Extra MustardSI On CampusFantasyPhoto GalleriesSwimsuitVideoFanNationSI KidsTNT
More MLB Coverage Chatter Up Blog Homepage
Talk baseball all season long with SI.com's Jacob Luft in Baseball Chatter, a journal for hot topic debates, Sabermetric ramblings and reader-driven discussions.
6/13/2006 05:09:00 PM

Get some PrOPS

Frank Thomas has been the unluckiest hitter in baseball this season.
Frank Thomas has been the unluckiest hitter in baseball this season.
A couple of years ago, economics professor and die-hard Braves fan J.C. Bradbury got sick of watching Chipper Jones rip line drive after line drive right into the gloves of the opposition. One of his beloved Braves was having a great season at the plate and it wasn't showing up in the stats because of something as simple as bad luck.

Bradbury decided to do something about it. In an effort to filter out the effects of luck -- both good and bad -- on batting statistics, he came up with a system called PrOPS (Predicted On-Base Plus Slugging Percentage.)

By now most of us are comfortable enough with regular old OPS. It combines on-base percentage and slugging percentage into a neat little number that Joe Morgan doesn't care for very much. Bradbury went one step further and took the play-by-play data provided by a company called Baseball Info Solutions to predict what a player's OPS should be based on the type of balls that he hits. A player who hits a lot of line drives, for example, would be rewarded, while a player who hits a lot of infield popups would be penalized.

"Sometimes balls that should be doubles are caught, and sometimes a popup falls for a hit," says Bradbury, who teaches at the University of the South and runs Sabernomics.com. "PrOPS takes all the things that players do that normally translate into success and judges them only on those factors."

Because breaks tend to even out as the sample size of data grows, PrOPS is a powerful tool in figuring out which players will benefit and which ones will suffer as their statistics regress to the mean. For example, Reds outfielder Austin Kearns posted a real OPS of .785 last season, but his PrOPS was .840, indicating that he was a better hitter than his statistics were giving him credit for. This season, Kearns' real OPS is almost identical to his PrOPS from last season: .852. If you look back at the top 25 underperformers for 2005 as calculated by PrOPS, you'll find others who have improved this season, including last year's leader, Jason Giambi, and Mike Lowell.

So who have been the luckiest and unluckiest hitters this season? Let's take a look (you can find these yourself at The Hardball Times):

2006 Top 10 Luckiest Hitters According to PrOPS
Player, OPS, PrOPS (Difference)
1. Gary Matthews Jr., .934, .762 (+.172)
2. Joe Mauer, .986, .856 (+.129)
3. Ichiro Suzuki, .876, .762 (+.115)
4. Alex Rios, 1.019, .906, (+113)
5. David Wright, .971, .863 (+.108)
6. Matt Holliday, .993, .886 (+.107)
7. Miguel Cabrera, .997, .895 (+.102)
8. Orlando Cabrera, .818, .724 (+.094)
9. Grady Sizemore, .921, .835 (+.085)
10. Scott Rolen, .970, .890 (+.080)

2006 Top 10 Unluckiest Hitters According to PrOPS
Player, OPS, PrOPS (Difference)
1. Frank Thomas, .877, 1.043 (-.165)
2. Phil Nevin, .729, .893 (-.164)
3. Adam Dunn, .970, 1.123 (-.154)
4. Jason Lane, .745, .882 (-.138)
5. Carlos Lee, .929, 1.052 (-.123)
6. Jason Kendall, .680, .800 (-.120)
7. Adam Everett, .570, .682 (-.113)
8. Jason Giambi, 1.087, 1.198 (-.112)
9. Carl Everett, .708, .819 (-.111)
10. Justin Morneau, .815, .921 (-.106)

Albert Pujols didn't make either list, but his PrOPS deserve to be singled out for admiration on their own:
OPS: 1.193
PrOPS: 1.279
Difference: -0.085

That's right. As scary good as Pujols was hitting before getting hurt on June 3, his numbers weren't doing him justice.

After coming up with the method, Bradbury crunched the numbers for every season since 2002. It turns out that his hunch about Chipper Jones' bad luck proved correct -- Jones' 2004 season ranks as the eighth unluckiest in MLB since 2002.
posted by JL | View comments |  


Posted: 8:06 PM, June 13, 2006   by Anonymous
I'd say the whole Braves team is having a bit of bad luck right now....
Posted: 6:59 AM, June 14, 2006   by Anonymous
I'm not even a recreational sabermatrician, but this is about as fascinating a statistic as I've seen.

Glancing around the internet, I can't find any career PrOPS numbers for Jason Kendall. Surely, he must be one of the unluckiest hitters in baseball. (Or one of the luckiest, depending on how you view OPS and PrOPS in relation to his contract value.)

Well played Joe Morgan reference, by the way.
Posted: 7:27 AM, June 14, 2006   by Anonymous
Phil Nevin isn't unlucky. He's just bad.
Posted: 7:46 AM, June 14, 2006   by Anonymous
While stats like that are fine and dandy, just remember the famous quote by Wee Willie Keeler when he stated his hitting philosophy.


Keeler, by the way, had a record 44 game hitting streak in the 1890's, which was broken by Joe DiMaggio in 1936.
Posted: 8:02 AM, June 14, 2006   by Anonymous
Aren't sports in general about luck? If a wide receiver doesn't jump high enough, your QB is unlucky, but it doesn't count as a catch. An out is an out, a hit is a hit. It's amazing that baseball has been working for a 100 years without stats like this, and it seemed to be fine.
Posted: 8:13 AM, June 14, 2006   by Jaybird
Please... just what we need ... another useless baseball statistic. By this nonsense, an guy like Ichiro is considered 'lucky' because he consistently gets infield hits (because of his hustle and speed).

Seriously, stop the madness. It's like baseball fans can't find enough reasons to celebrate mediocrity, especially at the plate.

Texas leaguers, seeing-eye singles, and infield hits are part of what makes baseball such a great game. It's not luck, it's the nature of the game. So now you want to call it luck ?

It's a flawed statistic.
Posted: 8:43 AM, June 14, 2006   by Anonymous
So the only factors this is adding is whether the player is hitting line drives, ground balls, fly balls etc? This doesn't seem to me enough to quantify an expected OPS. For example Jason Giambi, we know he hits the ball hard but when he hits into the shift I don't think we can call that bad luck
This reeks of subjectiveness.

If in truth it is measured somewhat accurately, this is a GREAT tool. But I have my doubts.
Posted: 9:37 AM, June 14, 2006   by Anonymous
I think it was Jack Nicklaus who said that he would rather be lucky than good. Somehow, though, it seems that the good players often turn out to be the lucky ones.
Posted: 9:42 AM, June 14, 2006   by Merle
I clicked through onto Sabernomics and one thing I did not see is how PrOPS rates compared to other measures of a player's "true" preformance. Couldn't we also measure the quality of the pitchers the batter faced, whether the opposing team plays a shift against them, the quality of the defense faced, number of time the player batted with runeners on base, any change in the protection in the lineup, etc. This innovation seems to be factoring in flyball ratio and line-drive ratio, and it's unclear to me that those are better prognosicators (or indicators of "luck") than any of the other available stats.
Posted: 9:53 AM, June 14, 2006   by Anonymous
Very interesting data. The one issue I would have with it is that it doesn't take into account things like speed and ability to hit to all fields. First, a quicker hitter will always be "lucky" compared to a slow hitter as he'll get more infield singles and stretch hits into doubles and triples.

Second, as we've seen with the shift that huge pull hitters face, they're going to have alot more hard outs as there are more fielders in their hitting zone.
Posted: 10:00 AM, June 14, 2006   by Anonymous
I wonder if they take into account how many base hits in the gap between first and second have been taken away from 'big papi' ortiz because of the shift.
Posted: 10:01 AM, June 14, 2006   by Anonymous
I'm a huge Tony Gwynn fan to this day and I don't think this stat would correctly represent his value. Tony made an art out of guiding bloop singles over the head of the shortstop. They weren't line drives, but they were hits that were achieved with skill. How would PrOPS treat Tony's stats?
Posted: 10:07 AM, June 14, 2006   by Anonymous
Is there a statistic for the number of useless statistics in baseball? How about a statistic that keeps track of great pitches that shouldn't have been hit? Or a statistic that shows how successful a team could have been if they won more games?
Posted: 10:13 AM, June 14, 2006   by Anonymous
What about the red sox players... they have been getting unlucky it seems like... Can you calculate Manny, Ortiz, and Covelli Crisp?
Posted: 10:21 AM, June 14, 2006   by Anonymous
Who cares what someone's PROPS number is? Is it going to give my team another victory? It's another wasted statistic that does nothing but give writers and analysts something to talk about.
Posted: 10:32 AM, June 14, 2006   by Anonymous
It seems fast players are "luckier" then slow players. Also, do we really need another weapon for agents to justify pay based on how productive the player "should" be? I say hit 'em where they aint.
Posted: 10:33 AM, June 14, 2006   by Anonymous
This needs to factor in the speed of the batter. Of course Ichiro is lucky - he can beat out plays that get Frank Thomas.
Re: the speed factor...
I asked Bradbury about this and he said he tried to add a speed component but found over a three-year period, the effect was negligible.
-- Jacob Luft
Also, Re: the factors he uses...
This is what he wrote in his article explaining PrOPS in the 2006 Hardball Times Annual:

1) Percentage of batted balls that were line drives
2) Ground ball to fly ball ratio
3) Walks per PA
4) Strikeouts per PA
5) HR per AB
6) HR park
7) Season adjustment
-- Jacob Luft
Posted: 10:56 AM, June 14, 2006   by Anonymous
Posted: 10:58 AM, June 14, 2006   by Anonymous
seeing as this stat says that almost anyone on the leader board for batting average is just lucky, even with people like Ichiro and Mauer who always have great at bats, the system obviously has some flaws. I like the Idea though.
It's not the guys are "Just" lucky or "just" unlucky. It's trying to measure how much luck is helping or hurting a batter, regardless of how good the batter is to begin with. Whether you are Manny Ramirez or Cristian Guzman, luck will play a factor in what shows up on the back of your baseball card.
-- Jacob Luft
Posted: 11:02 AM, June 14, 2006   by Anonymous
This is just another baseball stat that is, in fact, a load of crap, and uses some arbitrary formula to guage performance. OPS is a load of crap as well as is OBP and slugging. All I want to see is tangible account of what actually happens on the field. All you need is AB, Hits, Avg, HR etc.
Posted: 11:13 AM, June 14, 2006   by Anonymous
The thing about baseball is that it is mostly a luck thing. Too bad for Frank Thomas. Now the first blogger said it well, the Braves are not having any luck.
Posted: 12:08 PM, June 14, 2006   by Anonymous
Any time someone tries to quantify an immaterial quality such as luck makes me wonder if that person really knows what they are doing. Players make the numbers; numbers don't make the player.
Posted: 12:12 PM, June 14, 2006   by Anonymous
You have got to get a life!
Posted: 12:45 PM, June 14, 2006   by Anonymous
Pujols like any great hitter, doesn't need luck. He creates his own. You find the holes in the defense and hit them there.
It could be a useful stat for fantasy, like guessing which guys who have hot starts will regress to their mean, or who might be better than years before, but I agree that there is probably more skill involved in getting hits than just hitting the ball hard somewhere
Posted: 12:59 PM, June 14, 2006   by Anonymous
David Wright a lucky hitter....big surprise
Posted: 1:04 PM, June 14, 2006   by Anonymous
Anyone notice that a majority of the "good luck" guys have speed and accordingly, get more hits? On the contrary, the "bad luck" guys are, for the most part, big and slow.
Posted: 1:05 PM, June 14, 2006   by Anonymous
i came up with a better stat then that by myself, it takes the 8 offensive categorys of a player(avg+hits+walks+hr+rbi+slugging%+on base%+SB) and divided them by 8.i used that statistic for every hall of famer and of course babe ruth was tops with an average of 270.375 but it really shows the difference between the 2 eras of baseball.all the dead ball players averaged out come to 203.46 for 32 players compared to 198.382 for 99 live ball era players.with an average of 199.623 for all 131 players. i feel its a good little statistic to judge purely the players offensive im pact on the game.and heres my joe morgan tidbit his modified shares(what i call this stat) is truly average for a hall of famer coming in at 198.25 compared to the average of 198.382. all stats were based on 162 game schedule averages
Posted: 1:10 PM, June 14, 2006   by Anonymous
I'm sure David Wright is the 5th luckiest hitter in baseball this year. The guy hits for power and average, goes to all fields consistently, hits mostly line drives, almost always seems to get a good piece of wood on the ball, is a great 2 strike hitter and has an eye like DiMaggio. If that somehow constitutes luck, then sure this system works!
seeing as this stat says that almost anyone on the leader board for batting average is just lucky, even with people like Ichiro and Mauer who always have great at bats, the system obviously has some flaws. I like the Idea though.

Usually the people who win batting titles are lucky. Ichiro (as big of fan of him as I am) is not a .360+ hitter. His true average is probably pretty close to his career average of .334. However, due to variance he's going to have years where he hits .360 and years where he hits .300 due to just luck.

The one thing that this stat doesn't take into account are park factors and speed factors. Frank Thomas might smoke a ball of the wall and only hit a single, where as ichiro hit's a double. You could probably adjust the batted ball types for the park (safeco supresses power numbers for right handers, etc) and use some sort of boost for the speed factor (Speed Factor = SB * (3B/(2B+3B)) and get it even closer.
Posted: 1:52 PM, June 14, 2006   by Anonymous
The fundamental premise (that some types of batted balls are ALWAYS better than others)is flawed. Depends on the batter, and the situation. Inning, score, runners on base, etc. all should influence what the hitter is trying to do. Good hitters will have the bat control to do this more often than bad hitters...even if the sitation calls for them to make an out, and thus drop their OPS. So what looks "unlucky" to PrOPS is actually BETTER HITTING.
The gap between OPS and PrOPS has nothing to do with luck (which as you say, will even out over time). It's purely a flaw in the calculation of PrOPS.
merle: "Couldn't we also measure the quality of the pitchers the batter faced, whether the opposing team plays a shift against them, the quality of the defense faced, number of time the player batted with runeners on base, any change in the protection in the lineup, etc."

We already have a stat that measures all of this - it's called who gets whom out. Baseball itself is the measurement of who has better players and pitchers; the numbers only describe parts of the process.

People who are good at baseball are going to produce over the long haul, and the numbers will show that. "Unlucky" for Chipper Jones means, what - .275 instead of .290? Big deal. Maybe he hits .305 next year. The real unlucky ones are the ones who are supremely talented but have their careers ruined by injury, and the lucky ones are guys like Joe Carbonneau who get that run where everything looks easy, and then fade to permanent obscurity when the league catches up and they can't readjust.
Posted: 3:10 PM, June 14, 2006   by Anonymous
It seems to me that many of the "lucky" players are fast, and many of the "unlucky" ones are slow... might Ichiro's supreme skill in reaching base on infield grounders be ignored here, while Frank Thomas's aged legs are getting conveniently overlooked?
Posted: 4:01 PM, June 14, 2006   by Jaremy
That "firejoemorgan" website was insane. Is that guy for real? WOW.

Anyway, I wonder how lucky a guy like Ichiro is every year... And how much his year-end average reflects it.
Posted: 8:47 PM, June 14, 2006   by Anonymous
Clearly this stat doesn't factor in speed. One only needs to glance at the names there to see that the luckiest top 10 are fast while the unlukiest top 10 are slow. Still, I think the main value in the stat would be to value those in the middle. Players who are not significantly faster or slower than the average.
Posted: 8:52 PM, June 14, 2006   by Anonymous
"All I want to see is tangible account of what actually happens on the field. All you need is AB, Hits, Avg, HR etc. "

What an idiot, are you serious? Your "tangible" stats don't tell you didly. Just because you may be too lazy or math deficient to understand some stats doesn't mean they aren't accurate and predictive. Anyone who thinks batting average tells you a player's value is a fool.
``Sometimes balls that should be doubles are caught...''
`Should be' in whose opinion?
And, the effect of speed is negligible -- really? Somebody better quick sendword to Tony LaRussa (among others).
The great thing about baseball is that Everything Counts.
PrOPS is just a bunch of isolated stats bunched together in a way that suits its inventor's personal preference.
Gary Player, the great South African golfer, once said: ``The more I pratcice, the luckier I get.''
divider line