A little-used strategy---why?

[Guest Yancy Street Gang]

It would be interesting to know what managers have done in important (such as postseason) games in which they've fallen way behind. One thing I can think of is using one of your better pitchers in relief, instead of your standard mop-up man. (Example: Dwight Gooden pitching in relief of Ron Darling in 1988 NLCS Game 7.)

But that doesn't address what I think is Bret's point: Is there a desperation way of increasing your offense, that you can switch to mid-game? When a team has a big lead and switches to a better defensive alignment, they're possibly sacrificing additional offense in an attempt to keep their big lead. (When you're up 12-1, scoring that 13th run isn't as important as avoiding the error that can lead to a big inning.)

But is the reverse true? If you're down 12-1, can you switch to your best offensive lineup? Is defense less important when you're down by many runs?

[Guest Bret Sabermetric]

Yancy Street Gang wrote: It would be interesting to know what managers have done in important (such as postseason) games in which they've fallen way behind. One thing I can think of is using one of your better pitchers in relief, instead of your standard mop-up man. (Example: Dwight Gooden pitching in relief of Ron Darling in 1988 NLCS Game 7.) But that doesn't address what I think is Bret's point: Is there a desperation way of increasing your offense, that you can switch to mid-game? When a team has a big lead and switches to a better defensive alignment, they're possibly sacrificing additional offense in an attempt to keep their big lead. (When you're up 12-1, scoring that 13th run isn't as important as avoiding the error that can lead to a big inning.) But is the reverse true? If you're down 12-1, can you switch to your best offensive lineup? Is defense less important when you're down by many runs?

You've gotten my point, and probably expressed it better than I did, Yancy.

[Guest Johnny Dickshot]

I understand your point too. But understanding your point and advocating that a manager remove the second baseman aren't the same thing.

I in fact think removing the second baseman would be a good strategy if you wanted to fall further behind.

[nymr83]

Bret Sabermetric wrote: ="Nymr83"]I still think resting the regulars is the best thing you can do when you fall down by that many runs, How is putting Diaz, a good offensive/poor defensive bench player, into the game instead of Joe McEwing, a worse offensive/better defensive bench player, not giving your starting 2Bman the same exact amount of rest?

you're trying to change my words around, its not going to work. i'm not putting diaz in the game INSTEAD of mcewing, i'm putting them BOTH in at different positions so as to rest MORE of my starters

[nymr83]

]But is the reverse true? If you're down 12-1, can you switch to your best offensive lineup? Is defense less important when you're down by many runs?

even if your sole objective is winning the game (ignoring everything about resting players etc) i still say no. your chances of scoring more runs may have gone up but your chances of allowing even more runs even if you do score has gone up as well, and i'd have to say that it has gone up even more than your chance to score, if not why isnt that guy starting in the first place?

[Guest Yancy Street Gang]

Well, the reason he may not be starting could be because when you're making out the lineup, the score is 0-0 and you're looking for a balance of offense and defense.

Once you're up 12-1, it's probably safe to jettison some offense and increase the defense.

I agree that the opposite doesn't appear to be true. Putting a bunch of hitters who can't field into the game (assuming you have many such players on the bench) might result in the score becoming 18-9 instead of 12-1. But either way you're on the losing end.

[Elster88]

Bret Sabermetric wrote: ="Elster88"][Do you really want to know what defensive positioning I believe in? Or are you just trying to avoid answering the question? I don't see why your defense, if it works when down by a lot of runs, wouldn't work all the time. Sure I want to hear your answer. The advantage to my defense is that it's geared for a high scoring game--you get all your best offensive players (usually outfielders) into the game and you raise the other team's OBP considerably (and lower their SLG pct). Since OBP is, over the long haul, the other team's primary offensive weapon, you'll get killed with this in the long run as an every day strategy. You will give up more up more that way, no doubt. But the runs will be distributed differently. With a high OBP offense, the other team will have a lot of big innings. They'll score 50 runs over ten full games, but they'll be distributed more like 8, 6, 1, 7, 0, 1, 9, 7, 5, 6, while with a more normal defense they'll score only 47 runs over the same 10 game period, but the distribution will be 7, 5, 2, 6, 1, 2, 8, 6, 4, 6. Obviously, you want to give up the 4.7 runs per game as a general policy rather than the 5.0 rpg, but in this special situation your goal is to maximize the chances of giving up 0 runs, which the first model provides.

An excellent explanation, if the numbers work out that way. I see your argument now. I just don't think removing the second baseman would work out this way. It'd be interesting to see in practice.

One problem I see is that once a batter reaches first you'd want to put your second baseman back in his normal position for the force play, right? And since OBP is being raised, odds are good on someone reaching first.

[Elster88]

Thinking about it some more: I just think that the rise in OBP would be so great that it would offset the decrease in SLG. Someone should try this out with their little league team.

[Guest Bret Sabermetric]

Elster88 wrote: [ It'd be interesting to see in practice.

The funny part is that you almost never get back a 7 or 8 run lead anyway, so you could experiment witih putting your players' wives on the field for all the difference it will make in your final W-L record--that's one of the reasons I wonder more managers don't try some experimental strategies like this. I'm not even arguing that this strategy is inevitably right. I'm arguing "What do you have to lose, really?" So it doesn't work and you get killed worse-- dead is dead, correct? But it could work, and you could have figured out a strategy that nets you a win or two per year.

nymr86--You're just talking about the same old "get your regulars out of the game and give up" philosophy that I'm trying to improve. Like I said, there's nothing wrong with that, and it's what's been done for centuries now, but that's the philosophy I'm calling into question. It's not much of a response to say "No, try this" when "this" is the status quo.

I wouldn't dream of trying to twist your timeless words around.

[Guest Bret Sabermetric]

Johnny Dickshot wrote: I understand your point too. But understanding your point and advocating that a manager remove the second baseman aren't the same thing. I in fact think removing the second baseman would be a good strategy if you wanted to fall further behind.

You're probably right. I don't disagree a bit. Over ten games, if you institute the defense and offense I'm suggesting, you might give up 50 runs and score 40 with a team that normally scores and surrenders equal amounts of runs.

But do you care how runs you give up, or how many you score? Not ultimately.You're interested in how many games you win. In this situation, you're just interested in turning these games into slugfests, because when you're down seven runs in the seventh, a slugfest is your only chance of getting back in it. If you set up ten slugfest scenarios, and you lose nine of them, which will happen if youre starting out 7 runs down, so what? You were going to lose those games anyway. It's the off-chance that one of those times, with Victor Diaz playing a ridiculously deep 2b and Piazza in for Mientkiewicz (with Castro in for Piazza), your offense might just explode to 10 runs in the last few innings, and you will have stolen a win.

That's my theory, anyway.

[Guest Rotblatt]

Interesting idea! It'd be really fun to run simulations (ala Baseball Prospectus) of this strategy a couple million times to see what happens.

I have a feeling, though, that the impact of the move would be negligible, even over the course of an entire season. I mean, we're talking about maybe 3 or 4 innings worth of offense at best, right? And that probably equates with maybe two extra at bats for our fourth outfielder per game. Now, we had only 17 games last year where we lost by 5 or more runs; even if we round that up to 20, we're only talking about 80 at bats.

Matsui contributed .09 RC every plate appearance last year, while Victor Diaz contributed .14, so in our best-case scenario, we would net 4 RC over the span of the season--not even enough to affect the outcome of one 5-run blowout.

And that's BEFORE taking into account whatever negative impact our 4th outfielder would have defensively.

If we had, say, Rey Ordonez playing short, his clone playing 2nd and HIS clone playing center, it'd be a different story, but that's a pretty unrealistic example . . .

[Guest Bret Sabermetric]

I get it that this has few, if any, practical ramefactions, most likely. It is funny, isn't it, that the exact opposite measures (putting in glove guys for your big offensive threats) is standard practice, practically a no-brainer, when you HAVE a huge lead, no? I'm not sure what the huge difference is here. We accept the one (largely, I think, because it happens all the time, and has for decades) and totally reject the other (because it's weird, it doesn't look like baseball, it's four outs and a keg on third base, as KC put it.) I'm not so sure that this isn't just blind idolatry of tradition.

[nymr83]

] It is funny, isn't it, that the exact opposite measures (putting in glove guys for your big offensive threats) is standard practice, practically a no-brainer, when you HAVE a huge lead, no? I'm not sure what the huge difference is here.

going to defensive replacements in the LAST inning has no downside, just as pinch-hitting in your last at-bat has none, but to do it earlier is something i have criticized in the past.... i can remember Piazza specifically being pulled from games where the Mets had a small lead and then him not being there when it suddenly got tied up again.

[Guest Bret Sabermetric]

Nymr83 wrote: [going to defensive replacements in the LAST inning has no downside.

Well, it does if the defensive replacements screw the pooch and the other side ties the game up, and you have to bat all through extra innings with your scrubinnies, doesn't it?

[Elster88]

Bret Sabermetric wrote: It is funny, isn't it, that the exact opposite measures (putting in glove guys for your big offensive threats) is standard practice, practically a no-brainer, when you HAVE a huge lead, no? I'm not sure what the huge difference is here. We accept the one (largely, I think, because it happens all the time, and has for decades) and totally reject the other (because it's weird, it doesn't look like baseball, it's four outs and a keg on third base, as KC put it.)

This is stretching it a bit. The opposite of subbing in your best fielders is done, too. I've seen pinch hitters used before. They just don't stay in the rest of the game the way a defensive sub.

[Guest Bret Sabermetric]

Well, pinch-hitters are used with a big lead, with a small lead, with a tie, with a small deficit, and with a big deficit. What I'm talking about is focussed on having a big lead, and saying "To hell with D, what we need is O," more or less analogous to putting in the gloves with a lead and saying "To hell with O, what we need now is D."

[Guest Johnny Dickshot]

How to come back from a big deficit late in a game you have to win:

5-0 after 5:
METS 6TH: Alfonzo doubled to center; Olerud singled to center
[Alfonzo to third]; Piazza out on a sacrifice fly to left
[Alfonzo scored]; Ventura doubled to right [Olerud to third];
Hamilton singled to second [Olerud scored, Ventura scored];
MULHOLLAND REPLACED MILLWOOD (PITCHING); AGBAYANI BATTED FOR CEDENO; Agbayani walked [Hamilton to second]; Ordonez lined into a double play (shortstop unassisted) [Hamilton out at second]; 3 R, 4 H, 0 E, 1 LOB. Mets 3, Braves 5.

7-3 after 6:
METS 7TH: SMOLTZ REPLACED HERNANDEZ (PITCHING); M. FRANCO BATTED FOR COOK; M. Franco doubled to center; Henderson doubled to left [M. Franco scored]; Alfonzo flied to right [Henderson to third]; Olerud singled to right [Henderson scored]; Piazza homered [Olerud scored]; REMLINGER REPLACED SMOLTZ (PITCHING); Ventura flied to right; Hamilton grounded out (first unassisted); 4 R, 4 H, 0 E, 0 LOB. Mets 7, Braves 7.

7-7 after 7:
METS 8TH: Agbayani singled to right; Ordonez out on a sacrifice
bunt (pitcher unassisted) [Agbayani to second]; MORA BATTED FOR
HERSHISER; Mora singled to center [Agbayani scored]; Henderson
forced Mora (shortstop to second); Henderson stole second;
Alfonzo walked; Olerud flied to left; 1 R, 2 H, 0 E, 2 LOB.
Mets 8, Braves 7.

I won't remind you how it ended, but seems like, in this game, the best chance for the O to come back came with strategic pinch-hitting when it was needed, and not before: For instance, had Agbayani replaced Ordonez as this strategy might have suggested, then Cedeno has to face Mulholland, or Mora has to PH 2 innings earlier, leaving a guy like Hershiser to hit for himself with the go-ahead run on base.

On obvious edit: Let's not overlook the importance of your starting lineup hitting a shitload of doubles & home runs.

[Zvon]

This is a most interesting thread.
^That game was a classic JD, and great post there.
And Yancys post of support was really well put.

Bret, I commend your yearning for new strategies in this old game.

But like in JDs breakdown above, so many things have to go right for it to payoff.
In either case, of going with a conventional strategy or a novel one like yours, so many things have to end up going right for you.
Baseballs a game where, when managed well, options used will create other options, and if you play them in the right order, it pays off.

But ill tell you what.
Id try that in just the right situation.
Say the 2nd baseman is slated for a rest, and slumping to boot.
And the pitcher shows that he is, on that day, getting pop ups and fly outs primarily.
If my gut said I had a 4th outfielder who had a better chance of continuing any offensive flow set in motion, Id give that strategy a shot. For a game that wasnt important.
Because truthfully, Id do it for the wrong reason, because i would only try that if i had actually given up on the game,....pretty much.
And if the O steps up and delivers those needed runs, and the opposing team cant take advantage of the holy defense, I look like a brainiac.
But soooo many things have to go just right.

I could more so seeing it done only against certain batters during an attempt to make that comeback, splitting the 4th outfielder from 2nd to OF when %s dictated. And like someone mentioned, if they do get afew seeing eye bingles thru that hole, you do eventually want a guy to be there for the force play at that point, i would think. If the other team is able to hit grounders off the fly ball pitcher, you have to be flexable and start thinkin conventional double play.

Anyone remember when Gil Hodges put on the 4 outfield deep shift vs McCovey in '69?

Most interesting thread, which more than likely shows im not such a good manager as well as G.M.

[Guest Bret Sabermetric]

="Johnny Dickshot"]How to come back from a big deficit late in a game you have to win: 5-0 after 5: METS 6TH: Alfonzo doubled to center; Olerud singled to center [Alfonzo to third]; Piazza out on a sacrifice fly to left [Alfonzo scored]; Ventura doubled to right [Olerud to third]; Hamilton singled to second [Olerud scored, Ventura scored]; MULHOLLAND REPLACED MILLWOOD (PITCHING); AGBAYANI BATTED FOR CEDENO; Agbayani walked [Hamilton to second]; Ordonez lined into a double play (shortstop unassisted) [Hamilton out at second]; 3 R, 4 H, 0 E, 1 LOB. Mets 3, Braves 5.

I'm not saying that catching up by conventional means is imposssible, and I'm not saying that flooding your lineup with early pinch-hitters is the only way to go. But please notice that in the example you chose NOT pinch-hitting for Rey (the key difference between your conventional strategy and my loopy radical one) got you the worst possible, most rally-killing, event. At that point in the game, the Mets were only down by two runs, with runners on first and second and one out. You don't know (none of us can, of course) whether pinchhitting for Rey in that unconventional spot would have simply caused the game to be tied up, or the Mets ahead even, innings before it actually happened.

Of course by that point you were only down by two, so maybe this strategy comes off the table anyway.

Good job tracking down a specific example, JD.