Shane Victorino batting with the bases loaded
Posted by Andy on October 7, 2010
I don't watch a ton of baseball on TV and during last night's Phillies-Reds game, I was reminded of one of the reasons why.
When Shane Victorino game to bat with the bases loaded in the second inning, TBS put up a graphic about Victorino's career with the bases loaded. It showed some of the data available from Victorino's splits.
It said something like he was 22-for-60 with a .367 batting average and 54 RBI.
My question is: what does the average baseball viewer think when they see these numbers? Someone who is really knowledgeable about stats (i.e. not the average fan) might have some sense of what these numbers mean, but to the average person the stat might as well say "Globally, there have been 74 earthquakes of at least 3.0 magnitude in the last 6 months." Is that significant? Is it a lot of earthquakes or very few? What does it mean? Who the hell knows without more information?
Here are my issues with the Victorino data as presented.
- First of all, I don't want to know his at-bats. I want to know his plate appearances. How many times has he walked with the bases loaded? That data is easily available. He's had 67 career (regular-season) plate appearances with the bases loaded. In 6 of those, he walked. In one of those he hit a sacrifice fly. His OBP with the sacks full is .418.
- Secondly, we need to know what the typical performance is for MLB. In 2010, for example, over 4,702 plate appearances with the bases loaded, the total major league performance was a .281 BA, .331 OBP, with 319 BB and 3,228 RBI. So that's 0.69 RBI per plate appearance with the bases loaded. Victorino, with 54 RBI in 67 plate appearances, is at 0.81 RBI per plate appearance, so he's doing quite well. Across MLB, players walk 6.8% of the time with the bases loaded. Victorino has done it 9.0% of the time.
So, yes, Victorino has done really well with the bases loaded. I just wish that when they showed his stats, they'd also show major-league averages--at least just batting average. That at least gives some context to Victorino's performance.
Beyond these basic comparative stats, there's one other thing I'd be curious about:
- What fraction of Victorino's plate appearances have come with the bases loaded? He's got 3,043 career PAs, so 67/3043 = 2.2% have come with the bases loaded. I suspect that's well below average since he has hit leadoff a lot, meaning that game-opening PAs are clearly never bases-loaded, plus hitting after the pitcher can't help either. In the majors in 2010, 4,702 out of 185,525 plate appearances came with the bases loaded, or 2.5%. So Victorino has been low, but now terribly.
October 7th, 2010 at 9:25 am
He's still no Pat Tabler.
October 7th, 2010 at 9:26 am
In that particular situation, even if we're not sure what the league BA is with bases loaded, we know .367 is very good. But I do agree that far too often the telecast presents numbers with no context. Like a switch hitter's HR/RBI numbers from each side of the plate. Well, we know he must be batting left-handed most of the time, but how often? How do those numbers compare? And a lot of the time I hear the broadcaster, who must know better but isn't choosing his minimal brainpower, state that Sam Switchy has shown a lot more power from the left side this season. Simply because the graphic says he has more HR and RBI that way. Which is obviously because he's had at least twice as many opportunities left-handed.
Don't get me started on all the problems with baseball telecasts! Well, you already did, but I'll stop myself for now...
October 7th, 2010 at 9:27 am
*but isn't choosing TO USE his minimal brainpower.....
October 7th, 2010 at 9:34 am
Excellent work and adding "batting after the pitcher" impact was a nice detail.
I think the TBS folks (in lieu of doing your level of work) want to give viewers a gut feel on two questions.
1. Has this guy "been here before?" Is he feeling the pressure? (60 at-bats sounds experienced, 5 would not)
2. Is he likely to "fold under the pressure?" .200 batting average = temper your expectations, the guy's going to fail. .300 = Yes! Anything less than a run is cause for disappointment!
Neither answers a meaningful question, as you did. They just frame expectations for fans at a basic level of stat awareness.
October 7th, 2010 at 9:36 am
Not to mention perhaps the biggest thing that's missing, which is, of course: small sample size. Not that his performance isn't interesting, or worthwhile to show to fans, or maybe even a little meaningful, but when we're told he's had 60 ABs with the bases loaded over 5 full major league seasons (plus parts of two more), well, that's a teeny tiny sample. Is he actually better with the bases loaded? Well, in practice, yes - he has succeeded at a higher rate than in all other types of ABs over his career. But does that actually mean anything in terms of this one particular plate appearance we're watching on TV right now? Probably not.
October 7th, 2010 at 9:59 am
You guys are nuts. Context? You mean to tell me without knowing what other batters have done, you can't tell that batting .367 in 60 abs with the bags full is pretty good? Why do you need to know pas? I feel bad that you can't enjoy a baseball game.
October 7th, 2010 at 10:10 am
Another example: Last night Smoltz stated that Liriano this season had a 1.56 ERA in wins and 7.97 in losses, so this shows he was either very good or very bad. But that is not unusual at all, it's perfectly normal. All AL pitchers this season had a 2.04 ERA in wins and 7.71 in losses. Liriano's spread is slightly more, but I bet it wouldn't be hard to find several pitchers this year who were more extreme.
When you present numbers with which most people are unfamiliar, it is quite easy to surprise them, because they have no real idea what to expect. There were a few situations in the little of the game I saw where Smoltz seemed to do this to himself. It's great he is curious enough to explore some numbers beyond the norm, to try increasing our understanding of the game, but he needs to go beyond that and find what these numbers really mean, so his viewers know what they mean as well.
October 7th, 2010 at 10:27 am
"I don't watch a ton of baseball on TV..."
Too busy in your mom's basement playing baseball games on computers?!?!?!
That isn't even the most egregious example of crap from TV broadcasts. But, where being wrong and stupid is one thing, simply being lazy is quite another. Don't just slap stats up there for the sake of doing so. Stats should inform. Those don't, as Andy points out.
October 7th, 2010 at 10:51 am
A little stat knowledge in the hands of a color commentator is potentially a dangerous thing, especially if the guy is a former player and perhaps not exactly a Rhodes Scholar when it comes to mathematics. I heard someone on the TBS broadcast of the Yanks-Twins game say that when the Twins put the ball in play, as a team, they're hitting .306, "So you know they're gonna make something happen, 3 times out of ten" if they make contact" or something like that.
Never mind that the league average BABiP is about .295 or .300 every year, or that the Yankees are also hitting .300 in that situation, or that basically you can't really control what happens much in that situation anyway.
Paul O'Neill had a similarly misguided comment about BABIP last month when the yankees were playing Detroit, talking about how (at the time) Austin Jackson was hitting something like .412 when he put the ball in play. Paulie took this to mean that if he could just cut down on the strikeouts, boy howdy, he'll be reaaly awesome, when what he should have taken from it is that Jackson is due for a lot less action in 2011. Oh well.
October 7th, 2010 at 11:28 am
In the third inning of the Yankee/Twins game, Sabathia got to a 3-2 count on the leadoff hitter. In the previous inning, he let up 2 runs. The announcers went on and on about how Sabathia had set the water mark for his team, and now they knew they only needed to get 3 runs to win, since he bears down and gets better as the game goes on. This, in spite of the fact, that the previous inning yielded 2 runs and he was yet to record an out in the current inning. The leadoff hitter eventually singled and scored. But, waht did that matter... the narrative was already constructed. Despite zero evidence to suggest that Sabathia was, in any way, "bearing down" compared to his 2nd inning work, the announcers went with that line anyway. Ah, what brilliance...
October 7th, 2010 at 11:50 am
JT @2,
I thought that exact thing last night when Texiera came up and they showed his Righty/Lefty splits. Higher avg righty, but way more HR lefty. I thought at the time we needed to see the HR/AB ratio also. They put the info there to let you know if he is better off batting righty. But you really can't tell.
October 7th, 2010 at 11:54 am
#9 Travis--at least O'Neill understood as much as he did. I don't hold it against him that he doesn't realize that BABiP tends to converge to the mean.
October 7th, 2010 at 11:57 am
I am not a Phillies fan. I am not anti-Phillies either.
Can anyone tell me why I do not like Shane Victorino. I have no idea what it is about him, but he might be my least favorite player. It is very odd because there is no reason. Usually it would be because they are jerks or they are really good and constantly beat my team (like several Red Sox - Manny, Ortiz, Youkalis, Pedroia, etc.) But with Shane those reasons just do not apply.
October 7th, 2010 at 12:01 pm
I have nothing against Victorino but always felt that he appears, at least from afar, to be a bit "gangsta" in the sense that he seems to be trying to look cool and aloof. I have no idea if that in any way resembles his actual persona...just my impression.
October 7th, 2010 at 12:13 pm
@Tmckelv:
It's because he looks like Yul Brenner.
http://t0.gstatic.com/images?q=tbn:ewXIPphuAjsobM:http://img148.imageshack.us/img148/4099/yulfe0.jpg&t=1
http://www.aroundhawaii.com/assets/articles/2009/09/1830/images/2009-09_vifvictorino010.jpg
That 1000-yard stare scares the hell out of you.
October 7th, 2010 at 12:17 pm
Another factor in determining the set of players hitting with the bases loaded is that poor hitters will bat with the bases loaded more often than simple chance might predict because teams will often walk superior hitters to load the bases (depending upon the situation) in order to face the inferior hitter. On the other hand later in the game a team might be more apt to pinch hit for the weaker player in the bases loaded situation. Other factors would include batting in a part of the lineup where singles hitters, high base on balls guys and slower runners are more prevalent.
Obviously some of these factors cut each way in terms of quality of batter, but I would be hesitant to assume that the league average bases loaded hitter is the same as the league average hitter in general.
October 7th, 2010 at 12:31 pm
#6 Frank.
Here is another example.
Let's say I tell you than in 2010, a certain player batted .260 with 2 outs and runners in scoring position. Let's say in 50 such at-bats, he had 20 RBI.
Did he do well or poorly?
Let's say I show those stats on the telecast. What do you think about that? Clutch player or not? Guy you want up in that situation or not?
Well here are the league-average 2010 numbers taken from here.
League average batting average with 2 outs and RISP is .240. So a a guy hitting .260 is doing a lot better than average. Across MLB, there were 6,223 RBI in 19,679 AB in these situations, an average of 0.32 RBI per AB. Our guy had 20 RBI in 50 AB, a rate of 0.40 RBI per AB, so better than average.
So the bottom line is that our guy did pretty damn well, despite what you might think is a low BA.
Don't tell me that context doesn't matter. You can't figure anything out with the bare numbers without considering the context.
October 7th, 2010 at 12:43 pm
Would love to see that analysis done for Pat Tabler in his time.
October 7th, 2010 at 1:23 pm
Andy, the example you give in #17 does need context. But I agree with Frank, the "incomplete stat" that set you off, does not really, especially in a braodcast where this information has to be conveyed-- and comprehended -- between pitches.
October 7th, 2010 at 1:30 pm
@15 Travis,
thanks for the well thought-out theory. the only problem is I like Yul Brynner.
"et CETera, et CETera, et CETera"
maybe that will help me to like Victorino better, though. 🙂
October 7th, 2010 at 1:37 pm
I disagree, completely. You guys think the Victorino one doesn't need context because .367 sounds like a good number. I just gave an identical example, albeit it in the other direction. A .260 average sounds like nothing all that good, except that it turns out to be way above average. You guys are making the assumption that .367 is a good number and while you're right, you didn't really have a great basis for assuming that.
Let's say a given team's starters batted .285 this year. Is that good? Overall batting average was .257 so it seems good. But all starters hit .297, so .285 isn't all that good.
Let's say a guy hit .350 when the count reaches 3-0. Is that good? Well league average was .409. But .350 sure sounds good, doesn't it? But it's below average.
A guy on my favorite team hit .305 with a runner on 3rd, less than 2 out. He hit .285 the rest of the time. Pretty good, right? Nope. League average was .318 in those situations.
There's a right-handed batter who hit .380 this year when he hit pulled the ball into left field. Really great right? Nope. League average was .434 in those cases this year.
All of these stats are available from the 2010 MLB batting splits page.
http://www.baseball-reference.com/leagues/split.cgi?t=b&year=2010&lg=MLB
Need I go on? ALL of these numbers need context.
October 7th, 2010 at 1:58 pm
Wow, 319 RBI walks this year in MLB?
October 7th, 2010 at 3:45 pm
@21
Andy -- Great reply. I was about to cite the example of the 3-0 count, then saw you'd nailed it.
One of my "aha!" moments in understanding baseball stats in context came a few years ago when I first found the splits on B-R. I had been noticing, anecdotally, that when a baseball announcer mentioned a player's BA with the bases loaded, it was usually over .300. When I finally had a chance to look at the MLB splits, I saw that the MLB average with bases loaded was consistently 20 to 30 points higher than with bases empty. (This year: .252 with bases empty, .281 with bases loaded.)
So I realized that bases-loaded BAs were often deceptive. In reality, if a guy is a .290 hitter overall and hits .300 with the bases loaded, he's doing *worse* than should be expected.
Once I saw that the rise in bases-loaded BA was an MLB-wide phenomenon, I naturally wondered why. I couldn't believe that the majority of hitters simply thrive in such pressure situations, nor that the majority of pitchers struggled. So what could be driving up the BA so much in that situation?
I kept looking at the Bases Occupied splits and thinking, and it finally hit me: It was nothing but the silly rule that sacrifice flies don't count as at-bats! When I added the sac flies as ABs and ran the True BA (as I call it), it was very close to the bases-empty BA. This year, True BA with bases loaded was .259.
So, by understanding the context, I was able to see that what seemed to be a higher frequency of hits with the bases loaded was just a mirage, an accounting trick.
Now, it happens that Victorino has hit just 1 sac fly with the bases loaded, giving him a .361 True BA in that situation. But the audience wasn't told that; nor were they told that you can expect BA to be about 25 points higher with the bases loaded. They were just given a number that sounds like a batting champion.
All numbers need context.
October 7th, 2010 at 3:57 pm
@22
Jimbo -- 319 bases-loaded walks this year sounds high, right? I kept noticing bases-loaded walks this year and thinking, too many pitchers just can't throw a strike when they need to.
But my conclusion was based on a false assumption. The bases-loaded walk rate this year, 6.8% of PAs, is perfectly normal compared to years past. For example, in 1960 -- after integration, before expansion, theoretically the point at which the ratio of top pitching talent to MLB roster spots was at its highest -- the bases-loaded walk rate was 6.9%. I've checked it for several years in between, and there's not much variance.
The walk rate in all other situations this year was 8.5%. It still seems to me that MLB pitchers should be able to do better than a 20% reduction in walk rate when a walk means a run. But the simple fact remains: Today's bases-loaded walk rate is not unusual in historical context.
October 7th, 2010 at 4:30 pm
I wrote about bases-loaded walks back here:
http://www.baseball-reference.com/blog/archives/4459
This year's total isn't all that different from the totals for the last few years.
October 7th, 2010 at 4:34 pm
The thing about the context is - and I'd have to watch to see - I don't think they put stats up there unless they actually tell a story. The network doesn't put up someone's stats when they hit .260 with the bases loaded because there's nothing significant about the number. The fact that Victorino hits .367 with the bases loaded - somebody said it earlier... It says he hits well given the opportunity. The fact that they're even putting a statistic up on the screen IS the context.
I honestly have to say I can't remember ever seeing them post a situational stat on the screen that was mediocre. I could be wrong - maybe I just notice the telling ones.
October 7th, 2010 at 5:08 pm
Here's my take on the Victorino bases-loaded stats. "Context" should include his BLAB/PA ratio (bases-loaded at bats over plate appearances) compared to the league-widee average, and how many years he's been playing ball. How many times he walked is irrelevant, because the pitcher is obviously trying to throw strikes - Victorino isn't trying to work a walk out of the guy, because with the count at three balls and however many strikes, Victorino is looking to swing at a hittable pitch. Averaging almost an RBI per BLAB over a five-year career, and BLAB totals of 50 or more (10 times a year), is something I would consider a workable sample size on which to judge his consistency. After all, focus groups are as small as five to ten people and they can predict national elections with that. (So how come the national weather service doesn't use focus groups instead of radar, seems like their 30-day forecasts would be more accurate? But I digress....) The guy has a earned a reputation for hitting with ducks on the pond. That's another stats RBI/DOP, which in Victorino's case comes out to 54/180 = 30% (60 at-bats with bases loaded = 180 runners on base). How many players drive in 30% of all runners on base across total career or season plate appearances? And that doesn't take into account the 4th RBI when he hits a grand slam, so there's another stat (GSRBI) to plug into the formula somehow.
October 7th, 2010 at 5:40 pm
Tom-
I disagree. I remember originally being impressed when I saw guys with RBI totals approaching the number of ABs they had with the bases loaded. I thought, "Wow, this guy gets an RBI nearly every time up!" Then, as I thought more about it, I realized this might not be as impressive as it seems. An out will often score a run with the bases loaded. And any hit with the bases loaded is likely to score more runs than the same hit with a different baserunner situation. But, I didn't know for sure and couldn't know for sure without context.
The tricky thing is we have to define what "good" is. Hitting .360 is good hitting in the sense that the batter is getting a lot of hits. He is likely hitting the ball well. But, in terms of value, it might not be very much if everyone else is doing better. Context is always important. The only reason we think of .360 as good is because we have the context of what BA typically ranges from.
October 7th, 2010 at 5:41 pm
Mistertmo-
I'd love to agree with you, but see far too many useless stats.
My personal favorite? Before Super Bowl 41 (I believe... I could be wrong), they showed that the coin toss had previously landed heads 20 times and tails 20 times. Really???