Monday, April 06, 2009

Palmer and Thorn and Denard's Decision

There could be a ton of great stuff to cover tonight, so let's lighting round through them quickly before we get to something that I've been meaning to study for several years.

Michael Cuddyer batting third - Until Joe Mauer comes back, the Twins are going to need to piece together the middle of the order, and I'm not sure how much it usually matter one batter versus another. But it mattered tonight.

Cuddyer got the lone RBI, but he also left a couple of other scoring chances on the field, and he looked overmatched. That can happen for right-handed hitters facing King Felix. So the logical question is "Why was a right-handed hitter batting third?"

It isn't because Gardy didn't want to bat two left-handers next to each other, because two lefties still did end up batting next to each other, in the fourth and fifth spots. It wasn't because Gardy wanted to guard against a late left-handed reliever being used, becaue Seattle's bullpen doesn't have a left-handed reliever.

It's a puzzling choice. On the other hand, it would make perfect sense tonight against the left-handed Erik Bedard, especially because Cuddyer has hit Bedard (3 for 12) better than either Delmon Young (2-15) or Joe Crede (o-11). Don't be shocked if it's Mike Redmond (2 for 5).

The Bullpen Meltdown - Like the Twins this offseason, I think I'm just going to ignore this for a little while. Let's leave it at this: Nice job, guys.

Sitting Delmon Young - In general, I like Gardenhire's approach to this. He's going to play them all and let it sort itself out. And I think it will. After all, let's not forget that there's a reason why none of them are above this little competition. They all could be capable starters, but they all also have their flaws.

Tonight I thought Gardy's choices made the team better. I'm not crazy about penciling Gomez's name above line 8 in my scorecard, but I need to admit the outfield defense really did shine. Gomez made a catch that Span probably doesn't make in center. And Span made three catches tonight that Young doesn't make in left. Of course, I wonder if he really should have made all three of those, so let's move onto our next topic. But first, a little shout out to AT&T....

My Tweeting - I'd like to give a shout out to AT&T, whose 3G network apparently couldn't handle 50,000 people in the area of the dome. So I couldn't shoot out a blog post or a single tweet. Nice job guys.

Denard's Foul Catch - The Mariners scored their first run tonight on a play one seldom sees. With Adrian Beltre on third base and one outs, Jose Lopez hit a fairly deep fly ball to left field that was going to be about 25 feet into foul territory. Span chose to catch it, Beltre tagged up and came home, and the throw from Span wasn't particularly close. Mariners led 1 to 0.

I winced a little, suspecting that "The Book" would say that ball should've been allowed to drop. I think it was obvious to most that by catching it, Span was conceding the run, and Span is heady enough that I suspect he knew the same thing. It was the leading run, so that's bad. But it's also early in the game, so that's good. So I wanted to research what various baseball mathematical models would say about the decision.

First, let's check out Palmer and Thorn's Expected Run Matrix. It's a neat grid that shows, given a certain number of outs and people on base, the average number of runs that should score that inning, based on 75 years of major league games. It was published in The Hidden Game of Baseball by Pete Palmer and John Thorn. It looks like this, and I've bolded the appropriate squares:

OUTSNone1st2nd3rd1st & 2nd1st & 3rd2nd & 3rdFull
0 Outs0.4540.7831.0681.2771.381.6391.9462.254
1 Out0.2490.4780.6990.8970.8881.0881.3711.546
2 Outs0.0950.2090.3480.3820.4570.4940.6610.798

- With a runner on 3rd and one out, one would expect to give up .897 runs. That's the situation if Span doesn't catch the ball.
- With two outs and nobody on, one would expect to give up just .095 runs, but of course a run has also scored, so that's 1.095 runs. That's the situation if Span caught the ball.
- By catching the ball, on average, Span gave up an extra tenth of a run.

So according to Palmer and Thorn, it's fairly close, but the smart move is generally to let the ball drop. Considering Lopez isn't an especially dangerous hitter, and the count would be 1-2, and Francisco Liriano was on the mound, I think the context supports that idea too.

But before we move on, I want to check one other model. Fangraphs.com has a beautiful method for tracking the outcomes of a game called Win Probability. The idea is similar to the one above. For just about every conceivable situation and find the probability of a team winning that game, based on what has happened historically in that situation.

From yesterday's game graph, it looks like the Twins chances of winning that game decreased from 45.3% to 43.2% with that catch. If it was a smart move (and it would have been if the game was later and the Twins held a decent lead) you would have seen the probability increase.

So, it's like we thought at the time. Probably not a good move, but not particularly damaging or stupid either. Which bring me to my next shout out....

Bridge Work on 35W last night at 10:00 - It's nice to know that a little thing like 50,000 fans in cars doesn't stop MNDOT from reducing the major artery from downtown to just one lane. Way to play through adversity guys. Nice job. If you really want to impress me, you should do the same thing tomorrow during rush hour. You know,if it's convenient for you.

Podcast - Which reminds me, if you would like to hear my live reaction to MNDOT's decision, or the 'shout out' I gave to the bassbowl determined to halt any thought of merging, you'll want to check out Seth Stoh's and my podcast last night at MNGameNight.com. And join me tonight, too, while I break down the Bedard/Blackburn duel.

8 comments:

Kris said...

I was having the same thought on the Span catch. Thanks for doing the analysis. That's good stuff!

skepticalMan said...

you should also check what is the % that a play like this would result as a out at the plate. I am guessing atleast one tenth of a run.
I completely disagree, he did the right thing. He should have gone for the catch. It was very early in the inning, and could have led to a really big inning, potentially putting twins chasing their tails from the 1st inning.

SethSpeaks said...

I had the same thought when Denard caught it, but in the second or third inning, you have to still take the out, right? If that's in the 8th or 9th, you obviously let it drop.

If he lets it drop and the hitter ends up with a double on the next pitch and they score 12 runs the rest of the inning, we would be talking about how he just needed to get the outs at that stage in the game.

And, for what it's worth, Michael Cuddyer is batting .333 with runners in scoring position. That's not bad for a #3 hitter.

KEN said...

at&t is too busy spying an americans illegally to upgrade their network.

CapitalBabs said...

Even though I picked the right players that Gardy would put in last night's lineup, I got the order wrong because I just wouldn't have guessed the Cuddy in the 3 hole thing..

I know this will make people scream (or would have) but I fully expected Gardy to put Red Dawg in the 3 hole because that seemed to be his default position last year...

This is how I pictured it:
Span
Casilla
Redmond
Morneau
Kubel
Crede
Cuddyer
Punto
Gomez

*shrug* who knows if that would have worked better last night but considering Cuddy's history against King Felix and how the actual performance came out? Perhaps mine was actually better... LOL Mind you, that was a prediction on what I THOUGH Gardy might do.

I fully expected last night to be Span, not Young, if just to emphasize the speed we have vs the "power" we have for the opening night. It's so much fun to watch Span in the field and I hope he didn't freak out the BP too much by joining them, however briefly.

Jack Ungerleider said...

John, your phone situation is a long standing issue. I remember when my in-laws brought a Cingular phone up from Oklahoma and the got spotty reception in town and none as they went north. (Accept for a few areas of Duluth.)

The odd thing about the Span foul ball play is we had a game here last year, Rays maybe, that we won because of a similar play.

P.S. - You need to teach them how to pronounce your last name on the radio. 8^)

Evan Nelson said...

You have to take the outs when you get them. Span should catch the ball.

In a pitch-count controlled context, think how many pitches were saved by Span's catch. Liriano had 97 pitches last night. Considering that there's little chance the Twins would let him pitch up to 100, if Span wouldn't have made that catch, Liriano might have been yanked earlier in the game, and those strike-phobic relievers would've been in the 8th.

Always take outs.

Further, you should retract the "Lopez isn't a dangerous hitter."

A) All MLB hitters are dangerous.
B) Jose Lopez had three RBIs last night, more than any other Mariner.
C) Jose Lopez is a better batter than 3/4ths of our infield. (The 1/4th ain't Punto.)

TT said...

Why do you assume the runner will score? Certainly that is the most likely outcome, but if he failed to score even once in 100 plays, the odds flip back in Span's favor.

The problem with the claim that you are considering "every situation" is that it isn't really true. These charts are averaging a lot of variables that are important.

For instance they assume every player involved on both sides is average. Something no one really believes.