Baseball Beat December 28, 2007
James on Raines

As a follow-up to 30 Rock, I thought it would be interesting to read what Bill James had to say about Tim Raines in the 1982-1988 Baseball Abstracts. The Abstracts hit the bookstores in the spring and were based on the previous season (e.g., the 1982 Baseball Abstract covered the 1981 campaign). As such, a look back at the Ballantine-published Baseball Abstracts gives us a glimpse of what James thought about Raines in real time during Tim's first seven seasons in the bigs.

I believe you will find the following commentary of interest with respect to both James and Raines.

1982

James ranked Raines third in his list of left fielders in the 1982 Baseball Abstract, behind Rickey Henderson and George Foster. In the "Introduction of the Player Ratings and Comments," James wrote: "This year's player evaluations, unlike the ratings I have presented in the past, are based solely on the player's performance during the 1981 season."

According to James, a player's offensive won-lost percentage is:

(Runs Created/Game)²
(Runs Created/Game)² + (League Runs/Game)²

The offensive won-lost method is adjusted for what James termed "park illusions." "If a player plays in a park which increases offensive production by 10%, his runs created are divided by 1.05 before his OWL percentage is figured." In the team section, James concluded that "Olympic Stadium reduces offensive production by approximately 4%."

If one wanted to convert the W-L % into a W-L record, you would divide the player's outs by 25 (which is the approximate number of outs per game rounded down) to get an equivalent number of games. The number of games multiplied by the player's W-L % equals the number of offensive wins. Games minus wins results in the number of losses.

With respect to defensive won-lost percentage, James opted to use two decimals. "To use three decimals here would imply a degree of accuracy which is entirely non-existent." You gotta love his candor.

3. Tim RAINES, Montreal (.691)

Offensive: .783
Defensive: .50
Playing Time: 90%

Home-road breakdowns (.347 and .265) shouldn't be taken too seriously on 150 at bats each place. Raines has already established 21% chance of breaking Brock's career stolen base record. I'm anxious to get the season started and see how many he can steal. His offensive won-lost percentage is the best in baseball for a left fielder, but his defensive stats were so-so; I suppose you know he never played the outfield in the minors.

1983

Switching to a rating system based on the previous two years, James ranked Raines fourth among all left fielders in the 1983 Baseball Abstract.

4. Tim RAINES, Montreal (21-12)

By the lead-off formula given in the Henderson comment, Raines ranks as by far the best lead-off man in the National League. But the formula says he should have scored 111 runs, and he wasn't anywhere near that (he had 90; the -21 is easily the largest discrepancy of the season). This suggests two things: 1) that the Expos lacked a decent #2 hitter, which it is pretty obvious they did, and 2) that all of the Montreal fans who wrote to me that Dawson wasn't hitting anything in the clutch probably weren't imagining it.

TRIVIA TIME – He came to the majors as an infielder, he was shifted to left field as a rookie, he had an outstanding rookie year in which he led the National League in stolen bases, he was then shifted back to second base, and he had a long and outstanding career in the major leagues as a second baseman. Who is he?

I'm not going to give you the answer, by the way. But I'll give you a hint: he is still active and at the major-league level in some phase of the game. (19-10; 3-2)

The latter two figures are offensive and defensive won-lost records.

1984

In "How The Ratings Are Derived" in the 1984 Baseball Abstract, James stated that "the parenthetical expression at the beginning of the player comment gives the combined number of wins and losses that the player has produced for his team over the last two years."

1. Tim RAINES, Montreal (28-15)

Missed becoming the first man to score 20% of his team's runs by only three runs scored. The record for scoring the largest share of your team's runs is held by Kindly Old Burt Shotton of the 1913 St. Louis Browns (I don't think he was "Kindly Old" at the time) who has to be one of the few St. Louis Browns to hold a single season mark of any kind. George Sisler's hit record of 1920 is the only other one to come to mind.

Raines did establish a new NL record, breaking the old one set by Mays in 1964 by a good margin.

Raines and Henderson are two of the few leadoff men on this list; Shotton was another. Mostly, they're sluggers.

An interesting thing is that most of the players come from relatively good teams; there don't seem to be too many cases of exploitation of a team's low production. Shotton is the only member of a last-place team on the list, although Chapman's, Davis' and Billy Williams' teams were under .500. – Jim Baker

In the Henderson comments, James wrote, "The ratings are mixed up here; Rickey should be #1, Raines #2 and Rice #3. This happens because of a small flaw in the rating system, having to do with rounding the won/lost records into integers. Henderson is listed at 26 wins, 14 losses, but his two-year winning percentage is actually .659, not .650. I'll try to get that problem straightened out by next year."

1985

In the 1985 Baseball Abstract, James broke down the rankings by league. Raines was listed as a center fielder.

2. Tim RAINES, Montreal

Strengths: Hitting for average, speed, range, line-drive power, strike zone judgment.
Weaknesses: See below.

The 1984 Montreal Expos, not meaning to slight Charlie Lee (sic) or anything, had essentially two strengths. In Gary Carter, they had one of the greatest catchers in the history of baseball. In Tim Raines, they had the outstanding lead-off man in the history of the National League. Raines hit .309, got on base almost 40% of the time, reached scoring position under his own power 130 times (with the help of 75 stolen bases and 38 doubles) and, playing center field, was second among National League outfielders in putouts. Raines scored 106 runs with a terrible offense coming up behind him, led the league in stolen bases and is now five years ahead of Lou Brock's pace as a base stealer. He doesn't throw real great, but if you've got to have a weakness that's a good one to choose, because it really doesn't cost the team a half-dozen runs a year. He is a great ballplayer, one of the ten best in baseball.

So what do they do? Of course: They trade off the catcher and worry about the center fielder's throwing arm. It's crazy, but if you're losing and you're frustrated, it seems logical. Losing ball teams focus their frustration on their best players in exactly the same way that a man who gets fired from his job and loses his house to the bank will then divorce his wife, who is the only thing in his life that's worth hanging onto. You know how the story goes from there. If Andre Dawson doesn't come back and play the way he did before his knees went, the Expos will lose ninety games this year.

1986

James continued to rate players within their league in the 1986 Baseball Abstract. However, the players were rated by "a poll of the scorers who participated in Project Scoresheet." James claimed there were two reasons for allowing these scorers to vote. "One was to reward, and thus encourage, participation in the project. The other is that I sincerely believe that it's the best way that I can devise to rate the players."

The parenthetical numbers next to the names represent the number of precincts in which the player finished first in the voting. The voters from Project Scoresheet ranked Raines second among left fielders.

2. Tim RAINES, Montreal (1)

Now clearly the greatest lead-off man in National League history. He hit .326 on artificial turf, the highest turf average of any player who played on turf in his home park. His .788 offensive winning percentage was third in the league, behind Guerrero and Strawberry. As mentioned in the San Diego comment, Montreal lead-off men – Raines, for the most part – scored 128 runs, by far the largest percentage of team runs scored by any batting position in the league. A great, great player.

1987

The players were once again rated in the 1987 Baseball Abstract by a poll of approximately 140 scorers participating in Project Scoresheet. The voters were divided into 26 precincts, one representing each major-league team. "The voters were asked to rank the players on the basis of present, clearly established ability."

Raines was ranked number one among all NL left fielders and given 1 1/2 pages of space. It may be a bit long, but it is well worth your time and, in my opinion, should be required reading for all HOF voters. Give yourself three units of credit in Sabermetrics 101 for tackling the following:

1. Tim Raines, Montreal (11)

He would have been a deserving recipient of the National League Most Valuable Player award last year, which is not to say that Schmidt wasn't.

If you compare them offensively, Schmidt and Raines are oddly similar in dissimilar ways. They went to the plate almost the same number of times, 664 for Raines and 657 for Schmidt. Offense in baseball consists of two things: getting runners on base, and advancing runners. Raines won the batting title, and with seventy walks also led the National League in on-base percentage, at .413. Schmidt hit .290 himself and drew 89 walks besides, so that he was on base a lot, too. With adjustments for getting caught stealing and grounding into double plays (we penalize the hitter for taking other runners off base), Raines is credited by the runs created formula with being on base 259 times to Schmidt's 246. Close, but the edge to Raines.

As to advancing runners, Schmidt because of his power, had 302 total bases, which is the largest factor in the advancement of runners. However, Raines had 54 extra base hits himself (35-10-9), and being the batting champion, he too had 276 total bases. In addition, Raines stole 70 bases, 69 more than Schmidt. Although the runs created method considers the value of this to be equivalent to only 36 batting bases, with an adjustment for stolen bases and miscellaneous stuff, Schmidt is credited with 326 "advancement bases" by the runs created method, while Raines is credited with 333. Again, it's very close, but again Raines has the edge. Raines did slightly more to advance himself or other baserunners than did Schmidt.

Putting the elements together, you get:

```                      Schmidt   Raines
On Base                 246      259
Plate appearances       657      664
Runs Created            122      130
```

They're very similar, oddly similar in the proportions, but Raines probably created about eight more runs for his team than did Schmidt.

Then you have to put that into a context of outs. Schmidt made 392 batting outs (552 minus 160) and 19 miscellaneous outs (9 sacrifices, 2 caught stealing, and 8 double plays). Raines made 386 batting outs (580 minus 194) and also made 19 miscellaneous outs (4 sacrifices, 9 caught stealing, and 6 double plays). The totals are 411 outs for Schmidt, 405 for Raines.

Putting the runs in a 27-out context, you have 8.01 runs created per 27 outs for Schmidt, and 8.66 for Raines. They are one-two in the league, but Raines's small advantages add up to a significant edge, making him pretty clearly the best offensive player in the league.

Except, of course, that Schmidt drove in and scored more runs than did Raines. I'll finish the MVP argument for Raines, and then I'll get back to that.

1) Raines created more runs than Schmidt despite playing in a much tougher hitter's park. Raines's batting and slugging percentages were 16 and 50 points higher on the road than they were in Montreal. Schmidt's batting and slugging averages were 16 and 94 points higher in Philadelphia than on the road. With adjustments for the statistical distortions of the parks, Raines was really a much better hitter. Further, the average Phillies game had 9.0 runs, whereas the average Expos game had only 8.2 runs, so the runs that Raines created were more valuable – had more of a win impact – than the runs that Schmidt created in his inflated environment.

2) Schmidt at 36 started the season at first base, and spent most of the season at third base, where he was an ordinary defensive player at a somewhat key defensive position. Raines played left field, where he is an exceptional defensive player at what is not a key defensive position. Raines was third in the league in outfield assists, and twice ended games by throwing out the potential tying run at the plate. We can call their defense a wash.

3) Neither player's team was ultimately successful, but Raines's early season streak of reaching base in 42 straight games helped greatly to keep the Expos in the pennant race. It wasn't ultimately meaningful, but it was very meaningful at the time. Schmidt piled up 67 RBI over the last three months of the season, when the Phillies were already dead and buried as a team. Not wanting to seem too eager to claim advantages for Raines, we'll call that a wash, too.

4) Schmidt has been a great player, but he has won the award twice before. Raines has been a great player for six years now, and he's never won it. In the interests of fairness, 1986 would have been an opportunity to balance the scales.

OK, then we go back to the issue of actual run and RBI counts. I would say this: that if Schmidt's advantage in runs scored and RBI resulted from his superior performance in run-production situations, then it is reasonable to consider this an advantage for Schmidt. If, on the other hand, the advantage resulted from offensive context (that is, having better hitters surrounding him), then it is unfair to penalize Raines because his teammates were not as good as Schmidt's.

Quite clearly, those differences resulted primarily from offensive context, and not from inidividual differences. Consider:

1) Schmidt drove in 25% of the runners that he inherited in scoring position. Raines drove in 26%. The big difference in RBI was that Schmidt came to the plate with 253 runners in scoring position, and Raines came up with only 173 ducks on the pond.

2) Schmidt drove in 49% of his runners on third base with less than two out, 20 of 41. Raines drove in 58% of his, 18 of 31.

3) Another situation in which at bats have a disproportionate impact on runs resulting is the first at bat in the inning, the leadoff spot. Raines had 190 such at bats, and Schmidt 142. In that game situation, power increases for most hitters but on-base percentage decreases, as pitchers concentrate on not allowing walks.

Schmidt's slugging percentage increased only nine points in that situation, from .547 to .556, while his on base percentage plummeted (all of this data is in the Great American Baseball Stat Book) from .390 to .340, 50 points. Adding the two together, he lost 41 points (+9, -50). Raines, on the other hand, increased his slugging percentage as a leadoff hitter by 35 points, to .511, while his on-base percentage dropped only to .404, 9 points, so he showed net gain of +26 (+35, -9). A big edge to Raines.

In the late innings of close games, pitchers strive to avoid the game-breaking homer, so exactly the opposite happens: Slugging percentages go down, while on-base percentages go up. Both players hit almost their averages in the late innings of close games – Schmidt .290 (+ or - zero) and Raines .339 (up 6 points). But again, Schmidt's slugging percentage went down 31 points while his on-base percentage went up only 21, a net loss of 10 points. Raines's slugging went down 24, but his on-base percentage went up 28, a net gain of four points. So again, Raines exploited the positives of a critical game situation more effectively than did Schmidt.

So it seems obvious that Schmidt's higher run counts resulted not from his own ability, but from the fact that he was hitting in the middle of Gary Redus, Von Hayes, and Juan Samuel, while Raines did not have comparable support. The Phillies scored 739 runs as a team, second in the league. The Expos scored 637, tenth in the league. This shouldn't be the criteria for who gets the MVP award.

I'm not criticizing anybody for his vote. I too thought, just looking at the statistics, that Schmidt had had the best year. Raines's remarkable base stealing (70/79) is easy to overlook, particularly when the run count doesn't reflect the advantage. But having looked at the issue more carefully, I now realize that Tim Raines was, in fact, the best and most valuable player in the National League in 1986.

Whew! If you're still with me, give yourself an extra unit of credit. But be prepared for a pop quiz down the road.

1988

In the "Introduction to Player Ratings" in the final Baseball Abstract in 1988, James wrote, "The players . . . will be rated by subjective judgment. Mine. For the past couple of years I've rated players by a poll of the members of Project Scoresheet, but this year I just decided to do it myself."

James changed the format, separating the rankings and comments for the first time. He also combined the two leagues and had one ranking for each position. Raines was rated as the #1 left fielder. The player comments were provided in alphabetical order and letter grades were given as follows:

TIM RAINES
```Hitting for Average:       A
Hitting for Power:         B
Plate Discipline:          B+
Baserunning:               A
OVERALL OFFENSE:                A
Defensive Range:           A
Reliability:               B+
Arm:                       C+
OVERALL DEFENSE:                B+
Consistency:               A
Durability:                A
OVERALL VALUE:                  A-
In a Word:  Brilliant
```

James printed a letter that he solicited from Neil Munro, comparing Raines to Wade Boggs. Munro provided a detailed explanation and chose Raines. "As a consequence, if you rate them pretty much even as hitters (with park adjustments) and fielders, you must give the nod to Raines for his baserunning ability."

Earlier in the book, under "Rain Delay," James penned one of his best essays, holding a conversation with himself in the search of the best baseball player in the game. It's a six pager with insightful comments on about 20 players. His conclusion? James ranked Boggs as the best player in baseball, followed by Raines, Ozzie Smith, Don Mattingly, Tony Gwynn, Darryl Strawberry, Dale Murphy, Roger Clemens, Rickey Henderson, and Kirby Puckett.

Although Raines' career lasted until 2002, the Baseball Abstracts were, unfortunately, retired in 1988. However, the Ballantine era books coincided with Raines' first seven seasons (which most would also view as the best seven-year stretch of his career).

Raines. James. Baseball Abstracts. Three of the very best of the 1980s.

* * * * *

Feel free to give James' trivia question from the 1983 comments a shot.

"1) Schmidt drove in 25% of the runners that he inherited in scoring position. Raines drove in 26%. The big difference in RBI was that Schmidt came to the plate with 253 runners in scoring position, and Raines came up with only 173 ducks on the pond."

It is truly amazing that 20 years after James wrote this, many writers still fail to grasp this very simple concept. Excellent piece.

Red Schoendienst!

If Raines had played in New York during the '80s, the "He's a Hall of Famer" hype would have begun long ago. The Montreal connection hurts his visibility in a big way.

Schmidt drove in 25% of the runners that he inherited in scoring position. Raines drove in 26%.

What difference does it make whether a runner was in scoring position? How many runs did each drive in who weren't in scoring position? Essentially you are saying if Raines drives in a runner on third with a sacrifice fly, that counts. But if Schmidt hits a home run with a runner on first, neither RBI counts. And neither do his solo home runs.

Proof that statistics can be used to prove whatever you want them to prove.

Ross Williams has got to be a pseudonym.

Who are you really? Plaschke? Conlin?

It drives me nuts that nuckleheads like that have kept Blyleven out of the Hall and that they may do the same to Raines.

Bill James was refering to "clutch" situations.

He covered what you are saying in "advancement bases".

He is trying to find the truth... a reason... to why the voters consider Schmidt more valuable in that particular year.

He could not. Schmidt was not.

That is just one data point that serves to narrow the perception gap. It serves to prove nothing.

James's point was simply that both were deserving candidates, despite Schmidt's MVP and Raines's 6th place finish. One had a 152 OPS+ with 1 SB, the other 145 with 70 SB's. I dunno, they sound comparable to me.

I´ll take the edge in OBP in the tougher park AND the 70 stolen bases over the + 7 points in OPS+... thank you very much.

Comparable... of course.

ABSOLUTE PROOF... hell, that is a whole other argument.

"25% of the runners that he inherited in scoring position. Raines drove in 26%."

"The big difference in RBI was that Schmidt came to the plate with 253 runners in scoring position, and Raines came up with only 173 ducks on the pond."

So let see:
Schmidt: 63 RBI's/253 Runners.
Raines: 45 RBI's/173 Runners.
Difference: 18 RBI's.

Schmidt: 119 RBI's
Raines: 62 RBI's
Difference: 57 RBI's.

"Quite clearly, those differences resulted primarily from offensive context, and not from inidividual differences."

To the contrary, the numbers show quite clearly most of the difference did not result from offensive context. Even if Raines had the same number of chances, he still would have been over 40 RBI's short of Schmidt's production.

Of course.

That's the point.

The difference is NOT clutch hitting. It's more people on base and hitting more Home Runs (Higher slugging %).

The big difference between hitting 37 HR versus 9 HR. It goes to show that RBI's is a TEAM oriented stat, not an individual stat.

Hence, RBI's are not the whole picture, they are a very small part of the picture. The whole picture contains total bases, on base percentage, slugging, park effects, stolen bases, defense, etcetera.

The big difference between hitting 37 HR versus 9 HR.

That's the point.

That wasn't James point obviously. He was apparently trying to prove the opposite.

It goes to show that RBI's is a TEAM oriented stat, not an individual stat.

To the contrary - it goes to show that the biggest difference between Raines and Schmidt in terms of RBI's had nothing to do with their teammates and everything to do with their own performance.

What this discussion shows is that stats, far from being objective, can be used to support whatever preconceived ideas people have.

I'd suggest that the true clutch situation is when nobody's on base. Consider a tie game. Leadoff man draws a walk, steals second, comes home on a single. Why is it more clutch to drive home the run than to get on base in the first place?

My favorite Raines memory was the year he missed three weeks at the start of the 1987 season -- some sort of collusion thing. The Expos re-signed him and put him in the lineup hitting third. No spring training, nothing. He triples on the first pitch and hits a grand slam in the 10th. Expos 11, Mets 7.

What this discussion shows is that stats, far from being objective, can be used to support whatever preconceived ideas people have.

What it shows is that bad stats (like RBI) can be manipulated every which way. It doesn't show anything about more legitmate stats that actually represent value.

James' point is obviously not that 37 homers and 9 homers (a stat you cite, by the way) are the same. His point more generally is that the numbers are not nearly as disparate as they seem, and in light of other differences in ability (Raines' advantage in average, on base percentage, baserunning and defense) that they are broadly comparable in value. Blanket assertions about how the statistical record can be manipulated are essentially stall tactics by people who lack stronger arguments.

The real and obvious reason for Schmidt driving in so many more runs than Raines is that Schmidt batted down the lineup and Raines led off. Nothing against Mike Schmidt, who had a great year, but some people just insist that RBIs are the be-all end-all of MVP voting. Consider Mattingly's MVP. He had a better year than George Brett in only one area: wisely choosing to have Rickey Henderson leading off for him instead of Willie Wilson. When your leadoff hitter is the first player in almost 40 years to reach 146 runs scored, it's easy to get more RBI than a guy with a mediocre leadoff hitter. Brett's 1975 OPS+: 178; Henderson's 157; Mattingly's 156; Wilson's 97. Henderson's OBP was over 100 points higher than Wilson's, which led to scoring 59 more runs and gave Mattingly an undeserved MVP.

Raines versus Schmidt is the same argument, with guys like Williams unable to accept that sometimes the RBI comes from the guy getting himself to third and not the guy who hits the sac fly or infield out.