In my previous post, I wrote about the 2010 best and worst fantasy baseball draft values. I thought it would be interesting to analyze the results of last year’s 21 Razzball Commenter Leagues to see how well these rankings lined up with fantasy baseball team success. The format for all these leagues was: 12 team, MLB, snake draft, 5×5.
For simplicity’s sake, I used draft results for assigning each player to a specific team. Obviously trades happen, players get added/dropped, etc. But sifting through the data to credit team A with said player because he had 290 AB vs. 250 AB on Team B is a much larger exercise. I focused only on players drafted in 50% or more of the leagues and went through the final stats per team to assign the following key free agent pickups: J. Axford, J. Bautista, T. Cahill, J. Garcia, A. Huff, C. Lewis, B. Myers, A. Pagan, B. Posey, A. Torres, and CJ Wilson.
I created three stats to rank players using the 2nd and 3rd only for tiebreakers:
1) % of Top 3 Finishes – How many times did a team with this player finish in the top 3 of their league?
2) Median Finish
3) Average Finish
The full results can be found here. The top players for 2010 were:
- Paul Konerko – 61.9% Top 3 Finish (Median=2nd)
- Dan Uggla – 61.9% (Median=3rd)
- John Axford – 57.1% (Median=3rd, Avg=3.7)
- Buster Posey – 57.1% (Median=3rd, Avg=4.2)
- Carlos Gonzalez – 57.1% (Median=3rd, Avg=4.4)
- Chris Perez – 52.9% (Median=3rd, Avg=4.8)
- Raul Ibanez – 52.4% (Median=3rd, Avg=4.4)
- Robinson Cano – 52.4% (Median=3rd, Avg 4.7)
- Rickie Weeks – 47.6% (Median=4th, Avg 4.9)
- Neftali Feliz – 47.6% (Median=4th, Avg 5.0)
Seven of the above players finished in the top 50 of the Best Values with Posey’s 72nd place finish a bit misleading since he started in May. Cano is 75th overall on the Best Value list but he is 3rd in value amongst top 50 picks (Halladay and Votto are ahead of him). The only fluke in the bunch is Raul Ibanez.
Here are the bottom 10 players:
- Brad Hawpe – 0% Top 3 finish (Median=10th)
- Yadier Molina – 0% (Median=9th)
- David Ortiz – 4.8% (Median=10th, Avg=8.9)
- Tim Lincecum – 4.8% (Median=10th, Avg=8.9)
- Scott Kazmir – 4.8% (Median=9th, Avg=9.0)
- Huston Street – 4.8% (Median=9th, Avg=8.3)
- Todd Helton – 4.8% (Median=9th, Avg=8.0)
- Stephen Drew – 4.8% (Median=8th, Avg=8.1)
- Adam Lind – 4.8% (Median=8th, Avg=8.0
- J.A. Happ – 4.8% (Median=7th)
Four of these players finished in the top 50 of Worst Values (of players with ADP<212): Hawpe, Kazmir, Helton, and Lind. Ortiz is the only player who finished in the top half of value – I think this may be because several teams who drafted Big Papi dropped him after his awful start so they didn’t benefit from his resurgence.
There are a myriad of theories that can either be created or tested from this data set. I looked into several that I have detailed below. Feel free to look at the data yourself and add your thoughts under comments. (If you own a blog, feel free to provide your post URL in the comments. Just make sure to add a link to this post and note the data came from Razzball Commenter Leagues.) Just one word of caution – a sample of 21 leagues is more directionally significant vs. statistically significant. Anomalies should be expected working with a sample like this (e.g., how could teams that drafted Rollins finish in the top 3 more than teams that drafted Tulo or H-Ram?)
You Can Overcome a Bad Top 3 Pick – It’s definitely possible to overcome a bad pick in your top 3 rounds – but it’s not recommended. I know it’s statistically anomalous but 8 of the 21 teams that drafted Jimmy Rollins finished 1st, 2nd, or 3rd. Only teams with V-Mart, Zimmerman, Mauer, and Pujols had more top 3 finishes among the top 36 picks. Even teams that drafted Jacoby Ellsbury or Mark Reynolds – the absolute bombs of the top 36 picks – managed 3 top 3 finishes.
You Can Overcome two Bad Top 4 Picks – Of the top 48 picks, one-third (16) were below average for their position based on Point Shares (Granderson, Rollins, Haren, Pedroia, Sizemore, Kinsler, Sandoval, Morneau, Utley, Ellsbury, Youkilis, Reynolds, Upton, Greinke, Bay, Lind). The average finish of teams that drafted two or more of these picks was 7.2 (average would be 6.5) with 6 first place finishes, 7 second place finishes, and 2 third place finishes. There were 15 top 3 finishes compared to 28 bottom 3 finishes. Granderson and Rollins were drafted on 2 winning teams. The other four combinations were Sandoval-Morneau, Haren-Pedroia, Kinsler-Sandoval, and a Granderson-Youkilis-Kinsler trifecta. So missing on two of your first four picks isn’t a suggested path but it’s not a death sentence either.
Draft Slot Not Important – The Razzball Commenter Leagues indicate that draft position does not have a major impact on team success. There were winners coming from all 12 draft positions and only one draft position (1st) managed even 3 winners. My favorite position of 4th finished worst with a median finish of 9th. This is despite Ryan Braun being the most common pick at that position which is a lot better than 5th (Utley) or 7th (Kemp). Below is a ranking of draft picks based on median team finish (with average finish as tiebreaker):
1 – 10th draft position (Median 4th place finish, 2 winners)
2 – 1st draft position (Median 5th place finish, 3 winners)
T-3 – 3rd draft position (Median 6th place finish, 1 winner)
T-3 – 11th draft position (Median 6th place finish, 2 winners)
5 – 6th draft position (Median 6th place finish, 1 winner)
6- 9th draft position (Median 6th place finish, 2 winners)
7 – 8th draft position (Median 7th place finish, 1 winner)
8 – 5th draft position (Median 7th place finish, 2 winners)
9 – 12th draft position (Median 7th place finish, 1 winner)
10 – 2nd draft position (Median 8th place finish, 2 winners)
11 – 7th draft position (Median 8th place finish, 2 winners)
12 – 4th draft position (Median 9th place finish, 2 winners)
Drafting SPs in Top 3 Rounds – As some of you may know, I am a bigger proponent of drafting pitchers in top rounds than most bloggers. This wasn’t a winning strategy AT ALL in 2010 RCL leagues though – at least for the first 3 rounds. There were 5 SPs taken in the first 3 rounds of RCL leagues: Lincecum (13), Halladay (19), F-Her (26), Greinke (28), and Sabathia (30). Halladay drafters managed 6 top 3s which is slightly above average (21 leagues * 25% of teams finish in top 3 = 5.25 top 3 finishes). Lincecum, F-Her, Greinke, and Sabathia drafts managed 7 TOTAL top 3 finishes (2 for all except 1 for Lincecum). The Greinke/Lincecum finishes might make some sense but F-Her?
I dug in deeper and found that, in 8 leagues, one or more teams drafted 2 of these SPs. These ten teams had the following finishes: 4th (twice), 7th (once), 8th (thrice), 9th (twice), 10th (once), 11th (once). So it’s certainly clear that you should NEVER draft two SPs in the first 3-4 rounds. But that doesn’t fully explain why teams that drafted SPs in the first 3 rounds (espec. F-Her teams) finished so poorly.
If someone wants to pore through the 21 leagues to figure out, that would be great. My hypothesis is that it is harder to draft SPs early and catch up on 1B/OF power than the reverse. I think it can be the more successful strategy if applied right in the typical league. But, if you’re going to deploy this strategy, I would suggest doing a number of mock drafts and seeing how well you can draft offense in later rounds. (For an example of how this could work, I did an expert NL-only snake draft last year and picked Lincecum and Reyes as my top 2. Obviously, those two players did not meet my expectations yet I still finished 2nd because I found power at a cheaper price later in the draft – Werth, Uggla, LaRoche, Garrett Jones, Cody Ross, Barajas, picked up Stanton as a free agent)
Avoid The SB Specialists In Top Rounds – There were 7 players drafted in the first 6 rounds that were safe bets for less than 20 HRs and more than 20 SBs: Crawford (16), Ellsbury (23), Reyes (39), Ichiro (43), BJ Upton (44), Roberts (55), Figgins (72). All of these 7 players finished below average (5 or less) in top 3 finishes. As you’d expect, Crawford and Ichiro did best (5 top 3s each) while the rest had 3 top 3s except for Figgins with one measly 3rd place finish (the worst success rate of all top 6 round picks).
Drafting 2 MIs In First 6 Rounds – I prefer to not draft more than 1 2B/SS in early rounds so I thought I’d check to see how teams fared that drafted 2+ MIs in the first 6 rounds. There were on average 3 teams per league that used this draft strategy and the distribution of team finishes is almost completely even between 1st through 12th (so very close to a 6.5 average)
Relief Pitcher strategy – Relief Pitchers represent 7 of the top 20 and 10 of the top 36 players ranked by their drafted teams’ performance. John Axford (I counted teams who had the most IP for him), Chris Perez, and Neftali Feliz all had 9 or more top 10 finishes. So is there a way to predict top value closers? Pretty much, no. I’d say that closer value is dependent on six variables: 1) Opportunity to Save Games, 2) ERA/WHIP, 3) Strikeouts, 4) Wins, 5) Saves, and 6) Health. The first three are easier to project than the other three. Strikeouts and ERA/WHIP are fairly correlated with high BB pitchers being what I’d think is the most likely exception. So I’d focus on targeting high K closers with some opportunity to save games. I also advise to overdraft this position in terms of quantity and include any set-up guy with 70+ K potential. Lastly, I would avoid drafting one of the first 3 closers unless you felt confident of 90+ K. I’d rather build up in other positions in those early rounds and go for volume vs. quality on closers.
Draft Top Catchers or Punt? – There are always a handful of catchers every year that go in the first 100 picks. I generally punt catcher until later rounds for a number of reasons, most notably that catchers have higher injury rates. (This is in snake drafts only. I don’t mind investing for catchers in auction drafts if I’m getting enough of a discount.). Last year, there were four catchers picked on average in the top 100 picks within Razzball Commenter Leagues: Mauer (18), V-Mart (48), McCann (49), and Wieters (92). While neither Mauer and V-Mart had sensational years, they provided very good value for their draft position and finished on 9 and 10 top-3 teams, respectively (note: one 2nd place team drafted both). McCann teams fell slightly below average with 4 top-3 teams while Wieters owners had 3 top-3s. While that seems to average out, just note that only teams that picked up Posey fared better, on average, than teams with Mauer and V-Mart. I’d argue drafting a top catcher was a good bet for 2010 but still wouldn’t advise it for 2011 unless you get a good discount – a good rule of thumb for one catcher leagues is to take Catcher ADP and multiply it by some multiple of 1.25-1.5 – e.g., if V-Mart is at an ADP of 50th and you think that’s his true value, don’t make a move on him until he drops to 63-75.
High Upside OFs Are Great, Right? - When a player like Carlos Gonzalez has a breakthrough year (2nd on estimated value vs. draft position, 5th place based on team performance in Razzball leagues), the natural reaction is to draft several high-upside players in next year’s drafts. Just a word of warning, though. Here are some of the other high-upside OFs coming into 2010 and their RCL team performance rank: Travis Snider – 96th, Jason Heyward – 108th, Colby Rasmus – 164th, Adam Jones – 237th, Jay Bruce – 239th, and Justin Upton – 269th. I’m not saying to avoid high-upside plays – just don’t overpay for them with the thought you’re going to get the next CarGo.
Free Agent Strategy – The 1st/2nd/3rd place finishers averaged slightly more than 1 of the top 10 free agents. But while Jose Bautista was the highest value free agent, he only appeared on 5 top-3 finish teams. On the other hand, John Axford and Buster Posey each appeared on 12 top-3 teams (including 8 1st place finishes for Axford!). Why? Many ‘good’ teams didn’t have room for Bautista at 3B, OF, 1B/3B, or UTIL. Poorer teams had more holes and were more likely to gamble on what appeared to be a fluke. Axford and Posey – on the other hand – are in positions where many teams are open to an upgrade. Starting pitchers seem to finish somewhere in between with Trevor Cahill and Jaime Garcia finishing on 8 and 7 top-3s while CJ Wilson, Colby Lewis, and Brett Myers finished on 4-5 top 3s. This seems to fit in with typically recommended free agency behavior to: 1) Pick up any closer and 2) Don’t be afraid to drop late draft picks – especially at catcher – if a higher upside play presents itself.