PDA

View Full Version : A Statistic I Made Up: Total Bases Advanced (TBA)



EvanJ
05-02-06, 06:34 PM
I made up a statistic called Total Bases Advanced (TBA). TBA equals the amount of total bases you got from your plate appearance plus the number of bases each of their teammates advanced because of you. The most TBA you could get in 1 plate appearance would be 10 for a grand slam because in that case you advanced 4 bases, the runner on first advanced 3 bases, the runner on second advanced 2 bases, and the runner on third advanced 1 base. I would allow a hitter to get TBAs on walks, hit by pitches, and productive outs none of which I hitter can get total bases from. If you hit into an error I would make your TBA for that plate appearance to be equal to what it would have been if the fielder made the play, which would sometimes be 0 such as if the play should have ended the inning. Another decision I would have to make is if you can get TBA for grounding into a double play which you cannot get RBIs from. If you put the ball in player, the lead runner gets out, and the end result is runners on the same bases with one more out, your TBA for that plate appearance would be 0. If you hit the ball in play and a runner who is not forced to run attempts to go to the next base and gets out, resulting in at least one more out and baserunners in worse positions than before you batted, your TBA would be 0 but I wouldn't charge you with negative TBA. TBA could be done as a total or as a percentage of TBA/Possible TBA (or PTBA) where PTBA for each plate appearance equals 4 + 1 if there is a runner on third + 2 if there is a runner on second + 3 if there is a runner on first (ranging from 4 to 10). An example:
An RBI single with a runner on second: 3 TBA (1 for you and 2 for the runner), 6 PTBA (4 for you and 2 for the runner, so TBA/PTBA = 0.500 in this case.
If you bat with the bases empty and have an at-bat, your TBA/PTBA for that at-bat will be 1/4th of your slugging percentage for that at-bat. If you bat with the bases empty and reach first base on a plate appearance that is not an at-bat, your TBA/PTBA for that plate appearance would be 0.250.
TBA/PTBA could also be known as TBA%
For TBA a hitter who bats with more runners on base will have an advantage over a hitter who bats with fewer runners on base (the same applies to RBIs). For TBA% baserunners when you come up to bat would affect it also, but your PTBA (Edit: I meant TBA% here) would be 1.000 for any home run whereas your TBA for a grand slam would be 2.5 (10/4) times higher for a grand slam than for a solo home run.

JeffWeaverFan
05-03-06, 02:10 AM
Cool stat. I like the TBA/PTBA better than just TBA as it will be more fair to hitters that don't come up with men on all the time. One thing that you didn't discuss is how faster runners on base will help your TBA. If Giambi is on base then he is pretty stationary so it would affect the guys behind him while if a guy like Damon is on base he'll help out the other guys TBA.

YankeeClemens22
05-03-06, 09:38 AM
Neat, I like it. Now, if only there was a way to calculate it easily during the seaosn....

Steph19
05-03-06, 12:18 PM
Interesting. When I saw the thread title, I thought it would be something different though. I thought you'd measure how many bases a player advances while already on base. This seperates it from slugging and measures baserunning ability and speed.

For instance, if Derek Jeter singles he'd get 0 TBA but if he stole a base, he'd have 1 and then if he scored on a single he'd have 3 as opposed to someone like Giambi (slower speed) who would probably only get one 1 TBA in that sequence (going to second on the single)

I don't know if that even works or tells anything but I'm no statistician.

JJazz
05-03-06, 02:17 PM
Interesting. When I saw the thread title, I thought it would be something different though. I thought you'd measure how many bases a player advances while already on base. This seperates it from slugging and measures baserunning ability and speed.

For instance, if Derek Jeter singles he'd get 0 TBA but if he stole a base, he'd have 1 and then if he scored on a single he'd have 3 as opposed to someone like Giambi (slower speed) who would probably only get one 1 TBA in that sequence (going to second on the single)

I don't know if that even works or tells anything but I'm no statistician.

To measure speed, I think you're better off just looking at steals and stealing percentage (and maybe bunt hits) because your way of looking at it is incredibly dependant on who hits behind you and who is clogging the bases in front of you.

gdn
05-03-06, 02:21 PM
Ooh, very interesting! I'll have to put some more thought into it, but at first blush, it seems good.

homer2931
05-03-06, 10:12 PM
I dont think TBA is very meaningful (team dependent) , but I really like TBA/PTBA

Epy7280
05-03-06, 11:09 PM
better than the Win Method™ :P

JJazz
05-04-06, 05:04 PM
Is there a play-by-play baseball database available? I could figure out some TBA% for various players.

JJazz
05-04-06, 06:19 PM
BTW, retrosheet has an article that examines this idea. Gary Hardegree calls it base-advance-average: http://www.retrosheet.org/Research/HardegreeG/Base-Advance-Average.pdf

It seems like retrosheet has play-by-play data for certain years, but I can't seem to find it on the site. I just see links to gamelogs and then aggregate information about players, ballparks, etc. If anybody could point me to a database of play-by-play data, I'd appreciate it [Edit: nevermind. I found it]. Thanks.

Mean Linguine
05-29-06, 11:57 PM
Before he died in 1984, former Yankees hitting coach Charley Lau wrote a really good book called "The Winning Hitter" in which he touted something similar to your runners-moved-up stat. I think you take it even a step further. I like what you've done.
Lau's feeling was that home runs were all well and good, but trying for something that is only successful once every 20 at bats for the top power hitters isn't really worth the risk. His feeling was that if a right-handed batter came up with a runner(s) in scoring position, it made no sense to try to pull a home run. Even if you got a single to left, you probably wouldn't score the runner from second. He thought it was better strategy to take the ball anywhere from around the second base bag to the right field foul line (he called it the AMV, Area of Maximum Vulnerability), where even if you made an out, the runners would move up.
Just curious, have you put this stat into practice yet with Major Leaguers? Was curious who stood up as good examples.

The Sox Kid
05-30-06, 06:40 PM
I hope I'm not intuding guys, but I have two statistics I would like to share as well. First off is SA or RA, Score Assist or Run Assist, however you want to call it. Well, say Ortiz is at the plate and Cora is on 2nd, there is one out. Now Ortiz hits a ball deep to center field and, say, Bernie Williams makes a great catch for an out and Cora tags up from 2nd to 3rd. Many Ramirez steps in now and singles on a very shallow ball to right, and Cora scores, but he wouldn't of scored if Ortiz hadn't flown out. Ortiz, due to this, would get a Run Assist. Kind of like an assist in fielding.

The second is a TS. Total scores. When you see runs and RBI's you see the runs scored and runs batted in. Well, how do you determine from that how many total runs/scores they have gotten in or have created? A TS would help show that, the total scores, sparing us statistic-lovers much work. Well, that's all for me :)

EvanJ
06-01-06, 03:33 PM
As for run assists, I think you could have given a better example than making an out the advances a runner from second to third. If a team hits three singles in an inning like the Yankees did in the first yesterday, frequently the first guy who got a hit scores and gets a run, the third guy who gets a hit gets an RBI, and the middle guy who hit a single adavncing a runner from first to second or first to third got a hit for the same number of bases as the other two guys but does not get an RBI or a run. I think there could be run assist stat for a guy who gets a hit that contributed to a run without getting an RBI or run.
There's already a stat called runs created which equals runs + RBIs - home runs Home runs are subtracted because you get a run and an RBI for a home run (plus RBIs for baserunners if there are any) but the run you score counts as only one on the scoreboard.

nyymt62
06-03-06, 01:17 AM
Very creative. I think it would be an informative statistic and would be a great stat to accompany the total bases of a player. It would be interesting to see how well the stat would hold up when some MLB players had their TBA's determined. The only problem I could see with it is that it may be slightly weighted toward the home run. But this is certainly an interesting stat to explore.

patrick.o
06-03-06, 03:10 PM
There's already a stat called runs created which equals runs + RBIs - home runs Home runs are subtracted because you get a run and an RBI for a home run (plus RBIs for baserunners if there are any) but the run you score counts as only one on the scoreboard.
That's a great measurement and I use it a lot, but it isn't runs created. The basic runs created is RC = ((H+BB)*TB)/(AB+BB).

I like the RA stat. It's a good measurement of a batter doing something good for the team that doesn't show up in the boxscore, sort of like an inherited runners standed stat for relievers.

I like the idea behind the TBA% numbers but the problem I see is that the values may be too close to clearly differentiate between a good TBA% and an average TBA%. First, it would have to be done as TBA%. If it's just TBA then a batter hitting a single with a man on first who goes to third (a TBA value of 3) would be equal to a batter who hits an RBI single with a man on 2nd (also a 3). Those two things are clearly not equal, so therefore the PTBA is neccessary. As TBA% the first scenario is a .438 compared to a .500 for the second, giving the second scenario the added weight it deserves. But is that enough of a difference to judge future performance on? I'd like to see some long term numbers crunched to find out how things play out, if players have consistent TBA% over their careers, if a 5 year sample is a good indicator of future performance, if the numbers differentiate enough to clearly tell what's great/good/average/poor/terrible, etc.

I'd also be curious to see how TBA% interacted with OBP, which I think has taken on a little too much importance over the last decade. No one ever drove in a man on third with a walk, you know? In that instance a single would be .400 while a BB would be .200.

Actually, looking at that last scenario TBA% would probably have to be combined with some other stat to be truly useful, since coming to the plate with a man on 3rd and hitting an RBI single is worth less (.400) than coming up with a man on first and hitting a single that puts him on 3rd (.438). And hitting a double that doesn't score the man from 1st is worth .500, a full .100 more than driving in the run. If the runner has to hold up to make sure the ball's not caught, or if the runner is slow, the batter gets penalized. Maybe advancing to home (scoring) could be given more weight than the other bases? Or maybe each base could be given accending values? Or some sort of combo of your TBA% and Kid's RA?

Cool idea, though, lots to chew on. Send it to SABR and ask one of their resident geeks to run it through the computer :P

The Sox Kid
06-11-06, 06:22 PM
I was thinking of how certain pitchers get it easy while others get it hard... Some pitchers face the best of the game and look horrible because of it, yet some pitchers who wouldn't even compare against them may look like heros if they only play the horrible teams. I'm pretty much sure there MAY be a statistic related to this, yet I'll just try my hand anyway. If you look at a pitchers OAVG and see it is shockingly low, you would think that that pitcher seems pretty great, but what if the batters he faced hit worse than how much he gave up? For example if the batters he faced overall batted .220 and his OAVG was .240, he would be doing worse, he wouldn't in fact be as good as you thought him to be. So, to add on with OAVG, ERA, etc. etc., you can add the OAP, OERAP, etc. What these stand for will be Opponments Average Prior, or the pitcher's opponent batter's overall average prior (not including) theirs against the pitcher. The OERAP would be the same thing except with ERA, The Opponents Earned Run Average Prior. From this you could judge how well they had done considering who they faced. I meen, even if a good pitcher were to face the Yankees every single game of a season he wouldn't be doing to well now would he :D ? Then... If he plays the 2006 Royals every game he'd be, well, all star worthy, judging by his stats. So if you were to double stats, which would be a problem, but having the doubles always be priors, another example is K/9 compared to OKP/9, Opponents Strikeouts Prior/9, you could judge stats. Pitchers nowadways have it much harder than pitchers in the dead ball era. Well, thats all I have to give for now. By the way, thanks for starting this thread Evan, it really gets you thinking :)

GIDP
06-20-06, 11:46 AM
i dont think id understand what you said even if it wasnt all in weird blue text

Workhorse
06-21-06, 11:29 AM
i dont think id understand what you said even if it wasnt all in weird blue text

My thinking exactly.

Thinking about this stat makes my head hurt but for it to be really useful, you'd probably have to build in some stats to determine how useful "advancing a runner by making an out" is versus through a hit or a walk -- As it correlates to runs scoring. For example, moving a runner from first to second through an out on average adds X runs, or from second to third adds Y runs.

Jetfanmack
07-16-06, 01:17 AM
I was thinking of how certain pitchers get it easy while others get it hard... Some pitchers face the best of the game and look horrible because of it, yet some pitchers who wouldn't even compare against them may look like heros if they only play the horrible teams. I'm pretty much sure there MAY be a statistic related to this, yet I'll just try my hand anyway. If you look at a pitchers OAVG and see it is shockingly low, you would think that that pitcher seems pretty great, but what if the batters he faced hit worse than how much he gave up? For example if the batters he faced overall batted .220 and his OAVG was .240, he would be doing worse, he wouldn't in fact be as good as you thought him to be. So, to add on with OAVG, ERA, etc. etc., you can add the OAP, OERAP, etc. What these stand for will be Opponments Average Prior, or the pitcher's opponent batter's overall average prior (not including) theirs against the pitcher. The OERAP would be the same thing except with ERA, The Opponents Earned Run Average Prior. From this you could judge how well they had done considering who they faced. I meen, even if a good pitcher were to face the Yankees every single game of a season he wouldn't be doing to well now would he :D ? Then... If he plays the 2006 Royals every game he'd be, well, all star worthy, judging by his stats. So if you were to double stats, which would be a problem, but having the doubles always be priors, another example is K/9 compared to OKP/9, Opponents Strikeouts Prior/9, you could judge stats. Pitchers nowadways have it much harder than pitchers in the dead ball era. Well, thats all I have to give for now. By the way, thanks for starting this thread Evan, it really gets you thinking :)
I got lost quickly here, but are you saying something like measuring how a pitcher (or hitter) does against what is expected due to averages and stuff? Like if you face the Royals every day, and say their average OPS or somethin is .750. Then say that the pitcher holds them to a .700 OPS, and he would get a rating based on that? Or if KC scores 4.3 runs per game, and the pitcher holds them to 3.7, he gets that? Kind of like a more specific ERA+?

If that's what you're saying, it would probably suffer from small sample size for each team. I guess if it all got averaged together, it wouldn't be a bad thing to add to adjusted ERA+, but it probably wouldn't make much of a difference.

Random
07-22-06, 09:53 AM
As far as TBA, TBA%, etc. I seem to remember an article in Sport Magazine in the late 80s that had this, or something very close.

I remember one distinction -- if you stole a base, you got credit for another base (another TBA), whereas if you were caught stealing you lost one. This assumes you were caught stealing second. If you were caught stealing third, you should be docked two, b/c you cost your team the two bases you had already achieved.

As I recall, the stat was also called TBA, but I think it stood for Total Baseball Average. It was used to measure hitters, but you could also measure pitchers by analyzing the TBA they gave up.

Anyone else remember this one?

The Sox Kid
08-13-06, 09:44 PM
No, no. Sorry for the confusion guys. What I am saying is you have a pitcher's OAVG and ERA. Say he has an ERA of 2.78 and an OAVG of .236, you would consider him a pretty good pitcher wouldn't you? But lets say the batters he faced usually score 2.70 earned runs a game and have an average of .228. In fact he would be worse than pitchers with a 3.64 ERA and an OAVG of .259 who had faces batters who would usually score 3.41 ER a game and hit .271. So in other words a statistic like OAVG can't really determine the skill of the pitcher, same with ERA. What they should have to add on to them is OAVGP and OERAP. So adding on to Opponent's Average you can have the Opponents Average Prior (Or the opponent's average if they didn't face the pitcher) and adding on to ERA you can have the OERAP (Or the opponents ER average if they didn't face the pitcher). So what this stat would determine is to show how well pitchers are relative to the batters they face... I hope this makes it a bit easier to understand. If I wasn't off the forum for such a long time I would of gotten to it quicker. Again, sorry about that guys. :(

P.S. How's the Boston Colors treating you instead of that bright blue??? :)

homer2931
08-13-06, 10:59 PM
No, no. Sorry for the confusion guys. What I am saying is you have a pitcher's OAVG and ERA. Say he has an ERA of 2.78 and an OAVG of .236, you would consider him a pretty good pitcher wouldn't you? But lets say the batters he faced usually score 2.70 earned runs a game and have an average of .228. In fact he would be worse than pitchers with a 3.64 ERA and an OAVG of .259 who had faces batters who would usually score 3.41 ER a game and hit .271. So in other words a statistic like OAVG can't really determine the skill of the pitcher, same with ERA. What they should have to add on to them is OAVGP and OERAP. So adding on to Opponent's Average you can have the Opponents Average Prior (Or the opponent's average if they didn't face the pitcher) and adding on to ERA you can have the OERAP (Or the opponents ER average if they didn't face the pitcher). So what this stat would determine is to show how well pitchers are relative to the batters they face... I hope this makes it a bit easier to understand. If I wasn't off the forum for such a long time I would of gotten to it quicker. Again, sorry about that guys. :(

P.S. How's the Boston Colors treating you instead of that bright blue??? :)

I doubt there is a big difference between pitchers in this regard, but I'm pretty sure VORP accounts for it

TwentySeven
08-14-06, 10:41 PM
I really really like the TBA%. It would say a lot about guys like Abreu who make a ton of productive outs.
I do have a question though. While you are batting, if a runner steals a base because of your AB, do you get a TBA?

EvanJ
08-15-06, 10:49 AM
Considering one or more runners could steal bases on a pitch that you strike out on, I wouldn't give you any TBA for stolen bases from other runners (or for yourself, I'm making this a batting statistic only).