Lineup Analysis

by Shoeless In SC on Mar 13, 2008 6:56 PM CDT

So must of us have used the lineup analysis tool over at baseball musings in the past. With a lot of talk on here about who Ozzie will bat where, etc, I thought I'd try and find out if this tool supports any one lineup heavily over another.

Granted, this is just one tool, and it's the only one of its kind that I am aware of. If someone knows a better one, please share.

For those who have used this before, you know what I did. I put in 9 players in a specific order, 1-9, and plugged in there OBP and SLG as projected by Bill James for 2008.

The first lineup I put in was:

Owens, Cabrera, Thome, Konerko, Swisher, Dye, Fields, Uribe, Pierzynski. This lineup resulted in 5.339 runs per game.

The best lineup it gave was Swisher, Thome, Cabrera, Fields, Konerko, Dye, Pierzynski, Uribe, Owens. This resulted in 5.479 runs per game, and increase of .140.

The worst lineup it gave was Uribe, Cabrera, Fields, Owens, Pierzynski, Swisher, Konerko, Thome, and Dye. This resulted in 5.025 runs per game, a decrease of .314.

Now... I decided to swap some players and see what would happen with the Runs Per Game.

First, swap Uribe with Richar. The same lineup order as before produced 5.410 runs per game, an increase of .071. The best order for the Richar lineup produced 5.568 runs per game, an increase of .209 or so.

Next, I put Uribe back in the lineup and swapped Quentin for Owens. Again, some order, with Quentin now leading off inplace of Owens, and this lineup generated 5.449 runs per game. The best lineup generated 5.565 runs per game. There was in increase of .110 and .226 respectively.

Now, for what we all may hope and pray for... Quentin and Richar starting in place for Owens and Uribe. What would that lineup produce, if Quentin led off? 5.520 runs per game The best order of lineup would produce 5.649 runs per game

So... a recap.

Owens and Uribe in lineup: 5.339 (Ozzie lineup), 5.479 (best) 5.025 (worst)

Owens and Richar in lineup: 5.410 (Ozzie lineup), 5.568 (Best)

Quentin and Uribe in lineup: 5.449 (Ozzie lineup), 5.565 (Best)

Quentin and Richar in lineup: 5.520 (Ozzie lineup) 5.649 (Best)

What this would suggest is that if Owens and URibe were replaced with Quentin and Richar, and put in the most efficient order, our lineup could produce .210 more runs per game. What that means over the course of a full season... I don't know. I would like to believe that Richar and Quentin would have a greater impact than that.

That was a fun way to kill time while I ate my dinner. Thoughts?

SouthSideSox is a community driven site. As such, users are able to express their thoughts and opinions in a FanPost, such as this one, which represents the views of this particular fan, but not necessarily the entire community or SouthSideSox editors.

Tweet Comment 10 comments | 0 recs |

Do you like this story?

Comments

Do lineups really matter?

Studies have shown that batting lineups are simultaneously one of the most controversial and meaningless aspects of baseball. I love that you took the time to do all lineup tinkering, and in fact, the difference between the best realistic (i.e. Ozzie) lineup and the one we have now is .18 runs per game. .18 runs per game comes out to about 29 extra runs for the year, which is certainly significant, and may well be worth a few extra wins.

What really gets under my skin is that Ozzie keeps saying he needs his leadoff hitter to "get on base and score runs." If Owens is leading off (as opposed to Quentin or Swisher), he will a) get on base less often, and b) always start off at first base (thanks to his stunning lack of power), making him score runs less frequently. I need a cold bath or something.

Joist

by Joist on Mar 13, 2008 7:16 PM CDT reply actions

There are certainly...

a few ways to get some extra wins out of tinkering with lineups, but I believe this shows that the biggest gain comes from not just the order of the lineup, but who is -in- the lineup.

Replacing Owens with Quentin sparked decent change, about as much as replacing Uribe with Richar.

With our current roster of players... putting Quentin and Richar in over Pods-redux and Colonel Sanders in the right order could net us in the neighborhood of 3-4 wins. That's certainly nothing to sneeze at.

Now, if Gavin and Danks and Contreras all perform above their projections, we might get another 6+ wins out of those three. And suddenly we are left with the same situation as 2005... we need guys to play above their capacity and have career type years and we can't afford to squander a win here or there because Owens is leading off.

"God knows I gave my best in baseball at all times and no man on earth can truthfully judge me otherwise."

by Shoeless In SC on Mar 14, 2008 2:19 AM CDT up reply actions

let's be clear

about what a rather unnuanced tool like this is useful for. It is not very good at predicting how many runs a team will score over a full season - or how many more runs - for reasons that should be rather plain, starting with the obvious observation that the lineup posted will not play 9.innings, 162 games. What a tool like this is useful for is finding an optimal lineup.

by larry on Mar 14, 2008 9:51 AM CDT up reply actions

Right

"God knows I gave my best in baseball at all times and no man on earth can truthfully judge me otherwise."

by Shoeless In SC on Mar 14, 2008 11:47 AM CDT up reply actions

Wow

So the optimal lineup realistic lineup will produce .210 more runs/game than the worst realistic lineup. That's big. That comes out to 34 more runs of the course of the season -- about 3-5 more wins. Could be the difference between 78 wins and .500!

...lol

by madvillian on Mar 13, 2008 8:50 PM CDT reply actions

Or...

it could be the difference between baseball in October and sitting home.

Lots of other things are going to have to happen for us to make the playoffs however, and tinkering with the lineup isn't going to get us there alone.

"God knows I gave my best in baseball at all times and no man on earth can truthfully judge me otherwise."

by Shoeless In SC on Mar 14, 2008 2:21 AM CDT up reply actions

or...

Put in Brian Anderson...I swear he must have raped someone in Ozzie's family.. I have never seen someone hated so much by an organization

by beanohatesthecubs on Mar 15, 2008 1:38 PM CDT reply actions

Any idea how these lineups

effect defensive efficiencies? I've always felt, with no basis other than gut feeling, that a run differential via better defense is better than the same amount of run differential via offense and would translate to more wins.
Starting pitchers are in longer, bullpens fresher, etc.
Would love to know anyone's thoughts on this. That could help explain OG's thought process other than the keed has speed.

Show mercy. I get to watch the Rangers.

by Tim from Texas on Mar 15, 2008 1:46 PM CDT reply actions

I tried a few lineups

using ZiPS numbers (swisher)

let's start with Ozzie's lineup:
Owens, Cabrera, Thome, Konerko, Swisher, Dye, Pierzynski, Fields, Uribe produces 5.043 runs

what if BA replaces Jesse?
Swisher, Cabrera, Thome, Konerko, Fields, Dye, Pierzynski, Uribe, Anderson produces 5.151 runs

what if Jesse was on #9?
Swisher, Cabrera, Thome, Konerko, Fields, Dye, Pierzynski, Uribe, Owens produces 5.151 runs

so no difference when using BA or Jesse #9
but BA's D will save us a few runs

what about putting Q in center?
Swisher, Cabrera, Thome, Konerko, Fields, Dye, Pierzynski, Uribe, Quentin produces 5.243 runs
(using Q's community projection numbers we get 5.284 runs)

so the difference between ozzie's lineup and quentin's lineup is 0.2 runs if my numbers are correct (big if)
0.2 is 2 in 10 or 32 in 160, that's 3 wins yes?

by The Wizard on Mar 15, 2008 7:59 PM CDT reply actions

if you make the impossible assumption

that your lineup will play all nine innings of all 162 games. interleague play, anyone?

historically a team will probably play the same/similar lineup about 30 times a season. so less than 20% of the time. as i pointed out above, lineup tools such as this are good for finding your best lineup. they are not good for figuring out how many runs or many more runs a particular lineup will score. i really, really do not want to hear for the next nine months how the sox would have won x + 3 games if only they had played a certain lineup.

by larry on Mar 16, 2008 12:02 PM CDT up reply actions