Upon further thinking, I realized I should probably restate why I created this WSG thing. First, SLG is an inherently flawed stat. Second, OPS counts certain elements twice, while only counting others once. Third, to create a single stat which considers all aspects of getting on base, as well as advancing existing baserunners in a non-team specific situation allows us to rank players using stats which are independent of team success.

First – SLG is based on the assumption that in all situations, a homerun is worth four times as much as a single. But in regards to advancing baserunners, that is only the cases when the bases are empty. When there is a man on 2nd, or men on 1st and 3rd. A homerun is only twice as effective at advancing baserunners. This means that a homerun is not always worth four times that of a single. With my current percentage distribution of baserunners (which again, is wrong, and needs to be fixed), a homerun is worth about 2.68 times that of a single. This seems counter intuitive, but re-read my earlier posts about this, and I think it makes more sense.

Second – OPS is a combination of OBP and SLG, both of which value hits, while only OBP weighs walks. So this means if someone has a high AVG component of SLG, this will be counted again in OBP, which will have a much greater effect than someone who has a medium AVG component of SLG, and a high walk rate which will only be counted in the OBP part of resultant stat. While I can’t think of any examples off the top of my head, I’m pretty confident in the above statement.

Third – This point deals mostly with the removal of RBI and Runs as a factor in considering player worth, and trying to place each batter in a team-neutral and environment neutral setting. While the final stat would most likely need to be balanced for park-effect, the success of the surrounding team on a player’s WSG would not be seen. This is because the percentages taken in evaluating the value of each type of base hit is taken from the neutral league context, and therefore, negates all team effect on an individual player’s success or failure in this stat. The argument that a player on a good team would get up more isn’t an issue here, since this stat is not a cumulative stat, such as 1B, or RBI, but one that is based on the percentage of time the player is successful when up to bat no matter how many times he is up to bat.

I think that is a good summary of why I felt a new stat was needed. Read below for more details on the nuts and bolts behind it. Comments, questions?

- posted by LT @ 8/08/2003 11:02:00 AM

 
A few more comments after reclaiming the computer. . .
I really liked the idea proposed by Baseball News Blog (who runs a good website, but is on vacation for a bit), to combine the AL East and West for the rest of the season in the standings and only the top three teams make it. So here is what he calls, the AL Coastal Standings:
Yankees : 69-43 - .616 - XX
Mariners : 69-45 - .605 - 1.0
Red Sox : 67-46 - .593 - 2.5
Athletics : 66-48 - .579 - 4.0
OK, so I may have screwed up the GB column, but you get the point. In case you were curious how close the AL Central leading KC Royals are, their record is 60-53.

Now to try and use my brand new WSG stat on this season's leaderboard, here are the top 10 in OPS, top 5 in AVG, OBP, SLG with their respective WSG.
OPS
1 B Bonds - 1.229 - .833
2 A Pujols - 1.120 - .810
3 T Helton - 1.084 - .778
4 J Edmonds - 1.042 - .741
5 C Delgado - 1.024 - .712
6 G Sheffield - 1.015 - .722
7 M Ramirez - .993 - .707
8 J Giambi - .992 - .684
9 T Nixon - .992 - .713
10 B Mueller - .978 - .693

AVG
1 A Pujols - .372 - .810
2 T Helton - .356 - .778
3 I Suzuki - .337 - .600
4 B Bonds - .330 - .833
5 E Renteria - .330 - .635

OBP
1 B Bonds - .510 - .833
2 T Helton - .449 - .778
3 B Giles - .441 - .683
4 A Pujols - .440 - .810
5 M Bradley - .429 - .670

SLG
1 B Bonds - .719 - .833
2 A Pujols - .680 - .810
3 J Edmonds - .648 - .741
4 T Helton - .635 - .778
5 C Delgado - .600 - .712

Interesting to see how Giambi is the only sub-700 player in the top ten of OPS in WSG. Obviously I must have screwed up big time, since we all know how amazing Giambi is. Now for a few Yankees, Robin Ventura, and for my Marlins and Phils fans in the crowd, Lowell and Thome.

Nicky J - .654
D Jeter - .621
J Posada - .613
Godzilla - .592
A Soriano - .589
B Williams - .556

A Boone - .567
R Ventura - .554

J Thome - .651
M Lowell - .684


I thought Sori's number was very low, but remember that Sori's OPS is only .823 and that's below Jeter, Posada and Giambi. This stat isn't amount of playing time dependent, so its good (like AVG/SLG/OBP) for evaluting partial season results (good for the injury prone Yanks). Also, should I be worried that in 6 games, Boone has to draw a walk in pinstripes.

Also ESPN.com's stats don't seem to match MLB.com. Maybe they're off by a day or so, but Ramirez's OPS is .993 on MLB.com, and .982 on ESPN.com which makes no sense. So I'm throwing that disclaimer out there post-initial post. Some improvements I think I should make to make this a more universal stat, is that I'd have to balance it versus the league averages in order to eliminate era differences in statistics (not to take away what these guys are doing out there, but just to try and equalize era effects, this might also be done with the % of baserunners in certain situations). But for now, I hope that holds you folks over, and gives you some modern day, and all-time reference points to work with. Enjoy. . .

- posted by LT @ 8/08/2003 12:21:00 AM

 
Well, AC was fun, but now, back to the fun of stats. . .
OK, what I did was figured out the average advancement for all hits as I did before with 1B and 2B and here are the final totals (I changed 1B totals to make them a little more optimistic about baserunner advancement).

BB = 15
1B = 28
2B = 40
3B = 48
HR = 56

So just using these figures, we have the following BB:1B:...:HR relationship based on a BB = 1;

1.00:1.87:2.67:3.20:3.73

These are misleading becuase they weigh all base-runner situations as equally likely to occur. Now I have no stats on the distribution of the times each situation happens, so I'm going to use an awful approximation and say that 40% of the time players come to plate with no one on, and 20% of the time they come to the plate with a man on first, and assume that the other 6 situations occur with an equal probability of 6.67% of the time. If anyone has any better statistics on this, or know where I might be able to find them, I'd sure appreciate that. But, with using those distributions, here are the new results (in the ratio form with BB=1):

1.00:1.63:2.50:3.13:3.75

Also we may need to balance for the number of offensive players on base (including the man up to bat) at the time of the walk/hit of your choice. Here are the results, once again in ratio form, but these are not adjusted for the percentages as the previous set was:

1.00:1.81:2.67:3.29:3.92

And now adjusted for the percentages:
1.00:1.45:2.38:3.13:3.88

So this may give us better numbers to work with for a new WSG stat we are trying to create. Here are the top 10 all-time OPS seasons converted into (this version) of WSG (which still has major faults, the main one I can see is the percentages of what happens when, but there may be others) and the top 6 all-time players' and their career OPS.

Season (OPS:WSG):
Bonds ('02) : 1.381 : 0.912
Ruth ('20) : 1.379 : 0.941
Bonds ('01) : 1.379 : 0.925
Ruth ('21) : 1.359 : 0.941
Ruth ('23) : 1.309 : 0.903
Williams ('41) : 1.287 : 0.896
Ruth ('27) : 1.258 : 0.910
Williams ('57) : 1.257 : 0.952
Ruth ('26) : 1.253 : 0.861
Ruth ('24) : 1.252 : 0.869

Career:
Ruth : 1.164 : 0.816
Williams : 1.116 : 0.791
Gehrig : 1.080 : 0.769
Foxx : 1.038 : 0.747
Helton : 1.032 : 0.742
Bonds : 1.023 : 0.728

A few short observations since the computer is in hot demand here, and it took me like 30 minutes to compute those numbers. . . I think the stat might show some promise, it does need a little tweaking regarding the percentages of which situation happens how often. Also, how sick is it that T Williams is on that list, 16 years apart? And how about Ruth showing up 6 times? Ok, that's it. . . for now. . .

No mo' Mando. But the Curse of Jeff Nelson has been lifted!
- posted by LT @ 8/07/2003 10:22:00 PM

 
OK, I've been reading Bill James, yadda yadda yadda. Here's my problem, the idea of batting average as the key indicator is stupid. Just dumb. Yah, I know its good for . . .blah blah blah. OBP is the same thing, except it tell you whether the person can draw walks which are almost as good as singles. For getting you on base they're equivalent to 1B, 2B, 3B so why don't we consider them in AVG? It doesn't make much sense if you think about it. This point has been made a million times over, but I felt repeating it would finally convince people.
Anyway, my problem really is that in game situations a walk is almost as good as a 1B. If no one's on base, they are the same. If a man's on third, a single will usually score him, while a walk will do jack except make a double play ball more likely. So now my problem is, in creating a slugging-esque stat which includes BB (WSG - walk slugging, I couldn't think of a better abbreviation), how do I weigh walks? They're obviously not equal to 1B, but if we set them equal to less than 1 (the value of a single) we run into a problem.
Example of Problem (Weight 1 BB=2 1B):
Disciplined Hitter (Guy A): 20 PA, 5 BB, 3-15 with 2 1B's, 1 3B.
UnDisciplined Hitter (Guy B): 17 PA, 0 BB, 6-17 with 5 1B's, 1 3B.
A: .200 AVG/.400 OBP/.333 SLG/.375 WSG
B: .353 AVG/.353 OBP/.471 SLG/.471 WSG
Obviously most teams would want a guy who has a OBP of 47 points higher, and in this situation, most teams would rather have Guy A (although this is an awfully designed example, and would make more sense if I thought it out longer). The ratio of 1 BB= 2 1B is also not feasible, but even if we value it as 1 1B = .75 BB, Guy A still would lose. Again, don't look too deep into this example, but think up your own, and you can see that if in this WSG stat you value a BB as less than 1, you could in theory punish someone for walking. The more I look at this the more I begin to think I shouldn't even include this example, but maybe someone out there will be able to show me a flaw in my logic, or offer supporting evidence.

Assuming my supposition is correct, that a walk needs to be valued at 1, how do we change the values of the other base hits to represent their true value. Are we even sure that a double is worth twice a single? Here's the possible man on configurations, and we'll see what the logic tells us. Again this won't be super-accurate, because there are cases where a single with a man on third won't score the man on third, but we'll deal with the majority of cases. Cases:
*Bases Empty. Double is worth twice a single.
*Man on 1st. Single gets man on 1st and moves other to 3rd (again, this might not be the most common case, but we'll assume it is - 3 Bases Gained). Double scores man on first and puts man on 2nd (5 BG).
*Man on 2nd. Both cases score the man. (Single 3 BG, Double 4 BG)
*Man on 3rd. Both cases score the man. (Single 2 BG, Double 3 BG)
*Men on (1,2)/(1,3). Single scores 2nd/3rd, and advances 1->2 (4/3 BG). Double scores both (7/6 BG).
*Men on 2,3. Both men score (Single 4 BG, Double 5 BG).
*Bases loaded. Single has 2,3 score, 1->2 (5 BG). Double scores all (8 BG).
Our totals leave us with 25 BG from singles, and doubles give us only 40 BG (these are with conservative estimates for singles also). While these cases may not, and do not, occur with the same frequency in games, if we assumed they did then we would have that a 1 1B = 1.6 2B, which is probably a low estimate, but still would be a number less than double the value of a single.

OK, that's it for now, chew on this faithful viewers, send in comments/questions/fan mail/hate mail and I'll get back to you as soon as I can. Once again, I am leaving the blog for a few days, this time a trip to Atlantic City for a pair of nights.
- posted by LT @ 8/05/2003 12:10:00 AM

 
SI.com - Power Rankings: "Not everybody is impressed with Billy Beane's latest trade-deadline pickup, well-traveled Jose Guillen, who has a rep as selfish and temperamental. 'Guillen is merely a younger version of Raul Mondesi,' one NL official told The Daily News of New York. Ouch, that left a mark."

That made my evening.
- posted by LT @ 8/04/2003 10:34:00 PM

 
While in Cooperstown I stumbled across an old book store and found two copies of Bill James' old Baseball Abstracts, which are now being talked about in a lot of new media, so I figured, why not, and bought them ('83 and '86 I think). I've really enjoyed them, and they've inspired a few ideas which I was going to write about here in some depth before I realized I hadn't examined the recent trades in any depth (either for you, my faithful reader, or for me, I had been going on my gut, which wasn't exactly accurate).


Watching this new set of Yankees (minus Raul and Robin), something reminds me of the style of ball that they used to play back in the "good 'ole days" of the late 90's. I know this is an 'uber-optimistic' claim, and it comes after watching three games (of which they lost 2 - but I don't want to talk about that right now, my language wouldn't be appropriate in mixed company). So this may be a bit of an overstatement, but the extra speed and youth seem like a throwback of sorts. First I'll disect the 3B situation, for both speed and power, then I'll look at The Raul vs the 3 Guys Who Are Replacing The Raul.

Boone this year alone has 16 SB (3CS, 84% Success Rate), while in Ventura's last 5 seasons he only has 9 SB (12 CS, 43%). Boone is 6 years younger than Ventura and even when Ventura was younger he never stole more than 4 bases in a season. While I'm not saying we need a switch to a hit and run team, it is nice to be able to move people over with a little bit more success which comes from extra speed on the basepaths. For the AVG/OBP/SLG stats I'll go with this year, and then the average of the last three years.
Boone: .270/.335/.462 -- .268/.333/461
Ventura: .246/.337/.384 -- .243/.355/.420 (his power has disappeared this year)
So Ventura has typically been better at getting on base, but his power which kept him in this race has gone away, and didn't appear to be coming back anytime soon. Boone's OPS was above .825 in both '00 and '01, but dipped to .753 in '02 but this year is back to .797 and Ventura had been below .780 in both '00 and '01. So this seems like a good trade had it been straight up, but unfortunately the Reds were not in that desperate a fire-sale. The real value of this trade appears to be in the difference in value of Claussen vs "Crosby and Proctor" who are both performing quite well at AAA, but as Larry from RLYW reminded me, are 25 and 26, while Claussen is 24 and has already had Tommy John. So this trade for now will receive an incomplete grade (with the Yankees getting a passing grade, but unsure of where it will fall in the C- to A- range) and will only be fully grade-able when any of the prospects reach (or don't) the show, OR if they are trade bait next year, OR if either of the two 3B's make a significant impact as the season goes into the homestretch.


Comparing Raul w/ his three replacements (I don't think it's clear yet who will replace him in anything more than a platoon role) I'll standardize all their statistics from the last three years based on a 162 game season, so from now on it will be Raul vs DGS (I like the name Dellucci the best, so that gets to go first). SB/CS will be based on the last three seasons, while the AVG/OBP/SLG will just be based on this years info (I'm not that motivated).
DGS: 4/3 (57%) - Note: None of these three have really tried to steal anything w/ the exception of Dellucci this year and Sierra early in his career.
Raul: 25/10 (71%)
Definate edge Raul.

DGS: .256/.316/.408
Raul: .255/.326/.465
Again, edge Raul.
Why do I still prefer this to Raul? Because he was a jackass. Always was, always will be. Same reason I have serious reservations about Sierra, but Torre said he had everything worked out with him from the first time around. I have to wonder though if the lack of Raul's arm will cost them any runs in the post-season. In today's broadcast before Matsui [screwed] up, they said "LF - below average arm, CF - below average arm, RF - average arm." That doesn't exactly inspire fear in the opposition when they're on the basepaths in a close game. But this trade (forgetting the rest of the details, and just looking at it as Raul and $2 mil for DGS) get the Yankees a B. With the stat-numbers alone I want to give this a C/C+ but a few of the intangibles come into play here... Raul's 32, Dellucci's 29, Garcia 27, Sierra 37 (avg 31). Not really a deal breaker, but I think if they're more interested in keeping Dellucci and Garcia, the few extra years may mean they stay on a little while as role/platoon players who can get accustom to the Yankee system while I think the writing had been on the wall since the day of the Raul trade last year that he was gone after this season (Vlad was mentioned as a possible replacement back then). The attitude also make this a better than just numbers deal. I know that the "gut instinct" is what sabermetrics tries to remove, but I think even Beane knows what a bad attitude can do to a clubhouse (rumors were that Jeremy Giambi got shipped off to Philly for exactly that reason). Also Raul had been in a major slide since the early part of the season and showed no signs of life. So these are a few of the reasons why I bumped this trade up a few plus-minuses.

So the trades seem to be a slight improvement, with the future impact being directly related to Claussen's sucess vs. the AAA players gotten by the Yankees. For this season, it depends on the DGS trio being adequate and Boone staying steady, but the trades seem to be good for the Yankees even w/ the loss of Claussen.


In the "Pettitte Free Agent Watch of '03"-related news, during today's game the YES broadcasters mentioned that he has said he can't imagine playing for any other team, and that he's more than comfortable waiting until the end of the season to work on those details. He also said he understood managment's position in waiting until the season finishes to watch his health which was a major question after last season's set of injuries. So hopefully the managment will realize they need the rock-steady Pettitte to help anchor the middle/bottom half of this team's rotation for at least the next few years.


My one sentence about cricket: Seriously, the whole sport is being held together with the same logic used to convince anyone to back the WNBA. (Try and read a cricket news story, and you'll understand)
- posted by LT @ 8/03/2003 09:59:00 PM

 
ESPN.com - Page2 - Trade deadline special: This will have to act as filler until I get a chance to write some stuff up here (hopefully later today). Just pretend I'm as funny as these guys and you'll keep coming back for years.
- posted by LT @ 8/03/2003 08:27:00 AM