Several weeks ago, as part of the Prospectus Idol competition, I penned an Unfiltered post regarding the use of statistics in baseball telecasts. In that missive, I noted that while technological advances have made the experience of watching a baseball game on television more visually satisfying than ever-in some ways better than attending the game in person-innovation in the use and display of meaningful statistics during a broadcast has lagged well behind. Not only does this make the broadcast less interesting than it could be to the more enlightened baseball fan, it misses the opportunity to introduce even a few simple sabermetric principles to a wider baseball audience-which in turn creates more devoted fans, higher ratings, and increased revenue.
To expand on that point, I’m going to review a fairly typical MLB telecast-Fox’s quasi-national broadcast of the Angels-Twins game on Saturday, July 25th-focusing on the display and discussion of statistics throughout the game. My purpose here isn’t to make fun of baseball announcers (in this case, competent front man Kenny Albert, with Eric Karros on lead guitar); that has been done, and you can’t improve on perfection. Nor is my goal to comment on the overall quality of a given broadcast – I’m not Tom Shales (though apparently neither is Tom Shales). The idea is rather to note whether there are any small changes that could be made to a broadcast that would provide better, more useful information in an unobtrusive manner, hopefully sparking awareness of and interest in a few basic sabermetric principles.
The streaking Angels, winners of seven straight, sent Matt Palmer out to face the struggling Minnesota Twins and starter Nick Blackburn. The Twins jumped on top in the third inning behind a two-run home run by Jason Kubel, while Blackburn was perfect through three. But the wheels came off in the bottom of the fourth, when the Angels put up a nine-spot despite a questionable 3-2 double play.* From there Los Angeles coasted to an 11-5 victory, with Palmer getting the win to run his record to 8-1.
Setting the Stage
A standard assortment of palate-cleansing graphics are almost always shown at the outset of a baseball telecast, and Fox’s version of lemon sorbet is typical. Batting orders are listed with player names and batting handedness, along with a single team note (e.g., “16 HRs in last 10 games” for Minnesota). Defensive alignment graphics show headshots of each player perched at their position on the diamond; the only additional information was a note that Maicer Izturis has only made three errors in 48 starts.
So… viewers have been introduced to the show’s cast, but given nothing else. Since each of these displays fills the screen for 10 or more seconds between pitches, adding a little extra information would be easy. Other telecasts list batting average when showing the batting order; adding on-base and slugging percentages, even unmentioned, might upgrade this display from “perfunctory” to “useful.” Imagine what would happen if viewers began to notice and announcers to observe leadoff men with low OBPs. As for defense, I would personally love to see some metric (Fielding Runs, UZR, Plus/Minus, PMR, pick your poison) listed for each player, but that’s probably a bridge too far.
During a batter’s first plate appearance, Fox flashes us nothing more compelling than the standard triple-crown stats. Other outlets, including WGN, have taken the bold step of also showing OBP, arguably the gateway drug to a raised baseball consciousness. Is there any rational reason we can’t get all three slash stats, or even OPS? During the game, Albert and Karros had no difficulty discussing rate stats without sounding like kids playing dress-up. This seems like an easy bone to throw-if your announcers don’t like it, they can just ignore it. Once everyone’s gotten used to this, perhaps you could dazzle us with platoon splits (but only when there’s a meaningful sample size).
When pitchers enter the game (except for the first pitching change; see below), Fox gives us a little more information: Games/Starts, Win-Loss record, ERA, IP, BB, K, HR, and Opponents’ batting average. There’s lots here that can be useful, but why ask us to assemble it ourselves? Anyone looking at the walk or strikeout counts is not doing so in a vacuum; they’re only really meaningful if you compare them to IP (or, better still, PA, but you gotta crawl before you can walk). Walk and strikeout totals are merely data; BB/9 and K/9 are information; BB% and K% are enlightenment.
And Now A Stat From Our Sponsors
On several occasions Fox broke out a product-sponsored stat display, and in each case the statistic was hardly more interesting than the product.
1.The first relief pitcher to enter the game triggered the “Zantac Relief Pitcher Profile,” which listed R.A. Dickey‘s Games, ERA, and Strikeout total; the least information provided for any new pitcher all night. Apparently Zantac is designed to relieve the discomfort of swallowing too much information too fast.
2.Taco Bell ponied up for the “Bring The Heat Pitch Speed Comparison” chart, which listed the fastest pitch thrown by each starter in each of the first four innings: Blackburn (92, 93, 93, 93); Palmer (90, 92, 91, 92). It’s like PITCHf/x, for eight pitches, and after every shred of kewlness has been removed. Nice.
3.Just For Men sponsored the “AL Hold Leaders” listing, topped by Bobby Seay, presented with absolutely no description of how Holds are calculated; off the top of my head I couldn’t tell you either. So much for the argument that you can’t display a stat unless Vin Scully can explain it in 10 seconds. And if you ask me, the sponsor for the Hold statistic should be some sort of fictional pomade, not a men’s hair dye.
Perhaps sponsored stats are deliberately uninteresting so as not to distract viewers from thinking about the product-if so, I say ‘well played.’
In the Unfiltered post, I railed somewhat against the proliferation of fairly meaningless, small sample-size splits. I have to say there wasn’t much of that in this broadcast, and almost all of it was defensible. Most splits were of the “recent performance” variety, e.g., Erick Aybar‘s scorching hot July numbers and Bobby Abreu‘s league-leading RBI total since June 1. We were shown that Jason Kubel’s .347 batting average against righties was the third-highest rate in baseball; while listing his gaudy 1000+ OPS might have been better, at least there was a valid point being made. Even Abreu’s RISP numbers seemed reasonable to show in context. But there were still two instances where the data presented could have easily been misinterpreted. Mike Napoli‘s platoon splits were displayed using standard triple crown stats (.400/5/13 vs. lefties, .242/9/28 vs. righties), which might have left the impression that he’s hit for more power against northpaws (hint: he really hasn’t). And Matt Palmer’s starter/reliever splits were especially deceiving.
Palmer had been moved back into the Angel rotation for this start after a series of relief appearances. Color analyst Karros did a good job early on explaining how Palmer’s 7-1 record owed a lot to his run support, and then mentioned that he had pitched better as a reliever than as a starter. To back up that assertion, Fox ran the left portion of this chart, but not the right:
Shown By Fox Not Shown By Fox Role ERA Opp.Avg. HR/9 OBP SLG OPS sOPS+* Starter 5.10 .263 0.96 | .342 .397 739 96 Reliever 2.92 .220 0.73 | .353 .341 694 93 *sOPS+ compares OPS between pitchers in that particular split, with 100 as the norm.
The stats on the left hint at a pitcher that’s been far better in relief. But if you were shown the numbers on the right, you’d see the difference isn’t that great. Sometimes the wrong stats can be worse than no stats at all.
The Big Decision
After pitching three perfect innings while staked to a 2-0 lead, Twins starter Nick Blackburn ran into trouble in the bottom of the fourth. Blackburn allowed a leadoff home run to Chone Figgins, followed by a string of hits that put the Angels ahead 4-2 with two outs and two on. Twins skipper Ron Gardenhire decided to pull Blackburn at that point, and replaced him with R.A. Dickey, who promptly threw enough non-knuckling knucklers to help the Angels plate five more runs before recording the final out.
The decision to go with Dickey in this situation was never debated by Albert or Karros, and since Fox ran the Zantac numbers (Games, ERA, and Ks) when he entered the game I couldn’t tell at first how questionable the move was. Here were Gardy’s main options at that point:
Reliever Throws OPS Matt Guerrier R 552 Bobby Keppel* R 664 Jose Mijares* L 681 Brian Duensing* R 753 R.A. Dickey R 772 *Pitched later in game
This was a pretty high-leverage situation for the fourth inning, and Gardy went with possibly the worst option. Dickey’s the veteran long man in the pen and is thus the knee-jerk selection to enter a game in the fourth, but he was pulled after the fourth anyway, while Keppel, Mijares, and Duensing all pitched later when the game was out of reach. Gardenhire’s decision had a major impact on the outcome of this game-yet nothing on the screen pointed to that fact in any way. Win Expectancy might be too esoteric for a national baseball broadcast, so maybe what we need is a “Bullpen Box” on the screen that lists the available relievers and their numbers, displayed each time a new pitcher enters the game. To help make the medicine go down, perhaps the fine folks who brought us FoxTrax could work up a graphic where the incoming pitcher bursts out of the bullpen “corral,” lowing aggressively. At least then the announcers may have noticed that the guy with the worst numbers was being tapped for such a key situation, and might have mentioned it.
After digging through this single telecast, I continue to feel that broadcasters are missing a golden opportunity to use slightly more advanced statistics to both enlighten and entertain baseball viewers, increasing their interest in and connection to the game. Broadcasters’ use of stats isn’t exactly bad, just disappointing when you think about how much more could be done. The small changes I’ve mentioned (e.g., adding rate stats for batters and pitchers) may seem almost quaint, but they would be easy to implement, and easy to either explain or ignore at an announcer’s discretion. Even a few baby steps in that direction would feel like a breath of fresh air.
*: If you’re not interested in reading an umpiring rant, just ignore this. But… once again, I had to witness a clever baserunning play go unrewarded. Kendry Morales was streaking towards the plate trying to beat Justin Morneau‘s throw. The ball beat Morales home, but the savvy Cuban executed a terrific slide, pushing his foot towards the plate to draw Mike Redmond‘s tag, then pulling it back. Redmond lifted his glove to make the tag further up the leg, but Morales deftly kicked his foot forward again and appeared to touch the plate well before Redmond could react. He was called out anyway. Perhaps umpire Randy Marsh saw it all clearly and felt the tag was in time, but more likely, it seemed to be another case of the automatic out call which occurs virtually every time a ball beats a runner to the bag. How many games of Hot Box did we play as kids, and how many trick slides did we work to perfection? All, apparently, for naught, since had our major league dreams come true we’d be called out anyway. I’ll say it again-either change the rules to make every play a force play, or make sure players are actually tagged.