SubscribeFor those of you who have clamored for an infield position (i.e. any of 1B, 2B, 3B, and SS) added to the Price Guide: Your pleas have not gone unanswered.
As expected, adding an IF slot will typically give the same results as just adding an extra CI.
A couple of other recent changes:
- Teams and positions have been added for lots of minor leaguers.
- Stats for any hitter who has never played in the majors are reduced by 25%.
- Replacement levels have been tweaked a little bit for CI, MI, Util, etc.
Related posts:
What’s your new formula for determining replacement level for the generic positions?
Let’s say these 12 players are used for CI:
Encarnacion 3B
Loney 1B
Stewart 3B
LaRoche 1B
Guillen 3B,1B
Figgins 3B
Konerko 1B
Glaus 3B
Kouzmanoff 3B
Butler 1B
Cantu 3B,1B
Gordon 3B
Previously, Butler’s value would be used as the replacement level for 1B, and Gordon’s value would be used as the replacement level for 3B. Now, Gordon’s value is used for 1B and 3B. (Basically, the generic position replacement level is used for all of the positions that would put players at that position.)
My thinking was something like this:
- The difference between Butler and Gordon is negligible, so using one replacement level is cleaner.
- There is enough overlap between 1B and 3B (Guillen, Cantu, etc.) that it makes sense to consider them essentially the same position.
Mays, can you remind me what to add on to the end of the web address to just have the players eligible at the position they played the most at the prior year?
I have a strange league where our roster is 30 players – one from each major league team. My dollar limit is $650 and I’m given a list of prices for each player based on their last 2 years of play. There are 7 categories for hitters and 4 for pitchers – several of which you’ve set up for me. There are 18 hitters, 2 at each position including DH, and 1 bench hitter. There are 4 starters, 2 relievers, 4 either and 1 bench pitcher. You can elect to swap out your bench players for a hitter and a pitcher at the all-star break and throw out the stats of the player you replace. This is a league with about 25 teams – and each team can have the same players.
All this to ask – is there a way to use the price guide in this scenario to optimize my roster? As it is currently, I’m just running the price guide based on a 12 team league, and comparing the value of each player to their assigned value and trying to take the ones with the greatest positive variance (within the $650 and position constraints).
Thanks, again, for all the great work you do.
quick suggestion – any way to not show or count the players with no teams when running an AL or NL league? Thanks.
I’m getting really goofy replacement levels now. Depending on what projections I use I am getting the following replacement levels.
C: -6.76
SS: -4.99
2B: -4.99
3B: -3.25
OF: -3.25
1B: -3.25
MI: -4.99
CI: -2.73
Util: -3.25
SP: -4.19
RP: -0.07
SS=2B=MI – Makes Sense
3B=1B – Makes Sense but > CI Doesn’t Make Sense
Also 3B/1B=OF – Seems Too Coincidental
Finally Util>CI – Util should always be smallest since Anyone can play it.
C: -5.88
SS: -4.15
2B: -4.15
3B: -2.64
OF: -3.04
1B: -3.04
MI: -4.15
CI: -2.64
Util: -3.04
SP: -4.65
RP: -0.21
Now OF=1B wich is > 3B – Seems wrong
Util is still > CI which never makes sense.
Do these make any sense?
In your first example, the CI level won’t actually be used by any positions, because the Util takes precedence. The Price Guide fills positions in this order:
3B
1B
CI
Util
So, initially, the replacement level for 1B is the 12th-best 1B. Then, the Price Guide looks for the next 12 1B or 3B and adjusts the 1B replacement level after that. Then it finds the next 12 best players at any position for Util, and lowers the replacement level for all of their positions.
Your second example indicates that there were no 3B found in the 12 players used for Util, only 1B and OF. While not typical, that does seem like it would be possible, and the Price Guide would have given you similar results before.
Hello,
Amazing site. Thanks for what you’re doing.
Quick question — I ran the model for my league settings and the output noted that it had used “IP per team: 813.25.” Is it at all a problem that in my league the teams tend to make sure they approach the league IP maximum of 1350 in order to be competitive in W, K, and QS?
Question #2 (sorry for such a basic question that I’m sure you’ve answered before): should I account for the number of bench slots in my league anywhere? Most teams use those bench slots to stock up on extra SP to get to the IP limit. Thanks very much again.
Regarding using 2008’s QS totals: you’re right that it’s better than nothing. But it causes some weird results for players who are projected to make many more starts than they did last year. It might be better to go through and figure that a good SP gets about 0.09 QS/IP (which is a rough average I made from last year’s stats) to create a very rough QS projection that’s more correlated with projected IP.
Ran the numbers slightly more carefully — looks like last year’s top 100 starters got 0.091 QSs per inning. The top 30 starters got 0.102 QS/IP, and the top 50 0.0988. Depends on how you define the universes, but maybe some good ballpark numbers to work with in there.
@Hegotgame: If your league using P (either SP or RP), I recommend adding SP instead to get the IP up. So for a league that starts 12 P, using 7 SP and 5 RP in the Price Guide gets the IP pretty close.
For other leagues, it’s common to keep a couple starters on your bench that you rotate in on off days. In this situation, you would want to count those spots as starters, since that’s how they function. So a 12 P league might then become 9 SP and 5 RP.