I am using my simulator that plays actual baseball games to find what it believes are the most optimal lineups for the 2015 San Francisco Giants. I started out with over 2,000 permutations of lineups, a group that was filtered to remove such things as left handed hitters hitting back to back or the pitcher not hitting 8th or 9th etc... and I slowly widdled it down to the top 25 lineups. I also compared the best lineup against the lineup that MLB Depth Charts (0.40 wins/162 games worse) shows as a likely lineup and also one of the worst lineups to see what kind of spread there is. The top 25 lineups all came within a half of a win per 162 games so really anything in that range is pretty good. When you start nearing a full win differential with the top lineup then you definitely have problems. You will see a few lineups that may seem strange in a most optimal list but keep in mind the simulator does not take into account psychological factors. It is not like Joe Panik's mom is going to be calling up Giants management in tears asking for Joe not to bat 4th because he is going to have to change his approach at the plate. The simulator just takes the Steamer projection inputs along with a speed rating for base running and gives you the best lineups. Sure, the Giants manager won't use many of these lineups because they might not be prototypical but that is not part of the exercise. I typically simulated each lineup over 1 million times which eliminated a large portion of the random noise and I did so only against a right handed pitcher. Below are a list of the top 25 lineups and a batting order spot frequency table, where you can see how many times each player appeared in each of the top 50 lineups.
Top 25 Optimal Lineups
| Rank | Lineup | Wins behind/162 games |
|---|---|---|
| 1 | Pagan-Aoki-Posey-Panik-Pence-Belt-McGeHee-Pitcher-Crawford | 0.000 |
| 2 | Pagan-Aoki-Posey-Belt-Pence-Panik-McGeHee-Pitcher-Crawford | 0.027 |
| 3 | Pagan-Aoki-Posey-Belt-Pence-Crawford-McGeHee-Pitcher-Panik | 0.060 |
| 4 | Aoki-Pagan-Panik-Posey-Belt-Pence-Crawford-McGeHee-Pitcher | 0.116 |
| 5 | Crawford-Pagan-Belt-Posey-Aoki-Pence-Panik-McGeHee-Pitcher | 0.239 |
| 6 | Pence-Aoki-Posey-Belt-Pagan-Panik-McGeHee-Pitcher-Crawford | 0.247 |
| 7 | Aoki-Pagan-Belt-Posey-Panik-Pence-Crawford-McGeHee-Pitcher | 0.258 |
| 8 | Aoki-Pagan-Belt-Posey-Crawford-Pence-Panik-McGeHee-Pitcher | 0.263 |
| 9 | Belt-Pagan-Panik-Posey-Aoki-Pence-Crawford-McGeHee-Pitcher | 0.277 |
| 10 | Pence-Aoki-Posey-Panik-Pagan-Belt-McGeHee-Pitcher-Crawford | 0.286 |
| 11 | Pagan-Aoki-Posey-Panik-Pence-Crawford-McGeHee-Pitcher-Belt | 0.294 |
| 12 | Aoki-Pagan-Panik-Posey-Crawford-Pence-Belt-McGeHee-Pitcher | 0.296 |
| 13 | Crawford-Pagan-Panik-Posey-Aoki-Pence-Belt-McGeHee-Pitcher | 0.301 |
| 14 | Aoki-Pence-Panik-Posey-Belt-Pagan-Crawford-McGeHee-Pitcher | 0.321 |
| 15 | Pagan-Aoki-Posey-Crawford-Pence-Belt-McGeHee-Pitcher-Panik | 0.325 |
| 16 | Pagan-Aoki-Pence-Panik-Posey-Belt-McGeHee-Pitcher-Crawford | 0.327 |
| 17 | Pagan-Panik-Posey-Belt-Pence-Aoki-McGeHee-Pitcher-Crawford | 0.357 |
| 18 | Belt-Pence-Crawford-Posey-Aoki-Pagan-Panik-McGeHee-Pitcher | 0.360 |
| 19 | Aoki-Pagan-Panik-Pence-Belt-Posey-Crawford-McGeHee-Pitcher | 0.361 |
| 20 | Crawford-Pagan-Panik-Pence-Aoki-Posey-Belt-McGeHee-Pitcher | 0.393 |
| 21 | Pagan-Aoki-Posey-Belt-Pence-Panik-McGeHee-Crawford-Pitcher | 0.400 |
| 22 | Pence-Aoki-Posey-Belt-Pagan-Crawford-McGeHee-Pitcher-Panik | 0.407 |
| 23 | Pagan-Panik-Posey-Belt-Pence-Crawford-McGeHee-Pitcher-Aoki | 0.428 |
| 24 | Crawford-Pence-Panik-Posey-Aoki-Pagan-Belt-McGeHee-Pitcher | 0.442 |
| 25 | Pagan-Aoki-Posey-Belt-Pence-Panik-McGeHee-Crawford-Pitcher | 0.442 |
| .. | ||
| Bad | McGeHee-Belt-Pagan-Aoki-Pence-Panik-Crawford-Posey-Pitcher | 2.438 |
Batting Order Frequency Table
| 1st | 2nd | 3rd | 4th | 5th | 6th | 7th | 8th | 9th | |
|---|---|---|---|---|---|---|---|---|---|
| Pagan | 19 | 17 | 0 | 0 | 5 | 9 | 0 | 0 | 0 |
| Aoki | 17 | 17 | 0 | 1 | 9 | 3 | 1 | 0 | 2 |
| Posey | 0 | 3 | 24 | 17 | 2 | 4 | 0 | 0 | 0 |
| Belt | 3 | 1 | 8 | 14 | 10 | 8 | 4 | 0 | 2 |
| Pence | 4 | 7 | 2 | 8 | 18 | 11 | 0 | 0 | 0 |
| Panik | 0 | 5 | 14 | 7 | 2 | 8 | 8 | 0 | 6 |
| McGeHee | 0 | 0 | 0 | 1 | 0 | 3 | 24 | 21 | 1 |
| Crawford | 7 | 0 | 2 | 2 | 4 | 4 | 13 | 8 | 10 |
| Pitcher | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 21 | 29 |
Synergy
20 Most Common Back to Back Occurrences in Lineup:
| Rank | Occurrences | Player 1 | Player 2 |
|---|---|---|---|
| 1 | 37 | McGeHee | Pitcher |
| 2 | 22 | Pagan | Panik |
| 3 | 22 | Posey | Belt |
| 4 | 19 | Aoki | Posey |
| 5 | 18 | Pitcher | Aoki |
| 6 | 17 | Aoki | Pagan |
| 7 | 17 | Belt | Pence |
| 8 | 17 | Pitcher | Crawford |
| 9 | 16 | Pence | Belt |
| 10 | 16 | Panik | McGeHee |
| 11 | 15 | Pagan | Aoki |
| 12 | 15 | Panik | Posey |
| 13 | 13 | Belt | McGeHee |
| 14 | 13 | Crawford | Pagan |
| 15 | 13 | Crawford | Pitcher |
| 16 | 12 | Pence | Panik |
| 17 | 12 | Panik | Pence |
| 18 | 12 | Crawford | McGeHee |
| 19 | 11 | McGeHee | Crawford |
| 20 | 10 | Aoki | Pence |
| 20 | 10 | Posey | Panik |
| 20 | 10 | Belt | Pagan |
| 20 | 10 | Belt | Posey |
10 Most Common Back to Back to Back Occurrences in Lineup:
| Rank | Occurrences | Player 1 | Player 2 | Player 3 |
|---|---|---|---|---|
| 1 | 17 | McGeHee | Pitcher | Crawford |
| 2 | 15 | Posey | Belt | Pence |
| 3 | 12 | Pagan | Aoki | Posey |
| 4 | 12 | Panik | McGeHee | Pitcher |
| 5 | 12 | McGeHee | Pitcher | Aoki |
| 6 | 12 | Crawford | McGeHee | Pitcher |
| 7 | 12 | Pitcher | Aoki | Pagan |
| 8 | 11 | Pagan | Panik | Posey |
| 9 | 11 | Aoki | Pagan | Panik |
| 10 | 11 | McGeHee | Crawford | Pitcher |
| 10 | 11 | Pitcher | Crawford | Pagan |
Analysis
Pagan: The simulator likes the switch hitting Pagan batting first or second in the lineup. There are a couple of permutations where the simulator likes Pagan batting sixth but he seems best suited for the top of the lineup which is where he is likely to bat in reality.Aoki: Like Pagan, the simulator pegs Aoki at one of the first two spots in the lineup with the occasional drop to fifth. There are even two appearances in the top 50 lineups with Aoki batting 9th.
Posey: The Giants will get the most bang for their buck with Posey hitting third or fourth which is really the only two spots the Giants would ever hit him in.
Belt: Shows up in many different lineup spots with second thru sixth being the best spots for him. Belt finds himself paired either in front of or immediately after Pence in the lineup in 33 of the top 50 lineups. The threesome of Posey-Belt-Pence show up as the second most likely consecutive trio in the top lineups.
Pence: Finds himself all over the radar on the first six lineup spots with fifth and sixth being his best spots. Which is pretty much in line where Pence expects to bat in the 2015 lineup.
Panik: Appears all over the place in the top 50 lineups. The only two spots he doesn't appear in is leadoff and eighth. Panik makes more than a few surprise appearances in the middle of the lineup in the top list which may come as a surprise. The Giants won't likely bat him in the middle of the lineup but the simulator does not take into account psychological reasons (like experience) for not batting a player in a certain spot. With the L-R-L requirement Panik ends up all over the place.
McGeHee: Looks to be a solid fit at the seventh and eighth spots as a late source of power in the lineup before the black hole of the pitcher bats. McGeHee-Pitcher is the number one most common consecutive batting duo in the lineup and McGeHee-Pitcher-Crawford is the number one most common threesome.
Crawford: He shows up all over the place but is a best fit hitting seventh, eighth, ninth or leadoff.
Pitcher: The pitcher shows up hitting eighth in 42% of the top lineups. Most MLB managers will not hit the pitcher eighth for psychological reasons but if done correctly with the right personnel and lineup construction it is a good thing.
No comments:
Post a Comment