2008 Ranger WPAs

by Adam J. Morris on Jun 2, 2008 1:01 PM CDT

Among the various and sundry other awesome things B-R offers, they now have the most influential plays of each game in terms of WPA. (What's WPA? Click here).

So I thought I'd go through and see what the 5 or so most crucial plays to Ranger wins, and to Ranger losses, have been so far this season...

Today is the most influential plays in losses. Later this week, we'll do the most influential plays in wins.

Sound exciting? Then follow the jump...

6th most influential play in a loss (tie):

May 14, at rare Wednesday day game, at home against the Mariners. A win means the Rangers are back at .500. Carlos Silva vs. Scott Feldman. Both pitchers pitch well, but the M's score 2 unearned runs in the top of the 7th to give them a 3-2 lead.

Bottom of the 7th, David Murphy doubles, Brandon Boggs grounds out. Brandon Morrow comes in and walks two batters in a row, loading the bases with one out and Ramon Vazquez due up.

The M's bring in Arthur Rhodes, and Ron Washington, in a hugely second-guessed move, brings in Marlon Byrd to pinch hit in what would be (all together now) Byrd's first major league AB in almost a month.

Byrd works the count, thinks he draws ball four, has that pitch called a strike instead, and then grounds into a 5-4-3 DP. The Rangers' chances of winning the game drop from 56% to 26%. Much flaming is done on message boards for a couple of days thereafter on whether Vazquez should have been lifted, and if so, whether Byrd should have been the one to hit.

6th most influential play in a loss (tie):

April 24, at the Tigers. This is the game that was the low-water mark for the Rangers this season, dropping them to 7-16 on the year.

Most plays that lead to big swings in probability involve late inning plays in close games. Not this one. Jason Jennings was going up against Jeremy Bonderman, who was lifted in the top of the 5th after walking in the tying run. With the game tied at 2, Jennings needed a shutdown inning.

Instead, Jennings allowed two walks and a single, and with two on, two out, and the game tied, faced Magglio Ordonez, who homered to deep left. The Rangers were left with a 14% chance of victory, compared to 44% when Ordonez came to the plate.

5th most influential play in a loss

April 19. The second game of a 4 game series in Boston. Another Jason Jennings game, but one where Jennings pitched well, allowing just 2 runs in 6 innings. The Rangers scored single runs in each of the first three innings, and it looked like the Rangers might eke out a victory.

Bottom of the 8th, Joaquin Benoit gets an out, then allows a double to Dustin Pedroia and a single to David Ortiz (which had a 23% swing in the chances of victory). With the game now tied, Benoit went after Manny Ramirez, who clobbered a home run to left field, giving the Red Sox a 5-3 lead, and turn a 39% chance of victory to an 8% chance of victory.

2nd most influential play in a loss (tie):

April 18. The day before the Ramirez homer off of Benoit. Still in Boston. Luis Mendoza against Daisuke Matsuzaka. The Rangers are up 1-0 in the bottom of the 3rd, but Mendoza is struggling. He had been cruising, retiring the first 8 BoSox batters of the game, before allowing a 2 out double to Jed Lowrie, which he followed up with a pair of walks.

Two outs, bases loaded, Mendoza had a chance to escape if he could retire David Ortiz. And Ortiz went opposite field with a grand slam home run that gave Boston a 4-1 lead, and a 53% chance of victory turned into a 20% chance of victory.

2nd most influential play in a loss (tie):

Again, that May 14 game against the M's. Despite the Byrd GIDP, the Rangers end up tying up the score, and the game goes to extra innings. Franklyn German comes in to pitch the 12th, and ends up facing Miguel Cairo with 2 outs and a runner on third.

Cairo singles. The Rangers chances of winning the game go from 53% to 20%. The hidden .500 mushroom strikes again.

2nd most influential play in a loss (tie):

May 19 in Minnesota. Another hidden mushroom game, as a win gets the Rangers to .500. Scott Feldman against Boof Bonser. Feldman struggles, but the Rangers come back to take a 6-5 lead into the bottom of the 9th. But Joe Mauer singles in a run in the 9th against C.J. Wilson to tie the score, the Twins end up having to have their pitcher bat in extra innings, and German ends up giving up a walk-off double to Howie Clark in the 12th.

Remarkably, though, neither the Mauer single (at 29%) nor the Clark double (at 28%) were the most influential plays of the game. Rather, it was Alexi Casilla's 2 out, 3 run homer in the 4th, wiping out a 2-1 Ranger lead, that turned a 57% chance of victory into a 24% chance of a win.

Most influential play in a loss

Believe it or not, it is from that same series in Boston. April 20, the Rangers have lost two in a row, but are up 5-0 to the BoSox heading into the 7th.

Kevin Millwood gets into trouble in the 7th, and Wes Littleton comes in to relieve him. Littleton also isn't on his game, and C.J. WIlson is brought in in the 8th inning in a 5-3 game, with 2 outs and a runner on, to put out the fire.

With the shift on, David Ortiz singles on a ball that Ian Kinsler knocks down but can't throw him out on. 5-4, and Dustin Pedroia comes up as a pinch hitter. Pedroia doubles on a blast to deep right-center, tying the game, and advances to third on a bad through from Josh Hamilton. Pedroia's double turns a 74% chance of victory into a 40% chance of victory, and set the stage for the winning run to be walked in later that inning.

Tweet Comment 8 comments | 3 recs |

Do you like this story?

Comments

Amazing tag

by Agreen07 on Jun 2, 2008 1:14 PM CDT reply actions

WPA stats

This is a neat writeup, though painful to relive some of those memories.

Something about WPA though strikes me as kind of wrong. I can’t really put my finger on it – maybe it is the fact that a 3 run double in the top of the third when your team is up by 2 runs already is worth much less than a texas leaguer single with a guy on base in the top of the ninth of a tied game. Its as if the designer of the stat has tried to sneak “clutch” in the back door?

I know that this all equalizes over the course of a year – regular players will have several high variance end-of-game plays that will eventually cancel one another out, and trends will emerge. It just seems like of all the highly variable stats that baseball uses, why go out of your way to find one that adds within-game noise to the calculation?

by JBImaknee on Jun 2, 2008 1:18 PM CDT reply actions

Maybe I misunderstood your post

but are you saying you are confused or disagree with why 3 runs in the beginning of the game means less than a run towards the end? I would think that the more innings left in the game, the more chances and opportunity a team has to come back or screw the game up. Its not really clutch.

by Gdawg on Jun 2, 2008 7:32 PM CDT up reply actions

I know why that is the case with this stat

I’m more questioning the validity of the stat as a measurement of a player’s value.

I just don’t really get why we should adjust our perception of some guys “value” more because he hits a single in the ninth as opposed to a 3 run double in the 4th.

by JBImaknee on Jun 3, 2008 11:12 AM CDT up reply actions

This is all very interesting

but I’m really not sure what, if anything, to make of it. It kind of makes it seem like a manager can’t win a game, but he can lose one.

How much did Tejeda’s blow up drop the win chance?

...and curse Sir Walter Raleigh, he was such a stupid git.

by t ball on Jun 2, 2008 3:41 PM CDT reply actions

Quite a bit, looks like

here’s the link

The bottom of the sixth ended with an 88.7 percent win chance for the Rangers. The top of the seventh ended with the Rangers having a 2.3 percent chance of winning. Wright’s four at-bats resulted in drops of 4.3 percent, 7.3 percent, 18.7 percent and 4.9 percent. Tejeda’s three at-bats resulted in drops of 14.7 percent, 21.9 percent and 7.1 percent. The WE dropped from 88.7 to 53.5 percent with the Electrician on the mound and from 53.5 percent to 9.8 percent with Tejeda on the mound. Even though Fx2 gave up a pair of home runs, the damage was done and he only dropped the WE from 9.8 percent to 2.3 percent… although the first batter Fx2 faced, Cust, homered, dropping the WE from 9.8 percent to 3.5 percent with one swing.

If we had WE graphs posted for every game like Lookout Landing’s, Tejeda would be the “biggest suckfest” at -43.7% while Bradley would have the biggest contribution at 30.6%. Rupe had the biggest contribution from a pitcher, 0.6%, while Byrd was the biggest non-pitcher suckfest at -9.9%.

by Inkara1 on Jun 2, 2008 5:15 PM CDT up reply actions