Monday, April 21, 2014

8 comments Local Announcers Think The Same Advanced Statistics They Don't Seem to Care to Understand Have No Place In a Baseball Broadcast

This article isn't really by Bruce Jenkins, it's more of a compilation of comments that Bruce has gathered. These are comments made by local announcers about the use of advanced statistics during a baseball broadcast and really the only opinion reflected here is the opinion that advanced statistics should not be discussed during a baseball broadcast. Includ anecdotes during a broadcast? Abso-fucking-lutley. Advanced statistics? Nah, because "the people" won't understand the message being conveyed. Bruce poses the question because the Houston Astros have decided they will encourage their announcers to use advanced statistics during the broadcast and Bruce needs local announcers for more successful teams to describe how stupid this is. After all, it's the Astros and they aren't a very good team. So obviously this has to be a very bad idea since it came from management of such a shitty team. There is obviously a correlation between a bad baseball team and bad ideas ownership of that team may have. It's just like good baseball teams never have any bad ideas.

My personal opinion is that advanced statistics absolutely has a place in a local baseball broadcast. I don't believe the announcers should bust out with a whiteboard and start breaking down these advanced statistics or talk down to their audience, but there is a place to at least mention advanced statistical theories and alternative methods of measuring a player's performance. It's important these advanced statistics aren't presented in a way that the announcer is talking over the audience's head, so the announcer will have to be well-versed enough in advanced statistics to make it simple for the audience. Therein lies the problem. These announcers aren't well-versed enough to help the audience understand exactly what WAR (since that seems to be the advanced statistic most of these broadcasters latch on to) is in the context of discussing a player's performance during a current or previous year.

The baseball season is upon us, and as we turn our gaze to the east, we find the Houston Astros taking a most peculiar stance.

Not an "alternative" stance or a "different" stance, but a "peculiar" stance which immediately shades the decision to include knowledge of advanced statistics by the announcers as out of the norm and possibly the incorrect decision. The bias is in the writing.

They not only like the idea of their broadcasters going deep into modern-day statistics, they consider it a prerequisite for the job.

It's insane to think the Astros would want announcers who are well-versed in the game of baseball to discuss methods of evaluating a player's performance outside of the traditional methods. Who would ever want to hear a broadcast as presented by a well-rounded announcing crew? Certainly not Bruce Jenkins or any of the other broadcasters that Bruce interviews in this column.

In other words, tuning into a typical Astros game: "Well, here's big Hank against a real tough lefty. You wouldn't believe how this guy's WAR stacks up against that guy's WHIP."

This example only goes to prove that Bruce Jenkins doesn't understand advanced statistics, so yet again we have the example of a sportswriter criticizing that which he doesn't understand. Bruce only shows his ignorance in this example.

At which point a torrent of statistical information unfolds, all about Wins Above Replacement - an effort to summarize a player's total contribution to a team in one statistic - and Walks Plus Hits Per Inning Pitched.

But see it doesn't have to be a torrent of information that is provided. WHIP can be summarized and given context by explaining what WHIP is and saying whether a pitcher with a 0.97 WHIP is good or not. The viewer doesn't need a calculator and there doesn't have to be a ton of information given. It can be very simple. Say a relief pitcher comes into the game against a left-handed batter. The announcer can say this pitcher's WHIP against lefties is 1.64 while his WHIP against righties is 1.00. This gives context that the relief pitcher is better against right-handed batters and gives the viewer the knowledge that this could be a rough at-bat coming up.

WAR can be more difficult to calculate during a game, but WAR is more of a macro-statistic as opposed to a micro-statistic anyway. WAR is better at giving the big picture over a larger statistical sample and wouldn't merit being mentioned constantly during a broadcast other than to compare two players and their performance over a period of time.

The Astros are sort of weird that way. Perhaps in an effort to steer folks away from the fact that their team is awful, losing a combined 324 games over the past three seasons, they are lurching earnestly into the unconventional.

Of course, because it's the Astros doing this then it means there's no viable reason to mention advanced statistics during a baseball broadcast.

For all we know, they're at the vanguard of progress. But they're facing a ton of resistance, and when it comes to filling the broadcast air with analytics,

You mean, resistance like this very article's introduction that describes using analytics as "unconventional," "peculiar," and suggests it is just a cover to make up for the fact the Astros don't have a very good team?

they won't find any allies on the Bay Area's broadcasting crews. We checked with all eight of them, and they have the floor:

What I find most interesting is that most of these broadcasting crews are like, "Oh well those analytics don't transfer to a discussion on-air," even though most of these broadcasters admit they either (a) have looked into advanced statistics and understood them but simply won't use them or (b) don't care to understand advanced statistics and clearly lack knowledge about them. Simply falling back on, "It won't work" is a lame excuse. It's just resistance to using a method these broadcasting crews either don't fully understand or don't want to fully understand.

Vince Cotroneo, A's radio: "If I'm out in the backyard barbecue, listening to a game, I want an entertaining broadcast.

Bring in puppies or kittens! Everyone loves puppies and kittens. Perhaps bears can wrestle during a broadcast. Wrestling bears are very entertaining.

Maybe I can tell 'em that Sonny Gray was a star quarterback in Tennessee, and as you watch him - leader, ornery, trusts his stuff - you see that he has those vital qualities.

I don't find the knowledge that Sonny Gray was a star quarterback in Tennessee to be entertaining at all. Why is it that an anecdote that has nothing to do with how Sonny Gray plays baseball is entertaining and explains why he's a good pitcher, but explaining Gray has the fifth best WHIP in the American League doesn't explain Gray's qualities sufficiently?

And most of all, what's happening on the field? That's what they want to know. I absolutely look at this stuff: FanGraphs, Baseball Prospectus, some other sites. I just don't think it has a place on the air.

Those aren't "stuff" those are baseball sites. That's like if I said I understand vaccinations don't cause autism, but I still believe vaccinations do cause autism because I look at stuff like WebMD and other sites. Vince Cotroneo visiting these sites doesn't mean he is in a position to make a judgment on whether advanced statistics have a place on the air.

"For all the ridicule you hear about traditional stats - RBIs, wins and losses, batting average - those things resonate with the listeners, and they are very relevant to the players. They don't want to hear that 'wins' means nothing.

No one is going to say that wins mean nothing, but staying ignorant about other methods to evaluate baseball players isn't the winning combination to have an informed audience. I don't want to hear the Braves are fucked and won't have enough hitting from guys like B.J. Upton and Dan Uggla to win the NL East, but it's not going to hurt my feelings if someone says it. Wins are a method of evaluating how well a pitcher is pitching. Giving the audience access to information that says whether a pitcher wins a game or not doesn't reflect entirely on the pitcher's performance is relevant and can resonate with the listener if explained correctly. Also, who gives a shit what the baseball players think? They are playing baseball. The broadcast is for the fans.

Bottom line: I think there are enough numbers in the sport. Give people the atmosphere. Let the game come to you."

Cotroneo can't even talk about talking about baseball without using cliches.

Dave Flemming, Giants radio/TV: "I find a lot of it extremely interesting, and there are things you can incorporate on the air. For example, the basic numbers don't tell you the difference between the Giants playing at AT&T Park and the Rockies at Coors Field - revealing that the Giants' offense really isn't that bad, and might even be above average. On the flip side, maybe we overrate the Giants' pitchers because the home park makes their numbers look so good.

Perfect. Then use a statistic to show the audience exactly what is meant by these statements. The audience will understand the point trying to be made, and if they care enough, choose to dig a little bit deeper. This is just the same as using traditional statistics to point out how two different ball parks are different from each other (by using runs scored/allowed in each park) except it goes a little bit deeper.

"Last year, Carlos Gomez came in with the Brewers, and everything the Giants hit, he caught. It was astounding. I did a little digging, and the fielding analytics said he was having a historically great season in center field - sort of confirmation of what I was thinking

I personally think this is an excellent use of fielding analytics to prove a thought Flemming had upon watching Carlos Gomez play. Contrary to what idiots like Murray Chass write when talking about Sabermetricians this is a good example of how advanced statistics should work. They can work to prove/disprove what the viewer is seeing while watching the game. Advanced statistics, traditional statistics, and watching a game are supposed to complement each other, not have one work against the others.

Still, for all of its value, you have to wonder, how much can an audience actually absorb? The answer is, really not that much. There's an arrogance behind the theory that your audience is hanging on every word, to the point where they can understand a long explanation of numbers.

Again, while I agree, there is also a way to convey this information to the audience without overwhelming the game and broadcast with a litany of numbers. Flemming could simply tell the story he just told Bruce Jenkins and explain what statistic he used to confirm what he was thinking. This is the same thing Flemming would do using traditional statistics if he stated on-air that Carlos Gomez seemed to be on a roll playing at home lately as compared to playing on the road. Then Flemming could explain he looked it up and Gomez has a .453 batting average at home and a .214 average on the road during the last month. He doesn't have to explain the statistic, but give the viewer the context and information required to prove the point made by the use of advanced statistics.

The analytics might say this guy's the best fastball hitter in the game, but maybe tonight he's late on somebody's fastball. The thing I always emphasize is to look up, not down. Don't read when you have a game in front of you."

But mention "this guy" is the best fastball hitter in the game according to a certain advanced statistic and this gives the viewer more information on just how well the pitcher is currently pitching. It's looking up, not down and is a statistic that enriches the viewer's enjoyment of the game.

Ken Korach, A's radio: "I think about this a lot, because the times have changed. Although I take interest in certain things - some of the fielding stats, or how often a pitcher tends to hit his target - my knowledge of advanced metrics is pretty rudimentary.

But the fact Korach's knowledge of advanced metrics is pretty rudimentary, this doesn't prevent him (of course) from giving his opinion on the subject as if he were an expert. I don't follow politics but let me tell you why President Obama is a crook...

I'll always feel the anecdotes, stories and human-interest elements are most important.

These are great, but these are not analysis nor are they any type of information that provides the viewer with additional information or context on what's happening on the field. The fact Freddie Freeman is engaged to a model doesn't help me understand why he is one of the better fielding first basemen in the National League.

"When I talk to young announcers, I tell 'em your pile of information is never as important as what's going on out there on the field. No matter how brilliant you think you may be, you'd better tell me if the guy's a left-handed hitter, if the center fielder is shaded over toward left-center or the third baseman's in on the grass. Unless you tell me that, I don't care about all the other information you're putting out."

Deep sigh. Obviously no one is asking announcers to simply spend an entire broadcast spouting off statistics and some description of what's happening on the field is required. Why can't the broadcaster and analyst say where the outfielder is shaded and then point out this certain batter when going against a left handed pitcher usually hits the ball to that side of the field?

It drives me crazy sometimes. I hear announcers say, "It appears the shift is on for Brian McCann" and then the inevitable overhead shot is shown that shows the shift, but no explanation is given for WHY the shift is on. Then if McCann gets a base hit to left field the announcers act like he discovered the Holy Grail. What if the statistics show McCann doesn't have pull-tendencies against a left handed pitcher and tends to spray the ball all over the field? Then the outcome was to be expected, right? Context is what I want sometimes.

Jon Miller, Giants TV/radio: "I'm fascinated with statistical information, going back to Bill James and his 'Baseball Abstract' in the '80s. I spend a lot of daytime hours poring through that stuff. But for the broadcast itself, particularly on radio, I feel it just sounds like a bunch of numbers. They don't tell you what's really happening.

I agree, advanced statistics don't tell a viewer what is happening. These advanced statistics can give context for what just happened or may happen though. Are these announcers too stupid to understand they can say what's happening on the field while also using advanced statistics to explain the action on the field?

"I grew up with Russ Hodges and Lon Simmons: 'Here's Willie Mays, hitting .317, 26 homers, and leading the league with 51 RBIs.' The whole point being, what kind of a season is he having? Most of the analytic stuff, like WAR, is more of a predictive thing, telling you someone's worth.

This is not wholly accurate.

Well, I want to know how he's doing right now, especially measured against the rest of the league. WAR is not only hard to explain, there are two different versions, and hardly anyone knows how to calculate it. How am I going to fit that into a broadcast?

Fair point, but there are other advanced statistics that aren't as complicated as WAR that could be used, even briefly, in a broadcast.

"I hate to say it, sort of like fighting words, but too often, the analytics people show a lack of understanding of the game. These aren't people who have played, managed or really experienced baseball.

I don't understand what "experienced baseball" means but I know Jon Miller isn't someone who has played or managed baseball. Jon Miller went to college to be a broadcaster and has been a broadcaster his entire life. He hasn't managed or played baseball at the professional level, yet he has the balls to act like "analytics people" who haven't played the game don't know what they are talking about, but he does because he's sat in a booth with Joe Morgan and called games. See, Jon Miller understands baseball because he's a broadcaster, that's his resume for being such a fucking baseball genius.

And yet, if you disagree with one of their statistical theories, you're stuck in the past, you don't know what you're talking about. So you take a bunch of coaches and managers, guys who have been successful for 20-30 years, and none of 'em know what they're talking about

This is the same boring, cliched argument every idiotic person who dismisses analytics makes. These are fighting words, but too often, the anti-analytics people show a lack of understanding of analytics. They aren't people who understand how analytics work.

There is nothing wrong with disagreeing with a "theory," even though much of analytics isn't a theory but statistical calculations (which shows just how little Jon Miller does truly know about advanced statistics), and nobody says coaches and managers don't know what they are talking about. Much of advanced statistics is about learning a different way of measuring player performance. It's so frustrating these guys like Jon Miller don't even care to understand what point the analytics crowd is trying to make.

"Hunter Pence had 22 infield hits last year and he was swinging as hard as he could, every time.

Jeff Francouer swings as hard as he can each time and he sucks no matter which metric you use to judge him.

How do you measure hustle in on-base percentage against a guy like (notoriously slow) Bengie Molina?

You can't because hustle is an intangible. That's the good part about on-base percentage, it doesn't factor in made up intangible bullshit like "hustle." A player either gets on-base or he doesn't.

Willie Wilson got criticized in Kansas City for his low on-base percentage as a leadoff guy, but one year he reached base 27 times on errors. Because he was so fast, he made people panic. None of that shows up on the sheet.

Oh, so Jon Miller knows for sure Wilson made it on-base 27 times on errors because the fielders were in a panic? It must be nice to read minds. Perhaps Wilson just had a lucky year. You know what, it doesn't even matter because no Sabermetrician is saying all old statistics are useless...except for Brian Kenny with the win, but he goes overboard with his criticism of that statistics. I hate the win statistic, but it does serve some sense of a purpose.

Glen Kuiper, A's television: "I think if I started telling people about WAR, most of them would have no idea what I mean. I don't use it, I don't think about it, and my partner (Ray Fosse), who's very old school, certainly isn't going to use it. You have to define it and then explain it, and I don't think fans want to work that hard.

Why does every single person who hates analytics, and seemingly every person interviewed for this article, grab on to WAR? There are other advanced statistics that can be used during a broadcast. It's like WAR is the only advanced statistic they know...they probably latch on to it because it's easy to spell.

"I'm not saying it doesn't have value. If a general manager makes a big decision based on analytics, that's very legitimate from a player-evaluation standpoint. Or maybe you do a feature on it during a break in a telecast.

See, they latch on to WAR. I'd love to read criticism of advanced fielding metrics, but these guys don't understand those metrics and therefore can't discuss why they don't have a place in a baseball broadcast.

But you can't just squeeze these things in between pitches, and personally, telling me somebody's WAR numbers isn't going to make any difference how I see that player.

Right, because the broadcast is about you and not about the fan sitting at home. Good to hear.

I'd rather tell a story, give fans an idea of what kind of a person this guy is. That's way more entertaining for a fan."

And I wonder why baseball seems to trend toward the older crowd. I hate to point this out, but I think the younger baseball crowd are entertained by analysis of a player and not whether he likes roller coasters and playing silly jokes in the clubhouse.

"I just think the game itself, as it's being played right then, is so much more interesting than how a player values out over a full season, or how he projects for the future."

(Shakes head sadly) Right, because it's fine to talk in generalities about intangibles such as hustle, but pointing out Kelly Johnson has bad range at third base isn't relevant in the 9th inning of a close game with a runner in scoring position.

Mike Krukow, Giants TV: "What the analytics people fail to realize is that we're talking about human beings.

Oh yeah, definitely. The analytics community thinks baseball is played by robots. This is absolutely true. Do all members of the same anti-stats crowd read from the same handbook or is there just a lot of groupthink in the community?

That stuff can't measure someone's passion, toughness or feel for the game.

That stuff can't measure the amount of bullshit shoveled by Mike Krukow either.

It doesn't recognize that, as a pitcher, you might come out with only two of your pitches working. That the wind's blowing out, the sun field is in right, this umpire's got a huge strike zone. I can see where WAR would be relevant in an arbitration case.

So it's okay to not talk about WAR when a player is on the field, but when determining how much a player should be paid WAR is relevant?

But in a ballgame it's B.S., and you can tell when a guy's force-feeding a bunch of totally inane facts.

Oh ok, but it's definitely useful when determining a player's compensation though? Plus, WAR is based on how a player performs during a ballgame so it's not B.S. in that aspect.

Ray Fosse, A's television: "This game is over 100 years old. And it's a generational thing, fathers teaching their kids - just like my dad taught me. You can always get around to the numbers, but basically, you're teaching the game.

And an announcer plays the role of the viewer's father, teaching the game?

"It's depressing to me that numbers are considered more relevant than, say, a really veteran scout.

And this isn't a straw man argument at all because veteran scouts are often put on a local broadcast to impart knowledge to the viewer, right?

Then Ray Fosse starts talking about a specific Phillies scout because he obviously has no interest in answering the question posed to him.

Duane Kuiper, Giants TV/radio: "I'm probably the worst guy to ask, because I think most of the new statistics are nothing more than a distraction to the people at home.

But don't let your personal opinion affect your answer. We want as biased of an opinion as possible.

I've never received a letter saying Mike and I should do more of it, and it's really tough if you don't really believe in it.

Probably because someone who would want to see more advanced statistics are going to be writing an email rather than writing a letter like it is 1961.

"I especially resent the discounting of traditional numbers. Stuff like RBI, ERA, wins or losses, those things tell you something. They're part of the fabric our fans were raised with.

I guess these statistics are being discounted because there are allegedly more accurate metrics that can be used? God forbid that should happen. Not that the traditional statistics crowd feels threatened or anything of course.

20 years from now the part of the fabric the fans of baseball were raised with will be advanced statistics, so doesn't that mean they have some place in a broadcast?

They are part of the players' language. And there are certain things numbers can't really describe, like Hunter Pence (laughter).

The numbers have traditionally described Hunter Pence as a pretty good baseball player, but good try. Nothing can describe Pence's approach at the plate, but his performance certainly is well-quantified with numbers.

"Just because agents and general managers use analytics, that doesn't mean I have to.

Well, at least you are being open-minded about it. Just because Duane Kuiper doesn't like advanced statistics and doesn't think they have a place in a broadcast, that doesn't mean they don't have a place in a broadcast. Of course Kuiper doesn't think of it that way because he's beholden to his own biases and thoughts which reinforce his world view.

What team did you say is putting this stuff on the air? Houston? Well, I don't think you have to say any more, do you?"

Right, because if a team isn't very good on the field then that means any ideas that team has which are not considered part of the norm must not be very good either. I think broadcasters like Kupier and Mike Krukow are better off talking about baseball and not opening their mouths to talk about anything that shows their ignorance and fundamental resistant to change.

8 comments:

HH said...

And most of all, what's happening on the field? That's what they want to know.

Not much. Not much is happening at all. The average baseball game has something like 12 minutes of action in three hours. Plenty of time to fill with "he was a start quarterback" and "who allows less than a baserunner per inning.

Bengoodfella said...

HH, I like how that one announcer thought something completely non-baseball related is important to relay to the audience. The real answer is these announcers don't want to use advanced statistics in any way, so they don't.

The Casey said...

I can almost understand, on radio, that you're having to do more just to paint the picture for the audience of what's going on. But on TV you can just say something like, "Jimmy Southpaw has an 0.76 WHIP, good for first in the NL", and then throw up a graphic showing the top 5 in the NL to give the viewer a frame of reference. Or in your McCann example, drop an overlay on the field showing how many hits he has to left, center, and right.

I think that there are a couple of main reasons for this. First, as on most sports broadcasts, your target audience is really the fan who tunes in for 5/10 games a year. And, honestly, that's the right thing to do if you're looking for viewership. I've quit watching pregame shows because of the lack of meat, but I haven't stopped watching the games themselves. And you don't want a casual fan to tune in for 10 minutes, decide that they don't know what these guys are talking about, and then change the channel.

I think this will sort itself out over time for a few reasons. One is that most of the people stuck on the old stats are older. That goes for both broadcasters and viewers. My dad feels more comfortable that he knows what it means for a guy to be a .300 hitter than if he know that same guy had a 3.4 WAR last year. As more people grow up with things like WAR and WHIP and FIPS, these stats will become more entwined in the game to the point where an announcer won't feel like he needs to explain what it means every time he uses one. But, it will take time.

The Casey said...

HH - the other stat I loved, that I read sometime around the Super Bowl, is that someone timed an NFL telecast, and of the 3+ hours, there were 11 minutes of actual live football being shown.

Bengoodfella said...

The Casey, basically the old generation will die out and the new generation will use stats. I don't think they should throw a bunch of statistics at a viewer, but I also believe if there's a chance to mention an advanced statistic it should be avoided in order to make sure no one gets confused.

Of course, I don't watch pregame shows and I don't pay attention to half the things announcers say anyway, mostly because what they are saying seems pretty useless to me.

Don't tell anyone there is as much action in a baseball game as there is in a football game. That's hearsay.

Slag-King said...

Advanced Stats have their uses when it is understood. I agree with the Casey about having frames of reference. For the most part, the average fan usually understand poor performances, below averages, averages, above averages, and greatness, so if the broadcasters frame any stats with those criteria, then the average fan will see those numbers in reality instead of just out there numbers.

For example,Fangraph has a nice chart that explains the value of WHIP (in general) where 1.00 is excellent and 1.60 is awful. Once that context is established, then it becomes easier to visualize how the pitcher will do against a batter. If the pitcher has high WHIP and shuts out the other team, then the average fan can see that this is an unusual day for that pitcher and enjoy his game a bit more than usual.

This is a chicken and egg thing, in order for the mass fans to embrace advanced stats, they (namely broadcasters) have to use it more often and perhaps be in a teaching role (I think this is the hatred by these announcers--they hate to sound like a professor on broadcast).

I have dual monitors on my computer and I watch MLB game of the night for free on one monitor and work on the other monitor. I do much work because the action is so far in between :)

Slag-King said...

Don't tell anyone there is as much action in a baseball game as there is in a football game. That's hearsay.

Heresy or hearsay? Oddly both works in this situation :)

Bengoodfella said...

Slag, I spelled it "heresay" originally and was told by Blogger that is incorrect. I have always spelled it "heresay."