clock menu more-arrow no yes

Filed under:

White Sox Report for 2005

New, 2 comments

Individual player comments will be posted tomorrow.

Don't get me wrong; I'm ecstatic that the Sox won the World Series (well, as ecstatic as an economist/sabermetrician can get). Having been a fan for close to 40 years, I am very happy. The Sox, although a good team, were lucky. They won more games than their stats would indicate (although they dominated statistically in the playoffs). But maybe they finally offset alot bad luck they had over the years. They've had good teams that did not quite make it or good teams beset by injuries. Explaining all that is another essay. This is about the 2005 Sox, so I'm sticking to that.

After last season, the Sox management (GM Ken Williams and Ozzie Guillen) decided that the team needed a more balanced offense (one that relied less on homeruns), better pitching and better fielding. Something obviously worked since they won the AL Central division and made the playoffs for the first time in 5 years.

But what was it that really worked? Let's start with the offense. Supposedly the Sox have been better off relying more on speed than power. They traded HR hitter Carlos Lee for base stealer Scott Podsednik. They declined to sign the high-priced Magglio Ordonez. They let Jose Valentine go, a shortstop with 5 straight 25+ HR seasons. Here are some offensive comparisons between the seasons:

I do not think a decline of 124 runs is what Williams had in mind. But, if you listen to TV announcers, it was really more consistency in the offense that the Sox wanted. They did not want to score 15 one day and get shut out the next. Were they more consistent in 2005?

Not really. The table below shows how many runs they averaged per game in their lowest 54 game totals, middle 54 games and the best 54 in each season.

In low and middle scoring games, the Sox of 2005 were not too far behind the Sox of 2004. But, there was a big difference in high scoring games. The 2004 Sox scored almost two runs better per game than the 2005 Sox in the best 54 games. Basically, the personnel changes the Sox made pretty much ruled out a lot of high scoring games. Having Frank Thomas miss even more games this season did not help. And hanging on to Ordonez and Valentin would not have increased scoring, since they missed a lot of games with injuries.

The next table shows the distribution of runs by how many games the Sox scored a given number in both years

The Sox were shutout less in 2005, but just one less time. In fact, the number of "low scoring" games (of 3 runs or less), was almost the same in each season. There were 57 such games in 2004 and 60 in 2005. But there were 14 fewer games of 6 or more runs (67 vs. 53). So the Sox did not really avoid low scoring games, which was one of the objectives of their personnel changes.

But White Sox GM Ken Williams said this about Scott Podsednik: "One positive is his speed, his ability to create havoc on the bases." Perhaps the increase in base stealing kept the runs from falling even more. Maybe it helped the hitters to have a stealing threat on first, distracting the pitcher. This is probably not likely (although economist Ted Turocy has done research suggesting that base stealers can sometimes help the hitter).

Just glancing at the Sox stats from the first table shows they scored much less. To look at whether or not base stealers do help the hitters, I looked at the top 10 teams in SB's from 1982-92 and the bottom ten. Then I determined how much their AVG, SLG, OBP and OPS differed between having a runner on 1st or no runners on at all. I determined the runner on first data by finding the difference between the runners on base and the runners in scoring position data (from Retrosheet).

The top ten teams in SB's had the following increases when there was a runner on first compared to no runners on (the average across the teams)

AVG-.025
SLG-.030
OBP-.008
OPS-.038

(OPS is found by adding on-base percentage plus slugging percentage. So it takes into account the ability to reach base and power hitting ability.) So, with a runner on first, these teams had a .025 higher batting average than they did when there were no runners on base. Slugging went up .030, OBP .008 and OPS .038.

The bottom ten teams in SB's had the following increases

AVG-.019
SLG-.028
OBP-.011
OPS-.039

The top ten teams averaged about 240 SB's and the bottom around 40. The one difference that is big is the AVG difference (.025 - .019 = .006).  But in general, the best stealing teams had little additional benefit over what the worst stealing teams had.

The best stealing teams had in the 900 range of AB's with a runner on first. The bottom teams resided in the 1100 range. This makes sense because the best stealing teams steal and they will not be on first as often. Also, who is most likely to be left on first base on those teams? The few guys who don't steal, like Jack Clark (5 of the teams were Cards). But those bottom teams must have rarely had a good base stealer on; at least less often than the best. I think if the runners bother the pitcher, we should see more of an effect here. After all, we are comparing the best stealing teams to the worst.

The change in OPS for both teams is just about the same. I am still skeptical that having a good steal candidate on first helps greatly. Maybe the change in AVG is simply a result of the hole opened up at first. There is little change in SLG. Maybe the fast guy bothering the pitcher and making it easier for the hitter is not really happening as much as we sometimes like to think.

So the Sox offense (whether hitting or stealing) is not a big reason for their success. In fact, the average AL team scored 768 runs (27 more than the Sox) even though the Sox play in a fairly good hitter's park that is especially homerun friendly (they did hit 200 homeruns though). Certainly base running or base stealing was not a big factor (especially since they had so many caught stealing -- their success rate was 67.2% while the major league average was 70.5%, below and above the break even point respectively). Using Pete Palmer's run values of .22 for a SB and -.38 for CS, the Sox stealing generated about 5 runs (Palmer is editor of the Baseball Encyclopedia).

And for all the talk of "small all" or "Ozzie ball" or "smart ball," the Sox scored just about the number of runs that we would expect based on their OBP and SLG. The equation below shows the relationship between those two stats and runs per game for all teams from 2001-03:

R/G = 17*OBP + 11.17*SLG - 5.64

It predicted 78 of the 90 teams to within 40 runs per season. With an OBP of .322 and an SLG of .425, the formula predicts that the Sox would score 4.58 runs per game. It was actually 4.57. So all the stealing and bunting did not squeeze any extra runs out the Sox production. In fairness to GM Williams, the decline in scoring was not due to his personnel moves. Centerfielder Aaron Rowand saw his OPS fall from .905 in 2004 to .736 in 2005. Shortstop Juan Uribe went from .833 to .712. DH Frank Thomas played only 34 games (74 in 2004). No hitter showed any huge improvement. So we have to turn to pitching and fielding to try to explain the Sox success.

The table below shows how the Sox improved in both pitching and fielding going from 2004 to 2005

A drop in runs allowed of 186 is the reason for the Sox success this year. Several factors contributed to this: walking fewer batters, giving up fewer homeruns, fewer errors and a lower average allowed on balls in play (non-HRs, strikeouts and walks). The fielding was very good, with the DER (Defensive Efficiency Rating-defined at end of paper) being the second best in baseball (if they had stayed at .7073, they would have been 19th best). If the ball stayed in the park, the Sox fielders did a very good job of running it down (the major league average was about .7085). Adding Juan Uribe at short for good, Iguchi at second and Jermaine Dye and Podsednik in the outfield probably helped here. A.J. Pierzynski, an experienced catcher, may have helped the pitching staff improve (Tom Hanrahan has done research showing that experienced catchers can help a pitching staff).

How many runs did the improvements in walks, HRs and DER save? Pete Palmer gives a walk a run value of .33, so 72 fewer walks means 24 runs saved. Then 57 fewer HRs at 1.4 per HR is 80 runs. The DER improved by .0179. That means a 1.79 percentage point increase in the proportion of in play batted balls turned into outs. The Sox had about 27 per game each of the last two seasons. Since 27 times .0179 is .4833, that is how many fewer non-HR hits they gave up per game. The weighted-average of the linear weights run values of singles, doubles and triples is about .55 (again, using Pete Palmer's run values). Multiplying .4833 times .55 leaves .266 runs saved per game. For the whole season, that is 43 runs. These improvements saved 167 runs.

Some of the Sox pitchers performed much better in defense independent stats in 2005 than they did in 2004. HR is per IP. SO/BB is strikeout to walk ratio.

Three of the big starters (Garland, Buehrle, and Contreras) all improved significantly in both stats. Same for two of the relievers, Neal Cotts and Cliff Politte. The last three listed, although their strikeout-to-walk ratios fell, showed big improvements in HRs allowed. The addition of Bobby Jenks at mid-season helped, especially since closer Dustin Hermanson developed back problems (Jenks struck out 50 in 39.1 IP, allowing only 15 walks, 3 HRs and 34 hits).

The White Sox won about 7 more games (99 vs. 92) than their Pythagorean projection (the Bill James formula that says that a team's winning percentage will be about equal to run scored squared divided by run scored squared plus runs allowed squared).

The Sox were 35-19 in one-run games. That will usually help a team exceed its Pythagorean projection. But at the halfway point, their record in one-run games was 22-8 (even better than their overall first-half record of 55-26). So they went 13-11 in one-run games in the second half (not as good as their second-half record of 44-37). This all points to the fact that the Sox may have had more than their share of luck this year (see the Indians essay for more on how teams don't exceed their Pythagorean projection year after year). And despite all the personnel changes the White Sox had, their record in 1-run games was also very good in 2004, 28-18. The 2004 Sox also were projected to win 84.25 wins, just one more than they actually did. So the "problem" that Sox management wanted to address of winning by lopsided scores and then losing closes ones did not really exist in 2004.

The "ability" to win one-run games seems to have evaporated in the second half.

Another indication of the luck the Sox had is that they won more games than their OPS differential indicated. OPS is on-base percentage plus slugging percentage. As a differential, it is a team's OPS minus the OPS they allow their opponents. The following formula predicts a team's winning percentage:

Pct = 1.21*OPSDIFF + .5

Using all teams from 1989-2002, it predicted 28 of the 30 teams to within 3 wins per season. Using this formula, the table below shows what the Sox record was, what their predicted winning percentage was, their predicted wins, their actual wins and what the difference was at both the All-Star break and at the end of the season.

So at the All-Star break, the White Sox had 10.05 more wins than expected. But when the season ended, it was at 10.16 (meaning that after the All-Star break they won just .11 more games than expected and their OPS differential was probably about .042, not much different than before the All-Star break). What ever the mysterious power or ability was that got them all the extra wins in the first half again seems to have disappeared in the second half.

One possible explanation is that in the first half the Sox had an OPS differential of .122 with runners on base (the highest in baseball). But when the season ended, it was down to .092, meaning it was around .060 in the second half or not much higher than their .042 overall OPS differential in the second half.

Here are some specific instances of luck that I recall. The Sox got to play four games against the Orioles in late July/early August after the Orioles had their midseason meltdown that included the Rafael Palmeiro steroids controversy (the Sox swept). Chicago played their first seven games against the Angels with Vladimir Guerrero injured and not playing (then lost the three games he did play in). When they played the Dodgers, ace closer Gagne had just gone on the disabled list. In one of those games against Dodgers, A.J. Pierzynski hit a game winning HR in the bottom of the 9th inning right after Hee-Seop Choi dropped a pop foul. Maybe all teams get this kind of luck. But all of these incidents happened in the first half of the season or not too long after the All-Star break.

The Sox certainly improved going from 2004 to 2005. Their run differential increased by 62. That would normally mean about 6 more wins. They won 83 games in 2004. So we could have expected them to win about 89 games (as their OPS differential predicted). The Chicago White Sox had a good year, but they also had more than their share of luck. Also, their good luck was mirrored by the Indians bad luck. The Indians had knocked the Sox 15 game lead in early August down to 1.5 games in late September and the Sox did not clinch the division until game number 159. The Indians won fewer games than predicted by the OPS differential and the Pythagorean projection. Cleveland had a terrible record in 1-run games while seeing their OPS differential fall quite a bit with runners on base. Cleveland played much better against teams outside the Central division than within while the Sox played much better against the Central (again, see the Indians essay).

The Sox had the best record in the AL. But they seem like a "weak" best team. The table below has some data on the best teams in both leagues from 2000-2004. RDiff is run differential.

The 99 wins were very close to what a typical best team has. The Sox run differential (96) and OPS differential (.040) were good, but nothing like what we normally see for the team with the league's highest winning percentage. Again, the Sox seemed to have been lucky (or a great clutch team or Ozzie Guillen has a fantastic new theory of managing). They may also have benefited from the Indians' bad luck (see their essay).

Having home-field advantage helped in their first round sweep of the Red Sox (although the White Sox did have the best road record (52-29), Boston had the best home record (54-27). What helped the White Sox get the best record in the AL? Playing in a weak division. They went 35-35 combined against AL East and West teams but 52-22 against AL Central teams. The table below shows the record each AL division had against the other two divisions. The last column shows the division's record against the other with the first place team's record removed.

The Red Sox were 44-26 against the Central and West teams, almost the same as the first-place Yankees (43-27). So if the Red Sox were removed from the East (instead of the Yankees), the last column would almost be the same (142-138). The AL Central is clearly the weakest in the league. And that is where the White Sox did their damage.

It also allowed them to clinch a playoff spot a few days ahead of the Red Sox and gets their pitching staff set up. Jose Contreras started game 1 of the divisional series. He really started to come on in the last two months of the season, with Eras in the last two months of 2.14 and 1.99. The Red Sox had to start Matt Clement, who had been shaky all year (getting hit in the head by a line drive did not help). Clement hit two of the first three batters he faced, helping the White Sox rack up 5 runs in the first inning, capped by A.J. Pierzynski's 3-run HR. Chicago won the game 14-2, hitting a total of 5 HRs.

In game 2, with one out and a runner on first and Chicago trailing 4-2 in the 5th inning, Red Sox second baseman Tony Graffanino let a slow ground ball go through his legs. That kept the inning alive and two batters later Tadahito Iguchi hit a 3-run HR. The game ended 5-4.

In game 3, Boston loaded the bases with no outs in the 6th inning trailing 4-3. Chicago brought in Orlando Hernandez to pitch. Varitek and Graffanino popped out. Damon struck out on a 3-2 pitch. He tried to check his swing, but he appeared to have gone just far enough. Chicago added a run in the top of the 9th and the final was 5-3. The big blow for Chicago was Konerko's 2-run HR in the top of the 6th.

So homeruns played a key role in the White Sox sweep. They stole 3 bases but none of them figured in the scoring very much and they were thrown out twice. They out hit (.289-.240), out homered (7-3) and out scored (24-9) Boston. The pitching and defense came through, holding Boston's high scoring team to just 9 total runs.

The luck seemed to continue in the Angels series. Bartolo Colon, the Angels best starter, hurt his shoulder in game 5 of the Divisional series and could not pitch at all in the LCS. Game 2 was very interesting. With an apparent strikeout with two outs in the bottom of the 9th suddenly being called a dropped third strike even though all the Angel fielders started walking off the field as catcher Josh Paul rolled the ball back to the mound, the Sox stayed alive and scored a run to win. This tied the series 1-1. It would have been tough for the Sox to go to California down 0-2. There was some luck in game 4, with a catcher's interference not being called against the Sox and Podsednik being called safe when he appeared to be picked off first base.

In the World Series, I don't think the Sox had any specific instances of great luck (there was the HBP for Dye that clearly hit the bat and not him followed by Konerko's grand slam). But Dye may have gotten on base anyway. He battled pitchers and drew some walks. When you win two 1-run games and a 14-inning game on a HR by Geoff Blum, another game on a HR by Podsednik (who hit none during the regular season), there may be some luck. But Podsednik hit 12 in 2004, so he does not totally lack power. And in general, the Sox dominated their opponents statistically.

In the entire postseason, they out homered the other teams 18-9, out scored them 67-34, out walked them 36-29, out hit them 113-81, etc. The table below shows how the Sox dominated in each series. They had a much higher on-base percentage and slugging percentage than their opponents.

A weighted average (weighted by games) of their OPS differential works out to .215! Using that formula from above, that works out to a .760 winning pct. The Pythagorean method would be .795. Pretty good evidence that the postseason was not lucky for the Sox.

Sources: Yahoo Sports, ESPN Website, mlb.com, USA Today Sports Weekly and the Bill James Handbook.

Defensive Efficiency Rating (DER) is the ratio of team defensive outs recorded in defensive opportunities. To determine Defensive Efficiency Rating for a team, divide the total number of hits in play allowed (not including home runs) by the total number of defensive opportunities (all balls hit into play, not including home runs) and subtract from one: 1-((H-HR)/(PA-HR-BB-HBP-SO)). From mlb.com