) in order for that to happen, Gardy will have to successfully guess when those situations will occur, which might happen multiple times in a given game, and be willing to warm him up but not use him if the high leverage doesn't develop. If those don't apply, then it seems to me you're just second guessing a specific situation in a specific game in which you disagree with a reliever choice. You have every right to do that, of course, but at least let's not pretend it's part of some larger philosophical disagreement.

[Mr. Brooks]

It seems to me you're more than willing to wear out his arm, if you're making the argument Perkins should be used in every game's "highest leverage situation." 1) There are an awful lot of high leverage situations in a season, and 2) in order for that to happen, Gardy will have to successfully guess when those situations will occur, which might happen multiple times in a given game, and be willing to warm him up but not use him if the high leverage doesn't develop. If those don't apply, then it seems to me you're just second guessing a specific situation in a specific game in which you disagree with a reliever choice. You have every right to do that, of course, but at least let's not pretend it's part of some larger philosophical disagreement.

 

Either you missed the post where I specifically said you don't have to use Perkins in EVERY high leverage situation, or you chose to ignore it.

I said you have to use him in this one because Burton is unavailable and Duensing has already been used.

[Mr. Brooks]

It seems to me you're more than willing to wear out his arm, if you're making the argument Perkins should be used in every game's "highest leverage situation." 1) There are an awful lot of high leverage situations in a season, and 2) in order for that to happen, Gardy will have to successfully guess when those situations will occur, which might happen multiple times in a given game, and be willing to warm him up but not use him if the high leverage doesn't develop. If those don't apply, then it seems to me you're just second guessing a specific situation in a specific game in which you disagree with a reliever choice. You have every right to do that, of course, but at least let's not pretend it's part of some larger philosophical disagreement.

 

I don't second guess anything. I defended Gardy in today's game for sticking with Correia.

I took this Perkins/Robertson stance before the results.

[Guest USAFChief]

I'm not sure that last was necessary.