2009 in Review (Part I)

For the past couple of seasons, I've compiled data from the BPGPL into a spreadsheet with the hope of using it to reach conclusions after the season. Each April, I lay our draft data out to tap the Baseball Prospectus projections to plot the expected results for all teams. In the past, I've forgotten about it by the time November rolled around. This year, I thought I would swap out the BP projections for the actual 2009 season results on a player-by-player basis to explore a couple of questions that are at the heart of fantasy baseball: (1) whose team 'outperformed' the most and which team 'underperformed' the most versus BP projections; and (2) how much does a good draft predict a good finish versus the intra-season roster moves. I will put some of the summary data on the league spreadsheet for anyone interested.

First, let’s review whose core team exceeded projections and whose team underperformed. I compiled projections using the BP data back in April under the assumption that the starting roster for each team would be the most expensive players drafted for each position and that each player would sit there the entire year. This fails to account for players taken at 'steal' prices that allowed paying more for a backup or for players who are very careful about substituting players on off-days, but I needed a simple methodology.

Let’s see where BP was simply wrong. PECOTA is based on playing time projections that are impossible to get exactly right. While they can use playing time and injury history to predict the future, this fails to account for new injuries, managerial decisions and trades, which drag down the results for select players. For our entire league, I found an average underperformance of 11% below the runs projection and 8% below the RBI projection and 6% below the stolen base projection . This is a simple average, but these are not rate stats, so it is applicable. Home runs were 3% lower than forecast and OBP was in-line with the forecast.




Table 1 lays out the average over/under-performance for the batting statistics of each team in order of projected finish. Across the majors, offense registered similar levels to 2008, so MLB as a whole was consistent with their recent history in terms of runs and HRs. As might be suspected, there were unheralded players who likely picked up some of that ‘missing’ offense. There are some pretty clear trends, however, on who got 'lucky' or 'unlucky' (or alternatively who was 'smart' or 'dumb') in picking players that outperformed the PECOTA projections.


The projections from the outset of the season featured the Boston Doodles of Cambridge and the Replacement Levelers tying for first place. The forecast included the Doodles dominating the hitting categories with Albert Pujols, David Wright, Hanley Ramirez and a number of 20-HR players, while giving up on Ks and Ws in favor of control-oriented pitchers who could keep the rate stats (ERA, WHIP, K/BB) competitive. No credit was given to major league-ready prospect Matt Wieters on the minor league roster since the team also had a $14 catcher in Geovany Soto. The actual results were disappointing – 18% fewer runs and 20% fewer HRs. For a team built around winning hitting categories, this was a devastating turn of events.. Despite another great year by Pujols who justified his $40 price tag, each of the other starters on the field provided less offense than could be reasonably expected. Absolutely no baseball analysts were suggesting that Wright ($29) and Soto ($14) would each hit less than half as many HRs as they did the prior season. Sub-par performances were also provided by Jay Bruce ($15 for 22 HRs, 47 Rs and 58 RBIs) and BJ Upton ($15 for 79 Rs and a .313 OBP).


Meanwhile, the Replacement Levelers kept hitting underperformance within the averages of the rest of the league. The aim was clearly to dominate in stolen bases, which the Levelers handily achieved with the league lead. However, the Levelers were principally a pitching-oriented team (based on the PECOTA projections). We'll take a look at the impact of the Leveler's pitching disappointments shortly.

Before looking at negative cases, take a peak at the second place Team Los Del Rio who managed to dramatically outperform the forecast. Team LDR started the season at a major disadvantage since Paul took over as manager and had slim pickings for keepers other than a handsomely priced Roy Halladay ($20). The team was built on some fairly high priced players such as Mark Teixeira ($35), Kevin Youkilis for ($22), Ichiro Suzuki ($20) and Jason Bay ($20). The bets paid off with career years by Teixeira (39 HR / 122 RBI / 103 R), Youkilis (27 / 94 / 99) and Bay (36 / 119 / 103) with Suzuki ( 88 R / 26 SB / .386 OBP) performing exactly as advertised. Russell Martin ($17) performed below expectations in every category despite playing a full season, but still had a better season than most catchers. Garrett Atkins ($17) significantly disappointed, but the price reflects widespread belief in our league that he was poised for a rebound, but clearly not every gamble must pay off to be successful. The hidden value came in the form of Troy Tulowitzki ($8) who doubled his projection with 32 HR and Raul Ibanez ($10) who BP envisioned hitting 21 HRs, but actually hit 34.


Meanwhile, league winner * TINSTAAPP * had a more disappointing offensive performance by its drafted players across most categories than the league average. Nonetheless, the team placed in the top ranks of every offensive category. Here, it is very clear that what the base roster failed to offer, Matt went out and found in free agency.


On the other side of the field, the league average pitching statistics did not waiver nearly as much as the offense. These were each rate statistics, so it is more difficult to calculate the averages as clearly, but it appears that each stat was within a 4% band of expectations for the entire league, but the standard deviation was much larger among the individual teams.



As their name might suggest, * TINSTAAPP * did not aggressively pursue young pitchers. Instead, the team went after unheralded veterans including Jamie Shields ($20), Jon Lester ($12), Ervin Santana ($11), John Danks ($10), and Ryan Dempster ($5). Santana (5.03 ERA / 1.48 WHIP / 107 K) was somewhat disappointing, but Lester almost exactly balanced that out by performing above expectations (3.41 ERA / 1.23 WHIP / 225 K). A look across the entire pitching staff reveals fewer than expected Wins, but a nice offset from higher than projected K/BB. Nonetheless, * TINSTAAPP * finished the season with the most points from pitching of any team. Similar to the team's offensive performance, the pitching stats imply shrewd use of free agency and the waiver wire.


The Replacement Levelers were expected to dominate all pitching categories with modest-priced pitchers including Brandon Webb ($23), Ricky Nolasco ($12), Chad Billingsley ($12), and Matt Cain ($9). Webb pitched all of four innings before his season ended with a shoulder injury, which can only be chalked up to bad luck. The abbreviated season of Chris Young ($3), a control pitcher who projected to pitch 140 innings with a sub-4.00, could not have helped either, as Young went down with a similar shoulder injury after just 76 innings. Meanwhile, Nolasco did not pitch like the #1 pitcher for a contender as he posted an ERA over 5.00. Further, Randy Johnson ($7) threw like a 45 year-old with under 100 innings and an ERA near 5.00. Cain's career year (2.89 ERA / 171 Ks / 14 Ws) was simply not enough. Taken the Levelers put 20% of their budget on four players who were far below the replacement level.


The pitching category suggests the difficulty of dealing with bad players when it comes to rate stats. Ong Rising sunk 40% of its budget into pitching, including Johan Santana ($30), C.C. Sabathia ($29), Chris Carpenter ($9), and Adam Wainwright ($8). Those four CY Young candidates each were expensive, yet performed at or better than expected. Add in a great year by Clayton Kershaw ($5), the youngest pitcher in baseball, and its a wonder that Ong Rising did not blow out the pitching category. It seems the drag of Kevin Slowey ($12), Brad Lidge ($4), and Fausto Carmona ($4), each of whom threw up horrible stats, was too much to bear, but its very difficult to drop pitchers who seem like they may turn it around after another start or two.


Circumspection of The Madison Mohels yields some interesting data. They put only 20% of their budget into pitching, with $33 of the $46 committed to Roy Oswalt ($23) and Justin Verlander ($10). I've always wondered if it would be possible to punt several categories entirely to ensure victory elsewhere. The Mohels hitting surely came through, but they were projected to be the worst pitching team, and they lived up to that billing. They stunk in pretty much every pitching category, except one, wins. Though I have not reviewed the roster additions and drops, this suggests a regular strategy of adding pitchers likely to win the next day's game, even if their rate stats could be poor. Combined with top stats in every hitting category, they achieved 5th place finish, which further suggests to me that our league may permit ignoring one category (SBs, Ks), but not several at once.


Lastly, let's review the worst of the worst -- PECOTA Monkeys. The team was projected by its namesake PECOTA to finish 8th in our league, but wound up in dead last by a wide margin. The team seems includes an all-star roster of every player that BP called out in print as a possible break-out all-star, but which PECOTA itself projected to be lackluster. The roster was stocked with crafty veterans who were supposed to have something left in the tank, such as Grady Siezemore ($37), Rafael Furcal ($13) and Carlos Delgado ($11) combined with up-and-comers including Stephen Drew ($13), Carlos Quentin ($15), Joba Chamberlin ($15), and Scott Kazmir ($15). Yet it would have been very easy to conclude that Dave's judgment was shot the second he bid $10 for Lastings Milledge.


This concludes Part I of the 2009 Review. This provided some insight into whose drafted players lived up to expectations and which team were bitten by bad luck. If I have time and there is interest, I will next take a look at the raw data of projections for the drafted team versus where each team finished to see if we can find any info on who had the best post-draft.