Bat Hardy 3rd and Jones 4th

[CA-ORIOLE]

Stats are true results..things that actually happened.

Simulations are projections that are likely to happen..There is a difference.

Look, this is evolving into a pretty dumb argument.

The bottom line is pretty simple. Some people believe that players can/will perform differently depending on where they bat in the lineup..Others don't. Its really that simple. There is no proof that either side is right because what is being discussed can not be measured with a number. This will never be proven one way or another because of how lineups will continue to be done.

I'm sure there are plenty of guys who have moved around in different spots in the batting order with large sample sizes. Jeter has bounced around pretty good from 1 and 2 and his numbers seam pretty even. I'd gess it's not looked at because it's far more ridiculous and marginal than the clutch or protection nonsense.

I'll even concede that there may be a few guys out there where this has some marginal effect. That being said, it doesn't mitigate the far greater value of the lineup optimization model.

Either way, a good lineup vs a bad lineup can mean 2-3 extra wins and because of that, you should look to optimize the lineup as much as possible

Which in your case means taking a poll of where they like hitting I guess.

You and Gordo may have the last word. I definitely agree with you that the argument has gotten pretty dumb.

[El Gordo]

I think I have totally lost track of this argument. The traditional notion of line up structure as I see it is: you want you best OBP/P/PA.guy and best base runner batting first. the 3,4,5, guys should be your best all around hitters, BA, OBP, SLG, combo, and the 6 and 7 guys should be 3TO types, with the last two guys being speedy OBP, table setter types. So with the O's we have a bunch of 5,6, 7, hitters and a bunch of #2, 8,9, types but, no true #3,4 hitters. What would the statistical model say to this analysis. Is it the AL simulation model of:

4.488 5 2 4 3 1 6 7 8 9? Well who on the O's are 5 2 4 3 1 6 7 8 9 ? Using tonights' lineup that would be:

Davis

Nick

Vlad

Jones

Hardy

Reynolds

Wieters

Pie

Andino

Is that our true optimal lineup?

[Phantom]

Stats are true results..things that actually happened.

Simulations are projections that are likely to happen..There is a difference.

Look, this is evolving into a pretty dumb argument.

The bottom line is pretty simple. Some people believe that players can/will perform differently depending on where they bat in the lineup..Others don't. Its really that simple. There is no proof that either side is right because what is being discussed can not be measured with a number. This will never be proven one way or another because of how lineups will continue to be done.

Either way, a good lineup vs a bad lineup can mean 2-3 extra wins and because of that, you should look to optimize the lineup as much as possible.

So you'd rather randomly guess at the weather than trust a weather simulation, simply because it is not 100% perfect?

So, assuming you put no stock in these simulations, how on earth would you attempt to optimize a lineup? Would you just ask players where they want to bat?

[ledzepp8]

But there really aren't a lot of variables. The only variable is lineup position. We're not able to create a parallel universe where everything is the same except lineup position so it requires some abstract thinking. Discounting speed there is little reason for me to believe the model isn't compelling statistical evidence especially when compared to a gut feeling or an undefined synergetic effect. The only thing I can really think of is a LH batting second as there is some evidence production can be improved there with a high OBP guy batting first and finding the hole more often.

This is the point. It basically comes down to the "Here are some projections based on stats" guys vs. the "Well...I just got a feeling" guys.

[ledzepp8]

Stats are true results..things that actually happened.

Simulations are projections that are likely to happen..There is a difference.

Either way, a good lineup vs a bad lineup can mean 2-3 extra wins and because of that, you should look to optimize the lineup as much as possible.

Do you discount pre season player projections as well?

And what do you mean by "a good lineup vs. a bad lineup"?

[DrungoHazewood]

Doubt it all you like, the math has been done.

http://fansofdmb.yuku.com/topic/1526#.TkFHGGFLcTk

So that is a .176 difference between the best and worst lineups. You really think just moving Hardy is going to jump the needle?

If I were Buck I'd be squeezing every last run out of this lineup. The difference between batting Hardy 1st and Hardy 3rd could be the 2 runs that are needed to finish the season with 67 wins instead of 66.

[Phantom]

This is the point. It basically comes down to the "Here are some projections based on stats" guys vs. the "Well...I just got a feeling" guys.

This is pretty key in baseball analysis currently. While most fans have come around to advanced hitting metrics (or at least OPS) advanced fielding and pitching stats are still a bit suspect. Some take this as a reason to discount them completely. For me though, nonperfect objective analysis will always trump subjective decisions with no evidence at all.

[ledzepp8]

This is pretty key in baseball analysis currently. While most fans have come around to advanced hitting metrics (or at least OPS) advanced fielding and pitching stats are still a bit suspect. Some take this as a reason to discount them completely. For me though, nonperfect objective analysis will always trump subjective decisions with no evidence at all.

Yep and the strange thing is that Gordo and Sports Guy are both in the camp of fans that believe in advanced stats. I get where they're coming from on this, but the point is that player preference is probably so negligible as to not matter.

[DrungoHazewood]

It doesn't discount it but it doesn't make it more credible either. It's just a theory that can't be proved. What I think SG is trying toi say is that in the abstract, line up order is not that significant, but the practical reality of baseball can't dicount factors like psychology of the players, their comfort levels, health, slumps, matchups, etc. Statistical models may hold for the long view, but a manager is trying to put the best lineup out there to win that day.

The manager is going to optimize a lineup on anything but vague traditional ideas and some idea of where his players prefer to hit (if they even have such a preference). It's kind of crazy to think that's really any kind of optimization at all. It's just guessing and bowing to tradition and trying to fit who's currently available into the game.

[CA-ORIOLE]

The manager is going to optimize a lineup on anything but vague traditional ideas and some idea of where his players prefer to hit (if they even have such a preference). It's kind of crazy to think that's really any kind of optimization at all. It's just guessing and bowing to tradition and trying to fit who's currently available into the game.

Not my quote. It's Gordos. Direct it at him and SG.

[DrungoHazewood]

The bottom line is pretty simple. Some people believe that players can/will perform differently depending on where they bat in the lineup..Others don't. Its really that simple. There is no proof that either side is right because what is being discussed can not be measured with a number. This will never be proven one way or another because of how lineups will continue to be done.

What we have is the results of a bunch of different simulations, as well as a bunch of studies on lineup protection and the like, all of which show that the difference between an optimized lineup and another decent lineup (i.e. not batting the pitcher 3rd, or Cesar Izturis leadoff) is very small. On the order of a few runs a season.

On the other side we have the vague idea that players believe they're much, much better or worse batting in one slot or another, with little or no data to back that up.

I think the baseline assumption has to be that lineup order matters a little bit, and that's it. There's essentially no evidence that optimizing the lineup does anything significant. Most everyone who's studied the issue comes to the conclusion that the most important factors in constructing a lineup are a) keep the players happy, and b) bat the best hitters near the top of the lineup.

Either way, a good lineup vs a bad lineup can mean 2-3 extra wins and because of that, you should look to optimize the lineup as much as possible.

We think it's 2-3 wins, but that's the difference between a good lineup and a terrible lineup. The simulations show that batting Cesar Izturis near the top of the lineup instead of 9th while batting Matt Wieters leadoff and Mark Reynolds 9th might be worth 2-3 wins. The difference between batting Hardy 1st and Hardy 3rd is probably 2-3 runs over 162 games.

I agree that the manager shouldn't do ridiculous things. But the most optimization he needs to do is to go ask guys where they want to bat and try to accomodate them.

[DrungoHazewood]

Not my quote. It's Gordos. Direct it at him and SG.

Sorry, fixed the quote, something was wrong with that post.

[Sports Guy]

So you'd rather randomly guess at the weather than trust a weather simulation, simply because it is not 100% perfect?

So, assuming you put no stock in these simulations, how on earth would you attempt to optimize a lineup? Would you just ask players where they want to bat?

IFirst of all, the weather comp is poor...very poor. The weather doesn't involve pyschological issues.

Secondly, I NEVER said that I don't put stock into the simulations. All I said is that there is more to this than just random simulations that can't take into account certain aspects that can not be measured.

My main point is that I do think lineups matter, even if it is only a difference in a few wins.

There is a reason that you want your better players to get more at bats over your worst players.

[Sports Guy]

What we have is the results of a bunch of different simulations, as well as a bunch of studies on lineup protection and the like, all of which show that the difference between an optimized lineup and another decent lineup (i.e. not batting the pitcher 3rd, or Cesar Izturis leadoff) is very small. On the order of a few runs a season.

On the other side we have the vague idea that players believe they're much, much better or worse batting in one slot or another, with little or no data to back that up.

I think the baseline assumption has to be that lineup order matters a little bit, and that's it. There's essentially no evidence that optimizing the lineup does anything significant. Most everyone who's studied the issue comes to the conclusion that the most important factors in constructing a lineup are a) keep the players happy, and b) bat the best hitters near the top of the lineup.

We think it's 2-3 wins, but that's the difference between a good lineup and a terrible lineup. The simulations show that batting Cesar Izturis near the top of the lineup instead of 9th while batting Matt Wieters leadoff and Mark Reynolds 9th might be worth 2-3 wins. The difference between batting Hardy 1st and Hardy 3rd is probably 2-3 runs over 162 games.

I agree that the manager shouldn't do ridiculous things. But the most optimization he needs to do is to go ask guys where they want to bat and try to accomodate them.

And 2-3 wins are important, especially if you are fighting for a playoff spot.

To say lineups don't matter is bs to me....A few extra wins could be very important.

[DrungoHazewood]

My main point is that I do think lineups matter, even if it is only a difference in a few wins.

There is a reason that you want your better players to get more at bats over your worst players.

Isn't that what almost all lineups do anyway? It's one thing if you argue that Vlad shouldn't be playing at all. But the difference between him batting 4th and him batting 7th is really only a couple of runs a year. It's really the difference between him getting 550 PAs instead of 510 or something, plus the impact of him not wanting to bat 7th (which IMO is negligible).