added 11/22/2005 by Scott Barzilla
Hey, you might laugh, but I always get excited when I get my updates on my favorite sabermetric tools. This time of year it happens to be the Lee Sinins Sabermetric Encyclopedia and the Bill James Handbook. Armed with these two sources (I've had the encyclopedia for a few weeks now) I can throw out a whole lot more when it comes to ranking players. Since we are not in the market for a first baseman, second baseman, third baseman, or centerfielder I decided to save them for last.
For the Astros and first base, the 1000 pound gorilla in the room is Jeff Bagwell. Will Bagwell be healthy enough to play everyday? Will he be healthy enough to hit for power? Will he be healthy enough to play in the field at all? Yes, we are all concerned about the seventeen million Bagwell is getting, but that didn't concern them when they signed Lance Berkman. Before we go too far into answering those questions let's take a look at all of the first basemen that have been regulars the last three seasons and their win share totals.
Win shares is a nice and tidy statistic that includes all offensive and defensive contributions. What we're going to do is not only provide an average win share total for the three seasons, but also factor in the percentage of available win shares. The one weakness that the win share model has is that it is directly tied to a team's win total. This can hurt players that play well on bad teams. Percentage win shares will help eliminate this bias.
2003 2004 2005 Tot AVG Jason Giambi 28 8 24 60 20.00 Kevin Millar 16 17 11 44 14.67 Rafael Palmeiro 19 12 11 42 14.00 Eric Hinske 12 6 11 29 9.67 Travis Hafner 7 21 26 54 18.00 Paul Konerko 4 20 24 48 16.00 Mike Sweeney 15 14 16 45 15.00 Carlos Pena 9 11 7 27 9.00 Mark Teixeira 13 24 33 70 23.33 Scott Hatteberg 14 17 8 39 13.00 Richie Sexson 26 3 24 53 17.67 Darin Erstad 3 15 15 33 11.00 Julio Franco 6 12 7 25 8.33 Jim Thome 30 20 4 54 18.00 Carlos Delgado 32 16 29 77 25.67 Nick Johnson 14 6 20 40 13.33 Doug Mientkiewicz 20 6 4 30 10.00 Albert Pujols 41 37 34 112 37.33 Jeff Bagwell 22 22 3 47 15.67 Lance Berkman 25 30 20 75 25.00 Derrek Lee 25 19 33 77 25.67 Lyle Overbay 6 17 17 40 13.33 Sean Casey 17 28 13 58 19.33 Ryan Klesko 13 18 15 48 16.00 Tony Clark 4 7 18 29 9.67 Todd Helton 34 30 25 89 29.67
For our purposes, 20 win shares is the bench mark of a good season. A decent regular at first base should have fifteen win shares on average. What we will do now is eliminate everyone that averaged fewer than 20 win shares a season and compare the top first basemen in the game on percentage win shares.
WS Team shares Pct Albert Pujols 112 870 12.87 Todd Helton 89 627 14.19 Derrek Lee 77 777 9.90 Carlos Delgado 77 708 10.88 Lance Berkman 75 804 9.33 Mark Teixeira 70 717 9.76 Jason Giambi 60 891 6.73
Some of you may be wondering what a win share is. Simply put, it is a measure of a player's contributions to his team's wins. Three win shares equal one win, so the team shares are simply the team's win totals over the three year period multiplied by three. What you see is that these players averaged a little more than ten percent of the team's win shares. So, what does this have to do with the Houston Astros?
Currently, the Astros have two players at first base (we know Berkman will be shifted to left field if Bagwell is healthy) making 14 million and 17 million. If the Astros have a 90 million dollar payroll they end up having two players that take up 15.55 and 18.88 percent of the overall payroll. This isn't about those guys underachieving. Berkman would likely come closer if we assume total health, but not one of the first basemen achieved that. Albert Pujols is the reigning MVP and he has only accounted for 12.87 percent of his team's win shares.
Part of this is a warning to those teams looking to throw huge money at Paul Konerko. Even if we erase his 2003 season, he still has averaged 22 win shares a season in the last two and will not even come out as good as the seven players listed above. It doesn't take much imagination to put Lance Berkman in third behind Pujols and Lee in the National League, but even then you don't get the bang for your buck with the big money contracts.
Ah, but I'm reminded of phrases of spilled milk and what not, so let's leave the economics lesson aside. How much would a healthy Bagwell improve the Astros? Well, we should start by looking at his three year average and assume he would collect those win shares in a healthy season. However, it's not as easy as inserting Bagwell and adding the win shares. He will be replacing the monolithic left fielder Burke/Scott/Lamb in the lineup. So, we have to find the difference between Berkman plus Bagwell and Berkman plus those three.
AVG WS AVG WS Berkman 25 Berkman 25 Bagwell 15 Lamb 6 Total 40 Burke 6 Scott 0 Total 37
So, we can estimate that a healthy Bagwell would account for one additional victory next season. So, the Astros would go from 89 to 90 wins if everything else were equal. As we know, everything is not necessarily equal. The Astros have four prominent players on the wrong side of 35 including Bagwell. Brad Ausmus, Roger Clemens, and Craig Biggio had seasons that were better than expected. We cannot expect them to repeat that next year.
Of course, players like Willy T, Adam Everett, Luke Scott, Zeke Astacio, and Wandy Rodriguez could be considerably better next year. However, the focus for us at this time has to be the above list. The notion that a healthy Bagwell would actually improve is based on two assumptions: Bagwell would be good enough to get to fifteen win shares in 2006 and the trio that played in 2005 do not improve. Five players approached that total last season. We can look at their numbers to see if that is realistic.
PA AVG OBP SLG HR R RBI Ryan Klesko 520 .248 .358 .418 18 61 58 Darin Erstad 663 .273 .325 .371 7 86 66 Sean Casey 587 .312 .371 .423 9 75 58 Lyle Overbay 622 .276 .367 .449 19 80 72 Mike Sweeney 514 .300 .347 .517 21 63 83 Average 581 .282 .354 .436 15 73 67
Jeff Bagwell has never put up numbers exactly like the average. He has always had a bigger differential between his batting average and on base percentage. However, Bagwell hasn't hit .280 or better since 2002. The .790 OPS is a pretty decent barometer for what we can hope for from Bagwell. Of course, those of you that are paying attention are saying, 'wait a minute Barzilla, didn't you say win shares accounted for both hitting and fielding. Jeff Bagwell can't throw the ball so wouldn't that negatively affect his projected win shares?'
Kudos to those of you who were thinking that. According to James' model, first baseman typically don't account for more than two and a half defensive win shares per 1000 innings. Bagwell can still do some things defensively (pick balls out of the dirt and field grounders) so he wouldn't be a complete basket case. So, we could generously say Bagwell would need an extra offensive win share to make up for the missing defensive win share. That would put him into the Lyle Overbay, Tony Clark, and Mike Sweeney territory. Excuse me if I have my doubts that Bagwell can be that productive.
The other assumption is harder to assess. Young players don't have the track record to project. We know Luke Scott can't do any worse than zero win shares and the stories of him tearing up the Winter Leagues are promising. The majority of the hope probably resides in Chris Burke and his .790 OPS in the second half. The odds that both of them will combine for six win shares in the same amount of plate appearances is slim.
All in all, Jeff Bagwell doesn't add anything to this team unless he returns to the 22 win share level he reached in 2003 and 2004. That's certainly possible, but the 15 win share mark is much more realistic. More and more, we return to the need to improve the catcher or shortstop position.